key: cord-0006968-1db721ph authors: Klech, Heinrich; Köhn, Horst; Huppmann, Martin; Pohl, Wolfgang title: Thoracic imaging with (67)gallium date: 1987 journal: Eur J Nucl Med DOI: 10.1007/bf00253288 sha: 185edf04582cf6052f625b7875b6ae2572cfe8c3 doc_id: 6968 cord_uid: 1db721ph In the past, gallium-67 imaging has undergone several ups and downs related to its clinical importance. After a period of initial enthusiasm, its role and indications are now established. At present, there are two fields of clinical interest for (67)Ga-imaging: (1) mediastinal staging in bronchogenic carcinoma and the staging of malignant lymphoma; (2) assessment of activity in interstitial lung diseases, especially sarcoidosis and inflammatory lung disorders. The advantage of (67)Ga-imaging is that it is highly sensitive for the detection of neoplastic and inflammatory processes, independent of anatomical barriers. Particularly with the challenge of AIDS, (67)Ga-imaging will gain increasing importance in the future. The low specificity of gallium for detecting underlying disorders precludes its use as a primary diagnostic tool. Therefore, and because of the cost and radiation load, the indications for application will have to be selected very carefully. Since its introduction by Edwards and Hayes in 1969 , the clinical value of gallium-67 scanning has remained controversial, despite the fact that numerous reports and reviews (Siemsen 1978; Bekerman 1980; Neumann 1984; Niden 1984; Yeh 1984; Waxman 1986 ) are available. However, 67Ga scanning of the thorax is used worldwide as well as in our institution as an adjunct to radiography for a number of specific indications: 1. To assess hilar and mediastinal involvement in pulmonary malignancies 2. To assess extent, location and inflammatory activity of diffuse lung diseases such as sarcoidosis and other granulomatous lung disorders, extrinsic allergic alveolitis, interstitial lung fibrosis of various etiology, pneumoconioses, etc. 3. To follow up progression or response to treatment in these inflammatory disorders 4. To detect disseminated interstitial disease that escapes visualization by chest radiography in selected cases. Controversy focuses mostly on the issues of reliability of 67Ga scanning in assessment of malignant disorders (McKenna 1985) and of interpretation of the result clini-Offprint requests to: H. Klech, M.D., II. Medical Department, Wilhelminenspital, Montleartstrasse 37, A-1171 Vienna, Austria cally in inflammatory lung disorders. Furthermore, the issues of restricted availability, costs and radiation load have been heavily disputed in the past (Whitcomb 1984) . This article describes the authors' experience with thoracic 67Ga scanning, summarizes the present literature, and critically reviews the current clinical use of 67Ga scanning as well as its limitations. The uptake mechanism is probably the most fascinating aspect of 67Ga scanning. Since here we have an "imaging of cells," in contrast to other means of imaging, the uptake of 6VGa is not restricted by anatomical barriers. However, the exact uptake mechanism in tumors or inflammatory lesions is unknown. 67Ga is bound to transferrin in the blood and accumulates in the lysosomal fraction of leukocytes, macrophages, reticuloendothelial cells and tumor cells (Hoffer 1980) . Interference with transferrin receptors on the tumor cells is suggested as the primary reason for 67Ga accumulation (Hoffer 1978; Larson 1978; Vallabhajosula 1983) . Increased capillary leakage on inflammation sites and a direct bacterial uptake may be other contributing factors. Table 1 lists the patients with various malignant thoracic disorders that have been investigated at our institution since 1979. Lung cancer 146 Mesothelioma 9 Pulmonary metastases 17 Hodgkin's lymphoma 23 Non-Hodgkin's lymphoma 33 Total 228 Gallium imaging was done 72 h after intravenous administration of 3 mCi 67Ga citrate. Large bowel clearing by use of laxatives preceded the administration. A whole-body hy- brid scanner (Scannicamera, CGR, spectrometer setting 150-400 keV) or a multipeak large-field gamma camera using the three major photon peaks of gallium and equipped with a high-energy collimator was used. For evaluation of mediastinal involvement, a 67Ga scan was considered as positive if the hilar and/or mediastinal region showed an abnormal uptake. In a prospective clinical study of 55 patients with lung cancer, sensitivity and specificity of 67Ga scanning in detecting mediastinal lymph node metastases were evaluated and compared to conventional chest radiography, including chest tomography as well as computed tomography (CT). Sixteen patients had squamous cell carcinoma, 12 had adenocarcinoma, 12had small cell carcinoma, and 15had other types of bronchogenic carcinoma. Chest radiographs, conventional chest tomographs, computed tomographs, and 67Ga scans were examined by two independent investigators. Interobserver variance was checked prior to evaluation and accordance was obtained in 95% of the selected cases. A third-generation tomograph [Somatom DR3, 8 mm sections, matrix 512 x 512, 100 ml contrast medium, 0.5 ml/ s, Angiografin (Schering)] was used for CT. Evaluation of mediastinal lymph nodes by CT considered size, extent, and location with regard to primary tumor location as previously described (Hajek 1985) . Mediastinal metastatic spread was confirmed in 21 patients and excluded in 34 patients by means of mediastinoscopy, transbronchial needle aspiration, thoracotomy, or necropsy (see Table 2 ). In our series of 146 patients with lung cancer, 119 cases revealed positive uptake on the tumor site (82%). Lesions smaller than 3 cm were more difficult to detect than those larger than 3 cm ( Table 3) . Uptake of 67Ga exceeded the size of the lesion in those cases where