CASE REPORT

MICROSCOPY ‘AIDS’ IN DIAGNOSING A FEBRILE INFANT

Annelize Crous, MB ChB, DA (SA)

Department of Haematology, University of Pretoria, and National Health Laboratory Services, Tshwane Academic Division

Adele Visser, MB ChB, PG (Dip) TM, DTM&H, MMed (Clin Path), FCPath (SA) (Clin Path)

Department of Microbiology, Division of Clinical Pathology, University of Pretoria, and National Health Laboratory Services, Tshwane
Academic Division

A 5-month-old South African girl presented to a casualty department with a short history of fever. General examination did not reveal organomegaly or neck stiffness. In keeping with local guidelines, malaria was excluded on antigen testing and microscopy (thick and thin smear with Giemsa stain).1 Rickettsia typhi, R. conorii and Coxiella burnetti were also excluded on the basis of serological testing.

The full blood count revealed pancytopenia with a haemoglobin (Hb) level of 7.9 g/dl (normal 10 - 15 g/dl), a white cell count (WCC) of 1.17×109 /l (normal 5.50 - 18.00×109 /l) with an absolute lymphocyte count of 0.80×109 /l, and a platelet count of 47×109 /l (normal 140 - 350×109 /l). Biochemical values were suggestive of renal dysfunction, with a serum urea level of 9.6 mmol/l (normal 1.4 - 5.0 mmol/l) and a creatinine level of 180 μmol/l (normal 14 - 34 μmol/l), as well as a markedly elevated C-reactive protein level of 243.1 mg/l (normal 0.1 - 7.5 mg/l).

On investigation of the patient’s immune status, it was established that she was HIV-1-positive as confirmed by HIV-1 DNA Amplicor assay version 1.5 (Roche Diagnostics, Mannheim, Germany), with an HIV-1 viral load of 2 738 930 RNA copies/ml by Abbott m2000 Real Time HIV-1 Assay (Abbott Diagnostics, Johannesburg) and a CD4+ lymphocyte percentage of 28%.

Initial standard paediatric blood cultures did not yield any growth after 7 days of incubation. A follow-up full blood count, performed 2 days after the initial investigations and assayed on the ADVIA 2120 (Bayer HealthCare, Diagnostic Division, Isando, Gauteng), differed markedly from the previous results, showing a WCC of 13.83×109 /l with an absolute lymphocyte count of 10.99×109 /l. On microscopic examination of the peripheral blood smear (Fig. 1, a) these counts could not be verified and the leucocytes appeared markedly reduced. In addition, fungal elements could be demonstrated (Fig. 1, a and b).

Subsequent blood cultures, performed in paediatric bacterial blood culture bottles, confirmed the aetiological agent to be Candida albicans. Fungal morphology clearly showed the presence of oval yeast forms, slightly smaller than a lymphocyte nucleus (Fig. 1, a), with blastoconidia and pseudohyphae (Fig. 1, b). The fungus was identified by demonstrating formation of a germ tube on 4-hour incubation in horse serum.2 Subsequent antifungal susceptibility, performed on the Vitek 2 system (BioMerieux, Randburg, Gauteng), demonstrated sensitivity to all antifungal agents tested (amphotericin B, fluconazole and voriconazole).

Discussion

HIV-1 infection has been established as a risk factor for fungaemia in both children3 and adults.4 The routine use of fungal blood cultures remains controversial, being advocated by some authors6 and discouraged by others.7

Microscopic examination of the peripheral blood smear revealed a significant leucopenia, despite contradicting automated counts. WCCs as well as differential leucocyte counts have been described to be influenced by high fungal loads of Candida spp.8 The ADVIA 2120 uses both a lobularity and a peroxidise channel for leucocyte identification and quantification.8 In this case a comparison error had been flagged, indicating the probability of falsely elevated counts. This error, however, does not reflect on the laboratory report issued to the clinician, and is an analytical issue that needs to be addressed by the technologist and pathologist in the laboratory setting. Owing to their small size as well as their peroxidise-negative properties, the yeast cells had been identified as lymphocytes by the automated analyser (Fig. 1, c).

Considering the increased prevalence of fungal infections among HIV-1-positive patients,3 full blood count evaluation is best accompanied by morphological evaluation to detect any discrepancies between automated counts and actual observation.8 In most laboratory settings, morphology is evaluated once the clinician requests a differential count. Clinicians should further request differential counts once any full blood count parameters show significant abnormalities, or when there is a clinical suspicion of disease affecting the haematological system, including bleeding tendencies, infection, hepatomegaly, splenomegaly, lymphadenopathy, etc.10

In this clinical case, the rapid change in results over a 2-day period should have prompted the clinician to suspect an analytical issue with the samples in question.

REFERENCES

1. Zurovac D, Rowe A. Quality of treatment for febrile illness among children at outpatient facilities in sub-Saharan Africa. Ann Trop Med Parasitol 2006;100(4):283-296.

2. Winn W, Allen S, Janda W, et al. Mycology. 6th ed. Philedelphia: Lippincott Williams & Wilkins, 2006.

3. Walsh T, Gonzalez C, Roilides E, et al. Fungaemia in children infected with the human immunodeficiency virus: new epidemiologic patterns, emerging pathogens, and improved outcome with antifungal therapy. Clin Infect Dis 1995;4:900-906.

4. Launay O, Lortholary O, Bouges-Michel C, Jarrousse B, Bentata M, Guillevin L. Candidemia: a nosocomial complication in adult with late-stage AIDS. Clin Infect Dis 1998;5:1134-1141.

5. Hung C, Hsueh P, Hsieh S, Liu C, Chen M, Luh K. Bacteremia and fungemia in patients with advanced human immunodeficiency virus (HIV) infection in Taiwan. J Formos Med Assoc 1998;97(10):690-697.

6. Tattevin P, Chevrier S, Gangneux J. Can we describe the epidemiolgy of candidemia without using selective blood culture bottles for fungus detection? Clin Infect Dis 2004;39:598-599.

7. Mess T, Daar E. Utility of fungal blood cultures for patients with AIDS. Clin Infect Dis 1997;25:1350-1353.

8. Kim H, Park B, Lee M. Effects of bacteria and yeast on WBC counting in three automated hematology counters. Ann Hematol 2008;87:557-562.

9. Branda J, Kratz A. Effects of yeast on automated cell counting. Am J Clin Pathol 2006;126:248-254.

10. Bain B. Diagnosis from the blood smear. N Engl J Med 2005;353(5):498-507.