LS/fluw #22131" Q1 IROZEN RED CELL OUTDATING (747’ f0” wig" Mic/m jgzfiaé? DEC 3 1. 1": U.S.S.D. DISCRIMINATION PROHIBITED--Title VI of the Civil Rights Act of 1964 states: I'No person in the United States shall, on the ground of race, color, sex, age, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.” Therefore, the Division of Heart and Vascular Diseases, NHLI, like every program or activity receiving financial assistance from the Department of Health, Education and Welfare, must be operated in compliance with this law. ERDZEN RED CELL OUTDATING March 14, l975 Bethesda, Maryland Sponsored by THE DIVISION OF BLOOD DISEASES AND RESOURCES NATIONAL HEART AND LUNG INSTITUTE NATIONAL INSTITUTES OF HEALTH Edited by: PETER BJL§HERER, M. D. Blood Resources and Transplantation Branch Division of Blood Diseases and Resources National Heart and Lung Institute Bethesda, Maryland 200l4 DHEW Publication No. (NIH) 76—1004 M7 _ CONTENTS INTRODUCTION Frozen Red Cell Outdating Peter B. Sherer Storage of Red Blood Cell Products: Background of 24-Hour Dating Lewellys F. Barker and Madge L. Crouch DEVELOPMENT AND TESTING OF STERILE CONNECTORS An Aseptic Fluid Transfer System to Maintain Blood Sterility Richard M. Berkman, James C. Arnett, Edward L. Cleland, and Michael D. wardle A Two—Stage, Heat Sterilizable Connector for Use in Processing Frozen Blood William N. Cooper, Charles E. Huggins, and Mark Weinberg Sterile Coupling Device for Deglycerolized Red Blood Cells Stanley C. Roberts, Dennis T. Reuss, and Gerald L. Moore A Pull Tab Connector for Sterile Transfer of Blood and Biological Solutions Arthur w. Rowe, E. T. Marshall, L. L. Lenny, and J. R. Graham Sterile Blood Connector Francis J. Tenczar A Dry Heat Sterilizable Blood Bag Tubing Connector Thomas R. Hektner and John G. Christensen, III €V82. lli Q P 96 F75 PUBL 17 31 43 53 61 71 BACTERIOLOGIC AND METABOLIC EVALUATION The Growth of Bacteria in Frozen-Reconstituted Blood and Methods for Detection of Contamination Byron A. Myhre, Yutaka Nakasako, and Richard Schott Cultural Studies of Previously Frozen and Washed Red Cells Paul T. Wertlaké and Shirley E. Wertiake Sterility Test of Rotating Disposable Seal on the IBM 2991 Blood Cell Processor Frederick R. Kronenwett Bacteriological and Biological Quality of Frozen Red Cells 72 Hours Following Deglycerolization H. T. Meryman and R. A. Kahn iv 81 93 101 105 INTRODUCTION FROZEN RED CELL OUTDATING Peter B. Sherer (Division of Blood Diseases and Resources, National Heart and Lung Institute, National Institutes of Health, Bethesda, Maryland) The 1949 discovery by Polge, Smith, and Parkes (2) that glycerol had the ability to preserve living cells during freez- ing, storage, and thawing marked the beginning of modern cryo- biology. In 1950 Smith extended this work, finding that human erythrocytes also could be successfully preserved in this fash— ion (3). Mollison and Sloviter in the following year discovered that glycerolized red cells could be frozen for several hours, thawed, and deglycerolized with little hemolysis and could then be transfused with normal in vivo survival (1). Since that time a wide variety of procedures and preservae tives have been employed. Cryoprotective agents investigated have been penetrating (intracellular) such as glycerol and glu— cose, and nonpenetrating (extracellular) such as dextran, lactose, polyvinylpyrrolidone (PVP), and albumin. A number of effective and clinically safe techniques for red cell cryopreservation have been developed, and the "low glycerol-rapid freeze," "high glycerol—slow freeze,“ and “cytoagglomeration” methods are the most commonly used today. All of these use glycerol as the cryo- protective agent but differ in the concentration used, the rate of freezing, the temperature of storage, and the method by which glycerol is introduced and removed from the cells. Indeed each method has its strong proponents and each its own merits. Although these methods vary considerably, they have two things in common: the necessity of adding glycerol prior to freezing and the necessity of removing the glycerol after thaw- ing. This requires multiple entries into the blood bag, with violation of the closed sterile system and the restriction of the post—thaw shelf life to twenty—four hours. This short shelf life has been responsible for the loss of many otherwise transfusable ,units of blood which had been thawed but not transfused. Be- cause of this, there is often a reluctance to thaw units until just prior to use, and sometimes blood is not readily available in emergencies since it was not thawed in time. These problems could be partially solved if frozen-deglycerolized red cells could be stored for longer periods. Clinical experience in several institutions has suggested several possible advantages of frozen—thawed red cells.. 1) Inventory Control. Vagaries in blood collection and un— certainties in demand contribute to episodic cycles of blood shortages and excesses. The prolonged extension of the shelf life offered by freezing may be an effective means of circum- venting these variables and insuring an adequate, relatively stable supply of blood. Excess blood could be frozen during per- iods of surplus for future use when inventories are depleted. 3 2) Long-Term Storage. Freezing permits the long—term stor- age of red blood cells, with good maintenance of both post- transfusion survival and hemoglobin function. This enables the blood bank to store units of cells with rare antigenic combina- tions for future transfusion into immunized recipients or to store an individual's own cells for possible future autologous transfusion. 3) Leukocyte Removal. Freezing, thawing, and washing have been shown to be an effective means of preparing leukocyte— (and plateleth) poor blood. Such leukocyte removal is important in the transfusion therapy of patients who suffer from febrile, non— hemolytic transfusion reactions, which are largely the result of immunization to leukocyte antigens. Leukocyte removal apparently decreases the incidence of histocompatibility sensitization in tranSplant recipients; and while recent controversy has been raised over the importance of sensitization in graft survival, the transfusion of possible transplant recipients continues to be a major use of frozen red cells. 4) Plasma Protein Removal. Frozen, thawed, and washed red cell preparations are essentially devoid of plasma proteins. Transfusion reactions resulting from previous immunization to plasma proteins, particularly IgA, and from protein—bound exogen- ous materials, such as drugs, may be eliminated by use of frozen- washed cells. Additionally, the total absence of anti-A and anti—B isoagglutinins makes group 0 cells truly universal donor cells. 5) Reduced Risk of Hepatitis. Several investigators have noted the absence of post-transfusion hepatitis in patients who had received frozen—washed red cells. Many of these reports have been of the anecdotal variety, and a well—designed study of the possible decreased incidence of post—transfusion hepatitis with frozen cells and an elucidation of the mechanism involved are needed. While many of these advantages must be investigated more completely, frozen-deglycerolized red cells, nevertheless, remain quite attractive, and the attendance at this symposium is testi— mony to the increasing interest in their use. The paper by Dr. Barker discusses the 24vhour outdating period more fully and traces the development of this regulation. The development and evaluation of a variety of connecting devices which would ensure sterility during the various stages of frozen blood processing are discussed by Dr. Berkman, Dr. Cooper and Dr. Huggins, Capt. Roberts, Dr. Rowe, Dr. Tenczar, and Mr. Hektner. Because these devices can maintain a closed sterile system, they may play a significant role in extending the outdating period past twenty-four hours. These connectors, furthermore, could be used for any sterile transfer procedure in the blood bank, such as component preparation, and their ultimate use most likely will not be limited to frozen cells. Dr. Myhre reports on the incidence of contamination in frozen cells, the fate of experimentally introduced bacteria, and a possible method for rapidly determining the presence of contamination. Several methods of deglycerolization are currently in use, many of them employing mechanical devices such as cell separators and centrifuge bowls. The development of sterile connecting de- vices would be useless in frozen blood applications if the wash- ing devices permitted contamination, and this aspect of the pro- cessing procedure is considered by Dr. Wertlake and Dr. Kronen— wett. Finally, Dr. Meryman reviews the Red Cross data on contami— nation and discusses the metabolic integrity of the processed red cells. REFERENCES l. Mollison, P. L., and H. A. Sloviter. Successful transfusion of previously frozen human red cells. Lancet 2:862, 1951. 2. Polge, C., A. U. Smith, and A. S. Parkes. Revival of sperma- tozoa after vitrification and dehydration at low temperatures. Nature (London) 164:666, 1949. 3. Smith, A. U. Prevention of haemolysis during freezing and thawing of red blood cells. Lancet 2:910, 1950. STORAGE OF RED BLOOD CELL PRODUCTS: BACKGROUND OF 24-HOUR DATING Lewellys F. Barker and Madge L. Crouch (Division of Blood and Blood Products, Bureau of Biologics, Food and Drug Administra- tion, Bethesda, Md.) Dating of refrigerated red blood cell products for trans- fusion has varied over a range from 24 hours to 30 days during the past 25 years. Determinations of satisfactory dating periods haVe been based on Ln vitro tests of red cell function, Ln vivo survival, and empirical observations of the safety and effective- ness of the infused red blood cells. In addition to integrity of functional red blood cells, a major consideration in assigning dating to red cell products has been the problem of bacterial contamination during collection and processing with its conse— quent severe, sometimes fatal, adverse reactions in transfused patients. The transition from the use of glass bottles which were either vented or under vacuum to pre—sterilized plastic con- tainers reduced the problem of bacterial contamination of red blood cell products to a relatively rare occurrence. In the case of frozen red blood cells, however, dating following the thawing and deglycerolizing process has been limited to 24 hours, pri— marily because of concern over the possibility of bacterial con— tamination during processing. Favorable experience with modern techniques for processing frozen red blood cells should pave the way for extending the dating of these products, a step which would greatly enhance their acceptance and utility in transfusion therapy. An attempt to trace the history and basis of rules and rec— ommendations for the storage of red blood cell products for transfusion has revealed a number of interesting vicissitudes, for which explanations are not aways readily available in the published literature or regulations. This report will first ad- dress chronological events leading to current regulations, and will then review some of the documented explanations of how the regulations developed. Finally, proposals will be made of some possible approaches for present and future consideration in making decisions on the dating of red blood cell products. The subject of storage and preservation of whole blood and of red blood cells suspensions is dealt with in considerable de— tail in the text on Blood Transfusion by DeGowin, Hardin, and Alsever whiCh was published in 1949. In introducing this subject they stated that: The outdating periods of various blood mixtures are al- ways arbitrary and empiric... Regardless of the pre- servative solution employed or the time of storage the rapid osmostic fragility test should be mandatory before transfusion of dextrose-citrate mixtures. Outdating periods can only be considered as average estimates. (6) 'The authors provided storage guidelines for fourteen different preservative mixtures which contained varying combinations of 7 citrate and dextrose and also for citrated blood without dex- trose. The average outdating period cited for all but two of the citrate-dextrose solutions was 21 to 30 days; the remaining two were assigned 14 or 16 to 21 days, whereas citrated blood was as— signed five to ten days. In another section of the same text, storage recommendations are provided for red blood cell suspensions which were prepared by aspirating the plasma. Following such aspiration, the red blood cells were to be aspirated into a new vacuum container, or the collection container was to be resealed with a new sterile stopper. Storage recommendations were as follows: 1) up to five days for citrated cells, 2) up to ten to fourteen days for ci- trated cells from 500 ml of whole blood with 250 ml of a dextrose preservative added, 3) as long as 21 days for citrated cells from 500 ml of whole blood with 250 ml of 10 percent corn syrup added, and 4) up to the outdating period allowed for cells from whole blood collected in one of the citrate-dextrose preservative solu— tions (maximum 20 days) (7). Laboratory studies of stored red blood cells reviewed in the above text included cytologic changes, spontaneous hemolysis, osmotic fragility, mechanical fragility, glycolysis, potassium levels, agglutinogens, and oxygen—carrying capacity. Although there have been many scientific advances in the study of red cell preservation, workers in the 1940's considered the criterion of at least 70 percent survival of red cells 24 hours after transfu- sion as the major criterion of acceptable preservation, and this test remains to this day the cornerstone for assessing function of stored red blood cells. Turning now to Public Health Service regulations for storage of red blood cell products, we find in Minimum Requirements, December 9, 1945, for Citrated Whole Blood (Human): "Once the closure seal is broken the contents of the bottle should be used promptly or discarded" (11). In the same document an expiration date of 10 days was assigned to packed red blood cells stored at 4° to 10°C, with the admonition that the red cells be used promptly in the event that the seal of the bottle was broken to obtain a sample for cross—matching. Turning to some of the facts and events responsible for the course of regulations regarding storage of red blood cell prod— ucts, the transition from collection of blood in glass bottles to plastic containers must be viewed as the single most important advance in the prevention of bacterial contamination of blood bank products. Glass bottles, which were reused, presented prob— lems in cleaning and sterilization, and they possessed hard rub— 'ber stoppers; the stopper seals had to be broken at the time of collection, and, for further processing, the stoppers were often removed. Plastic sets, on the other hand, consist of sterilized collection and satellite containers connected by integrally at— tached tubing. It is possible to do extensive processing in these sets in a so—called "closed system," i.e., without breaking 8 the seal after the blood is collected. Plastic containers were first introduced into blood banking in the early 1950's and the 1960's, so that glass containers, although still used in a few blood banks in this country, have become somewhat of an historical anachronism. Although admonitions to use blood promptly after breaking the seal of the container were in minimum requirements in the mid—1950's, it is surprising that the 24—hour dating rule for red blood cells prepared in a system that required breaking the seal during processing was first promulgated in Additional Standards published in 1961 (1). By that time it was both practi- cal and desirable to prepare red blood cells in a pre-sterilized closed system in accord with current optimal practice. A striking feature of the literature describing bacterial contamination of red blood cell products, which was recently re- viewed by Myhre et a1. (12), is the cluster of publications in the 1940's and 1950's and the relative paucity of such publications in the 1960's and 1970's; one is certainly tempted to conclude that the switch to collection and processing in closed plastic systems had a great deal to do with the sharp drOp in reports of severe and fatal septic transfusiOn reactions. Death caused by bacterial contamination of transfused blood was ascribed to a wide variety of microorganisms; twenty examples of such case reports are listed in Table l (2,3,13-16). The majority of serious reactions due to bacterial contamination were caused by gram-negative bacilli which TABLE 1 Bacteria Associated with Fatal Transfusion Reactions Organism(s) Cases Reference Escherichia coli, Streptococcus and anaerobic bacilli (1) Officer, 1942 (13) g. 99;; (1) Whitby, 1949 (16) E. coli, P. aeruginosa (1) Whitby, 1949 (16) A. aerogenes (4) Whitby, 1949 (16) A. chromobacter (l) Borden and Hall, 1951 (3) Paracolobactrum aerogenoides (1) Borden and Hall, 1951 (3) Pseudomonas sp. (1) Stevens et al., 1953 (15) Alcaligenes faecalis and E. 991i (1) Pittman, 1953 (14) g. aerogenoides (3) Pittman, 1953 (14) "Hemolytic coli" (1) Pittman, 1953 (14) Pseudomonas and/or Escherichia freundii (3) Pittman, 1953 (14) Pseudomonas (l) Pittman, 1953 (14) Serratia marcescens (1) Black et al., 1967 (2) were capable of multiplying at 4° to 8° C. Consequently, a number of studies were performed on the growth of these cryophile bac- teria, and these studies appear to provide the best scientific basis for the 24-hour dating which is assigned to red blood cell products prepared in vented or open systems. Several studies doc- umented either a prolonged delay in the logarithmic phase of bac— terial growth or a slow but steady increase in bacterial counts in refrigerated blood (4,5,8-10,15). In addition, several of these studies demonstrated a capacity of stored whole blood to either limit bacterial growth or eliminate contamination. of particular interest regarding this phenomenon is the recent report by Myhre et al. of bacteriocidal activity in platelet concentrates (12). The early studies of antibacterial activity of whole blood as well as this recent study of platelet concentrates have shown the value of the long-recognized capacity of both white blood cells and plasma to limit or stop bacterial growth in enhancing the safety of blood for transfusion. In the case of the closed systems currently used for collec- tion and processing, the possible sources of bacterial contamina- tion include bacteremia in the donor, unsterile air to which the phlebotomy needle is exposed, and the skin at the phlebotomy site. Each of these sources would appear to be a potential source of rather low level contamination, and the same can probably be said for the numerous sites of connections which must be made in a typical system for processing frozen red blood cells (Figure 1). An additional potential source of bacterial contamination in the post-thaw washing of red blood cells is the rotating seal in the systems which employ centrifugation of cells and wash solutions. Despite these many potential sources of low level contamination, the experience of several blood banks which have made large num- bers of units of deglycerolized red blood cells for transfusion has been quite favorable with regard to reactions in patients, as summarized in Table 2. It should be pointed out, however, that approximately three quarters of the bacterial cultures on the units summarized in this table were made at the beginning of the 24—hour dating; that is to say, immediately after washing the cells. Also these results were not obtained in a manner that would discriminate between laboratory contamination incurred at the time of culturing the units as opposed to intrinsic contamin- ation of the units themselves. Finally, it seems desirable to consider what steps might be most useful in establishing the suitability of dating longer than 24 hours for frozen red cells after they are thawed for transfu- sion. Possible measures to minimize the risk of introducing bac- teria during processing include the development of "sterile docking devices," a project supported by the National Heart and Lung Institute's Division of Blood Diseases and Resources, de- velopment of centrifuge seals which will exclude bacteria, and connecting cells and solutions to centrifuge containers in a sterile environment. The advent of vertical and horizontal lami- nar flow hoods, which are now widely used in infectious disease and cell culture laboratories, provide a practical means of 10 collection l"89 (—7 Glycerol *Connoctionpoinu — pro-and pon-fmezingmdbloodcolb FIGURE 1 The single open point (x) in collection of red blood cell products for transfusion in a "closed system" (left) contrasted with the mu1tip1e open points or connections required for processing frozen and deglycerolized red blood cells (center and right). TABLE 2 Combined Experience of Three Blood Banks with Transfusion and Sterility Testing of Frozen Red Cells Total Units Processed 19,835 Total Units Transfused 19,355 Cases of Septicemia in Transfused Patients 0 Total Samples Cultured 1,193 Positive Cultures 79 (6.6%) aInformation provided by Massachusetts General Hospital (1973-1974), University of Pennsylvania Hospital (1973-1974), and Greater New York Blood Center (1972-1975). 11 working in a sterile environment. Environmental monitoring by a regular air sampling program provides a valuable quality control check on the function of such equipment. In documenting the safety of red cell storage for greater than 24 hours after thawing and processing, it would be desirable to perform bacterial cultures at both the beginning and end of the storage of such cell suspensions and also to evaluate the ability of these cell suspensions to support bacterial growth at refrigerator temperatures. Since red cells that have been sub- jected to a freeze-thaw-wash cycle will be essentially devoid of plasma and functional leukocytes, which are the components of whole blood known to possess bacteriostatic and bacteriocidal properties, it is not possible to extrapolate from the data for whole blood or platelet concentrates to deglycerolized red cells with regard to their capacity to limit bacterial growth. A final consideration in extending the dating of frozen red cells after thawing is, of course, the ability of such red cells to survive in vivo and transport oxygen. Employment of well es- tablished in vitro and in vivo assays as well as clinical studies of effectiveness should provide data to resolve this issue. In View of the advances in red cell freezing and processing tech— nology that have occurred in recent years, it seems appropriate that there is considerable interest in attempting to extend the post—thaw dating of red cells made in this manner for transfusion. It also seems desirable to approach any decision in this area by building a sound scientific base that will either support or fail to support a decision to extend the current 24-hour dating; the public comprising potential consumers of such red cell products have clearly come to expect something more than the approach sum- marized by DeGowin, Hardin, and Alsever's statement that "The outdating periods of various blood mixtures are always arbitrary and empiric" (6). REFERENCE S 1. Additional Standards: Packed Red Blood Cells (Human). Fed— eral Register, August 11, 1961. 2. Black, W. A., A. Pollock, and E. L. Batchelor. Fatal trans— fusion reaction due to Serratia marcescens. J. Clin. Path. 20:883, 1967. 3. Borden, C. W., and W. H. Hall. Fatal transfusion reactions from massive bacterial contamination of blood. New Engl. J. Med. 245:760, 1951. '4. Braude, A. I., F. J. Carey, and J. Siemienski. Studies of bacterial transfusion reactions from refrigerated blood: The properties of cold—growing bacteria. J. Clin. Invest. 34: 311, 1955. 12 10. ll. 12. l3. 14. 15. 16. Chaplin, H., Jr., E. Chang, and R. W. Kolb. Report of rou- tine tests for psychrophilic and mesophilic contaminants in banked blood. Appl. Microbiol. 