Risk factors that predict herpes virus infections in patients with malignant lymphoma receiving conventional chemotherapy

Author(s)：

Hiroto Tanaka¹⁾, Masahiko Nezu²⁾, Reiko Matsui³⁾, Kunihiro Tsukasaki²⁾, Kuniaki Itoh⁴⁾, Shirou Katayama¹⁾

Affiliation(s)：

1)Department of Pharmacy, Nippon Medical School Hospital
2)Department of Hematology and Oncology, National Cancer Center Hospital East
3)Department of Pharmacy, National Cancer Center Hospital East
4)Department of Breast and Medical Oncology, National Cancer Center Hospital East

Background: In most guidelines, prophylaxis against herpes virus infections (HVIs) is recommended for patients with malignant lymphoma (ML) receiving immunosuppressing therapies after hematopoietic stem cell transplantation, purine analogues or bortezomib. However, HVIs also occur in patients with ML receiving conventional chemotherapy. Identifying those patients at risk of developing HVIs might help physicians to consider prophylactic treatment in order to decrease serious complications including unbearable pain and secondary infection. Methods: Data were obtained for 184 ML patients started conventional chemotherapy from November 2010 to May 2012, and analyzed for 164 patients without prophylactic administration of anti-herpes virus drugs. To analyze the risk factors for the development of HVIs, we investigated retrospectively the patient characteristics at the initiation of chemotherapy. Logistic regression model was performed for univariate and multivariate analysis. Results: Median age of the patients was 68 years (range, 21 to 84 years); the majority of patients were male [n=96, 59%]. Median count of neutrophil, platelet and lymphocyte was 4,120 (range, 570 to 439,300), 20.2 x 10^4 (range 2.4 to 79.9 x 10^4) and 1,210 (range, 120 to 100,600), respectively. Diseases of the analyzed patients included diffuse large B-cell lymphoma [n=81, 49%], follicular lymphoma [n=31, 20%], marginal zone lymphoma [n=6, 4%], mucosa associated lymphoid tissue lymphoma [n=3, 2%], adult T-cell lymphoma [n=3, 4%], Hodgkin lymphoma [n=13, 8%] and other lymphomas [n=27, 16%]. Most patients were newly-diagnosed [n=117, 71%]. Prior chemotherapeutic regimens were as follows; CHOP+/-R [n=107], ESHAP+/-R [n=11]; ABVD [n=10]; CHASER [n=5]; rituximab monotherapy [n=19]; bendamustine [n=4]; and other regimens [n=8]. Rituximab and steroids were used in 125 (76%) and 129 (79%) patients, respectively. The median follow-up time was 419 days (range, 22 to 865 days). Number of the total regimes was one in 102 patients and two or more in the remaining 62. HVIs occurred in 15 (9%) patients; herpes zoster [n=13], herpes simplex [n=2]. The median time of onset of HVIs was 129 days (range, 8 to 476 days) after the initiation of chemotherapy. Among the 15 patients, number of the total regimes until HVI-development was one in 9 patients two or more in 6 patients. In univariate analysis, age > 75 years (with vs. without HVI: 53% vs. 23%, P=0.025) and febrile neutropenia (FN) during chemotherapy (47% vs. 24%, P=0.045) were associated with the development of HVIs. Based on multivariate analysis, 2 variables were independent risk factors of HVIs: age > 75 years (P=0.019, OR=4.953, 95% CI: 1.30-18.90), the onset of FN (P=0.010, OR=6.301、95% CI: 1.57-25.37). In our analysis, female sex (53% vs. 40%, P=0.412), lower absolute lymphocyte count (median; 1,130 vs. 1,220, p=0.527), relapse (33% vs. 28%, P=0.766) and number of prior chemotherapy regimen > 2 (53% vs. 36%, P=0.264), which were previously reported as risk factors for HVIs, were not identified. Conclusion: ML patients develop HVIs infrequently after chemotherapy without prophylaxis, and age > 75 years and the onset of FN during chemotherapy might predict the occurrence of HVIs. A risk model for HVI development in ML patients using a larger prospective cohort of patients with validation is needed to guide the effective use of anti-herpes virus drugs for prophylaxis.