Thyroid function can be affected during the course of immune checkpoint inhibitor (ICI) therapy, according to a recent study published in The Oncologist.
The study researchers searched medical records from 2010 to 2016 at the University of California San Diego Moores Cancer Center for patients (n=285) who had used pembrolizumab, ipilimumab, and/or nivolumab.
A total of 61 patients had thyroid function test abnormalities (TFTAs) at baseline vs 218 who did not. Another 6 patients had previously had thyroidectomies and were excluded from this retrospective analysis of thyroid function test results over the course of ICI treatment.
Continue Reading
The researchers found that 35% of the patients developed new TFTAs while on ICI treatment (median onset 46 days). Additionally, 70.5% of those with baseline TFTAs showed worsening of abnormalities on treatment (median 33 days).
New TFTAs occurred for 64.5% of patients on ipilimumab-nivolumab combination treatment. Rates were 31.3% with ipilimumab-only, 31.5% with nivolumab-only, and 26% with pembrolizumab-only treatments.
Among patients treated with the ipilimumab-nivolumab combination, 35.5% developed primary overt hyperthyroidism and 16.1% developed central hypothyroidism. For single-agent treatments, hypothyroidism dominated new-onset TFTAs, with primary, subclinical, or central types all represented. TFTA treatments, as appropriate, included thyroid hormone replacement, methimazole treatment, hospitalization, or recovery without treatment.
Because of the prevalence of TFTAs found in this study, the researchers recommended that thyroid function be checked frequently in the first 8 weeks of ICI therapy, and particularly for those presenting with TFTAs at baseline.
The authors noted that whether abnormal TFTAs remain long-term or if affected patients regain euthyroid status is unclear.
Reference
Patel NS, Oury A, Daniels GA, Bazhenova L, Patel SP. Incidence of thyroid function test abnormalities in patients receiving immune-checkpoint inhibitors for cancer treatment. Oncologist 2018;23:1236-1241.