3:213, 1955. DeGowin, E. L., R. C. Hardin, and J. B. Alsever. Blood Transfusion. W. B. Saunder Co., Philadelphia, 1949, p. 327. Ibid, pp. 439-440. Edsall, G., and L. H. Wetterlow. Missed contaminations in biologic products: The role of cryophile bacteria. J. Bact. 54:31, 1947. Geller, P., L. Chandler, and E. Jawetz. Experimental stud- ies on bacterial contamination of bank blood. III. The sur- vival and growth of bacteria in bank blood incubated at 37°C for short periods before refrigeration. J. Lab. Clin. Med. 49:573, 1957. James, J. D., and E. J. Stokes. Effect of temperature on survival of bacteria in blood for transfusion. Brit. Med. J. 2:1389, 1957. Minimum Requirements: Citrated Whole Blood (Human). Federal Security Agency, National Institutes of Health, lst issue, December 9, 1945. Myhre, B. A., L. J. Walker, and M. L. White. Bacteriocidal properties of platelet concentrates. Transfusion 14:116, 1974. Officer, R. Blood storage on active service. Austral. New Zeal. J. Surg. 12:111, 1942. Pittman, M. A study of bacteria implicated in transfusion reactions and of bacteria isolated from blood products. J. Lab. Clin. Med. 42:273, 1953. Stevens, A. R., J. S. Legg, B. S. Henry, J. M. Dille, W. M. M. Kirby, and C. A. Finch. Fatal transfusion reactions from contamination of stored blood by cold growing bacteria. Ann. Intern. Med. 39:1228, 1953. Whitby, L. Blood Transfusions. Williams and Wilkins Co., Baltimore, 1949, p. 467. 13 DEVELOPMENT AND TESTING OF STERILE CONNECTORS AN ASEPTIC FLUID TRANSFER SYSTEM TO MAINTAIN BLOOD STERILITY Richard M. Berkman, James C. Arnett, Edward L. Cleland, and Michael D. Wardle (Jet Propulsion Laboratory, California Insti- tute of Technology, Pasadena, California) An Aseptic Fluid Transfer System (AFTS) developed for the National Institutes of Health appears to eliminate the problem of microbial contamination when fluid transfers are made from one container to another. The problem of maintaining asepsis is particularly notable, for example, in the processing of frozen blood and blood components. The AFTS concept was derived from sterile insertion techniques for planetary spacecraft. The present system employs Teflon tubing attached to blood bags or other containers. The tubes are melted and fused with a heat sealer—penetrator designed for this purpose. A small Kapton sheet bonded to the inner walls of the Teflon tube provides support and restricts the fusion and pene- tration operation to only the desired sites. In operation, heat is transmitted through the non-sealable Kapton layers, causing the fusion of the bare Teflon layers of two fluid transfer tubes. The heat sealer contains several raised surfaces which, under pressure, direct the flow of the Teflon such that openings are produced in the fused common walls of the transfer tubes. Ste- rility at the site of transfer is assured by the high tempera- tures of the sealing process (300°C). Test data indicate that aseptic fluid transfers are possible, and that adequate flow rates, low particulate contamination, and operational simplicity of the AFTS process are achieved. A clin- ical evaluation of the AFTS is now in progress. An Aseptic Fluid Transfer System (AFTS) has been developed to allow aseptic transfers of fluids in and out of blood con- tainers. These fluids must remain sterile since they are ulti- mately transfused into humans. Fluid transfer is a requirement in the rapidly expanding use of frozen blood and blood compon- ents. The AFTS, developed under contract to the National Heart and Lung Institute, represents a direct technology transfer from the NASA Planetary Quarantine Program. The present system was in- spired by Rothstein and Arnett (2), who devised a means to remove or replace spacecraft components aseptically by heat Sealing and penetrating plastic envelopes. A schematic representation of the new fluid transfer system is shown in Figure 1. To accomplish the transfer, tubes from two sterile containers are juxtaposed as shown in the figure and are fused together, thereby forming a sterile common wall. The com- mon wall is then penetrated, allowing the sterile transfer. The essential components of the transfer system (Figure 1) include 1) provision for adapting to conventional blood contain- ers, 2) transfer tubes which can be fused and heat sterilized, l7 r' ---------------- 1 I TRANSFER ASSY \ - I . ADAPTOR \ ’ Ill-II \ m i » ~~ \ ‘———OIlI-II , L... l ________ __._.__/ (—5.; — _ — _-| I HEAD ‘—1 I FLUID F LUID PENEIRATOR CONTAINER ACTUATOR l 6. CONTROL CONTAINER POWER 5. l CONTROL [— SEAtER ASSY J FIGURE 1 Aseptic fluid transfer system. The component parts of a system which allows sterile fluid transfers between containers are shown schematically within the outer dotted lines. 3) a sealer mechanism consisting of one or more heat sealer heads and a control unit, and 4) a penetrator system which produces a passageway between the fused tubes. For blood—related applica— tions, requirements in addition to that of sterility include 1) the absence of toxic materials and particulate matter, 2) the provision for unrestricted flow, and 3) a reasonably low cost. While a mechanism for penetration is not evident from Figure 1, there are in fact several ways to penetrate the common wall formed by the fusion of two transfer tubes. Figure 2 shows the essential features of an internal penetration system, incorpor- ating penetrators attached to the internal walls of the plastic transfer tubes. Prior to use, each tube would be connected to a fluid container and the entire unit sterilized. For restoring sterility to any recontaminated surfaces at the time of fluid transfer, the tubes would be placed into a heat sealer, depressed (Figure 2b), and heated to sterilizing temperatures, causing the common walls to melt and flow as shown in Figure 2c. The end re- sult would be two tubes sealed together, with an opening in the fused walls as shown. Considerable time was spent in defining an optimum approach to the problem of sterile transfer. We explored about a dozen different melting and penetration configurations and also evalu- ated various methods of applying heat, e.g., microwave and ultra- sonic systems. Before any developmental work was initiated, a "weighted comparison matrix" was developed to aid us in selecting 18 PENETRATOR ' +CONTAMINATED OUTER SURFACES HEAT SEAL TOOL FIGURE 2 Fluid transfer system utilizing an internal penetrator concept. Two containers are con- nected as shown: (a) Tubes from two containers are aligned. (b) The tubes are compressed and heat-sealed. (c) The tubes are “ww»wg fused and penetrated. %Vn ummé2%%%mmm the most promising approaches. The matrix contained a series of selection criteria (e.g., ease of sterilization, materials avail— ability, fabrication cost, mechanical strength) which could be applied to each concept. The values assigned for each criterion were obtained from feasibility tests, from engineering judgment, and from prior experience. 0f the fluid transfer concepts conceived, two basic designs were selected for development. In the final analysis, however, the system most effective from the standpoints of cost and reli— ability was a simple dual-wall material tube employing a pene- trator external to the tube. The External Penetrator Dual—Wall System The tube system employed in the final design prototype is shown in Figure 3. For this design, we employed thin-wall (0.25 mm) FEB Teflon tubes having a lO-mm I.D., with thin layers of Kapton [Kapton (polyimide) Film-Type F is a laminate of FEP Teflon and Kapton. Type F and pure Teflon FEP films are produced by E. I. duPont de Nemours & Co., Inc., Wilmington, Delaware 19848.] attached internally to the Teflon tube walls as shown in 19 KAPTON INNER LINER A <— IllllllllllllLlIlllllllllllllllllllllillll 1||IIA1|LLLH l' . . V. ,1 ,3 Y41H IIIIIlHlIIlllIIIIIIllIIIIIIllIlllllllIlllIIlIlIIl 1|“: ELCE 'T—l—rivIIIIHIIIIIII‘IIHr IllllllllllllIlllllllllllllllll FIGURE 3 Fluid transfer tubes utilizing an external penetrator system. The tubes are made of FEP Teflon having heat-resistant Kapton liners (stippled surfaces). VIEW A-A the longitudinal and cross sections of Figure 3. The FEP Teflon, a thermoplastic material, is molten over a temperature range of about 270° to 370°C. In the current AFTS, temperatures of 290° to 315°C are employed to achieve the connection. Thus the melt- ing, fusing, and subsequent penetration of the Teflon walls occur at temperatures high enough to assure sterility. Kapton, a ther— mosetting material, is physically and chemically inert at temper- atures well in excess of 350°C, and is employed to maintain the integrity of the tubing during the melting and fusion operation. The method for melting and penetrating the fused walls with an externally situated penetrator is illustrated in Figure 4. Figure 4a shows tubes 1 and 2 compressed by two brass heating plates. Two sheets of Kapton (stippled) are_seen lining the in- ner walls of the tubes. Tubes 1 and 2 are composed of FEP Teflon having wall thicknesses of 0.25 mm, labeled AlBl and A2B2, re- spectively. The Kapton layers, K1 and K2, have wall thicknesses of about 0.1 mm, and are mechanically attached to layers A1 and A2. When the heating plates are compressed and heat is applied, a direct Teflon-to-Teflon contact and fusion will occur only be— tween layers B1 and B2. Layer Al will not fuse to B1 because of the intervening Kapton layer, Kl. Similarly, A2 and B2 are pre- vented from fusing by virtue of an intervening Kapton layer (K2). At temperatures above 270°C, the FEP Teflon surfaces begin to melt and flow. A small protuberance brazed directly to the lower heater, the penetrator, distorts the Kapton film (K1) some— what and forces adjacent regions of Bl and B2 to flow laterally 20 .. ,. . .. ”ll/Ill I [III III], II A%fi\\\\\\\\\\\\\\\\\\ l”””l “\ Figure 4a VIEW B - B Figure 4b FLAP FORMED BY PENETRATING COMMON WALL Figure 4c FIGURE 4 Sequence of events for fluid transfer (external penetrator) . In operation, the tubes are positioned and (a) compressed, (b) heated and fused, (c) penetrated. when melting temperatures are attained. The total flow pattern under pressure is shown in Figure 4b (note arrows). Finally, the heaters are air cooled, the jaws are opened, and the fused tubes assume the configuration shown in Figure 4c. The fused transfer tubes have the following features: l)the outermost Teflon layers which were in direct contact with the heaters are thinned considerably, 2) the upper Kapton layer is un- changed, 3) the lower Kapton layer remains unchanged except for indentations produced by the raised penetrator, and 4) the Teflon layers (B1 B2) have been fused and penetrated, forming a passage- way for fluid flow (large arrow). 21 The key to producing an opening which allows adequate fluid flow is dependent on the choice of a suitable penetrator config- uration. Figure 5 shows a preferred penetrator design, with the H-shaped penetrator mounted on the heated block. A longitudinal section (B-B) through the penetrator and block was shown in Fig. 4a and b. Figure 6 shows a cross-sectional view (representing C-C, Fig. 5) of the fusion and penetration process. With an "H" penetrator design, a double flap is produced in the fused Teflon layers. The double flap produced by this configuration allows for bidirectional flow and prevents plugging. FIGURE 5 Heating block with "H" penetrator attached. LOWER HEATER FIGURE 6 Cross-sectional View (representing C-C of Fig. 5) of heat sealing process. VIEW C -C AFTS Hardware The present laboratory device consists of both the dual- wall tubing described above and a heat sealer/penetrator system. The hardware associated with the sealer/penetrator system con— sists of l) a sealer frame, 2) sealer/penetrator heads, and 3) a controller. The entire system is shown in Figure 7. The 22 FIGURE 7 Hardware for sealing and penetrating transfer tubes. The control unit for regulating time and sealing temperatures is seen at left, the heat sealer at right. sealer frame is seen to the right and houses the sealer/penetra— tor heads. The controller unit is contained in the cabinet to the left. It controls the heating and cooling cycle and indi- cates when the cycle is complete. The unit controls automati— cally a pre—set time and temperature cycle for heating, sealing, penetrating, and cooling. The sealer frame contains two movable jaws to which the heat sealer elements are attached (Figure 8). The jaws open and close FIGURE 8 Sealer frame with attached heaters. hydraulically. Figure 9 shows the jaws closed, and Figure 10 is a closeup of the closed jaws with the transfer tubing in place. The smaller-diameter tubes attached to the upper and lower seal- ers direct compressed air to the heater blocks for rapid cooling. The cooling process requires about 30 seconds. Figure 11 shows the raised "H" penetrator brazed to the lower heater. F IGURE 9 Heat sealer with jaws closed. FIGURE 10 Closeup view of Fig. 9. The small—diameter tubes entering the sealer heads provide for rapid air cooling after the seal has been completed. 24 FIGURE 11 Closeup View of lower heater showing raised "H" penetrator brazed to its surface. Once the entire heating cycle is complete and the jaws are opened, the fused tubes appear somewhat flattened (Figure 12), but fluid flowing through the fused tube system is essentially unrestricted. If the outer Kapton layer is cut open, the fused FIGURE 12 Two fused transfer tubes after penetration. Teflon walls can be seen (Figure 13), with the double flap pro- duced by the penetrator clearly visible. Scanning electron micrographs were prepared of the lower Kapton layer, which comes in direct contact with the penetrator (Figure 14). Note that while the Kapton film is creased following a seal/penetration, the surface remains intact. A micrograph of the fused and FIGURE 13 Cutaway View showing flaps produced in fused Teflon walls. FIGURE 14 Scanning electron micrograph of intact Kapton layer after sealing and pene— tration are complete. 26 penetrated Teflon layers (Figure 15) shows the desired opening produced by the external penetrator in the molten Teflon. FIGURE 15 Scanning electron micrograph showing dual flaps produced during seal penetration cycle. F—l rum—4 Effectiveness and Prognosis To evaluate sterility, several sheets of Teflon were inocu— lated with Bacillus subtilis var. niger spores. Subsequently, the sheets were heated to sealing and sub-sealing temperatures and were assayed for spore survivors. In a second test proce- dure, several flasks containing culture media, and a set of empty flasks, were fitted with transfer tubes and the systems were sterilized. The outer surfaces of the transfer tubes were then contaminated by spraying with suspensions of spores and vegeta— tive bacteria. Following sealing and penetration, media were transferred through the connected tubings into the empty flasks. The transferred fluids were incubated at 35°C for up to 10 days to verify sterility. These studies indicate that vegetative organisms at any con- centration will not survive heating above 290°C for 15 seconds. Killing data for spores revealed statistically only two spore survivors per ten fluid transfers in the presence of 107 viable spores. Thus the probability that spore contamination could occur in a relatively clean laboratory environment, assuming an upper limit of 10 spores per transfer site, is about 2><10'7. 27 Tests to evaluate the physical integrity of the tubing be- fore and after heat sealing have included various pull tests, leakage tests, tensile strength tests, and others. The evalua- tion on the basis of our current data indicates an acceptable level of strength and reliability for the intended use. Test results of liquid flow rates were well within tolerable limits. When pure glycerol was flowed through the transfer tubing, the flow rate was 55 :3 ml/min, adequate for the intended blood bank applications. The generation of particles at the seal/penetration was measured by passing clean water through the system and performing microscopic counts on filtrates. For these fluid transfers, fewer than five particles were obtained per 150 ml water. The few particles recovered have diameters of from 0.5 to 400 um, but the largest of these are believed to be derived extrinsically, either from the PVC tubing or the adaptors, and will thus not have a direct impact on the functional aspects of the system. The possibility of toxic outgassing from the molten Teflon was dismissed initially on the basis of prior knowledge that serious Teflon outgassing occurs only at much higher temperatures. Mass spectroscopic analysis was performed however, and the data indicated a trace, low molecular weight peak (MW42), not likely a fluorocarbon residue. Only slight increases in outgassing were observed when sample temperatures were raised to around 360°C, a temperature above the nominal operating range of the system. The AFTS employs materials all of which appear to be suit- able for blood freezing and processing (1). The present system seems to solve the problem of sterile fluid transfer success— fully, and is now being evaluated at the clinical level. If clinical data support our findings, industrial production of this system will be sought. A commercial version of the AFTS should be comparatively inexpensive both in terms of the transfer tubing and the sealer system, and should be relatively simple to operate. This paper presents the results of one phase of research carried out at the Jet Propulsion Laboratory, California Institute of Technology. The work was sponsored by the National Heart and Lung Institute, National Institutes of Health, and, by agreement with the National Aeronautics and Space Administration, conducted under NASA Contract NAS7-lOO. REFERENCES l. Akerblom, O, and C. F. Hogman. Frozen blood: A method for low-glycerol, liquid nitrogen freezing allowing different postthaw deglycerolization procedures. Transfusion 14:16—26, 1974. 28 2. Rothstein, A. A., and J. C. Arnett. Sterile insertion - An aerospace application of gnotobiotic technology, pp. 65-77. In Developments in Industrial MicrobiolOgY' Vol. 9. Society for Industrial Microbiology of the American Institute of Biological Sciences, Washington, D.C., 1968. DISCUSSION DR. SHERER: You mentioned that up to five particles were ob—. tained by passing clean water through the system. Is this significant? DR. BERKMAN. I'd like to emphasize that no particulate release was observed at the penetration site itself. Furthermore, par— ticulate release under the conventional method for entering a closed system with a stylet is as much as tenfold higher. 29 A TWO-STAGE, HEAT STERILIZABLE CONNECTOR FOR USE IN PROCESSING FROZEN BLOOD William W. Cooper, Charles E. Huggins, and Mark Weinberg (Abcor, Inc., Cambridge,'Massachusetts and Massachusetts General Hospital, Boston, Massachusetts) The objective of this work was the development of a two-stage, heat sterilizable connector which would assure sterility when entering a blood bag system to add or remove agents required in processing blood for freezing. Because sterility would be as— sured, use of the connector could potentially extend the present 24-hour outdating period for frozen-thawed red cells. The connector has male and female sections which can be joined in the first stage of union without breaking the sterile seals between the systems being united. While in this stage, the con- nectar is heat steriliied to remove any contamination on the mating surfaces. After sterilization, the male and female sec- tions are completely joined in the second stage of union and the sealing septa broken in the process. Blood compatibility and toxicity tests have been carried out on all materials to be used in the connector. The results show the materials to be completely satisfactory for this use. Six hundred and eighty- one sterility tests with B. subtilis spore discs placed in the connector show that dry heat of 525° F held for three minutes is adequate for complete kill. This time is satisfactory from the viewpoint of integrating the system into present blood banking activities, yet is conservative in terms of killing B. subtilis. Two hundred and fifty flow- -through sterility tests with B. subtilis spore solution placed on the septa surface demonstrate the effectiveness of the heat sterilizable connector system in transferring sterile solutions from one environment to another. Additional work is proceeding to decrease required heating times, prevent presterilization mating and poststerilization un- coupling, and examine red cell viability and function after pro- cessing with the connector. Sterile blood bag systems are now entered by puncturing a diaphragm with a hollow needle. Although the components of pres— ent connectors are maintained in a sterile condition by protec- tive covers up to the time the connection is made, these covers must be removed to effect the actual union. This removal ex- poses the mating surfaces to possible airborne contamination and the blood bag components to the risk of loss of sterility. Because absolute sterility cannot be assured with present methods, there is a 24-hour time limit within which blood prod— ucts must be used after their containers are entered. This re— striction results in the loss of any materials not used within this time limit. 31 One specific case of particular interest is frozen-thawed red cells. The frozen cells are maintained in a viable condition throughout low temperature storage by a cryoprotective agent, usually glycerin. This material must be added to the cells prior to freezing by entering the blood bag system. Subsequent— ly, when the cells are thawed the cryoprotective agent must be rinsed out, again by solutions added from outside the system. With present entry techniques, all such frozen-thawed red cells are subject to the 24—hour outdating restriction. The practical problem is that red cells thawed for particular procedures and not required are discarded unless another properly matched re— cipient is available. With rarer blood types, the loss is more than just economic. The two-stage connector reported here proposes to extend the 24-hour dating period by assuring sterile entry into the blood bag system. Like present devices, this connector has a puncturable diaphragm; but the two-stage feature permits a first step union without diaphragm rupture. While thus partially joined, the mating surfaces can be dry heat sterilized and at the same time shielded from external recontamination by the walls of the connector. After this sterilization is complete, the septum may be broken and union completed. A dry heat sterilizer is used with the two-stage connector because heat is the most consistently reliable method of killing microorganisms. Vegetative and spore forms of bacteria, yeast, fungi, and viruses can all be destroyed by heat sterilization procedures. Other modalities may be effective with one type of organism, but not with other types. Description of Coupler—Sterilizer System Design and Operation The system to provide assured sterile entryconsistsof a two— stage coupler and a dry heat sterilizer. The male section” as shown in Figure 1, includes a plastic body and a silicone rubber sep— tum. The septum design is shown in Figure 2. The septum fits .500 FIGURE 1 Sterilizable blood connector -- male section. 32 4 L“ L FIGURE 2 Male septum. over the front portion of the plastic body and slides with it into the female section. A slight taper on the body gives a tight fit and prevents the male section from slipping out of the female piece. At the rear, the body is designed to connect tightly with standard blood set tubing. The female section, as shown in Figure 3, also has a plas- tic body and a rubber septum. The septum design is shown in Figure 4. However, this septum fits inside the orifice and can L 2,550 W r—Tm m \J L FIGURE 3 Sterilizable blood connector -— female section. i FIGURE 4 Female septum. 33 slide within it as the male portion is inserted. "0" rings molded into the side of the septum ensure a tight fit. At the rear of the center cavity, there is a hollow needle which can puncture both septa as the parts slide together. To avoid flow limitations in the connector, the inside diameter of the needle is fixed at 0.120 inches, the same as standard blood set tubing. The design of the female portion's septa and the limited length of the needle allow the male portion to be inserted part way without the needle puncturing either septum. Figure 5a il— lustrates the coupling in this first stage of union. Since the coupler is together, the mating septa are protected from further external contamination; but the systems to be joined are not yet in communication because the septa are intact. FIGURE 5a Connector partially FIGURE 5 Connector together, during sterilization. during operation. 