pneumonia due to bronchial obstruction was observed (n = 3). We did not find essential differences of 67Ga uptake among the various histological types of tumors. In patients with alveolar cell carcinoma, the 67Ga uptake sometimes did not match all the lesions demonstrated on chest radiography. The results of our series of 55 patients are shown in Tables 4 and 5. In patients with confirmed metastatic spread, conventional chest radiography had a true-positive rate (sensitivity) of only 66%. Conventional chest tomography increased the sensitivity to only 71%. In contrast to this, both CT and 67Ga scanning had a sensitivity of 90%. When used in combination (CT plus 67Ga), the sensitivity reached nearly 100%. An example is given in Fig. 1 . In patients where mediastinal involvement could be excluded, both chest radiography and conventional tomography of the mediastinum were negative in 32 of 34 cases (specificity=94%). In three cases (13%), CT yielded falsepositive results (specificity=87%). In these cases mediastinal CT revealed enlarged lymph nodes (> 1.5 cm) within the usual lymphatic drainage area of the tumor but malignancy could not be confirmed by mediastinoscopy or necroscopy. In most cases with centrally located primary tumors, CT was better than 67Ga scanning in discriminating between metastases and primary tumors. Thus, differentiation of centrally located primary tumors from mediastinal metastases was more difficult with 67Ga scanning. Falsepositive results with 6VGa, the lesions mimicking mediastinal tumor involvement, were observed in six cases (16%), yielding a specificity of 84%. In five of these cases no reason for the mediastinal 67Ga uptake was identified. In one case 2 . 67Ga camera scan of a patient with primary non-small cell carcinoma of the right medial lower lobe with hilar (NI) metastases. False-positive mediastinal 67Ga uptake was due to recent mediastinoscopy (2 days prior to injection of 67Ga). Mediastinal (N2) metastases were excluded by thoracotomy Fig. 1 . Computed tomography and whole-body 67Ga scan of a 46-year-old patient with primary non-small cell carcinoma of the middle lobe. Neither conventional chest X-ray nor radiographic chest tomography showed any signs of hilar or mediastinal metastases. 67Ga scanning documented pathological uptake in the right hilum and in the mediastinum. Hilar (N1) and mediastinal (N2) metastases were confirmed by thoracotomy mediastinal 67Ga uptake due to nonspecific inflammatory reactions was observed after mediastinoscopy, performed 2 days prior to 67Ga injection (see Fig. 2 ). Gallium-67 uptake in the primary lesion is reported to occur in 64%-100% of patients (Edwards 1969; Vaidya 1970; Grebe 1971; Ito 1971; Higashi 1972; Van der Shoot 1972; Deland 1974; Hjelms 1975; Waxman 1986 ). Earlier reports suggest differences among various tumor types regarding 67Ga uptake (Higashi 1972; Van der Shoot 1972; Deland 1974) . Recent reports demonstrate that most histologic types of pulmonary carcinoma are associated with positive uptake of 67Ga into primary lesions, provided the lesion is larger than 2-3 cm in diameter (Beckerman 1980; Waxman 1984 Waxman , 1986 . Our series demonstrate that detectability depends on tumor size rather than on histology. Almost all lesions larger than 3 cm in diameter were gallium-avid. Lesions that are larger than 3 cm on radiography and show no 67Ga uptake have been reported to be suspicious of metastatic origin (De Meester 1978) . Many reports suggest that radiation therapy may drastically decrease gallium uptake in lung cancer (Edwards 1970; Vaidya 1970; Higashi 1972; Kinoshita 1974; Hjelms 1975; Bitran 1978) . In conclusion, there is little clinical value in investigating 6VGa uptake in primary lung cancer. Uptake depends on size. Lesions smaller than 3 cm are hardly detectable. Previous radiotherapy alters the 6VGa uptake of the tumor. Therefore, a therapeutic effect cannot reliably be monitored by 67Ga scanning. Evaluation of mediastinum and hilum is essential for prediction of operability and prognosis in patients with lung cancer. Detection of involved mediastinal lymph nodes precludes curative resection. Mediastinoscopy, a conventional invasive technique, fails to document malignant mediastinal lymph nodes in 10%-30% of cases, despite its high specificity (Goldberg 1974; McKenna 1985) . In the past decade, most emphasis was therefore put on using gallium scans for detecting hilar and mediastinal metastases in patients with lung cancer (Deland 1976; Alazraki 1978; De Meester 1979; Fosburg 1979; Waxman 1982; Neumann 1981; Lunia 1981; Lesk 1982; McKenna 1985; Klech 1986a ). However, there is considerable variation in the figures reported so far. The study by Alazraki et al. Alazraki 1978; Lesk 1978; Curto 1980; Neumann 1981 ; Lunia 1981 ; Waxman 1984; McKenna 1985) (see Table 6 ). In contrast, De Meester et al. (1978) reported 56% sensitivity and 94% specificity. These results were subsequently confirmed by Richardson et al. (1980) , who demonstrated sensitivity of 56% and specificity of 97% in the detection of lymph node metastases in the mediastinum. These discrepant results reported by different investigators can be explained by the fact that basically different criteria were used to determine the sensitivity and specificity of the gallium scan. Alazraki et al. (1978) considered the scan positive if the hilar and/or mediastinal region showed increased gallium uptake. With this approach, the authors observed high sensitivity but relatively low specificity. We used the same criteria as Alazraki et al. in our study. The result is relatively high sensitivity of 90% and comparatively low specificity of 75%. De Meester et al. (1978) and Richardson et al. (1980) used a different set of criteria. Only an increase of 67Ga uptake