LETTER

We recently reported on a new commercial cytomegalovirus (CMV) antigenemia assay, the CMV Brite Turbo kit (Biotest Diagnostics, Denville, N.J.), which can be completed in only 2 h (4) instead of the 4 to 5 h required in the standard assay (1, 3, 6). The reduction in assay time is accomplished by using direct erythrocyte (RBC) lysis to separate leukocytes from whole blood instead of dextran sedimentation and by using shorter incubation times.

We found that the 2-h assay provides quantitative results equivalent to or slightly better than those of the standard test (4). However, we also noted that direct RBC lysis appeared to reduce our ability to isolate CMV from blood leukocytes. Therefore, we conducted a small prospective study in which 115 blood specimens collected in EDTA and submitted to the clinical virology laboratory for CMV antigenemia and culture were divided into two aliquots. Leukocytes were isolated from one aliquot by dextran sedimentation and from the other aliquot by direct RBC lysis using the solution provided in the Biotest CMV Brite Turbo kit (0.8% ammonium chloride with <0.1% sodium azide). Of note, dextran sedimentation is routinely followed by an RBC lysis step but uses only 5 ml of lysis solution instead of the 30 ml used for direct lysis. Virus isolation was performed in MRC-5 monolayer cultures (3). Seventeen of 115 specimens (15%) were positive by antigenemia, with the number of CMV-positive cells ranging from 1 to 445 per 200,000 cells examined. Eight of the 17 antigenemia-positive samples (47%) were also culture positive after separation by dextran sedimentation, but only 6 (35%) were culture positive following direct RBC lysis. In addition, fewer cytopathic effect (CPE) foci were evident in monolayers inoculated with leukocytes recovered by RBC lysis than in those recovered by dextran sedimentation.

Nevertheless, we decided to utilize the CMV Brite Turbo kit beginning in October 1999. Since that time, our CMV isolation rate from blood has declined from 47 to 21%, and for the first time since we began doing antigenemia testing in 1992, we have had no samples positive by culture alone (Table 1). To our knowledge, this finding has not been previously reported. In the report on direct RBC lysis by Ho et al. (2), culture was not performed.

The ability to provide a quantitative CMV viral load test result within 2 h of sample receipt in the lab is of great benefit to patient care. In contrast, isolation of CMV from blood 7 to 21 days after culture inoculation is much less useful. Thus, the CMV antigenemia assay is the mainstay of CMV detection in blood in our laboratory, and culture is primarily a backup to antigenemia assay in case of technical error or failure of the monoclonal antibodies to detect a patient's CMV strain (5). Occasionally, culture is performed in order to recover an isolate for drug susceptibility testing.

The reason for the reduced ability to culture CMV following direct RBC lysis with the CMV Brite Turbo kit is unclear. However, as a result of our experience, we would recommend separation of leukocytes by dextran sedimentation when obtaining a CMV isolate by culture is a priority.