Objective: Hypophysitis occurs in up to 25% of patients with melanoma treated with ipilimumab1,2. We aimed to determine if serial routine cortisol measurements could predict hypophysitis.
Methods: We performed a retrospective audit of metastatic melanoma patients with pituitary function tests performed during ipilimumab treatment (3mg/kg IV 3 weekly for 4 doses) at the Melanoma Institute Australia for 16 months to December 2014. Patients were included if they had two or more cortisol measurements prior to sequential doses of ipilimumab.
Results: 46 of 78 treated patients were included, receiving a total of 163 ipilimumab cycles (median 4 per patient, range 2-4). Nine (20%) developed hypophysitis, a mean of 12.3±1.6 weeks (median 13, range 7.7-18) from the first ipilimumab dose. There was no difference in cortisol levels between those who did and did not develop hypophysitis, after excluding measurements after 11 am, or on glucocorticoid therapy, prior to cycles 1, 2, 3, 4 and after cycle 4 (p=0.88, 0.98, 0.64, 0.91 and 0.23 respectively). Interestingly, TSH prior to cycle 4 was significantly lower in those who developed hypophysitis (0.31 vs 1.44 mIU/L, p=0.002), but not cycles 1, 2 and 3 (p=0.12, 0.88, and 0.11). If TSH fell by ≥80% from baseline at mean 9.1+/-0.26 weeks from first ipilimumab, prior to cycle 4, odds for developing hypophysitis were 136 (95% CI 6.2-2947, p<0.0001) compared to TSH fall <80%. The fall in TSH occurred a mean of 3.4±1.4 weeks (range -1.4-9.7) prior to hypophysitis diagnosis.
Conclusions: Pre-infusion cortisol levels did not differentiate those at risk of hypophysitis. TSH is less affected by external factors and a ≥80% fall in TSH level during ipilimumab therapy is a strong risk factor for concurrent or future development of hypophysitis. Whether TSH levels measured in the weeks between cycles 3 and 4 could predict hypophysitis earlier is unknown