in the mediastinal region was considered positive. Hilar increase alone was considered negative for mediastinal involvement. Thus, patients with mere ipsilateral hilar or perihilar increase were considered negative, as were patients in whom no activity was demonstrated within the hilus. With this approach it is evident that when the mediastinum is evaluated selectively, the sensitivity of detection is low, while the specificity is high. Waxman et al. (1984) nicely demonstrated the divergent results with regard to sensitivity and specificity by applying these two different approaches to the same set of patients. Therefore, the criteria used to define whether the gallium scan is positive or negative are the most important considerations when determining sensitivity and specificity of the test. It is probable that the results reported in the literature are in agreement; however, due to differences in the criteria for sensitivity and specificity they appear to be divergent. In 1985, McKenna et al. presented their study of 35 patients. The sensitivity was only 23% and the specificity was 82%. In contrast to previous studies, each pa-tient underwent thoracotomy with total mediastinal node sampling. By these means, malignant mediastinal lymph node involvement was observed in five of 19 patients with negative mediastinoscopy prior to thoracotomy. In contrast to previous reports, McKenna et al. (1985) concluded that because of the low sensitivity, 67Ga scanning cannot be recommended for preoperative staging. Although the authors' attempt to obtain the most exact diagnosis by complete mediastinal lymph node sampling is appreciated, the methods they applied, as well as their conclusion, may be subject to criticism: 1. Only some of the patients were scanned with a multipeak gamma camera, which is known to yield optimum resolution. 2. The authors evaluated only mediastinal 67Ga uptake, which is known to result in comparatively low sensitivity. Since the route of metastases from the primary tumor to the mediastinal nodes is almost always via the hilar nodes, most patients will demonstrate hilar abnormalities before or coincident with detection of mediastinal disease (Nohe 1956 ). If the authors had defined hilar 67Ga uptake as positive, the number of false negatives would perhaps have been smaller. 3. False-negative mediastinal 67Ga scans were observed predominantly in patients in whom the mediastinal lymph nodes were not enlarged or not larger than 1 cm. Today many surgeons consider these patients as operable, and improvement of survival has been reported when surgery is combined with adjuvant therapy (Daly 1986) . CT scanning reportedly yields highly satisfactory results in staging the mediastinum in lung cancer (Daly 1984; Breyer 1984; Glazer 1984; Heelan 1985; Hajek 1985) . Experience in our series of patients matches the published results. In centrally located tumors, CT scanning may discriminate better than gallium scanning between primary tumor and hilar or mediastinal lymph node metastasis. CT scanning provides valuable information when extent and size of lymph node metastasis is investigated with regard to localization of the lymphatic drainage of a tumor (Hajek 1985) . In our series combined use of 6VGa and CT yielded nearly 100% sensitivity. In conclusion, mediastinal 67Ga scanning should not be regarded as a routine investigation and should be used only in selected cases. Due to its high sensitivity, it is of clinical value in the management of lung cancer, provided the primary tumor shows uptake and both hilum and mediastinum are evaluated together. In case of negative mediastinal uptake together with negative radiography and negative CT, the patient can be referred to thoracic surgery without explorative mediastinoscopy. A positive hilar or mediastinal uptake suggests mediastinoscopy prior to thoracotomy. CT scanning and 67Ga scanning supplement each other very effectively: their combined use increases sensitivity to almost 100%. In the report of the cooperative study group (Johnston 1977) , 88% of untreated Hodgkin patients had positive 67Ga uptake at the site of the lesions. There is no marked difference in 67Ga uptake with regard to the histologic type of Hodgkin lesions. Lesions smaller than 1 cm are difficult to assess, and once a patient has been treated the sensitivity decreases. Given these limitations to prove the involvement, tous disease after therapy (Henkin 1974; Turner 1978; Yang 1979; Yehn 1979) . A positive gallium scan may be the first or only objective evidence of persistence or recurrence of treated diseases. Based on our experience and that of other authors, 67Ga scanning can be of great value to detect recurrent disease in the mediastinum and lung, since postirradiation fibrosis can diminish the reliability of other imaging techniques as compared to gallium scintigraphy (see Figs. 3, 4 a, b) . The 67Ga scan is less sensitive than chest radiography in detecting lung metastases of tumors that have a tendency toward pulmonary dissemination, such as breast cancer (Richman 1975 ) and soft tissue sarcoma. A sensitivity of 84% is reported for metastatic melanoma. Kirkwood et al. (1982) suggested 67Ga scanning as a routine procedure in diagnostic staging and follow-up of patients receiving therapy for metastatic melanoma. Fig. 3 . 