Background. T-cell lymphomas are often associated with abnormal T-cell responses. A viral etiology for certain subsets of cutaneous T-cell lymphoma (CTCL) and peripheral T-cell lymphoma (PTCL) has been proposed, but viral isolates from fresh malignant T-cells are rare. Recently, a significant association between cytomegalovirus (CMV) seropositivity and CTCL was observed (

Herne et.al.
Blood
101
(6):
2132
), suggesting a pathogenic role for CMV. However, no study to date has examined the incidence of CMV viremia in untreated CTCL and PTCL. Over the last 12 months we analyzed CMV viral loads in patients with CTCL or PTCL prior to initiation of systemic therapy to determine the frequency of CMV viremia and study the relationship between CMV viremia and cellular immune status, as defined by the absolute PB CD8+ T-cell count.

Methods. A total of 56 patients with either CTCL (n=41) or PTCL (n=15) were assayed for whole blood quantitative CMV viral loads using a hybrid capture assay for CMV DNA (Digene). PB immunophenotyping by flow cytometry was used to determine absolute CD8+ T-cell counts. In-situ PCR was used to detect CMV DNA in tumor samples.

Results. Only 2/56 patients had symptoms consistent with CMV reactivation at the time of the assay. Elevated CMV viral loads were seen in 15/56 (27%) of patients with a patient/cutoff ratio range of 1.0–107.6 and a median of 1.52 for CTCL and 2.14 for PTCL (>1.0 = positive). While all 15 patients with CMV viremia had positive serology for CMV, only 16/41 patients without elevated CMV viral loads were seropositive for CMV. Thus, although CMV viremia paralleled CMV seropositive status, the overall seroprevalence of CMV in this cohort 31/56 (55.3%) was not significantly different from that of the general population. The absolute CD8+ T-cell count was <200/mm3 in 8/15 (53.3%) patients and there was a significant relationship between lower CD8+ T-cell counts and higher CMV titers (p<.003). None of the 4 tumor samples tested from patients with elevated CMV viral loads (3 CTCL and 1 PTCL) was positive for CMV DNA.

Conclusions. We found that CMV seroprevalence in this cohort of PTCL and CTCL was not greater than that reported in the general population (55.3%). However, we report a previously undescribed high frequency (27%) of CMV viremia, before systemic therapy, and a significant correlation between higher CMV viral loads and degree of CD8+ T-cell lymphopenia. Whether this association reflects a yet to be defined pathogenic role of CMV in T-cell lymphoma or is simply the result of decreased T-cell immunosurveillance remains to be established. The absence of CMV DNA in a limited number of tumor samples from patients with elevated viral load does not support a direct pathogenic role of CMV, although a larger study is needed to confirm this hypothesis. The correlation between lower CD8+ T-cells and CMV viral load suggests defective T-cell immunosurveillance against CMV. A prospective quantitative monitoring study of CMV-specific T-cell responses, CMV DNA viral loads and expression in tumor cells from patients with T-cell lymphoma and appropriate controls (normal volunteers and patients with other hematological malignancies) is in development.

Author notes

Corresponding author

Sign in via your Institution