4&23maeaxxx\ A l" 4u§=a\\\\\\\’\\\\\\\m ..- 111111 1111111111 "'7 117/ ‘1’ /1/ //'/ FIGURE 5b Connector completely together, after sterilization. The sterilizing unit, shown in Figure 6, is designed to hold the male and female parts in the first stage of connection. Although the sterilizer top and bottom halves are hinged for Sterilizer with Connector in Place . “ Cartridge Heaters IE Temperature Controller FIGURE 6 Heat sterilizer. Sterilizer Body Illustrating Cavity for Connector 34 easy removal of the connector, the pieces are inserted with the sterilizer closed. Since the interior is machined to fit the connector pieces in the first stage of union, they can slide in only the proper amount. Cartridge heaters provide the tempera— ture needed for sterilization which may commence as soon as the pieces are in position. These heaters are regulated to maintain a preset temperature by a controller indicator system with moni— toring by thermocouples in the block. When the sterilization cycle is complete, the sterilizer unit is opened and the connector removed. After it has cooled, the two sections are pushed completely together, causing the hollow needle to pierce both septa. It is only at this time that the systems to be joined are actually entered. Because the septa surfaces are heat sterilized prior to puncture by the needle, possible contamination of the systems is avoided. Figure 5b shows the coupling in the second stage of union with punctured septa and direct communication between the systems being joined. Utilization This coupler system can be easily integrated into present blood processing methods. For example, in the preparation of frozen—washed red cells, the blood would be collected into a multiple bag system as is presently done. The bag containing the red cells after removal of the plasma would be equipped with a female coupler. The blood freezing unit would have mating male couplers. Using the technique just described, the cells would be transferred through the coupler to the freezing unit. Glycerol would then be added through another identical sterile coupler as prescribed by the usual agglomeration procedure (1). Couplers would be used only once and the lines joining them would be tied or clamped after use to prevent any contamination. Similarly, after freezing and subsequent thawing, the diluent reconstitution set would be attached through the new coupler. Another potential application is reducing the number of multiple bag sets used for blood collection. Blood drawn into a container fitted with this connector could be used directly as whole blood, or the plasma could be transferred to other bags for separation of platelets or preparation of cryoprecipitate. This procedure would increase the flexibility of blood-banking opera- tions because a single collection system could be used. There would also be an economic benefit from eliminating wastage of multiple bag sets. Materials and Fabrication Candidate materials for the connector had to fulfill a num- ber of important criteria. These included: 1) thermoplasticity so that the parts could be formed relatively cheaply by injection molding; 2) thermoresistance to the high temperatures required 35 for sterilization; 3) nontoxicity; 4) blood compatibility; and 5) low thermal conductivity to prevent damage to the low melting blood tubing. Kel-FG, Astrel0 , and Torlonfiiwere three initial candidate ma- terials for the plastic body of the coupler. Kel—F is a poly— chlorotrifluoroethylene marketed by 3M. It maintains its struc— tural integrity up to temperatures of 400°F. Astrel is a poly- arylsulfone, also marketed by 3M. It maintains its integrity up to 525°F. Torlon is a polyamide imide marketed by Amoco. It maintains its integrity up to 525°F and is significantly cheaper than any of the other materials. All are injection moldable, inert, and have low thermal conductivities. Toxicity testing of these materials was performed by Dr. John Autian of the Toxicol— ogy Laboratory of the University of Tennessee Medical Units. He performed standard USP extraction and implant studies as well as cytotoxicity and hemolysis tests. The results are presented in Table 1. All three passed all toxicity tests. Initial evaluation with prototypes machined from Kel—F showed this material to have inadequate thermal resistance. Consequently, all work reported here was done on connectors whose bodies were injection molded from Astrel. Torlon, a lower cost material but more difficult to mold, has not yet been eval— uated. Preliminary investigations have been conducted for several additional materials. These are KinelG, a polyimide made by Rhodia; EkkcelQ, a linear aromatic polyester made by Carborundum Co.; and PFA®, a perfluoroalkoxy fluorocarbon made by duPont. All of these materials show the required heat resistance and are less exepensive than Astrel. Toxicity data on these are included in Table l. The septa have to meet the same temperature resistance and nontoxicity requirements as the body material. In addition, they must be elastic and puncturable by the needle in the female section. No thermoplastic known to the authors meets these cri— teria, so the septa were cast from temperature—resistant silicone rubber. This material is supplied as a viscous liquid which upon the addition of a tin octoate catalyst cures to a resistant elastomer. Septa are formed by pouring the catalyzed mix into molds. Two types of RTV silicone elastomer made by General Electric were evaluated. Both had adequate physical characteristics, but the number 634 material was selected because of lower toxicity in the very sensitive culture tests. A comparison of the two ma- terials is shown in Table l. The sterilizer itself is simply an aluminum block machined out to fit the connector in the first stage of union. As shown in Figure 6, there are four lOO-watt cartridge heaters. The 36 LE TABLE 1 Toxicity Test Results Potential Body Materials Potential Septa Materials Ekkcel Kel-F Kinel 5515 Astrel 360 Torlon 4203 1-2000 PFA 9704 RTV-615 RTV-634 Tissue Culture Non— Cytotoxic Non— Non- Non— Non— Non- Test cytotoxic (very slight) cytotoxic cytotoxic cytotoxic cytotoxic Cytotoxic cytotoxic Rabbit Implant (USP) Nontoxic — Nontoxic Nontoxic — - Nontoxic - . a . . Hemoly51s l) 7% hemolySLS (Rabbit Blood) 1% 1% 3% 2% 2% 2% 2) 20% hemolysis (1% USP Extraction 0/5 0/5 0/5 with Saline & (No deaths _ (No deaths _ _ _ (No adverse _ Systemic Tox- nor adverse nor adverse effects or icity in Mice effects) effects) death) aMaterials with hemolysis values of 5% or less are considered to be nonhemolytic. hinged cover provides for easy removal of the connector after sterilization. Test Results Sterility Testing -- Static Tests g. subtilis both in the vegetative and the spore form was selected as the best generally recognized test organism. This test system is used standardly for sterilization controls for both heat and gas sterilization. Consistent kill of these spores is generally taken as demonstration of sterility. In the first series of tests, the objective was to determine if g. subtilis spores at the silicone rubber interface between male and female connectors could be killed by external heat. For these tests 6-mm diameter discs were punched out of commer— cially prepared g. subtilis spore strips intended for sterility controls. Each of these 50 X60 mm strips was certified by the manufacturer, Raven Biological Laboratories, Inc., Omaha, Nebraska, to have 100,000 to 200,000 spores/strip. Assuming even distribution of spores over the strip, each disc would thus have 9,425 to 18,850 spores. The following procedure was then used. 1) Place the spore- impregnated disc in the female section of the connector so that it is positioned on the septa. 2) Insert the male piece into the female so that they are together in the first stage of union and are positioned in the sterilizer. 3) Run the heat sterili— zation cycle for the allotted time. 4) Aseptically remove the disc from the coupler and culture it at 37°C in a thioglycollate medium. 5) Follow steps 1-4 with control spore discs placed be- tween unheated connectors. 6) Check after nine days for signs of bacterial growth. E. subtilis spores grown out in thiogly— collate broth have a distinctive growth pattern that is readily apparent to experienced personnel. In order to maximize compatibility with present blood bank— ing practice, the temperature of the sterilizer was raised high enough so that the sterilization time required would be small. At an external block temperature of 525°F, the results were as presented in Table 2. Although the results indicate that two minutes is adequate for complete sterilization, three minutes was used for most of the tests to provide a safety factor. Three minutes is suf— ficiently short that the sterilization cycle can be easily as— similated into blood banking procedures. 38 TABLE 2 Sterility TeSt Results with Disc Spore Stripa Time at 525°F No. of Cultures (min) Positive Negative 1 18 10 2 0 52 3 1b 681 4 0 29 5 0 22 a248 controls were positive, 1 control was negative. Dropped after being sterilized. Sterility Testing -- Flow—Through The objective of these tests was to determine whether heating of the junction in the first stage of union would con- sistently permit sterile transfer of fluid through the couplers. The following procedure was used. 1) Attach the female end of the connector to a stiff plastic tube. Place this in a glass test tube sealed with cotton. 2) Assemble the male end of the connector with Vinyl tubing attached to an lS-gauge needle with a protective cover. After steps 1 and 2, the system is shown in Figure 7a. 3) Autoclave both assemblies for 60 minutes at 15 psi Cotton Plug Transfer Needle , with a Transfer TUblng Protective Cover__\. ,»// in Glass Protective Tube FIGURE 7a Flow-through test system -- prior to sterilization. pressure to sterilize the inner fluid pathway. 4) Paint the septa faces with a previously prepared paste of E. subtilis or— ganisms. This paste was made by growing out spore strips in broth for ten days and concentrating the growth products by cen- trifugation. 5) Place the connectors in the first stage of union, heat for three minutes at 525°F, remove from the heating block, and completely push the couplers together, piercing the 39 septa. 6) Aseptically lower the tube attached to the female piece to bottom of the tube containing sterile thioglycollate broth culture medium. 7) Insert the needle attached to the male side through the previously flamed stopper of an evacuated tube. The culture medium is then sucked from its tube, through the joined couplers, and into the evacuated tube. This situation is shown in Figure 7b. 8) Run steps 1—7 with contaminated septa in unheated connectors as controls. 9) Incubate the medium for nine days and observe for growth. Male and Female Septum Ruptured after Heat Sterilization Sterile Sterile Ia Thioglycollate Evacuated _ Transfer Tube \r‘ Culture Medium FIGURE 7b Flow—through test system -- after sterilization. The results are presented in Table 3. These show that while all 80 control samples led to bacterial growth, samples taken from all 250 connectors put through the sterilization cycle showed no growth. TABLE 3 Flow—Through Studies (525°F X3 min) No. of Cultures No. Condition Tests Positive Negative Contaminated 250 0 250 Controls (unheated) 80 80 0 In all flow tests, the connector functioned smoothly, easily making the puncture for the second stage of union. Under the vacuum conditions used, one would not expect to see leakage from the connector. A separate series of tests was run to de— termine the tightness of the seal. In these tests, connectors in the second stage of union were placed under pressures up to 50 inches of water. No leaks occurred. 40 Discussion The data from these two sets of experiments indicate that the connector system gives an effective and consistent means of assuring sterility when transferring solution from one sterile environment to another. This device thus has potential applica- tion in blood banking and in the preparation of biologics. Use of these couplers promises to permit extension of the dating period for frozen red cells and to increase flexibility in the handling of all forms of blood for transfusion. Further work is proceeding in redesigning the pieces to de- crease the plastic mass and thereby cut both the required heating time and material costs. Several ideas are under study to pre— vent presterilization mating and poststerilization uncoupling. These include using materials of different coefficients of ther— mal expansion, heat shrinkable tubing, and mechanical clips. Additional work to show that the connector will permit pro- cessing of blood cells through freezing, storage, thawing and post-thaw processing with both sterility and acceptable meta- bolic functions is also under way. Results of this effort will be the subject of a future publication. These studies were supported by a contract from the National Heart and Lung Institute, Contract Number NIH—NOl-4—2927. REFERENCE 1. Huggins, C. E. Reversible agglomeration — A practical method for removal of glycerol from frozen blood, p. 138—155. In W. Spielmann and S. Seidl, eds. Modern Problems of Blood Pre— servation. Fischer Verlag, Stuttgart, 1970. 41 STERILE COUPLING DEVICE FOR DEGLYCEROLIZED RED BLOOD CELLS Stanley C. Roberts, Dennis T. Reuss, and Gerald L. Moore (U.S. Army Medical Research Laboratory, Ft. Knox, Kentucky) A device was constructed for connecting two sterile systems in such a way as to assure the maintenance of the sterility of both systems. The approach used was the connection of two sterile systems mechanically followed by sterilization of the area between them using ultraviolet radiation prior to final joining and mixing of the two sterile fluids. An evaluation of the effectiveness of sterilizing the connection demonstrated the feasibility of the device. The data presented suggests that if a sterile connector can be adopted for use with frozen cells, a post-thaw storage period of 3 days would be physiologi- cally acceptable. Blood preservation in vitro requires that particular atten- tion be given to the collection and preparation of components for clinical use. Asymptomatic, reputedly healthy, individuals are selected as blood donors and the phlebotomy site is carefully prepared for venesection. By convention, shed blood collected in accordance with applicable regulations (6) is sterile, but, once collected, any violation of the hermetic seal renders the unit "open" to bacterial contamination. Once "opened" the unit must be used within 24 hours (1,6). With the advent of the cryogenic preservation of red blood cells, the addition and removal of the preservative requires violation of the sterile system at least twice. For frozen red cell preparations, the 24-hour dating starts after the cells are thawed, since it is assumed that bacteria do not grow while the unit is frozen (1). Possible Contamination of Thawed Red Cells The contamination of thawed, deglycerolized red blood cells is always a possibility using the techniques currently available. This is especially true as the number of units processed in— creases and the number of people processing the blood increases. At The Blood Bank Center at Fort Knox, Kentucky, 10 ml cultures were obtained on all units prior to transfusion. The cultures were obtained in accordance with Food and Drug Administration's procedures for culturing bulk material (1). A 3.8% contamination rate of thawed, deglycerolized red cells was observed. The total number of cultures evaluated was in excess of 500 units. The organisms identified were all gram positive and usually a species of staphylococcus was isolated, e.g., S. epidermidis, S. aureus. All cultures required incubation for longer than 48 hours at 32°C to demonstrate growth. In order to eliminate even this small amount of contamina— tion, a method was sought to connect two sterile systems in 43 such a manner as to ensure the sterility of both systems after joining. The following sterile coupler and techniques were de- veloped to meet this requirement. This device was designed with blood banking applications in mind so the selection of tubing was compatible with that available to blood bankers. Description of Proposed Sterile Coupler The coupler is composed of two interlocking components. One of each component is essential for prOper closure. The disas— sembled, disposable coupler is shown in Figure 1. The part la- beled A is a flexible tube, made of polyvinyl chloride (PVC) tubing in the model, with a membrane closure at the distal, exposed ends. The part labeled B is an elastomeric sleeve which is used to join part A to part D. Part D is an ultraviolet per- meable tube of slightly larger diameter than part A. Part C is a movable, bevel—edged cannula. FIGURE 1 The disassembled coupler. Part A, polyvinyl chloride tubing with end membrane closure. Part B, elastomeric sleeves for securing part A to part D. Part C, a movable, bevel-edged cannula for piercing the membrane closure of part A. Part D, an ultraviolet permeable conduit. Figure 2depicts the assembled separate portions of the cou- pler, Unit 1 and Unit 2, as they would appear prior to coupling. For an example, assume that Unit 2 is attached to a container of packed red blood cells. At the time of manufacture, all parts of the coupler system would be sterilized, and the connector ends would be covered with plastic protective sheaths (parts E) until just prior to connection. The sheaths will protect the connector while not in use. When the sheaths are removed, certain areas of the connector are exposed to possible contamination as depicted in Figure 3. It is these areas, indicated by shading, that must be resterilized after the connection and coupling of Unit 1 and Unit 2. FIGURE 2 The assembled units prior to coupling with sheaths (E) in place. 44 . 'O’QVQV ".'« f%%$’¢%§& .. .9 99 FIGURE 3 The sheaths removed, possible areas of contamination due to environmental exposure are shaded. In Figure 4, the two units are joined together by pushing Unit 2 into Unit 1 and snapping the resilient sleeve B on Unit 2 over the end of part D on Unit 1. This provides an airtight seal from Unit 2 to Unit 1. This process isolates and defines the area to be sterilized in the next step. :21. - _ _ _ Ova 59.9.. FIGURE 4 The units are coupled and the airtight passageway (non—sterile, shaded area) separates the sterile units. Ultraviolet radiation is an accepted method of producing sterility in a limited area. Ultraviolet radiation was selected since it is rapid, dry, effective, easily monitored, and has been used for sterilization of medical products in the past. While Dr. Stiegner in Germany has an extant patent for a sterile cou- pler using formaldehyde as a localized sterilizing agent (11), it is readily apparent that the inconvenience of a liquid steril— izing system is effectively eliminated by the use of an ultra— violet radiation method. As mentioned previously, part D of the coupler is permeable to ultraviolet radiation. Vycor [Corning GLASS WORKS, Corning, N.Y. 14830] glass tubing (No. 7910), made of 96% silica glass, was used for this model, which characteristically transmits 90% of the incident ultraviolet (UV) radiation at 254-nanometer wave- length. A special device was constructed to hold the ultraviolet meter close to the assembled units of the coupler during the ir- radiation procedure. The relative position of these components is shown in Figure 5. The high intensity, shortwave ultraviolet source was appressed to the coupler, minimizing the dissipation of UV radiation while traveling to the coupler, thereby giving the coupler the greatest radiation exposure possible with this lamp. The shortwave ultraviolet radiation meter was shielded so that only UV radiation emerging from the Vycor portion of the coupler would be measured. This ensured that the meter was re— gistering the minimum dose of radiation experienced by the cou- pler. 45 high intensity, short- wave ultraviolet sourc 330 C WWW x 1 . short -wave ultraviolet radiation meter FIGURE 5 Sterilization of coupler using ultraviolet radiation. After sterilization of the coupler was completed, the can— nula, part C, would be manipulated to break the membrane seals at the ends of the PVC tubing A, as shown in Figure 6. After the cannula has bridged the space between tubing A of Unit 1 and tubing A of Unit 2, the glycerol solution or other sterile fluids could flow between Unit 1 and Unit 2 while never contacting any non-sterile surface. n ‘ 99'900‘4‘ ‘ “ ‘ >9 ~ . . { A (7 mm; ) c ) . , afi—l— .7 , came/779(— FIGURE 6 After sterilization is completed. The cannula is forced through the membranes to form a bridge through which solutions may flow. A minimum number of two couplers would be required to pro— cess a single unit of frozen red blood cells. The first would be to couple the packed cells and glycerol solution prior to freez- ing, and the other would be to couple the thawed cells to the deglyCerolization harness prior to processing. Elimination of airways and vents to the closed system is possible if plastic bags containing sodium chloride solutions for the deglyceroli— zation process were attached to the deglycerolization harness at the time of manufacture. This would require standardization in deglycerolization solutions which has already been demonstrated to be feasible (8,14,15). Evaluation of Coupler Model The evaluation of the effectiveness of this coupler was not performed by testing solutions running through the cannula as others have done, but rather by testing the ability to produce sterility of the contaminated areas after coupling. This was accomplished by heavily contaminating the exposed areas and site of connection, after removal of protective sheaths, with solu— tions containing known organisms and allowing these to air dry before coupling. After the Units were joined, they were exposed 46 to ultraviolet radiation for 60 seconds, and the total dose in micro watts/cm2 was calculated. After the UV exposure, 1 m1 of sterile saline was injected into the closed area which had been formed by joining the two units. Using a cannula, the saline was removed and transferred to a broth medium for culture. Table 1 shows the results of this series of experiments using vegetative cultures of g. subtilis, E. coli, g. epidermidis, g. aureus, and g. subtilis spores. Control couplers not exposed to the UV treatment had heavy growth. Most vegetative bacteria are killed by 5,000-15,000 u watts/cmz, and spores, both bacter- ial and mold, are killed by 20,000-4oo,ooo u watts/cm2 of UV radiation (9). Therefore, if the meter indicated that the cou- pler had been exposed to a minimum of 400,000 u watts/cmz, it was assumed that all bacterial organisms had been killed and steril- ity produced. The test results verified this supposition. TABLE 1 Sterilization of Contaminated Couplers Using Ultraviolet (UV) Radiation Exposure Time Total Dose Organism u watt/sec/cm2 sec u watt/cm2 Results E. subtilis 9,500 60 570,000 No Growth Control not exposed to UV Heavy Growth E. coli g. epidermidis 7,800 60 470,000 No Growth g. aureus Control not exposed to UV Heavy Growth g. subtilis (spores) 10,000 60 600,000 No Growth Control not exposed to UV Heavy Growth These results indicate that it is feasible to sterilize small isolated areas rapidly with the use of modified existing equipment. The question is, will a manufacturer find this design practical, efficacious, and easy to produce, or will another de- sign be more adaptable? Post-Thaw Storage of Red Cells Regardless of which method of ensuring sterile connections is ultimately selected, the question of the stability of thawed red cells must be answered. If it is assumed that the bacterial problem is solved by one of the devices discussed at this sym— posium, it may be appropriate to ask how well deglycerolized cells withstand post—thaw storage. Extensive work has been reported showing the viability and function of deglycerolized cells stored for 24 hours or less 47 (4,13). Dr. Valeri has studied the post-thaw storage of deglyc— erolized red cells for longer than 24 hours and has reported the successful transfusion of deglycerolized red cells stored for 6 days at 4°C with normal post-transfusion survival (3). Gibson has reported the transfusion of deglycerolized cells after 14 days of storage at 4°C when resuspended in autologous plasma (2). At our laboratory, several in vitro parameters of blood deglnd erolized with the IBM Blood Cell Processor [IBM, Information Records Division, Princeton, N.J. 08540] or the Elutramatic System [Fenwal Laboratories, Morton Grove, Ill. 60053] were studied. The final wash and resuspension solution contained 0.2% dextrose in 0.8% NaCl buffered to a pH of 6.9 with approximately 25 mEq of phosphate buffer. The cells were stored for up to 9 days at 4°C. It was observed that the pH remained approximately 6.9 throughout the 9-day period when measured at 4°C, and the intracellular potassium remained constant at 86.0 $3.8 mEq/l of red blood cells. Determination of free hemoglobin and extracellular potassium were performed at various times during storage and the results are shown in Figure 7. The mean for the 8 units tested indicated 200 -‘ 7004 —25 _ coo- '0 \ 2’ 500- —20 _ E E Q 0 w 9 400 +E 0 -IS g 0 , T m s . s. I 300— u a u. 1 -' w —10 d m u u < m .