67Ga scan of a 28-year-old patient with immunoblastic non-Hodgkin's lymphoma. Pathological 67Ga uptake is seen in the mediastinum (matching a huge lymph tumor) and in the spleen the 67Ga scan is at least as sensitive and probably more sensitive than radiography in detecting mediastinal disease (Turner 1978) . The sensitivity reported for Non-Hodgkin lymphoma is apparently lower, ranging from 50% in lymphocytic type Non-Hodgkin lymphoma to nearly 100% in Burkitt lymphoma (Turner 1978) . Gallium-67 scanning as a simple noninvasive procedure is helpful in follow-up studies of the course of a lymphoma-67Gallium scanning in "inflammatory disorders" of the lung Since 1979 we have studied more than 500 patients with various inflammatory thoracic disorders by means of 67Gaimaging. In our series, special emphasis was placed on patients with interstitial lung diseases in order to examine the clinical usefulness of 67Ga in these disorders (Table 7) . Sarcoidosis is defined as a multisystem granuloma-forming disorder that nearly always involves the lung. Various authors have proposed 6VGa-imaging as an aid in the diagnosis, evaluation of disease activity, and the assessment of therapy (Heshiki 1974; Niden 1976; Siemsen Fig. 4a, b . Chest radiograph (a) and 6VGa scan (b) of a primary lymphocytic plasmacytoid Non-Hodgkin's lymphoma of the right lung which obstructs the fight middle lobe and lower lobe bronchus, resulting in partial atelectasis of the lung $29 Nosal 1979; Line 1981 , Schoenberger 1982 Beaumont 1982; Gupta 1982; K6hn , 1983a Keogh 1983; Baughman 1984; Klech 1984 Klech , 1985b Hollinger 1985; Mishkin 1985; Niden 1986; Israel 1986; Rizzato 1986 ). 6 7 Ga-Imaging as a diagnostic adjunct 67Ga-Uptake depends on the extent and severity of the granulomatous inflammation in involved organs. Patients with active sarcoidosis usually accumulate 67Ga in the hili and/or lung parenchyma. This may be accompanied by in-creased 67Ga-uptake in the salivary glands in up to 75% of patients . Although this finding is nonspecific, the combination of pulmonary and salivary gland uptake strongly suggests sarcoidosis (Klech 1986a) . Interestingly, most cutaneous sarcoid lesions will not take up 67Ga, obviously due to insufficient granuloma mass (Lopez-Majano 1982) . Only with extensive cutaneous involvement can significant 67Ga-uptake be observed (Fig. 5a, b) . Some authors consider 67Ga a useful guide to optimal transbronchial tissue biopsy (Niden 1976) , whereas others are not convinced of this strategy (Ackart 1982) . In summary, due to its low specificity, gallium imaging should not be used as a single diagnostic test for sarcoidosis. Diagnosis must be based primarily on histologic, laboratory, and clinical findings. In established sarcoidosis, 67Ga-imaging is today widely used as a supplement to evaluate the extent and activity of the granulomatous process. The clinical activity of pulmonary sarcoidosis was assessed by different means in 60 patients with sarcoidosis. In 94% of cases, 67Ga demonstrated pathological uptake, whereas conventional chest radiography indicated active disease only in 80% . In a small series of seven patients who deteriorated from type I (bihilar sarcoidosis) to type II disease, as verified by transbronchial biopsy, 67Ga was positive in all seven patients while chest radiography failed in four (Klech 1983 a) . Experienced clinicains and radiologists are well aware that chest radiography is problematic for the assessment of interstitial lung involvement (Epler 1978; DeRemee 1983) . It is not possible to distinguish between active inflammation and irreversible fibrosis by means of chest radiography. To Fig. 5a, b. A 55-year-old female patient with chronic sarcoidosis, type-III, and a large cutaneous lesion on the right side oJ the face. 6VGa-Scan reveals abnormal uptake in the lung, spleen, and at the site of the cutaneous lesion $30 the clinician, the detection of pulmonary inflammation by 67Ga can be of great importance. Some authors emphasize the value of computer-based quantitation to improve the evaluation of diffuse pulmonary uptake, as well as to make therapeutic decisions (Line 1981; Rohatgi 1983; Unnik 1983; Bisson 1983; Wesselius 1983; Duffy 1986 ). This approach might be valuable in patients with diffuse homogeneous pulmonary granulomas, but the majority of pulmonary sarcoid lesions are not distributed homogeneously. We feel that if a semiquantitative score takes into account the regional distribution and intensity of gallium, sufficient information is provided for clinical routine care. In patients with biopsy-proven sarcoidosis, we found a close correlation between a positive 67Ga scan and the disease activity (Table 8 ). The 67Ga-scan was negative in 94% of patients with inactive or regressive disease. In recent years, other investigational means, such as serum angiotensin-converting enzyme (S-ACE) and bronchoalveolar lavage (BAL), have been widely used to assess disease activity. S-ACE has been found to be elevated in about 60% of patients with sarcoidosis (Studdy 1981). Comparing S-ACE to 6VGa in terms of sensitivity and specificity, we found 6VGa to be more sensitive than S-ACE Klech , 1983a Klech , 1984 . Both tests correlated significantly. Although other authors have reported similar findings, not all of them were able to demonstrate this correlation (Nosal 1979; Line 1981; Beaumont 1982; Gupta 1982; Schoenberger 1982; Rizzato 1986) . Other biochemical activity markers such as lysozyme, endopeptidase, procollagen-III-peptide, etc. are currently under investigation. We recently demonstrated a significant correlation between a 67Ga-score and serum levels of gamma-interferon and neopterin (K6hn 1986 ). Activated T-lymphocytes, predominantly found in the alveoli of sarcoid patients, secrete gamma-interferon which, in turn, stimulates macrophages to release neopterin. In recent years, considerable attention has been focused on BAL to determine whether it is possible to evaluate alveolar cellular changes earlier and more accurately with Fig. 6 . Markers of disease activity corresponding to different features of granulornatous lung disorders this method than with chest radiography. Line et al. (1981) have suggested that an activity index combining BAL (% lymphocytes) and 