Blood Journal
Leading the way in experimental and clinical research in hematology

HIV-associated lymphoma

  1. Alexandra M. Levine

The role of infusional chemotherapy and the value of rituximab with chemotherapy have not been clearly defined in patients with HIV-associated lymphoma. Recent data have provided insights into these questions.

Although the optimal therapeutic strategies for patients with HIV-associated lymphoma have remained controversial, recent studies by Dunleavy et al and Sparano et al in this issue of Blood have brought clarity to several contentious issues.1,2 First, should rituximab be included within the regimen? Second, does the benefit of a prolonged infusion, such as the EPOCH regimen,3 outweigh the inconvenience and cost? Third, should antiretroviral therapy be suspended during chemotherapy? Fourth, are there subtypes of lymphomatous disease that should be treated differently than others? Fifth, is FDG-PET scanning a useful restaging and prognostic tool in HIV lymphoma?

The addition of rituximab to the CHOP regimen has resulted in remarkable prolongation in long-term disease-free survival among patients with HIV-negative diffuse large B-cell lymphoma (DLBCL),4 and early results among patients with HIV-associated lymphoma appeared also to show benefit.5 Unexpectedly, a randomized phase 3 study of CHOP with or without rituximab, performed by the AIDS Malignancy Consortium (AMC), found only a trend toward greater efficacy in patients on the rituximab arm, with a statistically significant increase in death due to infection.6 When analyzed more carefully, these infectious deaths occurred primarily among patients with severe immunodeficiency. If patients with CD4 lymphocyte counts less than 50/μL were excluded from the analysis, no significant difference in infectious death was seen. Severe HIV-related immunodeficiency in itself has been associated with an increased risk of septic death, even in the absence of chemotherapy.7 Nonetheless, rituximab has been associated with various viral infections, as well as progressive multifocal leukoencephalopathy and hepatitis B reactivation, and could theoretically be a concern.8 The results of the AMC study served to emphasize these concerns, and was in all likelihood responsible for a change in treatment paradigm away from the use of rituximab in HIV-infected patients. In this issue of Blood, a subsequent randomized phase 2 study from the AMC did not show an increased risk of infectious death among patients receiving concomitant chemotherapy and rituximab.2 The study by Dunleavy et al, also reported in this issue, has again confirmed the efficacy of rituximab, allowing an abbreviated course of infusional EPOCH while noting no treatment-related deaths or new opportunistic infections. Whereas it behooves the clinician to be aware of the potential for rituximab-related infections, especially in patients with severe immunocompromise, it is time to put this question to rest. Similar to the case in HIV-negative lymphoma, patients with HIV-associated DLBCL clearly benefit from the use of concomitant rituximab and chemotherapy.

Results of infusional short-course EPOCH with double-dose rituximab (SC-EPOCH-RR), even when given for only 3 or 4 cycles, were remarkably good in the Dunleavy study, with 85% progression-free and 68% overall survival at a median follow-up of 5 years. CHOP-based regimens have not been associated with these kinds of results, and the AMC studies of EPOCH2 and of CHOP6 would seem to confirm that infusional EPOCH is the superior regimen in the setting of HIV-associated DLBCL. Greater ease of administration would be nice, but a shortened total treatment time may partially compensate for the inconvenience of a prolonged infusion. In the final analysis, the patient deserves to receive the regimen associated with the best chance of long-term survival.

Tumor histogenesis was found to be the only factor associated with lymphoma-specific outcome in the Dunleavy study,1 with non–germinal center (GCB)–DLBCL patients faring significantly worse than their GCB-DLBCL counterparts. Although optimal techniques to allow confirmation that these non–GCB-DLBCL cases did, in fact, represent the activated B-cell type of DCLBC, still, this conclusion is quite likely and indicates that new treatment paradigms will be necessary in these persons, be they HIV-infected or not.

Although multiple aspects of the pathogenesis and optimal therapy of DLBCL appear similar among HIV-negative and -positive patients, the Dunleavy study has shown that the utility of FDG-PET scanning in the 2 populations is quite different.1 Whereas FDG-PET after 2 cycles of SC-EPOCH-RR had an excellent negative predictive value, a positive scan was not necessarily clinically meaningful. This would be consistent with the fact that FDG-PET may be positive in the setting of inflammation or infection, conditions that are not uncommon among HIV-infected patients. With the Dunleavy data in mind, the role of FDG-PET in patients with HIV-associated lymphoma will clearly require additional evaluation prior to clinical adoption.


  • Conflict-of-interest disclosure: The author declares no competing financial interests. ■