— x III-I 100 -‘ FIGURE 7 Serial determinations of free hemoglobin and extra— 1 1 1 1 I cellular potassium values of o 2 4 7 deglycerolized red cells stored at 4°C for 9 days. DAYS AFTER THAWING that the free hemoglobin value started at 68 :25 mg/dl and after 9 days increased to 540 i142 mg/dl. The increase in extracellu- lar potassium seemed to parallel that of free hemoglobin. The values for extracellular potassium increased from 1.5 i0.76 mEq/l on the day of thaw to 17.6 $5.1 mEq/1 on the ninth day post—thaw. This would appear to indicate that the increase in potassium re— sulted from hemolyzed red cells rather than leakage from intact cells. 48 The values for red cell 2,3-diphosphoglycerate (2,3-DPG) and adenosine triphosphate (ATP) for the deglycerolized units were determined as previously described (5,12), and the results are depicted in Figure 8. The initial 2,3-DPG value of about 5.80 $0.64 um/ml of red blood cells was within the normal range. The reduction of 2,3-DPG during storage was similar to that ob- served in ACD preserved red cells, which is essentially nil after 14 days. The ATP values started at 1.30 $0.20 um/ml of red blood cells, which is normal, and slowly decreased to 0.94 $.22 um/ml after 9 daYS' POSI-IHAW STORAGE or RED BLOOD cms (HUMAN) DEGLYCEROLIZED L) 7-— S _ 6— E L) \QS“ m 12 I: 1 4_‘ _2._E o. .\ a. 3" __ l 9 IE ' Yd —I°=' °° 2 FIGURE 8 ISerial O; l A _- <1 determinations of 2,3—diphosphoglyc— erate (2,3—DPG) and f ] I I l l l l I I l adenine triphosphate (ATP) values of de— 3 6 8 10 l cerolized red glgod cells stored DAYS AFTER THAWING at 4°C. It would appear that from a biochemical point of view deglyc- erolized red cells are stable for a substantially longer period of time than 24 hours. Of concern to Valeri (10) and also observed by the authors was the rapid increase in free hemoglobin during the post-thaw storage at 4°C. It was suggested that the post—wash hemolysis could be retarded if mannitol were added to the final cell suspension (3). Further work in this area would be profit— able if 7—10 day post—thaw storage were desired. For shorter post-thaw storage periods of about 3 days, it would appear un- necessary to add anything to the washed cell suspension to reduce hemolysis, especially if the initial free hemoglobin value were 100 mg/dl or less. Discussion It has been suggested that a minimum quality assurance pro- gram for each unit of thawed, deglycerolized red blood cells should include a milliosmolality and free hemoglobin determina- tion (7). It is essential that these tests be performed if 3—day storage is planned. Units which have initial values of free hemoglobin greater than 100 mg/dl should be used as soon as pos- sible, preferably within 24 hours, and units with less than 100 49 mg/dl may be stored for up to 72 hours. As is apparent from Figure 7, if units have an initial free hemoglobin value of less than 100 mg/dl, then within 3 days the supernatant hemoglobin value would be less than 250 mg/dl. This is less than the aver- age value of free hemoglobin found in deglycerolized samples of Huggins agglomerated red cells, currently the only 1iCensed frozen cell product. Careful attention to quality control for Red Blood Cells (Human) Deglycerolized in conjunction with a sterile coupling device will ensure a safe, effective product having a useful shelf life of at least 3 days. The opinions or assertions contained herein are the private views of the authors, and are not to be construed as official, or re- flecting the view of the Department of the Army or the Department of Defense. REFERENCES 1. Federal Register 35(171):13930. Department of HEW, Washington, D.C. Government Printing Office, 2 September 1970. 2. Gibson, J. G., II, J. L. Tullis, R. J. Tinch, W. R. Ryan, and S. DiForte. Transfusion 12:198, 1972. 3. Greenwalt, T. J., ed. The Human Red Cell in Vitro, p. 281- 343. Grune and Stratton, New York, 1974. 4. Meryman, H. T., and M. Hornblower. Transfusion 12:145, 1972. 5. Moore, G. L., M. L. Failla, B. H. Blake, J. L. Gray, and F. W. Manalo. Transfusion 14:249, 1974. 6. Oberman, H. A., ed. Standards for Blood Banks and Trans- fusion Services, p. 19—23. American Association of Blood Banks, Washington, D.C., 1974. 7. Roberts, S. C., D. T. Reuss, and F. R. Camp. Transfusion 15:278, 1975. 8. Rowe, A. W., L. L. Lenny, G. Dayian, and K. Mayer. Trans- fusion l3:355, 1973. 9. Rubbo, S. D., and J. F. Gardner. A Review of Sterilization and Disinfection, p. 84-92. Lloyd-Like, Ltd., London, 1965. 10. Schmidt, P. J., ed. Progress in Transfusion and Transplanta- tion, p. 352. American Association of Blood Banks, Washing— ton, D.C., 1972. 11. Steigner, K. F. Transfusion/infusion line coupler. Deutsches Patentamt. 1 300 635, 7 August 1969. 12. St. John, J. B. Anal. Biochem. 37:409, 1970. 13. Szymanski, I. 0., and C. R. Valeri. Vox Sang. 21:97, 1971. 50 l4. Valeri, C. R. Transfusion 13:356, 1973. 15. Valeri, C. R. Red Cell Freezing, p. 1-25. American Associ— ation of Blood Banks, Washington, D.C., 1973. DISCUSSION DR. BERKMAN: How many organisms Were on the surface of your device? CPT. ROBERTS: For vegetative organisms the inoculating fluid contained 5X108 bacteria/m1- This suspension was used to coat the exposed surfaces of the coupler. 51 A PULL TAB CONNECTOR FOR STERILE TRANSFER OF BLOOD AND BIOLOGICAL SOLUTIONS Arthur W. Rowe, E. T. Marshall, L. L. Lenny, and J. R. Graham (The Lindsley F. Kimball Research Institute of The New York Blood Center, New York, New York) When blood or solutions are transferred from one receptacle to another for washing, centrifuging, or other processing, the ste- rility of the system is violated each time an external connection is made. Blood subjected to potential contamination during trans- fer has, by federal regulation, a useful shelf life of 24 hours instead of El days. ' The sterile transfer problem could be solved with a suitable fail-safe sterile connector. This paper describes the concept of a simple sterile connector or ”docking device" and the successful testing of the concept with a prototype connector. The principle involves adjoining connector fittings with two pull tabs (dia— phragms) each of which is folded back on itself. When the con— nector fittings are clamped in an airtight manner, the flexible pull tabs are pulled in upon themselves (invaginated) when they are withdrawn from the assembly and leave behind a sterile fluid path which has not been exposed to the exterior nonsterile en- vironment. The feasibility of the concept was demonstrated with a proto- type model with tests involving: 1) dye exclusion to measure intactness for screening purposes, 2) exclusion of radioisotopes as a sensitive indicator of system integrity, and 3) bacteriologi- cal studies to prove that the sterility of the internal fluid path remained intact. A sterile connector based on the invagination pull tab concept can eliminate the potential contamination hazard associated with fresh and frozen blood processing. In many fields such as medicine, pharmacy, and chemistry, it is necessary to transfer fluids from one receptacle to an— other while maintaining these fluids absolutely uncontaminated by extraneous matter, even those as elusive as airborne spores or gaseous air pollutants. This is particularly important in blood banking where blood, once collected, must be maintained in a sterile environment. However, it is frequently necessary and desirable to separate blood into a number of its components or to freeze the blood in order to store it for long periods (2,3, 4,5,7). Such connections are accomplished conventionally by in- sertion of a sterile needle or coupler attached by tubing from one plastic bag into a sterile diaphragm connected to the other bag. Whenever a new connection must be made to transfer the blood or blood component into another receptacle for washing, centri- fuging, or other processing, the sterility of the system is vio- lated, and the possibility of contamination, even from airborne material, is present. As a consequence of this possibility, blood or blOOd components which have been subjected to a trans- fer operation by means of the conventional type of connector 53 have been limited by government regulations (1) to a useful life of 24 hours as opposed to a normal shelf life of 21 days for blood which has not been so handled. Our approach to the solution of this sterile transfer prob— lem is to provide a simple system for connecting two blood re— ceptacles in a sterile, fail-safe manner so that the useful life of the transferred blood or blood components could be extended beyond 24 hours. This paper describes the concept of a simple sterile connector or "docking device" and the successful testing of the concept with a prototype connector model (6). Materials and Methods Sterile Connector Device An illustration of the sterile connector design is shown in Figure l. The prototype connector model used f r testing was fabricated from the following materials: DELRIN was used for the rigid support housings; KRATON-G was used for the compressible face gaskets; and a coated polyester film was used for the flexi— ble pull tabs that were heat sealed onto the compressible face gasket. The final selection of design configuration and materials is not yet complete however. STERILE CONNECTOR PROTOTYPE ADJUSTABLE FIGURE 1 Prototype model of the sterile connector. PULL TABS The principle of the device is to adjoin two pull tabs (diaphragms), each of which is folded back upon itself. One tab is attached to each end of two similar connector fittings, and while they are held together (clamped) in an airtight manner, the flexible pull tabs are pulled in upon themselves (invaginated) when they are withdrawn from the connector assembly. In this way the outer nonsterile surfaces of the two diaphragms are removed from the assembly without having been exposed to the internal (sterile) surfaces of the connector. 54 The method of carrying out the closure can be more fully understood from the drawings in Figure 2. The top drawing (Fig. 2a) shows the annular flange fitting (#2) that is attached to the flexible tubing (#1) leading from a receptacle, e.g., blood bag, wash solution, etc. Attached to the face of the rigid support flange is a soft compressible face gasket (#3) made from an elastomeric or plastic compound of suitable duro- metric properties. A diaphragm (#4) consisting of a thin but strong piece of plastic is attached to the compressible face gasket. The diaphragm has a tab (#5) which, when folded back upon itself, extends beyond the gasket to act as a draw pull. 3 FIGURE 2 Cross section of the invaginating pull tab connector. Fig. 2a. Single fitting showing: (#1) flexible tubing leading from a receptacle; (#2) annular flange fitting or housing; (#3) soft compressible face gasket; (#4) diaphragm attached to the face gasket; (#5) extended pull tab 2 c. 4 2 f 1d d b k 't if ‘\ Ezbé o e ac on 1 se to act as draw pull. 1 5 D\\\ ‘///44 { %A 7 Fig. 2b. Two fittings brought together: (#6) pull tabs super— imposed on each other; (#7) clamp. Fig. 2c. Illustration of partial withdrawal of pull tabs showing internal diaphragms invaginated or pulled in upon themselves. When the two fittings are brought together as in Figure 2b, the folds in the respective pull tabs of the diaphragm are to— gether (#6). The pull tabs are superimposed upon each other and extend out beyond the gasket. The assembly of the two fittings are clamped (#7) together to compress the two gaskets against each other with sufficient pressure to form a tight seal. The adjustable clamping device was used in place of a positive lock- snap fitting clamp to permit repetitive testing. The extended pull tabs of the diaphragm are held together and pulled out manually as one unit from between the compressed face gaskets as in Figure 2c. In so doing, the internal dia— phragms are invaginated as they are pulled in upon themselves and withdrawn from the connector assembly while the compressible 55 elastic gaskets continue to maintain a tight seal. Figure 2c is an illustration of the withdrawal when it is partially complete. The pull tab diaphragms are discarded when withdrawn, leaving a tight sterile connector assembly ready for transfer of fluid. Testing Procedures The feasibility of the concept of the pull tab connector was tested by submersing the device under various aqueous solu— tions in a small glass container to determine if leakage occurred. Testing of the prototype device was divided into three phases. Phase I testing involved immersion of the device into a glass container of dye (0.1% Trypan Blue) to look for gross leakage when the pull tabs were removed to render the internal tube path continuous. In some instances a negative pressure (vacuum test) was applied to the device as the pull tabs were re- moved. The negative preSsure was used to simulate the environ- ment encountered in certain receptacles such as evacuated solu- tion bottles. Phase II testing of the connector involved extraction of the pull tabs from the device while it was immersed in a solu— tion containing a radioactive isotope of high specific activity. The solution contained sufficient 51Chromium (Chromitope, Squibb) to detect a leak as small as 0.1 microliter. After with- drawal of the pull tabs to render the internal tubing path con— tinuous, 2 ml of Water were flushed through the connector device, collected at the opposite end, and counted in a scintillation counter (Autogamma, NuClear Chicago). If contamination had oc- curred when the pull tabs were withdrawn, the external radio- active solution would have entered the system and could be de— tected in the solution subsequently used to flush the system out. Phase III evaluation of the prototype connector involved sterility testing with microorganisms using the setup shown in Figure 3. Each flange of the connector was attached with tubing to a culture flask, one of which contained growth medium. Both T TABS REMOVED BY PULLING STERILE GROWTH MEDIUM CULTURE FLASK FIGURE 3 Setup used for sterility testing of the connector. (BACTERIAL SUSPENSION) 56 halves of the entire system were sterilized separately in an autoclave at 121°C, 15 1b pressure, for 15 minutes. Ethylene oxide sterilization was not used because of the time involved for degassing. The sterilized but separate devices were then mated and locked together with the clamping device. After im- mersion of the connector in a medium containing bacteria, the tabs were pulled to open the sterile fluid path. Once open, the system was flushed with the sterile broth by passing it from one flask to the other. The flask now containing the broth was in- cubated to determine if bacterial growth had occurred. The or- ganisms Used for this phase of testing were Pseudomonas fluores- cens (ATCC #El3525) and Escherichia coli (ATCC #Ell775). Results The results of the evaluation of the pull tab sterile con— nector device are summarized in Table 1. When immersed in Trypan Blue solution (Phase I) the device did not show any visible leakage when testing was performed at ambient atmospheric pres— sure. When a negative pressure was applied to the device, all tests showed a transitory leakage which occurred at the moment the pull tab was removed but not thereafter. TABLE 1 Pull Tab Connector Device: Evaluation of Integrity and Maintenance of Sterility Test Phase Immersion Solution No. Negative Tests I Trypan Blue 12/12 Trypan Blue (Vacuum Trial) 0/6 II Radioisotope (“Cr) 20/20 III Pseudomonas fluorescens 6/6 Escherichia coli 6/6 When the device was tested in a radioactive solution (Phase II) containing 51Cractivity, no leakage of the isotope into the system was evident in any of the trials. The determination of presence of radioactivity in the solution that was flushed through the device resulted in complete absence of counts, which indicated that the system did not leak during testing. Phase III testing involved immersion of the device in a so- lution containing bacteria, during which time the pull tabs were removed to render the sterile internal path continuous. When sterile broth was flushed through the system, and collected at the other end of the connector, the incubated flasks did not show any evidence of growth of either of the two microorganisms 57 studied, Pseudomonas fluorescens or Escherichia coli. After one to two days of incubation, the broth was also streaked on nutri- ent agar plates to further confirm that no growth had occurred. Discussion Conventional connectors currently in use for processing of blood and blood components violate the sterility of the system each time a connection is made by introducing the possibility of contamination. Potential contamination is a particular problem with frozen blood because the prefreeze and post-thaw processing procedures require multiple entries into the system. Depending upon the method of processing, as many as ten individual entries can be made into the system, Violating its integrity and steril- ity each time. For this reason, federal regulatory agencies consider the blood to be potentially contaminated and require that it be used within 24 hours provided it is stored at 4°C (1). This is currently a formidable problem involving large-scale routine use of frozen blood because the short outdating period invariably causes difficulties in logistics and often results in loss of the blood units through outdating. A simple system for connecting two separate blood recepta- cles in a sterile, fail-safe manner is a necessary requirement for extending the useful life of the transferred blood or blood component beyond 24 hours of storage. The purpose of this paper was to present a sterile connector concept and with a prototype model to demonstrate its feasibility as a simple system for con- necting two receptacles in a sterile, fail-safe manner so that no contamination of the internal surfaces from extraneous matter can occur (6). The feasibility of this concept was demonstrated with tests involving 1) dye exclusion to demonstrate visual in- tactness for screening purposes, 2) exclusion of radioisotopes as a sensitive indicator of system integrity, and 3) bacterio— logical studies to prove that the sterility of the internal fluid path remained intact. Based on these results, further development of this concept appears warranted. Development of a production model connector device will also require further testing to meet the sterility standards established by the Food and Drug Administration. The sterile connector device used in this study was a proto- type model fabricated to demonstrate and test the invaginating pull tab concept as well as to test various plastic materials. Further development is necessary to have a sterile connector based on the invaginating pull tab concept. The system requires that the rigid support flange have a compressible face gasket of soft elastomeric or plastic compound with suitable durometric properties and which will adhere firmly to the face of the flange. The pull tab diaphragms sealed onto the compressible face and folded back upon themselves must extend beyond the com- pressible face and be strong enough to act as a draw pull. The clamping device served the purpose of a positive lock-snap 58 feature and was used here only to permit repetitive testing of the connector. A production model device will have a positive lock-snap feature as an integral part of the connector which will prevent separation of the two halves of the device once they have been mated. The results of the tests for integrity and sterility under ambient pressure conditions indicate that this connector will allow the union of two tubings in a simple, sterile, foolproof manner, and also demonstrates the feasibility of the pull tab concept for invaginating the sterile internal diaphragms by pulling them in upon themselves to form a sterile fluid path. It would appear, therefore, that with a simple, sterile, foolproof connecting device based on the pull tab concept de— scribed here, the potential contamination hazard associated with frozen blood processing will be eliminated, thereby allowing the post—thaw storage period of frozen blood to be extended to some period in excess of 24 hours. The post-thaw storage period would then depend upon the stability of the cells and mainten- ance of their variability rather than the potential contamina- tion or sterility factor, as is currently the case. This same connector system could also find application in other fields where complete freedom from contamination is re— quired for the transfer of fluid from one receptacle to another. These studies were supported by grants from the National Heart and Lung Institute (HL09011) and Union Carbide Corporation. REFERENCES 1. Code of Federal Regulations - Food and Drugs: Red Blood Cells (Human) Frozen. 21 CFR 610.53b (revised April 1, l974L 2. Huggins, C. E. Reversible agglomeration - A practical meth- od for removal of glycerol from frozen blood, p. 138—155. In W. Spielmann and S. Seidl, eds. Modern Problems of Blood Preservation. Fischer Verlag, Stuttgart, 1970. 3. Meryman, H. T., and M. Hornblower. A method for freezing and washing red blood cells using a high glycerol concentra- tion. Transfusion 12:145, 1972. 4. Rowe, A. W. Preservation of blood by the low glycerol-rapid freeze process, p. 55—71. In Red Cell Freezing. American Association of Blood Banks, Washington, D.C., 1973. 5. Rowe, A. W., E. Eyster, and A. Kellner. Liquid nitrogen preservation of red blood cells for transfusion: A low glycerol-rapid freeze procedure. Cryobiology 5:119, 1968. 6. Rowe, A. W., and E. T. Marshall. Sterile connector for con— duits. United States Patent No. 3,865,411. 59 7. Valeri, C. R., and C. G. Zaroulis. Rejuvenation and freez- ing of outdated stored human red cells. New Engl. J. Med. 287:1307, 1972. 