67Ga-uptake be used to define lowand high-intensity alveolitis. This concept has been confirmed by other authors who also found a correlation between the percentage of BAL lymphocytes and 67Ga-uptake (Huchon 1981; Fayman 1984) . However neither we nor other authors have been able to confirm this finding (Beaumont 1982; Costabel 1983; Havranek 1983; Klech 1984; Baughman 1984) . The majority of centers contributing to a recent international survey on BAL in patients with sarcoidosis have reported a relationship between 67Ga and the percentage of BAL lymphocytes rather than S-ACE (Klech 1986b) , which reflects how difficult it is to compare data subject to variable modes of evaluation. The results of various activity markers probably correspond to different stages of the granulomatous reaction and are thus complementary rather than concordant. Each marker should be interpreted in light of its advantages and limitations regarding the assessment of pathophysiological processes in the course of the disease (Fig. 6 ). Although the efficacy of corticosteroid therapy in altering the ultimate course of sarcoidosis and preventing progressive pulmonary fibrosis remains controversial, it is generally accepted that corticosteroids improve symptoms and induce remissions. In the past, chest radiography and lung function tests were used as standard criteria for the long-term follow-up of patients on cortocosteroid therapy. However, for shorttime monitoring lung function tests in particular have proven to be too insensitive (Winterbauer 1980; Klech 1983b Klech , c, d, 1984 . In a prospective clinical trial, we investigated the impact of different activity markers such as S-ACE and 67Ga for monitoring therapeutic effects and assessing prognosis, in comparison with lung function tests and the information they may provide for the clinician. Twenty-seven patients (20-62 years of age, mean 38.8 years) were included who had histologically proven sarcoidosis. Of those, 16 were female and 11 male. Upon entering the study, the patients were evaluated by clinical findings, chest $31 radiography, lung function tests (total lung capacity, forced vital capacity), blood gas analysis at rest and during exercise S-ACE, and 67Ga-imaging. Each patient underwent serial follow-up assessments (mean 3, range 2-7) when significant disease activity changes were observed clinically; 69 follow-up assessments were classified as either deterioration, improvement, or as stable. All patients gave their informed consent. The clinical symptoms were recorded by the attending physician. At the same time, laboratory data were assessed. Chest radiographs were scored according to standard criteria: type 0 = normal; type I = bilateral hilar adenopathy; type II=bilateral hilar adenopathy with pulmonary involvement; type III=pulmonary infiltration without hilar adenopathy. Deterioration was defined by the following criteria: marked deterioration of the chest X-ray, with or without clinical symptoms such as dyspnea, cough, fever, chest pain, arthralgia, etc. These criteria were met in 15 follow-up assessments. Improvement was only defined when there was marked amelioration of the chest X-ray, together with improvement in or lack of clinical symptoms. Criteria for improvement were met in 29 follow-up assessments. In all other instances, the disease was regarded as stable. The majority of patients did not receive corticosteroid therapy during the study. If steroids had to be prescribed, treatment was stopped I week prior to follow-up assessment. Lung function was measured by means of a body-plethysmograph (Jaeger, FRG) in terms of vital capacity (VC) and total lung capacity (TLC). Blood gas analysis was measured by arterial puncture (radial artery) at rest and after exercise (75 W/5 min bicycle or treadmill exercise) and expressed as the alveolar-capillary oxygen difference (AaDO2) using the formula: -Fp _//PaCO2"~q aaDO2-L io2 ~ RQ jj-PaQ 6VGa-Scans were scored independently by two experienced observers and graded 0-3, comparing the lung activity with that in the liver and shoulders . S-ACE was measured photometrically according to the method of Cushman and Cheung as modified by Lieberman (1976) (normal range < 24 U/ml). For statistical analysis, Spearman's rank correlation coefficient was used as well as Wilcoxon's nonparametric test to compare follow-up assessments. AaDO2 at rest and exercise showed the most significant changes. AaDOz at rest increased by more than 10% in 87% and during exercise in 60%. S-ACE increased in 85% and showed no change or even decreased in 15%. The 67Ga-score increased in 66% and remained unchanged in 34%. In no case was 67Ga-uptake diminished compared with previous scans (Fig. 7) . We could demonstrate no significant correlation between changes of 67Ga and ACE or lung function and AaDO2. The 67Ga-score decreased significantly with clinical improvement (P< 0.001). Improvement in the 67Ga-scan was observed in 73%. In 27%, the gallium score remained unchanged. ACE decreased markedly in 64% of instances, but remained unchanged in 18% and increased in 16%. Changes in lung function were insignificant, VC improved in only 31%, and TLC only improved in 20% of cases. Changes in blood gases were also insignificant for the documentation of improvement (Fig. 8) . Loose, but significant correlations were found between the decrease of S-ACE and increase of VC (rs=0.59, P<0.01) or decrease of AaDO2 during exercise (rs = 0.54, P < 0.05). The results of this study confirm the limited clinical value of lung function studies (Winterbauer 1980) , particularly for follow-up assessments of patients with pulmonary sarcoidosis (Klech 1983b (Klech , c, d, 1984 . Obviously, in the majority of sarcoidosis patients deterioration of blood gas analysis is due to