Background
In East Africa, one-year mortality following a diagnosis of HIV-associated Kaposi sarcoma (KS) in the Treat All era was recently estimated, in a community-derived sample, to be 41%. To date, there is limited information on why these patients die. Understanding causes of death among adults diagnosed with KS might improve management and, hence, survival or suggest other strategies.
Methods
We identified adults (age ≥ 18 years) who died after a new diagnosis of HIV-associated KS between October 2022 and June 2024 and who were receiving care at one of 5 clinical sites in Kenya, Tanzania and Uganda. KS was confirmed by histopathology except where lesions were unsafe for biopsy. Adjudication of each death was performed using information obtained, when possible, from review of medical records and interview of attending clinicians and next of kin. A contributing cause of death was defined as suspicion that the condition mechanistically/causally contributed to death as guided by Rothman’s sufficient component cause (“causal pie”) model. Certainty of contribution to death was defined as highly (95%-100% certainty), likely (80%-95%) or possibly (50%-80%).
Results
Of 163 deaths that were adjudicated, 63% were men, the median (IQR) age was 36 (29-42) years, and duration from diagnosis to death was 4 (1-14) weeks. Many (40%) deaths occurred outside a health facility, and only 4% had an autopsy. KS was deemed to have at least possibly contributed to all deaths. Other potential causes of death, especially tuberculosis, were also present (Table). In 6% of cases, an unexplained sudden onset of malaise occurred following improvement of KS on chemotherapy, often with fever and respiratory symptoms, which was followed shortly thereafter by death.
Conclusions
Among adults who died after a new diagnosis of HIV-associated KS in East Africa in the Treat All era, KS was deemed to be at least possibly contributing in all cases. Other causes of death were also suspected, especially TB. Some scenarios beg the question regarding non-KS KSHV-related disease that is difficult to diagnose. Yet, without autopsy, it is impossible to know all causes or sequence of events. The prominence of the contribution from KS per se, combined with the inherent difficulty treating KS, imply that early detection of KS may be the most feasibly achievable intervention to reduce KS mortality.