60 STERILE BLOOD CONNECTOR Francis J. Tenczar (Interstate Blood Bank, Inc. of Chicago, Illinois and Northwestern Memorial Hospital Blood Bank, Chicago, Illinois) Fabrication of a ”sterile connector" for joining a supply source to an independent delivery location through a contam- inated environment requires, sequentially, four design steps. These are: l) alignment of connector elements; 2) exclusion of the contaminated environment; 3) sterilization of the excluded region; 4) penetration of the sterilized area and coupling of the protected internal conduits. Protective covers for connectors now in use may be modified to provide penetratable terminal barrier membranes that can be joined to exclude the general environment. An adhesive containing an antimicrobiologic agent would set the stage for a controlled mechanical penetration of the fused barriers. In another design concept barrier membranes of plastic are simultaneously fused, sterilized, and penetrated by a nichrome wire heated to 550°F. A system can be designed that is compatible, at the option of the user, with either of-these methods. The more economical mechanical approach could then be used for paren— teral fluid therapy and the preparation of simple blood com- ponents for immediate transfusion, while the ”hot wire" technic would be reserved for complex processing methods (frozen erythrocytes) or when storage of the resulting com- ponent is a factor. The first step in this project was the definition of a sterile connector as a device assuring, to an infinite degree of probability, the exclusion of exogenous microbiologic contam— ination of material transferred from a supply source to a de- livery location, irrespective of any potential or actual contam- ination through which the connection is made. Reflecting on the above and the usual connectors, illus- trated in Figure 1, one can identify the time and source of such potential contamination. The former is simultaneous with re— moval of the protective covers and the latter is the immediate environment containing the potential, heretofore exogenous, con— taminants. Both can be influenced by making the connection with- out removing the protective sheaths. First, the connectors must be aligned. Picture the illus- trated connector covers (Figure l) with, in lieu of the tapered terminal plastic seals, flat ends. These could be adhered to achieve alignment, but only at the expense of reinforcing a for- midable barrier in the intended fluid path. This problem is overshadowed by another effect of paramount importance -- the source of potential contamination is now limited to the adhesive in the barrier "sandwich" and the total environment is effec- tively excluded. If thislimited area of adhesion can be succes- 61 FIGURE 1 Plastic transfer bag illustrating "ports“ (A) and "spike" (B) currently used for transfusion and blood component preparation. sively sterilized and penetrated, the protected internal con— nectors, still sterile, can join without danger of contamination. There are, then, four basic sequential requirements for a sterile connection: 1) alignment of connectors, 2) exclusion of environment, 3) sterilization of excluded area, 4) penetration within sterile area. Each step can be accomplished in various simple ways. Two different concepts are presented. The first, a mechanical—chemical approach, subsequently evolved into the second, electrical—thermal, method. The components of the mechanical connector are shown in Figure 2. The "A" cylindrical housing contains a central in— ternal connector, arbitrarily male, recessed within a concentric cutting penetrator. This assembly, illustrated in greater de- tail in "C", moves longitudinally in a guide channel. The housing is integrally sterilized with an attached bag and pro- tected from contamination by the "O" ring seal and a barrier membrane (plastic film, metal foil, etc.) covering the free end. Cylindrical housing "B" contains a recessed central con- nector, arbitrarily female. The space in front of the connector is designed to receive the hinged flap formed during penetration of the fused barriers and to provide sufficient clearance so that this flap does not touch the connector. This unit is also closed at the free end by a barrier and sterilized integrally after attachment to a container. In operation, protective covers are removed from the outer surfaces of the barrier exposing an adhesive on at least one surface. The housings are aligned and the barriers adhered. Sterilization can be accomplished by an antimicrobiologic agent in the adhesive. Immobilization of particulate matter by the adhesive also contributes to sterility. After a suitable lapse 62 FIGURE 2 Schematic drawing of connector design with abUtting baFriers (2) surfaied 1. Housing 4. Connector-—Penetrator fizczgniggiségfinzgioi gzzzigaior 2. Barrier membrane a. Recessed connector ‘ II n - - assembly (4). 3. 0 r1ng b. 0b11que penetrator 5. Connector of time, the penetrator is pushed through the fused membranes. The flap formed is folded neatly to one side. The penetrator, being larger in diameter, allows the internal connector from "B" to enter and join the recessed male connector to establish the desired fluid path. The internal connectors cannot touch the fused margins of the barriers, the formed flap, or the cutting penetrator. This concept was presented to Travenol Laboratories, Inc. in August 1973. Reservation was expressed as to the adequacy of chemical sterilization of the adhesive obviously exposed around the opening in the barrier and flap. Since mechanical penetra- tion without such exposure seemed impossible, another penetrat- ing means was sought. Several methods of thermal sterilization were considered. These included fuse, induction, ultrasonic, and microwave tech- nics. Finally the mechanical penetrator was simply replaced with a loop of nichrome wire. When energized by lye-volt igni- tion battery —— or even a single "D" flashlight cell -- this device sealed and penetrated samples of polyethylene and poly- vinyl film. The preliminary use of an adhesive is no longer necessary for environmental exclusion since sealing, sterilizing, and penetrating are virtually simultaneous. This expanded concept was demonstrated to Travenol in September 1973. Travenol fabricated a test unit and conducted sterility experiments using Bacillus globigii to contaminate barrier films prior to penetrating and transferring culture 63 media from one container to another. Sterile transfers of media and recoveries of the test organism from the barriers adjacent to the penetrating seal were achieved. Shortly thereafter, Travenol elected to pursue development of another method and our connector projects went separate ways. This connector was resigned along the lines indicated in Figure 3. While the principles remain the same, there are the following differences and reasons. ©®©©§9© 9C9. . / v : x \ If, ..-' ‘ EAL-151%.— .-Tivf. _ u.__ _‘ . - » - -» _‘.- L.-li“~.tz.~)i."_“‘ ..- FIGURE 3 Schematic drawing of "telescoping" connector showing female (A) and male C (B) housings individually and the coupled unit (C). Either penetrator —— mechanical 1' HOUSing 6' Heater loop or electrical __ can be used in 2. Connector 7. A11gnment regess this or the previous (Fig. 2) 3- COHdQCtOFS 8. P1ast1c barr1ers configuration. 4- BU5h1n9 9. Connector 5. ”0“ rings 1) Housing "B" is reduced in diameter. The ability to in- sert a male into a female housing facilitates a more positive alignment of the internal connectors and good approximation of the plastic barriers without the use of an adhesive. Also, one can begin to speculate on the possibility of substituting this reduced configuration for one of the entry ports presently pro- vided on donor and transfer bags. 2) The internal connectors and penetrating device are sta- tionary. Relative movement for barrier penetration is provided by the "O" ring seal and bushing. The functional moveable parts are thus protected from inadvertent manipulation before use. Additionally, molding of the "A" housing and internal connector as a unit is now possible and this should reduce the ecst of manufacture. 3) The "hot wire" is substituted for the mechanical pene- trator and, together with the conductive wires, is mounted electively to the centered internal connector in the "A" housing. 64 This arrangement can be reversed and the otherwise same config- uration used with a mechanical penetrator. 4) The barrier membrane in "A" is recessed for reasons men- tioned above. A protective cover would now be placed over the open end of this housing. The model components are illustrated in Figure 4 and Figure 5. FIGURE 4 Disassembled model of telescoping connector: (A) female housing with nichrome loop penetrator; (B) bushing and "O" ring; (C) male housing and internal connector. The linear markings on the collar bushing are grooves to facilitate escape of air during coupling. FIGURE 5 Nichrome wire device assembly. The nichrome wire loop (A) is crimped to copper conductors (B) imbedded in a high temperature epoxy for adhesive mounting to the internal connector. (D) is the fixed wall of the housing. The external power connector (E) obscures the fluid conduit coupler. rswu’uinr'ww In operation, the "B" housing is inserted into the "A" housing and the plastic films approximated. A current of 2.0- 3.5 amperes is passed through the loop of nichrome wire. After the loop reaches operating temperature in about 15 seconds, the telescoped housings are slowly compressed until the central con— nectors join and establish the fluid path. The coupled connec-\ tors are shown in Figure 6. Figure 7 illustrates the hinged flap that is formed during penetration. Mechanical penetration produces a similar flap. Figure 8 shows the relative positions of the concentric internal connector, nichrome wire loop, and bushing. 65 FIGURE 6 Coupled connectors. Note that the male internal connector (A) passes through the center of the nichrome wire loop (B) to engage the female internal connector (C). The fluid path connectors (A and C) touch only each other. FIGURE 7 Bushing with penetrated plastic barrier. Note the clearance between the inner wall of the bushing and the margin of the plastic "window.“ The flap (A), hinged, is diverted to one side by the epoxy mount at the base of the nichrome wire loop and the internal connector of the male housing passes through the "window" without touching the flap or the margin of the plastic cutout. FIGURE 8 A View of the barrier end of the female housing illustrating the concentric alignment of components in- ternally. The components, centripetally, are: (A) wall of housing; (B) busing and "O" ring; (C) nichrome wire loop; and (D) internal female connector. The clear— ance between the inner wall of the bushing and the outer wall of the internal con— nector is approximately .070 inch. With the above physical configuration reasonably set, there remained the selection and control of the operating temperature and choosing the plastic film best suited for the barriers. Temperature control is achieved simply by regulating the current passing through the nichrome wire using a l—ohm rheostat and an ammeter connected in series with a 3—volt battery power supply. Current can be regulated in a 2.0-5.5 ampere range. The temperature of the nichrome wire is directly proportional to the applied current. 66 A likely sterilizing temperature was defined by Rothstein and Arnett (6) in their work on sterile re-entry technics for spacecraft. They reported sterilization within the sealed area of plastic films heated to approximately 550°F (288°C) for one second or less during the sealing procedure. Figure 9 shows plots of temperature vs. current for the nichrome wire device. Curve "A" was obtained by Mr. Larry Kehl of the Amphenol Corporation. Mr. Kehl indicated the recorded temperatures are erroneously low because the thermocouple attach— ment of the nichrome wire (.016—inch diameter) was large enough to act as a heat sink. With this in mind, Mr. Owen Young of Zenith obtained curve "B" using a monel—chromel wire .0005 inch in diameter as the thermocouple attachment. A preliminary qual- itative check with an optical pyrometer showed no color differ— ence at the site of thermocouple attachment at incandescent temperatures. Curve "B" is thus probably more accurate. Both are shown to emphasize the difficulty of this temperature mea- surement. Since the plastic films also serve as a heat sink during penetration, an actual temperature may not be obtainable. Both curves, however, show that the nichrome wire device can operate in a reproducible temperature range acceptable for prob— able sterilization in one second or less. The optimum tempera- ture (current) and time must be set by microbiologic studies. 'C ‘F NICHROME WIRE DEVICE B 600 FIGURE 9 A graph showing the linear relationship of -'°°° the observed temperature 500-900 and current input. Of the T A two curves (A and B) , B is E -500 probably the more accurate 2400- because of a smaller thermo— R '7°° couple attachment. The oper— ? 430° ating range suggested is R300 STERILE-Iuo or m. 550': Panama; theoretical and subject to E —500 modification by contemplated + microbiologic studies. The ”0-400 + tabular insert shows the .300 * behavior of fluoronated we» FEPuw. 407-54017 ethylene propylene (FEP) in ‘2°° VllroRIuTlmo-Ieuc contact with the nichrome 40° Rummfluwrofl-OS ohms wire at selected current mushy. ox) o; .. 2,0 3.. .5 a. .0 .9. Values - L'O Y 2:0 ' 3:0 ' 4:0 50 CURRENT (AMPS) Earlier studies using polyvinyl and polyethylene films clearly yielded a heat seal that excluded the environment and, in the Travenol experiments, sterile connections through con- taminated membranes. Unfortunately, considerable "smoke" was generated by pyrolysis of these low—melting films at the tem- peratures (unknown) used for a sterilizing penetration. For this reason a more suitable plastic barrier film became a' necessity. 67 Discussion with the Plastics Department of the Dupont Cor— poration led to the selection of fluorinated ethylene propylene (TeflomD, FEP) as a candidate material. This film also per— formed well in the study of Rothstein and Arnett. The thermal range (4,5) for continuous use of FEP is -425° to +400°F (—255° to +200°C) and the melt range is 487° to 540°F (253° to 282°C). These thermal characteristics suggest FEP can be penetrated by a relatively slow melt without significant particulate decom- position. In addition, FEP will tolerate autoclaving and expo- sure to liquid nitrogen. The tabular insert in Figure 9 documents the naked-eye effects of various current labels on the penetrability (melt), fusibility, and transparency of two FEP films, .005 inch thick, approximated without adhesive. Above 2.0 and below 3.5 amperes penetration appears "clean" —— without visible smoke or shrink— age, retention of transparency, and good fusion. Temperatures in this current range are adequate for rapid sterilization. Since 2.0 amperes produces a temperature approximating 550°F (288°C) (Figure 9, curve B) and the films act as a heat sink, the validity of curve "B" is supported. The fact that pene— tration requires more than 2.0 amperes suggests a "fail—safe" mechanism in that no penetration will occur unless a temperature adequate for sterilization is attained. A consideration of the safety of FEP in the proposed appli- cation merits noting at the outset that the suggested design precludes direct contact between the internal conn ctors or the fluid path with FEP. Also, the safe use of Teflo for the transport of fluid, implants, and accessory equipment in hospital devices is increasing in the medical field. FEP is completely stable (3) up to 400°F (205°C). The chief concern, then, is the small amount of FEP exposed to a higher sterilizing temperature. The volume of FEP .01 inch (.254 mm) thick displaced by the nichrome wire 0.875 inch (22.2 mm) in length and 0.016 inch (0.406 mm) in diameter is 2.29 mm3. Since FEP has a specific gravity of 2.1, this volume is equivalent to 4.82 mg of FEP. This is a relatively small amount of a very stable material. The thermal decomposition of fluorocarbon resins in air forms only trace amounts of carbonyl fluoride (COFZ) and per- fluoroisobutylene at 750°F (400°C). The principle toxic ingredi- ent up to 842°F (450°C) is particulate matter (2,8). Exposure of FEP to 700°F (370°C) for one hour results in an initial weight loss of 0.3%. Assuming that the weight loss occurs at a constant rate, the weight of FEP decomposed in the one second required for sterilization at the temperature would be 4 Xlo'smg. This quan- tity is probably insignificant. Moreover, it represents a maxi— mal value since the sterilizing temperature used would undoubt— edly be less than 700°F. 68 Summary and Conclusions A definition, criteria, and two design concepts for a ster— ile connector are outlined. The mechanical and electrical con— cepts are compared in Table 1. TABLE 1 Comparison of Methods L Angn AbuT Tebscope . Exclude Adhesive N0 adheswe While the "hot wire" technic may seem the more positive, both concepts should be subjected to definitive microbiologic studies. The nichrome wire has withstood a preliminary bacteri- ologic challenge, but there is no data indicating that chemical sterilization could not perform as well and the latter should be by far the more economical. A means of monitoring user adher— ence to the time interval required by an antiseptic for steril— ization before penetration would seem to be the major intangible factor in chemical sterilization. The hot wire would adapt well to semiautomation with rigid control of important variables such as temperature rise time and rate of penetration. Manual operation is, however, simple and should also be entirely adequate. In field use or emergency situations, simple dry cells or an automotive battery are likely available as a source of power. While this conference is titled "Frozen Blood Outdating Workshop," the true potential of a sterile connector, even in this limited area of application, will be realized only when this technic is available for every unit of blood collected. Only then can proposed rejuvenation technics (7) be applied ef— fectively to routinely collected donor units after conventional twenty-one days' storage. A connector element similar to Figure 3, B, can easily be made for optional use with either the electrical or mechanical complementary unit. If such a "universal" connector replaced one of the conventional ports now provided on donor and transfer bags, then design choice could be a user Option. The more eco- nomical method would suffice for simple component prepared for immediate transfusion, while the hot wire technic would be re- served for complex processing methods or when storage is a fac— tor. 69 In addition to blood transfusion practice, there is also potential application to the administration of parenteral fluids as well as many other uses in microbiology, pharmaceuticals, and artificial organs. Particularly exciting is the possibility for research in blood metabolism and preservation with the imminent promise of safe, virtually unlimited, access into a formerly untouchable unit of blood. The author is grateful to Mr. John Gove of Amphenol for advice and providing connectors for the nichrome wire device; to Mr. Larry Kehl and Mr. Owen Young for the nichrome wire temperature curves; to Dr. Leon LeBeau for the photographic illustrations; and to Mr. George Waring for the schematic drawings. REFERENCES l. Buch, L. Bacterial Challenge of Tenczar Sterile Connector Concept. Travenol Laboratories, Inc. Research Laboratory Report RD-08—103, 15 February 1974. 2. Coleman, W. E., L. D. Scheel,.and C. H. Gorski. The par- ticles resulting from polytetrafluoroethylene (PTFE) pyrol- ysis in air. Amer. Indus. Hyg. J. 29:54, 1968. 3. Coleman, W. E., L. D. Scheel, and C. H. Gorski. "TEFLON" Fluorocarbon Resins -- Safety in Handling and Use, pp. 5—10. E. I. du Pont de Nemours & Co., Wilmington, 1970. 4. Coleman, W. E., L. D. S heel, and C. H. Gorski. FEP ... a melt processible TEFLO (Product Specification Brochure). E. I. du Pont de Nemours & Co., Wilmington. 5. Coleman, W. E., L. D. Scheel, and C. H. Gorski. MODERN PLASTICS ENCYCLOPEDIA, pp. 730-734. Modern Plastics, Highstown, 1974. 6. Rothstein, A. A., and J. C. Arnett. Sterile insertion — An aerospace application of gnotobiotic technology, pp. 65—77. In Developments of Industrial Microbiology, Vol. 9. Society for Industrial Microbiology of the American Institute of Biological Sciences, washington, D.C., 1968. 7. Valeri, C. R., and C. G. Zaroulis. Rejuvenation and freez— ing of outdated stored human red cells. New Engl. J. Med. 287:1307, 1972. 8. Waritz, R. S., and B. K. Kwon. The inhalation toxicity of pyrolysis products of polytetrafluoroethylene heated below 500 degrees centigrade. Amer. Indus. Hyg. J. 29:19, 1968. 70 A DRY HEAT STERILIZABLE BLOOD BAG TUBING CONNECTOR Thomas R. Hektner and John G. Christensen, III (Travenol Labora- tories, Inc., Deerfield, Illinois) A system for the interconnection of tubing, which will enable hospital or blood bank personnel to make sterile fluid transfers between blood bags, is described. In this connector system, the exposed, contaminated sections of the blood bag tubing connectors are isolated from their respective tubing and containers by means of internal membranes. A hollow, aluminum, double-ended spike links the two connectors, but does not puncture the membranes. Following sterilization by dry heat, localized at the connector site, the membranes are ruptured by the spike to create a sterile channel for transfer of sterile fluid between containers without risk of contamination. Sterilization capability was investigated by applying a challenge of Bacillus subtilis spores to the con- nector and subjecting it to various temperature-time cycles. Challenges of l XlO6 spores were reliably destroyed at a maximum temperature of 325°C for sterilization times of both 65 and 85 seconds. Introduction The American Association of Blood Banks has established a 24-hour outdate limit for stored blood following any entry into its sterile sealed plastic bag (1). It may be possible to ex— tend or bypass this limit if it can be shown that a procedure and/or device can assure the sterility of all entries into or fluid transfers to or from a container, i.e., the sterile en- vironment within a container is entered only from another ster— ile environment. The design of a device to connect two such sterile environments, and the dry heat sterilization of this device, which can be performed in situ within the hospital or blood bank, are the subjects of this report. The connectors from two blood bags are preliminarily joined to form a closed, nonsterile area that is isolated from the sterile areas by mechanical barriers or membranes. This non— sterile area is then sterilized by dry heat. Thereafter, break— age of the membranes completes the sterile connection. Materials and Method Connector Configuration The connector system consists of two silicone rubber mem— brane tubes, a hollow, aluminum, double-ended spike, and two brass safety clips (Figure 1). For this study the connector system was not attached to blood bags; in use, the silicone rub- ber tubes are attached by means of internal bushings to the vinyl tubing leading from the blood bags. 71 SAFETY CLIP\ SILICONE MEMBRANE TUBE ALUMINUM SPIKE MECHANICALLY CONNECTED TO BLOOD BAG TUBING FIGURE 1 Sterile connector test configuration. The membranes in the silicone rubber tubes isolate the sterile portions of the blood bag system from the connector sec— tion exposed to air when tip protectors are removed. In actual use, one of the membrane tubes is already fitted with a double spike (Figure 2a). When two blood bags are to be joined, the I/ . 1 — - - — - - —-—- - - V////////////////////////”I7iiil711I/I/I/l/I/I/A I Will/Illlllllllll/l/[I/lll////4 (a) TIP PROTECTORS, DISCARDED WHEN CONNECTORS ARE TO BE JOINED 7 / %%%%%%%%%%szafifififizzzzzuz afikazzxzzxz 4,1CL,,,,,,,,,,.,,,mum”, ”II/”Z; ¢,;,;;;,,;,;,,2,,,,,,//,_//>///////////////////, ; /. WWW”%)1/Illllllilll}ll;l//I A (b) SAFETY CLIP TUBE BELLOWS AS PEELED BACK SPIKE PUNCTURES MEMBRANE FIGURE 2 Assembly sequence for sterile connection. . (c) free end of the spike is then inserted into the other membrane tube. Longitudinally rigid safety clips support the thin-walled, silicone rubber tubes and prevent premature penetration of the membranes by the spike (Figure 2b). It is the contained air volume, the exposed membrane sur- faces, and the inside walls of the spike that require steriliza— tion prior to forming the fluid path. After sterilization, the safety clips are partially peeled back and the spike is forced through the membranes (Figure 2c). An internal taper within the silicone rubber tubes, plus the external support of the safety clips, ensure easy joining, but difficult separation, should the tubes be inadvertently pulled in opposite directions. Heater Assembly and Controls A small, nichrome wire air heater provides the source of dry heat. Air, directed through the heating oven to the steril- ization chamber, is monitored by a rotometer and manually ad— justed flow controller. The temperature sensor, an iron— constantan thermocouple, is situated within the sterilization chamber in the hot-air stream directly under the connector. An automatic control system, consisting of timers, switching cir- cuitry, and OgdenQ§ temperature controllers linked to the thermo— couple, adjusts the heat input on the basis of the difference between the sensed temperature and a predetermined set point. Because of the close proximity of the oven, sensing element, and connector, there is rapid feedback control within the heat transfer path. Energy losses to the sterilization chamber and oven walls are reduced by the close coupling and by minimizing the mass of all components. The heat transfer rate to the con- nector is thus maximized. Test Procedure The test series consisted of six different temperature-time cycles: maximum temperatures of 260°, 305°, and 325°C, each for 65 and 85 seconds. Ninety—three connector assemblies were di— vided into six groups and each group was subjected to a differ— ent cycle. Membrane tubes and metal spikes were thoroughly cleansed in a Freon ultrasonic bath, subjected to a 12- hour ethylene oxide sterilization cycle, and air vented for 24 hours. Each compon— ent was inoculated with 0.01 ml of water containing 106 spores of Bacillus subtilis var. niger, a species commonly used as a challenge in determining dry heat sterilization efficiency. The spore suspensions were applied to the membrane faces and the inner wall of the spikes and dried for Eight hours at 55°C. Using aseptic techniques under a laminar flow hood, each spore- contaminated spike was inserted between two contaminated silicone tubes without breaking the membrane barriers. 