reversible air-flow limitation and an inequality in the distribution of ventilation caused by mucosal swelling and endobronchial granu- Restitution of Sorcoidosis (Spontaneously or by steroids) Fig. 9 . Decrease of activity markers according to improvement of sarcoidosis loma formation, rather than by irreversible fibrotic changes in the pulmonary interstitium. Therefore, blood gas analysis at rest and during exercise is a sensitive indicator of actual clinical deterioration, but has no prognostic significance for the long-term course of the disease. Lung function tests, particularly VC, reliably document irreversible impairments, but have no value for making short-term therapeutic decisions. Diffusing capacity (DLCO) has been proposed to monitor lung function in sarcoidosis. However, concomitant airflow limitation restricts reliable interpretation and reversible airflow limitation is a common feature in patients with pulmonary sarcoidosis. We used chest radiography as the standard criterion for assessment since it is widely used in clinical practice. Only those instances that clearly showed deterioration or improvement in the chest X-ray were selected for further assessments. At present, based on experience with new activity markers, the clinician's opinion on the value of chest radiography has been modified DeRemee 1983; Rizzato 1986 ). However, the clinical value of S-ACE, 67Gaimaging and BAL for follow-up assessments is controversial (Keogh 1983; Lawrence 1983; Baughman 1984; Hollinger 1985) . This may be due to the variable modes of evaluation or to the fact that these markers measure different sarcoidosis features. Therefore, changes in disease activity are not reflected in the same manner by different markers. If there is improvement, the S-ACE values decrease first, followed by an improvement in 67Ga; however, lymphocyte counts in BAL remain elevated for a longer period of time ( Fig. 9 ). At our institution, the following strategy is used for clinical follow-up and therapeutic decisions (Fig. 10) . Chest radiography and S-ACE levels are still the standard criteria. 67Ga Imaging is initially commonly performed in patients with pulmonary involvement (type II or III), but later on is limited to selected cases. In the past there has been much concern about the costs, availability and radiation load of 67Ga (Whitcomb 1984) . Using 3 mCi 67Ga-citrate, the radiation load does not exceed that of a chest tomography or a CT study. However, in young patients there should be a strong indication. The successful use of low dosage gallium imaging (1.5 mCi) was reported by Rizzato et al. (1986) . In conclusion, 67Ga-imaging provides information about the spread and extent of the disease and permits the progression to be followed at each site of involvement. 67Ga is more sensitive than chest radiography and when used selectively, is of reasonable value for follow-up and making therapeutical decisions in pulmonary sarcoidosis within the framework of other activity markers. Idiopathic pulmonary fibrosis (IPF) Line et al. (1978) have used 67Ga, together with BAL, to stage IPF and have reported a close correlation between a 67Ga index and the inflammatory cell reaction in the lung, particularly the neutrophil counts in BAL. However, the clinical features correlated poorly with the 67Ga indices. The results were essentially corroborated by Niden et al. (1984) . In our group of patients (n = 32), only a minority had an abnormal 67Ga uptake that was unrelated to BAL cell counts. In IPF, 67Ga probably accumulates only when there is acute neutrophil alveolitis at a stage where therapeutic improvement can be expected. In our series of patients (n=24), 67Ga-uptake was only observed after recent exposure to the causative antigen (Fig. 11) . Acute exogenous allergic alveolitis is characterized by T-helper cell alveolitis, often associated with interstitial granuloma formation, which can be very similar to . 67Ga-Scan shows abnormal lung uptake. Imaging was performed immediately after extensive exposure to moldy hay dust that seen in sarcoidosis (Klech 1986a) . Absence of antigen exposure for 4-6 weeks will normalize the gallium scan. Siemsen etal. (1974) first described diffuse pulmonary 67Ga-uptake in 110 patients with pneumoconiosis and reported the 67Ga scan to be more sensitive than chest radiography to document lung involvement. In a small group of patients (n= 16), we found abnormal 67Ga uptake only in patients who had been exposed to the antigen the 3 months before imaging. Talc and beryllium pneumoconiosis shows a 67Ga uptake that is similar to that of sarcoidosis (Brown 1984; Niden 1984) . We studied 11 patients with collagen vascular disorders. Four of 6 patients with lung involvement due to rheumatoid arthritis had an abnormal 67Gauptake. Only 1 patient out of 3 with pulmonary lupus erythematodes (LE) showed lung uptake. However, pulmonary manifestations of LE are generally mild. 67Ga has been used to assess pulmonary damage caused by pharmaceuticals such as nitrofurantoin (Crook 1982) or bleomycin (Richman 1975 ) and other antitumor drugs (McMahon 1978) . Lung damage could be localized better by 67Ga than by chest radiography in most cases. Pulmonary lesions associated with adult respiratory distress syndrome (ARDS) accumulate gallium in the acute stage due to interstitial neutrophilic inflammation rather than to a nonspecific increase in the alveocapillary permeability of gallium (Passamonte 1984) . As soon the neutrophilic inflammation subsides -mostly with the development $33 of interstitial lung fibrosis -abnormal 67Ga-uptake is rarely seen. Therefore, four of six patients with ARDS who were scanned later than 8 weeks after the lung injury no longer had abnormal 67Ga uptake. Siemsen et al. (1978) and Bekerman et al. (1980) have reported 67Ga to be useful for the diagnosis of virtually all pulmonary inflammations, including pneumonia, lung abscess, tuberculosis, and Pneumocystis carinii infection. Thadepalli et al. (1978) have found a close correlation between bacteriologically confirmed pulmonary infections and abnormal pulmonary 67Ga-uptake. However, Walsh et al. (1985) have recently raised the question of whether 6VGaimaging can be recommended for the assessment of tuberculosis activity due to its low specificity (only 27%). They concluded that 67Ga cannot be used reliably to distinguish between active or inactive tuberculosis. Our results are in agreement with these findings. Serial chest radiographs, together with sputum smears or bronchoscopy, provide sufficient information. 