Discordant detection of human cytomegalovirus DNA from peripheral blood mononuclear cells, granulocytes and plasma: correlation to viremia and HCMV infection.

BACKGROUND: There exist only few data about the HCMV infection of single positive leukocyte subtypes in immunosuppressed patients. Most reports describe HCMV coinfection of cells of the myelomonocytic line or even T- and B-cell populations. Correlation of positive PCR findings from two major leukocyte fractions and plasma to viremia and HCMV infection in general should contribute to select suitable sources of HCMV DNA for diagnostic purposes.
OBJECTIVE: The diagnostic value of qualitative leukoDNAemia of simultaneously isolated peripheral blood mononuclear cells (PBMC), granulocytes as well as plasmaDNAemia was evaluated by comparing the positive results of nested PCR from blood with virus isolation either from leukocytes or from any other sources, with serology and the clinical status of immunosuppressed patients. STUDY
DESIGN: PBMC, granulocytes and plasma were prepared of a total of 220 blood samples of 75 immunosuppressed patients with clinically suspected primary or recurrent HCMV infection. In a collective of 35 patients consisting mainly of recipients of marrow or solid organ transplants positive results of leuko- or plasmaDNAemia were correlated with data from HCMV screening and the clinical status. For standardization, HCMV IE Exon 4 DNA was amplified from 100 ng cellular DNA of each leukocyte population. Cross contamination can be excluded. DNA from plasma was extracted by phenol/chloroform. Using this experimental design, HCMV DNA was not detectable in PBMC, granulocytes and plasma of 23 healthy HCMV seropositive blood donors.
RESULTS: Leukocyte separation in a collective of 30 patients with positive leukoDNAemia revealed in only 12 cases (40%) double infection of PBMC and granulocytes. In the majority of cases (18 patients, 60%) however, HCMV DNA was detectable in only one leukocyte fraction, either in PBMC or granulocytes. LeukoDNAemia did not correlate to viremia. HCMV DNA amplified from plasma was shown to be cell free. Infectious virus from plasma was not isolated. The predictive value of qualitative nested PCR from blood to detect HCMV infection was high for plasma and decreased in the following sequence: plasma (0.92) > PBMC (0.83) > granulocytes (0.65).
CONCLUSIONS: Qualitative nPCR from plasma and PBMC seems to be sufficient to detect (an ongoing) HCMV infection of immunosuppressed patients. However, the rate of single positive leukocyte fractions is approximately 60%. Therefore, viral leukoDNAemia in 40% of cases seems to be restricted to either PBMC or granulocytes. For diagnostic purposes the whole leukocyte population should be used for PCR analysis.