73 The connector was then placed in the sterilization chamber. Once power was applied to the heater system, the air passing into the chamber reached an operating temperature of 330°C with— in 13 seconds. The chamber temperature was maintained at 330°C until the internal sections of the connector reached the maximum desired temperature, at which time the air temperature in the chamber was lowered to that value. In this way, the heat-up rate was consistent throughout the test program, regardless of the maximum temperature achieved. Thermal lag, both in heat up and cool down, between the air outside and inside the connector is presented in Figure 3. The heat-up time can be reduced by making available a constant supply of heated air. 1. ------ TEMPERATURE INSIDE CONNECTOR 3, ---- AIR TEMPERATURE WITNOIJT CONNECTOR 2, AIR TEMPERATURE OUTSIDE CONNECTOR 4. —-— OVEN TEMPERATURE 400 STERILIZATION CNAMIER ,V‘m.’\"~“‘.v.t’»3/M§ 300 ,/ w m!» " ' -I~/ W552“, ‘3’?!“ - I . I fl“ . MVVVK ° C / .' Temp 1/ / '3 ‘\ I \\ 100 ‘7' x ‘ \\ zx/ -- \ x I I ‘\ 0 20 40 60 80 100 120 140 160 TIME (Seconds) .\- ~\‘ ’__.— FIGURE 3 Controlled temperature histories at various locations within the sterilization system. Sterilization time was defined as that duration when the connector was at or above 250°C- This time began approximately 40 seconds after power was initially applied to the oven. The heat envelope charted in Figure 4 is representative of 65— and 85-second sterilization procedures at 260°C. The shaded area shows the range of several temperature-time histories de: rived from thermocouple measurements at each membrane and at the center of the spike. Temperature range throughout the connector assembly during heat up and cool down was i20°C; during sterili- zation (temperature 2250°C), the range was t10°C. ‘The total heat input, represented by the area under the temperature curves, 74 350 I I I I I I I I “— STERILIZATION TIME(S) —" 300 250 200 TEMP °C I50 I00 50 ,. ., “H O 0 20 40 60 80 I00 I20 I40 I50 TIME (SECONDS) FIGURE 4 Heat envelope representing temperature range within connector for sterilization times of 65 and 85 seconds. Tmax= 260°C. varied with the sterilization time and maximum temperature. Heat energy, applied after the connectors had reached 250°C, was the primary variable in this study. When the temperature—time cycle was completed, power to the oven was cut off. The connector was cooled to 60°C by directing ambient air into the chamber for 40 seconds. The connector was removed from the chamber and disassembled under the laminar flow hood using sterile gloves and flamed for- ceps and scissors. The components were cultured in 10 ml of soybean casein digest (SCD) broth and incubated at 30°-35°C for seven days. As each component of the three-part connector as- sembly was cultured separately, each connector test actually con- sisted of three separate challenges (units). Results The objective was to kill 100% of the spores applied to each spike or membrane. If a positive Spore culture was ob- served, the test was considered a failure. Because the basic functional requirement was the ability to kill one million spores applied to the connector surface, no attempt was made to determine’efficiency. All units tested at 325°C were sterilized at exposure times of both 65 and 85 seconds. Lower temperatures gave inconsistent results, indicating that the range between 260° and 305°C, for 75 either 65 or 85 seconds, would not provide reliable steriliza- tion of 106 spores for this connector design and method of heat- ing. The number of connector units successfully sterilized and the positive controls for each temperature-time cycle is shown in Table 1. TABLE 1 Sterilization Results Sterilization Timea Maximum Temperatureb (5°C) 260°C 305°C 325°C 65 48/48C 68/75 45/45 Positive control 0/6 0/9 0/6 85 22/24 63/63 24/24 Positive control 0/3 0/6 0/3 aSterilization time is defined as the total time that the con- nector is at or above 250°C. Time variation was :2 seconds. bTemperature range throughout the connector was t10°C of the temperature shown. cSterile cultures/total tested. Discussion Plester (5) has outlined the following recommendations for dry heat sterilization made by the Medical Research Council, United Kingdom: 160°C for 45 minutes, 170°C for 18 minutes, 180°C for 7—1/2 minutes, and 190°C for 1-1/2 minutes. Mullican et a1. (4) have reported the inactivation of aerosolized E. subtilis var. niger spores in 0.02 seconds at 230°C. However, test results from the study reported in this paper indicate that higher temperatures and longer time cycles are necessary to achieve sterility by dry heat for this practical application. In the air sterilization system reported by Elsworth et a1. (6), spores were exposed to a sudden influx of heat. In the application of dry heat to this connector system, however, the spores were heated gradually because of the relatively slow thermal diffusion rate through the assembly. This slower rate of heating created a different temperature—time history through- out the challenge, producing a different requirement for steril- ization. Also, the safety margins imposed on the connector system contribute to this different time-temperature cycle. Each con- nector assembly contained three challenges of a million spores. In his exposure time formula involving the decimal reduction time, Stumbo (7) has indicated that longer sterilization times may be expected for such a large population. 76 Another property of the challenge population was the high thermal resistance of the spores. Low water content within spores has been demonstrated to be a factor producing greater thermal resistance (3,4). To create a dehydrated challenge popu- lation in this test, the connector components were dried for eight hours following inoculation. Both simplicity of operation and suitability for steriliza- tion were objectives in designing the connector and in the se- lection of materials. Silicone rubber has the desirable proper- ties of heat resistance and moldability, and, by itself, the ability to form a good seal against the metal spike. The double spike is a simple and effective means of providing an in—line cannula for the silicone membranes. The brass safety clips per— form a multiple role. Of major importance is their ability to prevent premature penetration of the silicone rubber membranes. Also, they lend support to the flexible tubing during the joining of the connectors. Lastly, the presence of bent clips, or their absence, is a tamper indicator. Sterilization, at the time and site of fluid transfer, by means of flaming, acids, other chemical agents, or irradiation was not considered practical. Inactivation of microorganisms using the steam of an autoclave is not acceptable because of the lengthy time periods involved. Physical removal of organisms using filtration is limited to liquids and gasses. Dry heat sterilization was chosen for the connector system because of low cost, size of the ancillary equipment, and simplicity of opera- tion, all desirable features in the blood bank or hospital en— Vironment. This connector system utilizes connection techniques al- ready familiar to the blood bank technician. The spike- receptacle tube concept imposes no new handling problems and it allows for a connector with no moving parts. Most important, such a connector system for sterile transfer of fluids may con— tribute to the extension of the 24—hour limit for stored blood, once its sterile sealed plastic bag has been entered. The authors would like to thank Diane Allmen, Linda Buch, Joseph Duffy, Georgia Hawxhurst, James Kulla, Chris Lindberg, Karen Unrath, and William Zynda for their assistance on this project. REFERENCES 1. American Association of Blood Banks. Technical Methods and Procedures, 6th edition, p. 181. Washington, D.C., 1974. 2. Brannen, J. P., and D. M. Garst. Dry heat inactivation of Bacillus subtilis var. niger spores as a function of rela- tive humidity. Appl. Microbiol. 23:1125-1130, 1972. 77 Hoffman, R. K., V. M. Gambill, and L. M. Buchanan. Effect of cell moisture on the thermal inactivation rate of bac- terial spores. Appl. Microbiol. 16:1240—1244, 1968. Mullican, C. L., L. M. Buchanan, and R. K. Hoffman. Thermal inactivation of aerosolized Bacillus subtilis var. niger spores. Appl. Microbiol. 22:557—559, 1971. Plester, D. W. Effects of sterilizing processes on plastics. Biomed. Eng. (London) 5:443-447, 1970. Elsworth, R., E. J. Morris, and D. N. East. Chem. Eng. (London) 137:A47, 1961. Stumbo, C. R. Thermobacteriology in Food Processing, p. 92— 104. Academic Press, New York, 1965. 78 BACTERIOLOGIC AND METABOLIC EVALUATION THE GROWTH OF BACTERIA IN FROZEN-RECONSTITUTED BLOOD AND METHODS FOR DETECTION OF CONTAMINATION Byron A. Myhre, Yutaka Nakasako, and Richard Schott (UCLA School of Medicine, Harbor General Hospital Campus, Torrance, California) If units of blood which have been frozen, thawed, and reconsti- tuted are to be stored for periods of time at 4°C, sterility of these units must be assured before they can be transfused. Studies have shown that the incidence of contamination is quite low. The mechanics of freezing and thawing introduce few bacteria into the blood. If bacteria are introduced artificially, the washing pro- cedure removes most of them and the few that remain grow very poorly at a 4°C storage temperature. Finally, the Bactec machine provides a sensitive and fast method of determining that the blood unit is contaminated and should not be transfused. Since the early days of blood banking, it has been estab- lished that bacteria grow well in stored blood. Many of our technical procedures are devoted to obtaining and preserving the sterility of a unit so that bacterial contamination will not occur. In 1951, Borden and Hall (1) showed that bacterial con- tamination of a unit of blood could produce serious or even fatal transfusion reactions in patients who received this blood. At .that time, most contamination was produced by repeatedly entering and sampling the blood unit, but contamination was also found to occur by using donors who were bacteremic when drawn or by using improper phlebotomy technic. Studies performed showed that many bacteremias were caused by dental manipulation (5,11). Studies by Braude et al. (2,3) and others (6,7) showed that the severe reactions were due usually to gram-negative bacteria which grew well at 4°C and even metabolized citrate. Gram- positive organisms were found to produce a fever but not usually to produce the severe shock and death characteristic of the gram- negative organisms. Following these studies, most blood banks established a rule that if a blood container has been opened or entered for any reason, the blood should be transfused within 24 hours, after which it must be discarded. This rule was usually applied uni- versally, even though the entry had been performed under the most aseptic conditions. Recent studies by Bucholz et a1. (4) have documented similar reactions due to contaminated platelets. Numerous subsequent studies (8,9,12) have shown that most platelet concentrates are sterile, and indeed work by Myhre et a1. (10) has shown that they have a limited bacteriocidal effect, yet the danger of bacterial contamination cannot be disregarded. 81 The use of frozen blood has been severely limited by the ad— herence to this somewhat arbitrary 24-hour limit. Extremely rare units of blood may be stored for years in the frozen state. If they are thawed in anticipation of use and then not used, they must be either transfused to someone who does not need this rare type or they must be discarded. To avoid this loss, most blood banks thaw blood only at known times of need, which causes a de- lay in the providing of this blood. It would be much better if the blood could be thawed before use, stored several days at 4°C, and perhaps even shipped to another blood bank for use by someone else if the initial patient does not require the blood. Bacterial growth in frozen-thawed cells is hard to predict. Bacteria grow well in normal blood stored at room temperature or above. They grow much less rapidly if the blood is stored at 4°C. Frozen-thawed red cells are usually stored in 0.8% NaCl and 0.2% glucose solution which might serve as a minimal growth medium for bacteria and might produCe either more or less growth than whole blood. Additionally, the washing process required for frozen red cells removes the white cells, platelets, complement, and bac— teriolysins, which provide some of the normal bacteriological de- fense mechanisms of whole blood. This conceivably could allow the organisms to grow faster. This research was begun to establish the growth rate of various organisms in frozen-thawed red cells and to determine methods for the rapid detection of contamination. Initially we decided to determine the incidence of contami— nation of units of blood frozen and thawed by different methods. Fifty—three units of blood were frozen with the high glycerin, mechanical freeze method and washed by the Haemonetics method. Two units showed organisms. Eighty—six units were frozen with the low glycerin, liquid nitrogen method and washed with an IBM cell washer. None were found to be contaminated by usual blood culture technics. An effort was made to determine a possibility of contami- nation at the revolving seal of the centrifuge bowl. Sample Haemonetich§ bowls were subjected to a vacuum of 5 cm of Hg for periods of up to one hour. They were kept in one of three posi— tions: sitting on the laboratory shelf, in the centrifuge with the central shaft locked in the holder in an extended position, and in the centrifuge undergoing rotation. Fluorescein solution was sprayed or painted in the area of the seal. Subsequently, the bowl was rinsed with distilled water and the water analyzed for fluorescence. None was found. This seemed to indicate that the leak would be.minimal. A second attempt was made. Sterile bowls were tested in each of the same three positions under the same vacuum but were kept empty. Subsequently, each bowl was rinsed with 100 ml 82 Thioglycolate broth and this was cultured. Contamination was found in one case —— the central shaft locked in the up position. When this experiment was repeated a second time, no contamination was found. We attribute the first contamination to faulty sam— pling technic. These experiments were repeated with the IBM cell washer bags but were unsuccessful. Each time a decreased pressure was applied to the bag, it began to collapse and a vacuum could not be developed. From these experiments, we determined that the amount of leak which occurs at the centrifuge seal is minimal. The next set of studies were devised to show the growth po- tential of organisms in units of blood, assuming that they had been inadvertently introduced before freezing. Previously identified wild strains of bacteria were grown in nutrient broth, diluted in saline, and stored overnight in the refrigerator. Previous studies (9) had shown that in these conditions the organisms neither died nor reproduced but instead remained constant in number. At the time of storage, several dilutions of the organism suspension were made and placed on nutrient agar plates. The next morning colony counts could be performed to determine the number of organisms present in the suspension. Units of bloOd then were inoculated with known numbers of these organisms and frozen. Later they were thawed and washed in the Haemonetics bowl. The red cells were reconstituted in equal volumes of the 0.8% NaCl, 0.2% glucose solution, and the resulting mixture was stored at 4°C for varying amounts of time up to 21 days. Samples were removed at intervals for bacterial counts, various chemical determinations, and cellular counts. Initially, fourteen units of blood were contaminated before freezing. A sample graph of two of these units is seen in Fig- ure 1. The blood units were contaminated with 2 x106 and l.lx 103 organisms,respectively,of Staphylococcus aureus. A precip- itous decrease in the number of organiSms occurred during the process of freeZing and thawing. Later studies showed that al— most all of this decrease occurred during the washing phase, and most of the organisms could be recovered from the wash solution. The number of organisms removed was between log 2 and log 3 organisms (to base 10). After a large number of the organisms were washed out, the remaining organisms either remained the same in number or their number decreased during the 10 days of storage. The experiments were terminated at the end of 7 to 10 days because a large number of red cells were hemolyzed by that time and the units were felt to be unusable. Control studies with other units which were stored sterilely showed comparable 83 Q o---- LARGE INOCULUM 8 l06d \‘ .— SMALL INOCULUM 3 \ m '05— “ E o--0...o--lo-an-- ---- --- - - - -- .0 \ I04— 0) 5 3 2 IO — (<9 2- FIGURE 1 Growth of large and g '0 small inocula of Staphylococcus . aureus organisms in frozen— 0 IOI _‘ reconstituted red cells stored 2 at 4° C. A significant removal of organisms by the washing g g g ; § % g g g g : process can be seen. 0 2 4 6 8 '0 DAYS POST WASH hemolysis during the same time. These chemical studies will be reported in a subsequent communication. The next series of experiments was designed to determine the growth rate in the saline—glucose solution alone, or saline-+ CPD anticoagulant. Sample results are shown in Figure 2. The curves, which are representative of three studies, show that there is minimal growth in either saline-glucose or saline-CPD, and the curves obtained are rather similar to those found when the organisms were introduced into the red cells suspended in O C) C) _l CD 0-— SALINE-GLUCOSE % IO4— o---- SALINE+CPD l— —- on<1 g . \ '03- _-—---—_--.o (f) .2 9 FIGURE 2 Growth of ;: |()Z._ Staphxlococcus aureus < stored in saline—glucose (.9 medium or saline +CPD O: anticoagulant at 4°C. 0 { l E § § % § § £ 5 O z 2 4 6 8 IO DAYS POST WASH 84 saline—glucose medium. The red cells therefore do not seem to change the growth of the organisms appreciably. The next study was designed to show the growth of large numbers of organisms in the red cell suspension. The previous studies had not been too informative on this point since most of the organisms were washed from the red cells during the washing process, and only a limited number of organisms remained. Lack of growth in these cases might be due to insufficient bacteria to count accurately, and the so-called lack of growth in reality might be a counting error. To test this hypothesis, units of blood were frozen, thawed, and washed as before, then divided into two portions: one was inoculated with a light suspension, and the other with a heavier suspension of the same organism. We were thus able to determine if organism density had any in- fluence on the growth rate. Eleven units of blood were analyzed in this manner. Sample results of the light inoculum are shown in Figure 3. Both in larger and smaller numbers the organisms either did not grow appreciably or diminished in numbers. Therefore, counting errors did not influence the growth rate. During this study it was found that the difference in the lysis of red cells, which were heavily contaminated with bacteria, did not differ significantly from the units which were lightly con- taminated. O- K-E O---- Ps o——-°—-O———o—__.___.__—o 0"-- ---. O 0“--0.-....o.-.-.o..---O 5 r FIGURE 3 Growth of gram— negative organisms inocu- lated in small numbers to units of frozen—reconstituted red cells stored at 4°C. No. ORGANISMS/ ml BLOOD 5 5 N 0| I l *— l I l l l T I I I I 7 '2 3 4 5 6 7 DAYS POST WASH So far all studies had been performed with red cells sus- pended in the saline-glucose solution. This is a minimal main- tenance solution, and the red cells begin to undergo rather severe hemolysis after 7 to 10 days. In an effbrt to determine whether the storage time could be extended further, red cells were frozen, thawed, washed, and then resuspended in their own plasma which had been previously frozen and stored. We found satisfactory storage of these red cells for a three-week period. 85 As a control on the growth of the bacteria, these units were contaminated and comparable counting was carried out while the unit was stored at 4°C. A sample growth curve from these studies is shown in Figure 4. In a few cases the organisms doubled in number during the storage period, but in many they decreased to less than the initial number. From this study we concluded that growth of most bacteria at 4°C is much less pro— nounced than has been inferred. .— STAPH. AUREUS 0-" ECOU 0—- Ps. AERUG No. ORGANISMS/ UNIT OF BLOOD I04— '03— FIGURE 4 Growth of various |02__ bacteria in frozen-reconstituted red cells resuspended in their own plasma and stored at 4°C. I 3 5 7 M 3 DAYS POST WASH Currently, culture studies are being done to determine the percentage of cold—growing organisms which occur. The results so far have shown that only a rare gram—negative and almost no gram-positive organisms grow in blood at colder temperatures and that the classic cold-growing organism is a rather unusual item. Bactec Analyses As the study progressed, it became clear that it was impor- tant to develop a method of determining bacterial contamination in a time period sufficiently short that the unit could be trans— fused if its sterility could be assured. Blood cultures per— formed by the usual technics did not seem to offer this speed. By the time the blood culture has demonstrated no growth, the unit of blood would have autolyzed sufficiently to be untrans— fusible. On the other hand, the BactecQ§ machine [Johnston Laboratories of Cockeysville, Maryland] seems to offer this speed. 