e24060 Background: Neoplastic fever (or “tumor fever”, TF), is a rare and challenging condition, notable for debilitating night sweats, psychological distress, and extensive efforts to rule out other causes of fever. While treating cancer is critical, these efforts may take time and not be successful. To inform clinical practice and future trial design, we conducted a systematic review and meta-analysis evaluating pharmacologic, non-cancer-related treatments for TF. Methods: See Prospero registration 1124707. CINAHL, PubMed, Web of Science and Cochrane databases were searched from inception to August 19, 2025 for studies of any type with human subjects of any age that included: 1. Primary data, 2. Current cancer diagnosis, 3. Fever of cancer origin, and 4. A non-cancer directed, pharmacologic interventions. We excluded studies: 1. In languages other than English, 2. Addressing fever of any other origin, and 3. With <5 subjects. References for included studies were searched. Trained abstractors performed screening, data extraction, and risk-of-bias assessment using RoB 2, and GRADE assessments, with conflicts resolved by consensus. The primary outcome was “complete response rate (CR)”, as defined by each study. Duration of response, adverse events including gastrointestinal toxicity, bleeding, and other complications were pre-specified secondary outcomes. Results: 3,752 articles were identified. After removing duplicates and reviewing relevance, 98 full-text articles were reviewed, resulting in 26 articles. 24 were case-studies and 2 were prospective clinical trials. 14 were conducted in the United States, 2 in the last 5 years. Defining and reporting of primary and secondary outcomes was lacking, limiting meta-analysis. Most studies were subject to bias and of low certainty. Most studies evaluated non-steroidal anti-inflammatory drugs (NSAIDs); more recent studies investigated dronabinol and tocilizumab. Of the agents that could be synthesized, meta-analysis demonstrated a complete response rate of 86.4% for naproxen (95% CI: 82.2%-90.0%) and 80.0% for indomethacin (95% CI: 70.5%-87.5%). Secondary outcomes were not presented or synthesized due to lack of study reporting. Conclusions: Current evidence for non-cancer directed pharmacologic treatment of neoplastic fever is limited and largely based on low-quality case studies focused on NSAIDs, which show high apparent response rates. Rigorous, prospective trials are needed to evaluate existing and novel agents for standardized primary and secondary outcomes.
McKenna et al. (Thu,) studied this question.