Protease Inhibitor Vs. Non-Protease Inhibitor Based Antiretroviral Therapy In HIV-Infected Patients With Hodgkin and Non-Hodgkin Lymphoma Receiving Chemotherapy

HIV-infected individuals are at an increased risk of developing AIDS-defining cancers like high-grade B-cell non-Hodgkin lymphoma (NHL). Hodgkin lymphoma (HL), although not an AIDS-defining cancer, also occurs at increased frequency in HIV-infected patients. One of the main challenges that face clinicians is the high likelihood of Highly Active Antiretroviral Therapy (HAART) and chemotherapy drug-drug interactions. Antiretroviral agents that inhibit the metabolism of chemotherapy drugs may result in increasing the risk of adverse effects, which in turn could lead to life-threatening toxicities and treatment delays. On the other hand, antiretroviral agents that induce the metabolism of chemotherapy drugs may reduce their efficacy, and hence compromise treatment outcomes. Therefore, the choice of appropriate HAART regimen is essential when treating HIV-associated lymphomas.

The aim of our study was to compare toxicities, response rates and survival outcomes in patients with HIV-associated lymphoma treated with chemotherapy according to whether patients received protease inhibitor (PI) or non-PI based HAART.

This was a retrospective study, which included HIV-infected patients who were on HAART and treated with chemotherapy for NHL or HL from January 1^st 2003 to December 31^st2010 at a single institution. Patients were identified from the Haematology and Immunology departmental databases. All patients had biopsy proven lymphoma and were HIV positive at lymphoma diagnosis. We compared differences in adverse effects between patients on PI and non-PI based HAART. Univariate and multivariate logistic regression analyses were used to test for factors associated with grade 3-4 toxicities. Differences in response rates and overall survival (OS) between patients on PI vs. those on non-PI based HAART were also investigated.

A total of 53 patients were included, n=37 with NHL and n=16 with HL. Of the 53 patients, 47% and 53% were on PI and non-PI based HAART, respectively. Median baseline CD4 cell count value for all 53 patients was 273 (range 14-970) cells/μL. Baseline HIV viral load (VL) was undetectable (<50 copies/mL) in 15 (28%) of 53 patients. By multivariate analysis, the use of PI-based cART was significantly associated with increased odds ratio (OR) of developing grade 3-4 neurotoxicity (OR= 11.42, 95% CI 1.7-76.49, p-value=0.012) and grade 3-4 infections (OR=14.08, 95% CI 2.13-93.07, p-value=0.006). By univariate analysis, the use of PI-based cART was significantly associated with decreased OR of completion of stage appropriate chemotherapy (OR=0.23, 95% CI 1.1-19.07, p-value=0.023). The complete remission (CR) or unconfirmed complete remission (CRu) rate at the end of treatment for 51 evaluable patients was 86%. The OS rate for all 53 patients was 77%, with a median OS of 53 (range, 3-121) months. There were no statistically significant differences in response rates or OS between patients receiving PI-based HAART and those receiving non-PI based HAART.

This study demonstrated that PI-based HAART was significantly associated with more treatment related toxicities compared to non-PI based HAART in patients undergoing chemotherapy for HIV-associated lymphoma. However, there were no significant differences in survival outcomes or response rates between the two groups. If appropriate, non-PI based HAART should be considered during chemotherapy for HIV-associated lymphoma in order to minimize treatment-related toxicities.