67Ga may be useful in patients with lymph node tuberculosis. In patients with miliary tuberculosis, 67Ga was positive in only four of six cases. Nonspecific lung disorders such as pulmonary infarction, bronchitis, etc. inconsistently accumulate 67Ga, depending on the size and nature of the underlying inflammation. 67Ga-imaging has been reported to be helpful in the differential diagnosis of pneumonias or pulmonary infarctions (Niden 1977) . However, specific diagnosis seems not to be possible (Brown 1983) . Opportunistic lung infections caused by Pneumocystis carinii, cytomegalovirus, and mycobacterium avium-intracellulare are common in patients with AIDS. Since these infections are known to accumulate 67Ga, even without evidence of lesions in the chest X-ray (Levenson 1976; Turbiner 1978; Hamed 1979) , at present 67Ga imaging is of particular value in the clinical management of AIDS. We saw an increase in diffuse 67Ga-uptake in both lungs in two cases of opportunistic infection due to P. carinii and in one with cytomegalovirus infection in spite of negative chest X-ray findings. 6VGa-Imaging has been recommended as a sensitive initial diagnostic test in AIDS patients suspected of having an opportunistic infection. It helps to localize the involved lung areas as a guide for diagnostic bronchoalveolar lavage and transbronchial tissue biopsy (Levin 1983 ; Coleman 1984) . Efficacy of 67 Ga-scintigraphy in predicting the diagnostic yield of transbronchial lung biopsy in pulmonary sarcoidosis Reliability of gallium scan, chest radiography compared with mediastinoscopy for evaluating mediastinal spread in lung cancer Comparison of gallium-67 scanning, bronchoalveolar lavage and serum angiotensin converting enzyme levels in pulmonary sarcoidosis Gallium 67 in the evaluation of sarcoidosis : correlation with serum angiotensin converting enzyme and bronchoalveolar lavage Gallium-67 citrate imaging studies of the lung Computer based quantitative analysis of gallium-67 uptake in normal and diseased lungs Clinicopathologic correlation demonstrating the failure of 67 gallium scanning in determining response to radiotherapy Computed tomography for evaluation of mediastinal lymph nodes in lung cancer:correlation with surgical staging 67 Gallium scanning in talc-induced pulmonary granulomatosis Positive gallium 67 scintigraphy associated with pulmonary embolism Correlation between gallium lung scans and fiberoptic bronchoscopy in patients with suspected Pneumocystis carinii pneumonia and the acquired immune deficiency syndrome Die Bedeutung der Helfer-T-Lymphozyten in der bronchoal-veol~iren Lavage ffir die Aktivit/itsbeurteilung der pulmonalen Sarkoidose Gallium 67 scanning in nitrofurantoin-induced pulmonary reaction Value of mediastinic scintigraphy with Ga-67 in the study of lung cancer extension (abstract) Computed tomography. An effective technique for mediastinal staging in lung cancer 67 Ga citrate imaging in untreated primary lung cancer. Preliminary report of the cooperative group The role of Ga-67 scanning in the clinical staging and preoperative evaluation of patients with carcinoma of the lung The roentgenographic staging of sarcoidosis:historical and contemporary perspective Semi-quantitative gallium-67 lung scanning as a measure of the intensity of alveolitis in pulmonary sarcoidosis Tumor scanning with 67 gallium citrate Scanning malignant neoplasms with gallium-67 Normal chest roentgenograms in chronic diffuse infiltrative lung disease Quantitative 67 gallium citrate lung imaging in sarcoidosis : comparison of three methods and correlation with bronchoalveolar lavage Evaluation of the mediastinum by gallium-67 scintigraphy in lung cancer The mediastinum in non-small cell lung cancer: CT surgical correlation Mediastinoscopy for assessing mediastinal spread in clinical staging of lung cancer Gallium 67 citrate scanning and serum angiotensin converting enzyme levels in sarcoidosis Pulmonary cytomegaly virus infection. Detection by gallium 67 imaging in the transplantat patient Die bronchoalveol/ire Lavage (BAL) bei pulmonaleer Sarkoidose Carcinomatous involvement of the hilum and mediastinum:computed tomographic and magnetic resonance evaluation Scanning treated Hodgkin's disease with 67 Ga citrate Gallium-67 citrate scanning in patinets with pulmonary sarcoidosis Uptake of 67 Ga in malignant lesions of the lung and lymphatic tissue Mechanism oflocalisation : In: Hoffer PB, Bekerman C, Henkin RE (eds) Gallium-67 imaging Gallium: mechanisms Prediction of therapeutic response in steroidtreated pulmonary sarcoidosis Comparison of bronchoalveolar lavage and gallium-67 lung scanning to assess the activity of pulmonary sarcoidosis Assessment of gallium 67 scanning in pulmonary and extrapulmonary sarcoidosis Diagnostic evaluation of 67 Ga scanning in lung cancer and other diseases Gallium-67 citrate imaging in Hodgkin's disease: final report of the cooperative group The alveolitis of pulmonary sarcoidosis: evaluation of natural history and alveolitis dependent