86 The instrument operates by providing a nutrient medium for the bacteria which includes either 1|*C-labeled glucose or 1"C— labeled amino acids. Culture vials are available which provide aerobic or anaerobic conditions. Finally, a third vial, which is made hypertonic by the addition of extra sucrose, has been shown to promote the growth of some of the very fastidious organ— isms. After the organisms have been introduced into the vial, it is incubated at room temperature or 37°C for varying amounts of time. Subsequently, the gas above the medium is sampled for the presence of 11+C02, which is used as evidence of bacterial metabolism. Cumulative growth curves were performed to show that the machine would indicate bacterial growth in the usual manner, and a sample of the results is shown in Figure 5. Little growth is seen in the anaerobic vial when the organism is S. aureus, but the usual S—shaped curve is derived with the aerobic vial. Sampling of the labeled gas at hourly or 4—hourly inter- vals seemed to distort the curve very little. (/3 m5 —1 U) 5 < S . VIAL sA-AERoanc RUN A1. = 8 5 O VIAL BA-AEROSlC-HYPERTONIC HOURLY 800 2 400 ® VIAL TA- ANEROBIC 'NTERVALS Lu T ' AAA ABOVE VIALs — 2133:; ’4 Z INTERVALS E .3: 600— 2,000- : ___——O 8 2 400— __ I,4oo- / Lg) , \ 300— E Looo~ D E 200— 800- I r- 600— 3 . g Ioo- 4oo~ ' ————— 0 _ O —— 20° 5 o -®--@"®"®-::_- ___——A o— o ---*- ----- iééiéé‘réélbl'll'zlilh 2'4 TIME (Hours) FIGURE 5 Cumulative bacterial growth curves performed with the Bactec instrument. Studies were performed in which the units of blood were kept at 4°C after being inoculated with a known organism. At periodic intervals, aliquots of blood were removed and intro— duced into Bactec vials and also into petri dishes for colony counting. The results are listed in Table 1. As can be seen, the higher density of organisms often shortened the time of Bac— tec incubation somewhat, but most cultures were detected within 16 to 20 hours. An occasional specimen required up to 30 hours 87 TABLE 1 Number of Hours of Culture Needed for a Positive Result with the Bactec Instrumenta Organism ' No. Organisms in Bag Hours for-+Bactec Staphylococcus aureus 1.13x107 16 1.32 x 107 16 5.87 x 106 16 3.6 x 106 12 3.3 x106 16 1.1 x 106 20 1.07x105 16 1.03x105 8 9.7 x104 18 6 x 104 16 6.9 x103 16 4.7 x103 16 4.0 x103 30 3.6 x 103 16 3.3 x 103 16 2.5 x 103 16 2.2 x103 16 6 x 102 21 5 x 102 30 3.6 x 102 24 Klebsiella-Enterobacter 1 Km7 1 Sp. 4.2 x 105 12 3.6 x 102 16 2.5 x102 24 Pseudomonas aeruginosa 107 1 8 x 106 10 7 x10 12 5 x 104 10 3 x 103 16 7 x 102 12 aUsing the designated concentration and type of organisms. All organisms were removed from previously inoculated units of blood. incubation before they were detected. The gram-negative organ- isms were detected as easily as gram-positive. However, the growth was usually found first in the hypertonic vial rather than in the aerobic one. Discussion Blood which is to be transfused needs to be sterile so that it will not produce a severe reaction in the recipient. The preservation of this sterility becomes extremely critical when the unit is subjected to freezing, thawing, washing, reconstitu— tion, and then storage at 4°C for varying amounts of time before transfusion. The many manipulations needed to free or thaw a unit of blood allow ample opportunities for the unit to become 88 contaminated. The solutions themselves, or the plastic equip- ment used in the process, could be contaminated. Further, each time a bottle of solution is entered, a possibility for the in- troduction of bacteria exists. If the continuous—flow washing procedure is used, the most obvious source of contamination is the introduction of air through the seal in the centrifuge bowl which connects the fixed central shaft with the rotating bowl. Any time the pressure on the solutions flowing through this is decreased, contaminated air could be aspirated into the solution. We assume that a de- creased pressure would normally be obtained by either dropping one of the solution bottles on the floor or accidentally lower— ing it. The maximum vacuum which could be generated by this method is about 4 feet of water and would normally occur for a small amount of time. When leakage was tested by using a vacuum of 5 inches of Hg (equivalent to 5.66 feet of water) for 1 hour, the seal did not leak. Therefore, the possibility of contami- nation from this source is not too likely. The incidence of con- tamination found in finished units confirms this study. Even if organisms are introduced during the bleeding or freezing process, most of them will be washed out during the washing phase. This is most likely due to the light bacteria being packed less densely than the heavier red cells. The wash solution either removes them by convection, or dilutes them. The presence of glycerin in the first wash solutions would raise the specific gravity of the mixture and probably remove the bacteria faster than the more dilute wash solution. Further contributing to the safety of the unit is the poor growth of most pathogens at 4°C. The data shows that most patho— gens grow in red cells resuspended in saline-glucose at about the same rate that they do in saline alone, if they grow at all. Finally, studies showed that this slow growth rate is not influenced by the density of the organisms which were introduced. For all of these reasons, it appears that the incidence of bacteria in units of frozen-thawed blood is low, and that most of the organisms grow slowly even if they are introduced. Longer storage of the reconstituted units of blood would therefore be- come feasible. Nevertheless, it is not wise to depend on statistics to show that the blood unit most likely is sterile. A fast method of proving sterility would be much better. This method should give results rapidly so that the blood is usable after the re- sults are obtained. We have found (to be published) that blood stored in saline—glucose for more than 7 days exhibits so much hemolysis that it is probably not safe to transfuse. Therefore, this seems to present an end point beyond which our technics cannot progress currently. The Bactec machine provides a rapid 89 method for sterility analysis which seems practical. In most cases the organisms were detected within 16 hours of incubation, and in the 15% of cases where small inocula were used they were detected in less than 30. In only one case was it necessary to incubate the Bactec vials for 36 hours to obtain a positive re— sult with a very small inoculum. From these studies we have devised a simple and practical method for detecting contamination in a unit of reconstituted frozen blood. It is as follows: As soon as the unit of blood is washed, samples of the unit are introduced into the three Bactec vials, and the culture of these vials is begun at 37°C. During the first 24 hours the vials remain untouched. If the unit is needed during this time, it is issued and trans- fused as always. At the end of 24 hours, the Bactec vial is sampled to determine whether contamination is present. If the vial is sterile, the unit is put on quarantine and could be available for emergency trans- fusion since the organisms present, if any, would be few. The vials are sampled again at 36 hours and, if still sterile, the unit is considered aseptic and could be transfused for up to 5 days. To date, none of these 5—day stored blood units have been transfused, but all of the blood cultures taken have shown the correctness of this hypothesis. The blood cultures took one week to 10 days to confirm, and so the information obtained from them, although scientifically useful, is clinically useless. By this method, we feel that Viable bacteria may be pre— vented from being transfused. One must remember, however, that it is conceivable that a unit of blood could have gram-negative organisms which have been killed, lysed, or rendered inactive, thereby showing no viable organisms, and yet their endotoxins could be present in the unit of blood and still produce a trans- fusion reaction (13). Therefore, it may be necessary to test these long—term stored units for the presence of endotoxins as well as viable bacteria. This phase of the work will be con- tinued subsequently. The work on which this report is based was supported by NHLI Contract NOl-HB—4—2927. REFERENCES l. Borden, C. W., and W. H. Hall. Fatal transfusion reactions from massive bacterial contamination of blood. New Engl. J. Med. 245:760, 1951. 90 10. ll. 12. 13. Braude, A. I., F. J. Carey, and J..Siemienski. Studies of bacterial transfusion reactions from refrigerated blood: The properties of cold-growing bacteria. J. Clin. Invest. 34:311, 1955. Braude, A. 1., J. P. Sanford, J. E. Bartlett, and O. T. Mallery. Effects and clinical significance of bacterial contaminants in transfused blood. J. Lab. Clin. Med. 39: 902, 1952. Bucholz, D. H., V. M. Young, N. R. Friedman, J. A. Reilly, and M. R. Mardincy. Bacterial proliferation in platelet products stored at room temperature. New Engl. J. Med. 285: 429, 1971. Elliott, S. C. Bacteremia and oral sepsis. Proc. Royal Soc. Med. 32:747, 1939. Geller, P., and E. Jawetz. Experimental studies on bacter- ial contamination of bank blood. I. The nature of "toxicity" of contaminated blood. J. Lab. Clin. Med. 43:696, 1954. Hall, W. H., and D. Gold. Shock associated with bacteremia. Arch. Int. Med. 96:403, 1955. Katz, A. J., and R. C. Tilton. Sterility of platelet con— centrates stored at 25°C. Transfusion 10:329, 1970. Mallin, W. S., D. T. Reuss, J. W. Bracke, S. C. Roberts, and G. L. Moore. Bacteriological study of platelet con— centrates stored at 22°C and 4°C. Transfusion 13:439, 1973. Myhre, B. A., L. J. Walker, and M. L. White. Bacteriocidal properties of platelet concentrates. Transfusion 14:116, 1974. Okell, C. C., and S. C. Elliott. Bacteremia and oral sep— sis with special reference to aetiology of subacute endo— carditis. Lancet 4:869, 1935. Silver, H., A. C. Sonnenwith, and L. D. Beisser. Bacteri- ologic study of platelet concentrates prepared and stored without refrigeration. Transfusion 10:315, 1970. Yoshikawa, T., K. R. Tanaka, and L. B. Guze. Infection and disseminated intravascular coagulation. Medicine 50:237, 1971. 91 CULTURAL STUDIES OF PREVIOUSLY FROZEN AND WASHED RED CELLS Paul T. Wertlake and Shirley E. Wertlake (Saint Barnabas Medical Center, Livingston, New Jersey) Data represents cultural results of 562 red cell units pro- cessed by open system. Approximately 90% of the red cell units were previously frozen red cells and 10% washed red cells. Approximately 95% were transfused to patients. Two hundred and twenty-two red cell units were serially cultured. Each unit was cultured twice weekly to 21 days fol- lowing washing of units. The following media and culture con- ditions were employed: Thioglycollate at 4°C, 23°C, and 37°C; and Sabouraud slants at 23°C. Total cultures were 6,216. Three hundred and forty red cell units were cultured by single sampling. Units were sampled l day to 108 days follow- ing washing of units. Units were maintained at 4°C until time of culturing. The following media and culture conditions were employed: Thioglycollate at 23°C; Trypticase Soy at 23°C, 37°C with 002; Brain Heart Infusion at 23°C, 37°C with C02; and Liquid Sabouraud at 23°C. Total cultures were 4,760. Total cultures of the 562 red cell units including the serial sampling and single sampling studies were 10,976. The 562 red cell units were processed using three different washing methods: 134 units were processed by manual washing, 277 units by the Fenwal Elutramatic, and 151 units by the IBM 2991. Bacterial growth was obtained from: 20 of 562 (3.6%) red cell units cultured; 9 of 222 (4.1%) red cell units cultured by serial sampling; 11 of 340 (3.2%) red cell units cultured by single sampling; 6 of 134 (4.5%) red cell units manually washed; 12 of 277 (4.3%) red cell units washed with the Fenwal Elutramatic; 2 of 151 (1.3%) units washed with the IBM 2991; and 2 of 428 (0.05%) instrument-processed units positive in first 7 days. The results indicate bacterial growth from previously frozen and washed red cell units comparable to closed system blood and platelet concentrates. The authors conclude that the 24-hour outdate period is unnecessarily restrictive, and that a new outdate should be determined which takes full advantage of the bacteriologic safety offered by current technologic advances and the poten- tial offered by biochemical support of red cell metabolism. Introduction The 24—hour outdate for blood processed in an open system proves to be a very restrictive and costly regulation. In the real world of active medical practice, a variety of factors con— tribute to difficulty in avoiding outdate of previously frozen red cells. Frequently, frozen red cells are thawed, washed, and resuspended at a central facility and transported to a hospital. 93 Time is consumed in transportation. Previously frozen red cells not used in the hospital for the patient intended or another suitable patient, if available, must be shelved. In the remain- ing hours until outdate, opportunity to transfuse the previously frozen red cells is reduced and the risk of outdate is substan- tial. Alternately, the previously frozen red cells might be transported to still another hospital in hope of using the pre- . viously frozen blood. This generally is neither practical nor economical. Some situations in which the previously frozen red cells ordered are not used include: the amount required to sup- port the patient being less than originally estimated, the unan- ticipated death of the patient, and the reluctance of physicians to release unused autotransfusion units postoperatively since unanticipated bleeding might require transfusion. These exam- ples illustrate that a very significant portion of the 24 hours may elapse with the unit not being used as initially envisioned. This, in turn, may leave only a short period of time in which to use the unit for another patient, and this shortened interval of time on occasion is insufficient to locate an appropriate recip— ient.> The results are loss of donated units and nonproductive cost to the medical care system. The 24-hour outdate period was empirically set because of fear of bacterial proliferation and contamination within an open system. The present study was undertaken to provide data indic— ative of the time period during which previously frozen and washed red cells may be safely used. The study was conducted in an BOO-bed community hospital, Saint Barnabas Medical Center, Livingston, New Jersey. Approx- imately 90% of the red cell units were previously frozen red cells and 10% washed red cells. Approximately 95% of the units studied were transfused to patients without ill effect. Methods General Culturing Process All primary cultures of units, inoculations of broths, and streaking of plates or slants were carried out within an en- closure limited to insertion of gloved hands. This was done to reduce possible contamination by room air and to reduce hazard of contamination by breathing, coughing, or sneezing of the per- son doing culture work. All surfaces were cleaned with alcohol followed by Betadine to dryness prior to perforation by syringe and needle for sampling of the unit. A new sterile syringe and needle were used for each sampling. Samples were taken after thorough mixing of resuspended red cells. Three drops of blood were entered into each tube of the various broths employed. All broths were 10 ml in volume. 94 Serial Culture Procedure Two hundred and twenty-two red cell units were serially cultured twice weekly to 21 days following washing of the units. These units were thus sampled at approximately 3—day intervals. Each sampling included inoculation of Thioglycollate Broth incu— bated 10 days at 4°C, 23°C, and 37°C with C02 and streaking of Sabouraud slants incubated 6 weeks at 23°C. Total inoculations were 6,216. Single Sampling Study Three hundred and forty red cell units were studied by single sampling. These units were variously sampled 1 day to 108 days following washing of the units. Units were maintained at 4°C in the interval between washing and culturing. The fol- lowing media and culture conditions were employed: Thioglycol— late at 23°C; Trypticase Soy at 23°C, 37°C with C02; Brain Heart Infusion at 23°C, 37°C with C02; and Liquid Sabouraud at 23°C. Cultures were incubated for 14 days except for inocula in Sabourauds which were incubated 6 weeks. All negatives were subcultured to blood agar plates for confirmation of no bacter— ial growth. Total inoculations were 4,760. Manual Washing Procedure Manual washing of previously frozen red cells was performed according to the method described by Rowe et al. (3). Samples for culture were obtained after the last wash by expressing thoroughly mixed blood from the transfer pack to the satellite bag which was isolated and separated by heat sealing. Manually washed units represented 134 of the red cell units serially cul- tured. The successive sampling of the satellite bag was accom- plished by cleansing and piercing the tubing leading to the satellite bag. Sampling sites were started distally and moved progressively closer to the bag. Fenwal Elutramatic Processing Two hundred and seventy-seven units were processed by the Fenwal Elutramatic: 88 of these units were included in the serial culture study and 189 were included in the single sampling study. The processing method employed was the simplified method reported by Valeri (4). Sampling of the Elutrapak was accom- plished by expressing a small amount (approximately 10 ml) of the final washed red cell preparation into a transfer bag. Units were identified as originating from the blue or white coded bags of the Elutrapak. At the time of culturing, contents of the transfer bag were well mixed. A representative sample was ob- tained from the bag by stripping the tubing multiple times to insure adequate mixing before cleansing and withdrawing the spec- imen with a sterile syringe. 95 IBM 2991 Processing One hundred and fifty—one units were processed by the IBM 2991. The procedure followed was that described by C. R. Valeri (referred to above). These units were included in the single sampling study. Samples were obtained by preparing segments of the tubing superior to the clamping device. Segments were pre- pared by sealing with IBM metal crimps. Control~Units Thirteen control units were employed. Entire units of thawed red cells resuspended to a volume of 300 ml in sterile normal saline were inoculated with 108 organisms (103 organisms/ ml). Organisms utilized were stock cultures maintained at 37°C. Control units were prepared in the microbiology section of the clinical laboratory as unknowns. After 12 hours of incubation at 37°C, the controls were provided as unknowns to the person (S.E.W.) performing the culture studies. Control units were cultured by the protocol described above for single sampling study. Results Serial Sampling Study Serial samples were obtained from 222 red cell units, of which 134 were manually washed and 88 were washed by the Fenwal Elutramatic. Nine of 222 red cell units (4.0%) cultured by the serial sampling technique grew microorganisms. Six of the nine units growing microorganisms were manually washed and three were processed by the Fenwal Elutramatic. Of the 6 positive units manually washed, 2 units exhibited bacterial growth the first day following washing of the units, 1 unit day 4, 1 unit day 10, and l'unit day 24. Of the 3 positive units washed by the Fenwal Elutramatic, 1 unit exhibited bacterial growth day 4 and 2 units day 10. Single Sampling Study Eleven of 340 red cell units (3.2%) cultured grew micro— organisms. The 11 positive units were positive on the following days after washing: 6, 15, 29, 75 (two units), 78, 81 (three units), 94, and 102. Manually Washed Units Six of 134 red cell units (4.5%) manually washed grew microorganisms. Of 6 positive units, 3 grew multiple micro- organisms. Summary of microorganisms grown appears in Table l. 96 TABLE 1 Microorganisms Recovered: Manual Processing —— 6 Culture Positive Units Staphylococcus epidermidis Microorganisms Number of Units Pseudomonas species 2 Diphtheroids 2 Penicillium 2 Rhodotorula 2 l 1 Mima polxmorpha Fenwal Elutramatic Process Twelve of 277 units (4.3%) washed with the Fenwal Elutra- matic grew microorganisms. Three of 12 positive units grew mul- tiple microorganisms. Microorganisms recovered from the 12 pos- itive units appear in Table 2. TABLE 2 Microorganisms Recovered: Elutramatic Processing —— 12 Positive Unitsa Microorganisms Number of Units Pseudomonas fluorescens Pseudomonas maltophilia Pseudomonas mallei Pseudomonas species Mima polymorpha Streptococcus faecalis Herelea vaginicola H H a la H H .A aSix positive units were obtained from the white coded bag of Elutrapaks and five positive units from the blue; bag of origin was not determined for one positive unit. IBM 2991 Procedure Two of 151 red cell units (1.3%) processed with the IBM 2991 grew microorganisms. The organisms recovered appear in Table 3. TABLE 3 Microorganisms Recovered: IBM 2991 Processing —— 2 Culture Positive Units Microorganisms Number of Units Mima polxmorpha Diphtheroids 1 Staphylococcus epidermidis 97 Control Studies Tabulation of microorganisms recovered from controlled units appear in Table 4. It will be noted that units #4, #6, #10, and #13 which were not inoculated were found to be sterile. Failure to recover Pseudomonas aeruginosa occurred in two units (units #1 and #2) in the presence of Staphylococcus aureus and Clostridium perfringens. S. aureus was not recovered on one occasion in the presence of Escherichia coli (unit #12). TABLE 4 Control Units (all units inoculated with 3X105microorganisms) Control Unit Inocula Recovered #1 Pseudomonas-aeruginosa - Staphylococcus aureus + Clostridium perfringens + #2 S. aeruginosa - S. aureus + S. perfringens + #3 S. aureus + #4 None - #5 S. aureus + #6 None — #7 S. aeruginosa + Streptococcus faecalis + #8 S. faecalis + Enterobacter aerogenes + #9 S. aureus + #10 None - #11 g. aeruginosa + S. aureus + #12 S. aureus - Escherichia coli + #13 None - Discussion Results indicate bacterial growth in previously frozen and washed red cells to be low. Incidence appears to be influenced by the washing procedure employed. The occurrence of bacterial growth was low in units processed with the IBM 2991 as compared to the two other washing procedures tested, manual and Fenwal Elutramatic processing. These washing procedures had bacterial growth rates significantly greater than that experienced with the IBM 2991. This appears to indicate that all washing proce- dures are not equally effective in providing protection from microorganisms. The employment of enriched media in the single sampling study did not result in a greater percentage of bacterial growth 98 than that afforded by the media employed in the serial culture studies. This suggests that employment of various enriched media may not be necessary for routine detection of bacterial contamination. Despite employment in these studies of protocols designed to maximally recover microorganisms by employment of enriched media, with extensive culturing including serial cultures of the same units and culturing of units over extended periods of time, low recovery rates of microorganisms resulted. The occurrence of bacteria in only 1.3% of previously fro— zen and washed red cells processed routinely in a hospital lab— oratory, with a commercially available semiautomated instrument indicates that bacterial risk comparable to whole blood (1) or platelet concentrates (2) is now achievable. Of the 20 culture positive units in these studies, only 5 units occurred in the first 7 days. Three of these units were manually washed. It is well known that risk of bacterial con- tamination is higher with manual washing. Only two instrument- processed units were positive in the first 7 days (day 4, day 6), an incidence of only 0.05%. These studies indicate that bacterial contamination is low, that 97—99% of red cell units remain sterile for protacted peri- ods of time, and that red cell units are bacteriologically safe for clinical use well beyond the present outdate period of 24 hours. It appears that techniques now possible with available instrumentation and other supportive equipment enable processing which is comparable with closed system blood insofar as bacterial risk is concerned. Extension of the current 24—hour outdate period is clearly indicated. The determination of a new outdate should not be based on present practices. In addition to bacteriologic con— siderations, a new outdate should provide for full utilization of the potential offered by biochemical support of red cell metabolism. In this way, maximal benefits relative to costs can be achieved. Authors' present addresses: P. T. Wertlake, Cedars-Sinai Medical Center, 4833 Fountain Avenue, Los Angeles, California 90029; S. E. Wertlake, Northridge Hospital, Northridge, California 91324. REFERENCES l. Braude, A. I., J. P. Sanford, J. E. Bartlett, and O. T. Mallery, Jr. Effects and clinical significance of bacterial contaminants in transfused blood. J. Lab. Clin. Med. 39: 902, 1952. 99 Buchholz, D. H., V. M. Young, N. R. Friedman, J. A. Reilly, and M. R. Mardiney, Jr. Detection and quantitation of bac- teria in platelet products stored at ambient temperature. Transfusion 13:268, 1973. Rowe, A. W., J. B. Derrick, W. Miles, F. H. Allen, Jr., and A. Kellner. The biochemistry and clinical use of red cells frozen by the low glycerol-rapid freeze technique, pp. 184- 198. In Modern Problems of Blood Preservation. Gustav Fischer Verlag, Frankfurt/Main, Germany, 1969. Valeri, C. R. Simplification of the methods for adding and removing glycerol during freeze—preservation of human red blood cells with the high or low glycerol methods: Biochemi- cal modification prior to freezing. Transfusion 15:195, 1975. 