Correlates of fever in patients receiving combined dabrafenib (GSK2118436) and trametinib (GSK1120212) for V600 BRAF-mutant metastatic melanoma

Background: Dabrafenib is an effective targeted treatment of oncogenic BRAF-mutant metastatic melanoma (MM). 20–30% of patients (pts) expe- rienced drug-related fever in phase I and II trials of dabrafenib but not trametinib. The combination of dabrafenib and trametinib reduces skin adverse events (AEs) seen with either drug alone but dabrafenib-related fever persists as a common AE. We investigated correlates of fever in pts from a single institution enrolled in the BRF113220 phase I/II study of combined dabrafenib and trametinib. Methods: Demographics, BRAF genotype, extent and burden of disease, Response Evaluation Criteria in Solid Tumours (RECIST) response, progres- sion free survival (PFS) and overall survival (OS) were collected on all pts enrolled into BRF113220 at Westmead Hospital between November 2010 and November 2011. Data were collected, including the number and sever- ity of febrile episodes, associated symptoms, haematologic and biochemical alterations, the treatment and prophylaxis of fever, and the effi cacy of these measures. Results: 23 pts were enrolled. Nine pts (39%) developed fever defi ned as ≥38.5°C, five had associated rash and six had rigors. Fever was not associ- ated with pt age, sex, BRAF genotype (V600E vs V600K), M-stage, sites of distant metastases, baseline lactate dehydrogenase (LDH) and baseline tumour volume (RECIST target lesions). Fever did not predict RECIST response, PFS or OS. Fever recurred in six pts and was repetitive (>2 episodes) in two (6 and 7 episodes). Neither dose reduction of either drug nor anti-pyretic therapy (acetaminophen, nonsteroidal anti-infl ammatory drugs [NSAIDs]) were effective in fever prophylaxis but corticosteroids were effective in all cases. Eight of nine pts had biochemical assessment at fi rst fever, and all had a rise in aspartate aminotransferase (AST) (median 100%), alanine aminotrans- ferase (ALT) (79%), alkaline phosphatase (ALP) (40%) and LDH (40%) compared to those without fever (p < 0.05). At time of fi rst fever, there was no signifi cant change in neutrophil or lymphocyte counts, gamma-glutamyl transferase (GGT) or C-reactive protein (CRP) yet lymphopaenia had devel- oped in the four pts subsequently requiring corticosteroid prophylaxis. Conclusions: Fever is not predictive of response or clinical outcome, and there are no clinical characteristics which predict fever. Febrile episodes are associated with transient elevation of ALT, ALP, LDH and especially AST. Corticosteroids are the only effective fever prophylaxis.