changes in lung function Scintiseanning of pulmonary diseases with 67 Ga citrate Tomographic gallium-67 citrate scanning: useful new surveillance for metastatic melanoma Sarcoidosis: differential diagnosis Assessment of activity in sarcoidosis Pulmonary function tests (PTF) compared to 67 gallium scinitgraphy and S-ACE for monitoring recurrent pulmonary sarcoidosis How to monitor recurrent pulmonary sarcoidosis. Value of serial lung function tests, S-ACE and 67 Ga scanning Verlauf der Lungenfunktion bei pulmonaler Sarkoidose Sensitivity and specificity of 67 Ga scanning, bronchoalveolar lavage (BAL) chest radiography and lung function tests (LFT) for assessment of activity in patients with pulmonary sarcoidosis Value of 67 gallium scanning, computed tomography and chest radiography for non invasive mediastinal staging of bronchogenic carcinoma World wide clinical survey on bronchoalveolar lavage (BAL) in sarcoidosis. Experience in 62 centers in 19 countries Serial S-ACE and 67 Ga scanning supplement best routine chest radiography for monitoring recurrent pulmonary sarcoidosis 67 Ga scanning for assessment of disease activity and therapy decisions in pulmonary sarcoidosis in comparison to chest radiography, serum ACE and blood T-lymphocytes Gamma interferon and neopterin for assessment of activity in sarcoidosis Mechanisms of localisation of gallium-67 in tumors Serial changes in markers of disease activity with corticosteroid treatment in sarcoidosis The application of 67 Ga scanning in determining the operability of bronchogenic carcinoma Abnormal pulmonary 67 gallium accumulation in P. carinii pneumonia Pneumocystis pneumonia: Importance of gallium scan for early diagnosis and description of a new immunoperoxydase technique to demonstrate Pneumocystis carinii Elevation of serum angiotensin converting enzyme (S-ACE) level in sarcoidosis Gallium 67 citrate scanning in the staging of idiopathic pulmonary fibrosis : correlation with physiologic and morphologic feature and bronchoalveolar lavage Gallium-67 scanning to stage the alveolitis of sarcoidosis: correlation with clinical studies, pulmonary function studies and bronchoalveolar lavage Gallium uptake in cutaneous sarcoidosis Noninvasive evaluation of mediastinal metastases in bronchogenic carcinoma Critical evaluation of the gallium 67 scan for surgical patients with lung cancer The diagnostic significance of gallium lung uptake in patients with normal chest radiographs Clinical application of 67 gallium citrate lung imaging in sarcoidosis Abnormal gallium scan patterns of the salivary gland in pulmonary sarcoidosis 67 Ga scintigraphy of the thorax Clinical value of gallium 67 in imaging patients with lung carcinoma and melanoma Diffuse infiltrative lung disease Gallium-67 citrate lung scans in interstitial lung disease gallium lung scan. An aid in the differential diagnosis of pulmonary embolism and pneumonitis Prognostic significance of gallium lung scans in sarcoidosis An investigation into the lymphatic and vascular spread of carcinoma of the bronchus Angiotensin-I-converting enzyme and gallium scan in non-invasive evaluation of sarcoidosis Pulmonary gallium concentration in the adult respiratory distress syndrome Preoperative noninvasive mediastinal staging in bronchogenic carcinoma Usefulness of gallium scintigraphy in primary and metastatic breast cancer b) Gallium accumulation in pulmonary lesion associated with bleomycin toxicity An European survey on the usefulness of 67 Ga lung scans in assessing sarcoidosis. Experience in 14 research centers in seven different countries Quantitative gallium scanning in pulmonary sarcoidosis Lung inflammation in sarcoidosis: comparison of serum angiotensin converting enzyme levels with bronchoalveolar lavage and gallium-67 scanning assessment of the T-lymphocyte alveolitis Pulmonary concentration of 67 Ga in penumoconiosis The use of gallium 67 in pulmonary disorders Correlation of microbiologic findings and gallium 67 scans in patients with pulmonary infections Abnormal gallium scintigraphy in Pneumocystis carinii pneumonia with a normai chest radiograph Hoffer PB, Bekerman C, Henkin RE (eds) Gallium-67 imaging A method of quantitative 67 Ga scintigraphy in the evaluation of pulmonary sarcoidosis Ohnuma T (1983) 67 Ga-transferrin and 67 Ga-lactoferrin binding to tumor cells. Specific versus non-specific glycoprotein-cell interaction Gallium 67 scintigraphy in lung diseases The value of gallium 67 scanning in pulmonary tuberculosis Gallium scintigraphy in bronchogenic carcinoma: the effect of tumor location on sensitivity and specificity The role of nuclear medicine in pulmonary neoplastic processes Computer-assisted versus visual lung gallium-67 index in normal subjects and in patients with interstitial lung disorders Gallium scanning, bronchoalveolar lavage and the national debt Clinical significance of pulmonary function tests on the management of sarcoidosis Serial Ga-67 citrate imaging in children with neoplastic disease: concise communication Yeh SD 0984) Tumor localization scans in primary lung tumors Gallium scan in recurrent Hodgkin's disease in children An interesting comment about the future of gallium scanning was made. As a result of the AIDS problem, Gascanning has become very popular in the US. It has proven very useful not only in the assessment of lung disease (pneu-mocystis carinii pneumonia the most common AIDS-related opportunistic infection), but also of lymphoma-type lesions or abdominal disease. A dramatic uptake of 67Ga in the abdomen caused by different types of infections within the gastrointestinal tract can be seen despite the fact that the colon normally excretes 67Ga. Therefore, it has been recommended to perform a whole-body scan as opposed to limited lung imaging in these patients.