100 STERILITY TEST OF ROTATING DISPOSABLE SEAL ON THE IBM 2991 BLOOD CELL PROCESSOR Frederick R. Kronenwett (American Biological Control Laborato- ries, Tenafly, New Jersey) The IBM 2991 Blood Cell Processor washes fresh or thawed- frozen red blood cells. The procedure takes place in a closed system wherein wash solutions are introduced and supernatants decanted automatically. A test was performed to determine if the contents of the disposable processing set remained sterile during processing. A challenge of l07 spores of Bacillus subtilis variant niger was given to the area around the rotating seal on the processing set. This was done eleven times plus one control and all tests were negative. Therefore, under this se- vere challenge, the disposable rotating seal of the processing set was able to prevent the contamination of the fluid being processed. A test was performed to determine if the rotating seal on the processing set of the IBM 2991 Blood Cell Processor forms an effective barrier against contamination from the environment. This was done by operating the Processor in a recirculation mode while challenging the seal by providing an aerosol of spores in the Vicinity of the seal. After the run, a concentrated culture medium was injected into the processing set and incubated for ten days. All tests were negative which indicates that the ro- tating seal does form a barrier against external contamination. Method An IBM 2991 Blood Cell Processor was modified as shown in Figure l. A box was constructed over the area which includes the rotating centrifuge assembly. Also, two 3.8-cm diameter flexible hoses were connected from the box to the space under the centrifuge. This allows air to circulate from under the centrifuge bowl to the space above the bowl which includes the location of the rotating seal under investigation. The details of the seal construction are shown in Figure 2. Ordinarily, the action of the rotating bowl causes air to be pumped from the space inside the Processor, around the bowl, and to the outside of the machine. By constructing the enclosure, this air is con— tinuously circulated around the bowl and seal. For each run the processing set was attached, using two of the lines on the set, to the two ports of a plastic bag [Travenol Laboratories, Inc., Deerfield, Illinois] containing 1000 ml of 0.9% NaCl. This attachment was done on a laminar flow bench. The set was then mounted on the Processor and the box assembled around the bowl area. The tubing on the set was connected through the valves as shown in the figure. This allowed the saline to drain from the bag, spin in the centrifuge bowl, and be decanted back into the original bag. As soon as the automatic 101 FIGURE 1 Schematic diagram of modification of IBM Blood Cell Processor to provide recirculation of air around centrifuge bowl and rotating sea 1 . FIGURE 2 Cross—section of IBM 2991 Blood Cell Processor rotating seal. ,J‘ ROTATING SEAL SALINE BAG HEMOSTAT——\\\\ jzfi SPORE INJECTION PORT law—l? «WP _ _ 1L1 RECIRCULATION HOSE ( : l /E;/ / / III 177 II \ J \L—J RECIRCULATION HOSE CENTRIFUGE BOWL \\\ \ \\ /7 / \\\\\W \ \ :Zm’l ‘ §§\\\\\\\\\\ \7 / / \/ ll gr Latex 4 j Upper Ceramic Seal 6 V/I Lower Ceramic Seal Felt Journal Bearing 4—— Latex Sleeve \ / é a a P -B rocessmg ag E __ __mc:[5 51‘5 /////’ Afi / / K A l 102 fl processing began, a cap was removed on one of the recirculation hoses, and an aerosol of spores of the organism Bacillus subtilis variant niger in an alcohol carrier was injected into the hose. A viable spore count was made on the spore suspension and 30 ml of the suspension aerosolized via a freon—dispenser mechanism. This was done over a timespan of about five minutes to give the spores time to distribute themselves throughout the volume of air. The machine parameters were set as follows: TIMER l : 5 minutes TIMER 2 : immaterial Centrifuge Speed : 3000 rpm Super Out Rate : 450 ml/min Pump Restore Rate: 450 ml/min Super Out Volume : 400 ml AUTO/ MANUAL : AUTO Each spin cycle of five minutes was followed by approxi— mately one minute of pumping out saline, followed by another minute of agitation during which the saline drained back into the processing bag which is inside the centrifuge bowl. This was repeated for seven spin cycles, giving a total processing time of about 45 minutes. After processing, the tubing was clipped off above the seal using the metal clips which are de— livered with the IBM processing sets. There was about 750 m1 of saline in the original bag and 250 ml in the processing bag. During the cycling, exposed petri dishes were placed in the chamber to measure qualitatively the fallout from the aerosol upon the surface of the chamber. The fluid that was cycled was physiological saline (1,000 ml), and after completion of the cycle, concentrated culture medium was injected to culture any spores that may have permeated into the fluid pathway. The blood bag and the uphill saline bag were incubated with the culture medium for 7 days at 35°C and then examined for growth. Positive and negative controls were also run concur— rently to validate the experimental asepsis and ability to sup- port growth of the spore inocula. The viable spore count on the alcoholic suspension of Bacillus subtilis var. niger spores showed the count to be at 2,000,000/ml of solution. For each run made on the unit, with the exception of the first run, a concentration of 60,000,000 spores was recirculated for 45 minutes. The fallout plates ex- hibited over 500 colony—forming units per plate. The culture medium was six strength Trypicase Soy Broth that had been sterilized in 60sec catheter syringes and then in- jected aseptically into the saline bag and the processing bag. A dilution of 1 part broth to 10 parts of physiological saline was maintained depending upon the volume in the respective bags. A positive control inocula of the physiological saline and the 103 broth exhibited good growth within a 48—hour period at 35°C. The level of spore inocula was at approximately 1,000 spores challenge by injection. A negative control on the technique em— ployed shown no contamination of the saline broth bag when it was manipulated according to the routine with no aerosol being produced in the chamber. This was obviously the initial run of the experiment since the chamber was contaminated after the first run by residual organisms from previous runs. In total, 12 runs were made with the aerosol exposure pat— tern duplicated each time, plus whatever residual remained in the tubing between runs. After inoculation with the media and incubation for 7 days at 35°C, none of the bags exhibited growth of g. subtilis var. niger. Under the inordinate conditions of this study, the rotating seal on the IBM 2991 Blood Cell Processor did not allow permea— tion of the enclosed microbial aerosol of Bacillus spores. Summary Under the conditions of this study, the IBM rotating seal did not allow the permeation of a bacteriological aerosol of Bacillus subtilis var. niger spores during a 45-minute time peri— od under a recycling of 60,000,000 Bacillus spores. 104 BACTERIOLOGICAL AND BIOLOGICAL QUALITY OF FROZEN RED CELLS 72 HOURS FOLLOWING DEGLYCEROLIZATION H. T. Meryman and R. A. Kahn (American National Red Cross, Blood Research Laboratory, Bethesda, Maryland) Data collected on 2273 units of frozen-deglycerolized red cells processed by five different laboratories indicated a 2.4% incidence of contamination. When units found negative on second culture are subtracted, l.2% were positive. When contaminations attributable to a contaminated water bath are subtracted, the contaminations were reduced to less than 1%. The contaminating organisms were predominately air and skin borne. Ten different organisms were inoculated into red cell suspensions. There was at least a three-fold reduction in titer following glycerolization, freezing, and deglyceroli- zation, and in no case was any growth observed during three subsequent days of storage at 4°C. After three days post-thaw at 4°C, no significant changes in 2,3-DPG were seen, ifl vivo survival of red cells was comparable to that at one day, free hemoglobin did not exceed 0.5 gm %, and changes in cell electro- lytes were modest. It is concluded that red cells are biologi- cally acceptable at 3 days post-thaw and that the increased risk from bacterial contamination is negligible. Since the Red Cross Frozen Red Cell Program was initiated in 1972, over 100,000 units of frozen cells have been deglycer- olized and transfused, and probably an equivalent number has been processed by other blood banks without report of adverse results. The safety and efficacy of this blood product appears to have been established and a number of direct and indirect benefits of frozen cells have become apparent with increasing use (3). On the negative side, other than cost, the 24—hour outdating period for frozen-deglycerolized red cells presents the greatest obstacle to their use. The studies summarized here were designed to aid in evaluating the added risk to the recipient which would result from extending the post-thaw stor— age to 72 hours. All data reported are for cells collected in CPD anticoagulant and frozen according to the American Red Cross protocol (4). Except where specifically noted, all frozen cell suspensions were deglycerolized in the Haemonetics Cell Washer [Haemonetics Corporation, 8 Erie Drive, Natick, Massachusetts 01760] using the disposable bowl according to the Red Cross pro— tocol. Two factors must be evaluated in considering an extension of the present 24-hour limit: the biological quality of the cells and the hazard presented by accidental bacterial contami- nation at the time of processing. 105 Biological Quality Several factors require examination, particularly in vivo survival, 2,3—DPG levels, cell electrolytes, and supernatant hemoglobin. Our own studies of cell survival after 72 hours of post—thaw storage have been limited to two autotransfusions of 1Cr-labeled cells conducted for us by Dr. Irma Szymanski. Both units were stored at -80° in excess of one year and showed in vivo survivals of 82 and 83% of the cells 24 hours post-trans- fusion. Valeri (6) has reported a series of 10 transfusions of cells 4 days post—thaw with a mean 24—hour post—transfusion in vivo survival of 85%. Figure 1 presents the results of assays of 2,3-DPG for up to seven days post—thaw. There is no significant difference be- tween the prefreeze and post-thaw levels, and no significant fall is seen up to three days post—thaw. Valeri (6) has reported an average 50% reduction 4 days post-thaw. 2.3-DPG IN FROZEN DEGLYCEROLIZED RED CELLS STORED AT 4°C 120 a 2100 80 60 40 petcent of prefreeze va 20F A A - ; g - n A O 1 2 3 4 5 6 7 8 days following deglycerolization FIGURE 1 Red Cell 2,3—DPG measured in 4 units of blood prior to glycerolization, following deglycerolization and at daily inter— vals thereafter showed no significant deviation from the prefreeze value through the third day post—thaw. 106 Modest changes in intracellular sodium and potassium are seen during post-thaw storage. Table 1 shows data for seven units processed according to the standard Red Cross procedure with deglycerolization in the Haemonetics Cell Washer. There is a mean rise of 50% in cell sodium and a 13% fall in potassium on the third day post—thaw. Valeri (6) has reported an average 12% fall in intracellular potassium. Valeri reports an approximately five—fold increase in extracellular hemoglobin after 4 days of post-thaw storage. Our experience has been that, 72 hours after deglycerolization, extracellular hemoglobin rarely exceeds 500 mg%. TABLE 1 Electrolyte Changes and Hemolysis During Post—Thaw Storagea Intracellular Electrolytes Supernatant K+ (me :L} Na+ {me fL} Hemoglobin (mg%) Time Mean Range Mean Range Mean Range Pre—freeze 93 (89—96) 14 (13—18) Post—wash 87 (82-90) 17 (14—20) 66 (41—88) Day 1 85 (79-90) 18 (15-22) 136 (89—245) Day 2 83 (78—87) 21 (18—23) 178 (103—347) Day 3 81 (78-84) 22 (19—24) 214 (118—441) aSeven units of red cells, all 3 days or less in age, derived from full units of blood drawn into CPD anticoagulant, were frozen by the standard Red Cross pro— cedure and deglycerolized in the Haemonetics Cell Washer. Units were assayed for intracellular electrolytes and supernatant hemoglobin before freezing, after deglycerolization, and at daily intervals during storage at 4°C in 0.8% NaCl, 0.2% glucose, phosphate buffered to pH 7.0. Final cell suspensions had an hematm ocrit ranging from 40 to 50% and a total volume of 375 ml. Bacterial Contamination The frequency of contamination and the contaminating organw isms reported by several investigators are listed in Table 2. Forty-one units (I) were deglycerolized in the reusable stain- less bowl with one contamination observed but not confirmed by duplicate sample. Seventeen of the units assayed (IIb) were the fourth of four units deglycerolized through the same bowl and washing harness. No contaminations were observed. Two hundred and thirty—one units (IIIa) were thawed in an unsterile water bath. Twelve of the sixteen contaminations observed were Mima polymorpha var. oxidans. This organism was also cultured from the water bath, identifying the bath as a major contamination hazard. Following this observation, a sterilizing agent, Bathkleer [Instrumentation Laboratory, Inc., Lexington, Massachusetts 02173] was added to the water bath. It is noteworthy that four of the contaminations seen in this group of 231 units were not confirmed on a second sample. Subsequent to the use of the sterilizing agent, 459 units (IIIb) were cultured by the same blood center with five contaminations detected. 107 TABLE 2 Bacterial Contamination of Deglycerolized Blood Duration No. Con- of Culture Source No. taminated, Confir— Before of a Units lst mation Positive Data Tested Culture Contaminating Organism Culture (days) I 41 l Staphylococcus epidermidis None 11a 89 2 S. epidermidis Negative 6 Acinetobacter calcoacetima Negative 6 IIb l7 0 111a 231 16 Mima polymorpha var. oxidans None 3 Mima polymorpha var. ox1dans None 2 Mima polymorpha var. ox1dans Positive 1 Mima polymorpha var. ox1dans Negative 2 Mima polymorpha var. oxidans Negative 2 Mima polymorpha var. oxidans Positive 1 Mima polymorpha var. ox1dans Negative 1 Mima polymorpha var. ox1dans Positive 1 Mima polymorpha var. ox1dans Positive 1 Mima polymorpha var. ox1dans Positive 1 Mima polymorpha var. ox1dans Positive 2 Mima polymorpha var. ox1dans None 2 Staphylococcus aureus Coag. Neg. Negative 2 g. aureus Coag. Neg. None 2 Not identified ' None 6 Not identified None 9 IIIb 459 5 Enterobacter cloacae Positive 4 S. aureus Coag. Pos. None 2 S. aureus Coag. Neg. None 2 Proteus mirabilis None 2 Diphtheroids None 2 IV 765 8 S. epidermidis (7) None Streptococcus Alpha Hemolytic None V 670 23 S. epidermidis All Not _ or known Propionibacterium Negative aThis table presents sterility data collected from five sources. (I) Blood pro- cessed and assayed for contamination by thioglycolate culture at the Red Cross Blood Research Laboratory, Bethesda, Maryland. (II) Blood processed for clinical use at the Red Cross Washington Regional Blood Center, Washington, D.C., and as- sayed for sterility by thioglycolate cuture at the Red Cross Blood Research Laboratory. (III) Blood processed for clinical use at the Red Cross Massachu- setts Blood Program, Boston, Massachusetts, data courtesy Dr. Irma Szymanski, screened by Bactec [JohnstonLaboratories, Cockeysville, Maryland]. (IV) Blood processed for clinical use at the Cook County Hospital Blood Bank, Chicago, Illinois, data courtesy Dr. M. Telischi, assayed by thioglycolate culture with parallel brain-heart infusion culture. (V) Blood processed for clinical use at the E. W. Sparrow Hospital, Lansing, Michigan, data courtesy Dr. w. E. Maldonado, assayed from segments by Bactec. Seven hundred and sixty—five units'(IV) were deglycerolized at random in the Haemonetics, IBM [IBM, P. O. Box 10, Princeton, New Jersey 08540], and Elutramatic [Fenwal Laboratories, Morton Grove, Illinois 60053] cell washers, with eight contaminations. Six hundred and seventy units (V) were deglycerolized with the 108 Haemonetics cell washer with 23 contaminations detected of which all were negative on second culture. In summary, 2,273 units were cultured by these several groups. Fifty-five (2.4%) were positive on first culture. When those negative on second culture are subtracted, the number of positives falls to 28 (1.2%), im— plying that contamination during sterility testing is a major problem. When the series of units thawed in the unsterilized water bath is eliminated in addition to units negative on second culture, the series consists of 2,042 units of which 16 (0.78%) were positive. Because of frequent entries into the system during the introduction and removal of glycerol, contamination must be ac— cepted as an inevitable risk, and the foregoing data confirms that contamination does occur with appreciable frequency. The 24-hour outdating period has been imposed to reduce the possi- bility that contaminating organisms might grow sufficiently to endanger the transfused patient. Since the risk of contamination has already been accepted provided proliferation is controlled, it is therefore reasonable to ask, what are the additional risks of bacterial proliferation in 72 hours as compared to 24 hours? Studies, reported in more detail elsewhere (2), have been conducted in our laboratory on ten organisms previously reported by others as contaminants of blood. Tables 3 and 4 are drawn from that study. It can be seen that at least a three-fold re- duction in titer results from glycerolizing and deglycerolizing (Table 3) and that in no case was there any significant growth of organisms during the subsequent three days of storage at 4°C. Even when deglycerOlized units were inoculated with high numbers of bacteria, no growth was observed during three days of storage at 4°C (Table 4). Discussion Evidence from our own studies and those of others (6) indi- cates that in vivo survival is essentially the same after 72 hours of post-thaw storage at 4°C as after 24 hours of storage. There will be some elevation in free hemoglobin with storage, occasionally rising as high as 0.5 gm%. It is debatable whether this presents any hazard to the patient. In any event, the ma— jority of the free hemoglobin can be removed by sedimenting the cells and decanting the supernatant solution prior to adminis- ' tration. Citrate in the resuspension medium has been reported (1) to prevent this post—wash hemolysis. Changes in 2,3-DPG and cell electrolytes are not sufficient to influence the effectiveness of the cells following transfusion. There is no evidence from our data or that of Valeri (6) that there are any contraindications to the use of cells stored 3 days at 4°C insofar as their clini- cal effectiveness is concerned. The present 24-hour outdating limit has not been impOsed on the basis of any deterioration of the cells but because of the. 109 TABLE 3 Number of Colony-Forming Units of Bacteria per Milliliter of Sample Before Freezing, Post-Thaw, Post-Washa and Following Storage at 4°C Post—wash and Following Storage at 4°C Pre— Post-thaw, Organism freeze Pre-wash 0 hr 24 hr 48 hr 72 hr E. coli 213 103 1 2 1 l - 180 95 1 2 2 4 167 67 8 24 33 30 221 76 8 10 ll 13 104 47 2 1 5 4 113 49 l 1 l 2 Average 166 73 4 7 9 9 §. egidermidis 16 1 0 0 0 0 84 11 1 0 0 0 80 3 0 0 0 0 74 20 1 0 0 0 91 27 2 1 2 2 47 17 0 2 l 0 57 25 0 l 1 l 46 10 0 O 0 0 20 1 0 1 0 0 17 1 0 0 0 0 Average 53 12 l 1 l 1 E. cloacae 303 90 1 l 1 1 _ 252 89 1 l l 1 75 37 1 1 0 0 89 41 0 1 0 1 52 18 0 0 0 0 95 34 1 0 l 0 155 74 1 2 1 1 89 34 1 0 0 1 77 39 0 1 1 0 91 4O 1 0 0 1 84 37 1 1 2 l 136 63 0 1 1 1 128 65 1 0 0 0 87 26 1 0 l 1 Average 122 49 1 l l 1 P. aeruginosa 58 6 1 0 0 0 _ 94 6 0 O 0 0 72 7 l 0 0 0 144 21 0 0 l 0 139 18 0 O 0 0 93 4 0 0 0 0 105 8 0 0 0 0 Average 101 10 1 0 l 0 aThe units were deglycerolized using the Haemonetics cell processor with the disposable bowl. 110 TABLE 4 Inoculation of Bacteria into Unfrozen or Frozen—Washed Red Cells No. of bacteria No. of bacteria/ No. of introduced/ ml of sample Organism Experiments Sample m1 of sample after 72 hr at 4°C 5. aerogenes 4 Frozen 101—10“ 101—103 Unfrozen 10" 103 Control 10“ 103 g. subtilis 3 Frozen 102—10” 101—102 Unfrozen 101—102 10 -101 Control 102-103 101-103 g. lutea 3 Frozen 10 -10“ 0 —103 Unfrozen 103—10“ 10 Control 10" Unchanged g. aeruginosa 3 Frozen 101—105 Unchanged Unfrozen 105 10" Control lO“-lO5 Unchanged g. marcescens 2 Frozen 101—105 Unchanged Unfrozen 105 101 Control 105 Unchanged g. aerogenoides 2 Frozen 10“ Unchanged Unfrozen 10” Unchanged Control 10” Unchanged E. cloacae 2 Frozen 101—10” Unchanged _ Unfrozen 10“ Unchanged Control 10“ Unchanged S. faecalis 2 Frozen 102-10” Unchanged _ Unfrozen lO3 Unchanged Control 103 Unchanged E. coli 2 Frozen 101—105 Unchanged ’ Unfrozen 105 Unchanged Control 105 Unchanged S. epidermidis 2 Frozen 101-105 Unchanged - Unfrozen 105 Unchanged Control 105 Unchanged hazard of bacterial contamination resulting from the many entries into the system during glycerolization and deglycerolization. The 24—hour limit is implicit recognition that accidental con- tamination occurs and constitutes an acceptable risk to the pa— tient provided there is a limited opportunity for proliferation of the organisms. Of the estimated200,000 units of deglycerol- ized cells transfused to date, there have been no reports of toxic reactions. In view of the apparent frequency of contamin- ation, this clearly implies that contaminating organisms do not proliferate to toxic levels within the 24-hour post-thaw storage period. Since in our studies there is also no proliferation at 72 hours, we can conclude that, for these organisms at least, 72 hours of post-thaw storage presents no additional risk to the patient. lll Presuming that the low 4°C post-thaw storage temperature is primarily responsible for inhibiting growth of the organisms tested, the risk of extending post-thaw storage to 72 hours therefore depends on the frequency with which organisms adapted to low temperature are encountered and the extent to which they will proliferate in the minimal medium provided by the glucose- saline-phosphate solution in which deglycerolized red cells are suspended. An answer to this question is difficult to extract from the literature. Bacteria that grow well at low temperatures have been described but are reported to be rare (5). Since the organisms reported as contaminants of frozen cells appear to be predominantly air and skin borne, the frequency with which low temperature strains might appear as contaminants of deglycerol— ized red cells should be extremely low. Although in the future the developmentof sterileconnectors may further reduce the risk of contamination, the inability of air and skin borne organisms to grow at 4°C within 72 hours suggests that the additional risk to the recipient from extending post-thaw storage to 72 hours will be very low and will be justified by the substantial savings in time, cost, and blood to be realized by an extension of the outdating period. Contribution No. 316 from the American National Red Cross. REFERENCE S 1. Akerblom, O. Citrate effect on postwash hemolysis in pre- vioUsly frozen blood, p. 40. Abstracts of XIV Cong. Blood Transf., Helsinki, 1975. 2. Kahn, R. A., H. T. Meryman, R. L. Syring, and L. J. Flinton. The fate of bacteria in frozen red cells. Transfusion. In press. 3. Meryman, H. T. Red cell freezing by the American National Red Cross. Amer. J. Med. Tech. 41:265—282, 1975. 4. Meryman, H. T., and M. Hornblower. A method for freezing and washing red blood cells using a high glycerol concentra— tion. Transfusion 12:145-156, 1972. 5. Myhre, B. A., Y. Nakasako, and R. Schott. The growth of bacteria in frozen-reconstituted blood and methods for de- tection of contamination, pp. 81-91. In P. Sherer, ed. Frozen Red Cell Outdating. Bethesda, Maryland, 1975. 6. Valeri, C. R. Simplification of the methods for adding and removing glycerol during freeze-preservation of human red blood cells with the high or low glycerol methods. Trans- fusion 15:195—218, 1975. 112 (LBERKELEY LIBRARIES CUEHLHHBHE U.S. DEPARTMENT OF HEALTH, EDUCATION AND WELFARE PubTic Health Service National Institutes of Health Bethesda, MaryTand 20014 DHEN Publication No. (NIH) 76-1004