ABVD for Hodgkin Lymphoma in Malawi; Excellent Outcomes but with Persistent Clinical Disparities Compared to High-Income Countries: A Prospective Cohort Study

Abstract

BACKGROUND: ABVD is a well-established curative-intent therapy for Hodgkin lymphoma (HL) but, to our knowledge, little has been published about therapy or outcomes in sub-Saharan Africa (SSA) where supportive care is more limited, infectious milieu is different, and in our case, radiation therapy was not available. We report mature data from one of the first prospective HL cohorts treated under real-world conditions in SSA. METHODS: Patients ≥15 years with newly diagnosed HL were enrolled in Malawi from Jul 2013 to Dec 2021 and followed for five years if remain in remission; patients not yet achieveing five years or with relapse were censored May 2022. All participants had diagnosis confirmed by immunohistochemistry during telepathology conference between the team in Malawi and UNC Chapel Hill. Two patients were excluded from survival analysis as they were treated with CHOP chemotherapy due to misdiagnosis early in the study period. Staging was done by chest xray, abdominal ultrasound, and bone marrow biopsy. Participants were treated with ABVD chemotherapy and, if HIV+, concurrent antiretroviral therapy (ART). Post-progression therapy is generally limited to salvage chemotherapy regimens without access to radiation or autologous stem cell transplant. RESULTS: 38 participants were enrolled with mean age 29 years (SD 11). Twenty-one (55%) were men, and eleven (21%) were HIV+, of whom eight (73%) were on ART at HL diagnosis. Twenty-five (65%) participants had advanced stage disease and thirteen (35%) had limited stage disease, of whom ten were unfavorable risk and three were favorable risk. Bulky disease (at least one lymph node >10cm) was present in 29 (76%) participants at enrollment. Median CD4 count was 138 cells/mL (IQR 102-298) and 10 (91%) had an HIV viral load <400 copies/mL. Participants received a median six cycles (i.e. twelve doses) ABVD. Grade 3/4 neutropenia occurred in 15 (42%) but there were only two (5%) episodes of neutropenic fever and one episode of grade 3/4 anemia. There were no documented episodes of any grade cardiac or pulmonary toxicity. One patient was lost to follow-up, one patient absconded and died early in therapy, and one patient absconded and died during salvage therapy with median follow up 36 months for patients still alive at administrative censoring. Among patients with evaluable disease (n=34), 27 (79%) achieved a CR, three (9%) achieved a PR, and four (12%) had progressive disease. Among patients with limited stage disease, 2-yr overall survival (OS) and progression-free survival (PFS) were 92% and 80%, respectively. Among patients with advanced stage disease, 2-yr overall survival (OS) and progression-free survival (PFS) were 75% and 61%, respectively (see FIGURE). There was no difference in survival by HIV status. Of eleven deaths during the study period, nine (82%) were due to disease progression, one due to therapy-related sepsis, and one due to other causes. CONCLUSION: Here, we present, to our knowledge, the most complete prospective clinical and outcome data of pathologically confirmed Hodgkin lymphoma in sub-Saharan Africa. Most patients with HL in Malawi, a low-income country, present with advanced or unfavorable risk HL. With standard chemotherapy in the form of ABVD, long-term cure is possible, and in fact, probable, for patients presenting with HL in Malawi. The majority of deaths are attributed to disease progression and not to adverse events. However, there remains a marked disparity between outcomes in Malawi and in high-income countries and improvements through both clinical and implementation research are urgently needed to accelerate diagnosis at a lower stage and to provide therapeutic options for those who fail ABVD. Figure Caption: Overall survival and progression-free survival of Hodgkin lymphoma patients treated with ABVD in Malawi. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal