Effect of immunotherapy time-of-day infusion on overall survival among patients with advanced melanoma in the USA (MEMOIR): a propensity score-matched analysis of a single-centre, longitudinal study

Background : The dependence of the adaptive immune system on circadian rhythm is an emerging fieldof study with potential therapeutic implications. We aimed to determine whether specifictime-of-day patterns of immune checkpoint inhibitor infusions might alter melanoma treatment efficacy. Methods : Melanoma Outcomes Following Immunotherapy (MEMOIR) is a longitudinal study of allpatients with melanoma who received ipilimumab, nivolumab, or pembrolizumab, or a combination of these at a single tertiary cancer centre (Winship Cancer Instituteof Emory University, Atlanta, GA, USA). For this analysis, we collected deidentifiedparticipant-level data from the MEMOIR database for adults (age ≥18 years) diagnosed with stage IV melanoma between 2012 and 2020. Those who received fewer than four infusions were excluded. Standard of care doses were used, with modifications at the treatingphysicians' discretion. The primary outcome was overall survival, defined as deathfrom any cause and indexed from date of first infusion of immune checkpoint inhibitor.We calculated the association between overall survival and proportion of infusions of immune checkpoint inhibitors received after 16 30 h (a composite time cutoff derivedfrom seminal studies of the immune-circadian rhythm to represent onset of evening)using Cox regression and propensity score-matching on age, Eastern Cooperative OncologyGroup performance status, serum lactate dehydrogenase concentration, and receipt ofcorticosteroids and radiotherapy. Treatment-related adverse events that led to changeor discontinuation of immune checkpoint inhibitors were also assessed. Findings : Between Jan 1, 2012, and Dec 31, 2020, 481 patients with melanoma received treatment with immune checkpoint inhibitors at the study centre, of whom 299 had stage IV diseaseand were included in this study; median follow-up was 27 months (IQR 14 to 47). Inthe complete unmatched sample, 102 (34%) patients were female and 197 (66%) were male,with a median age of 61 years (IQR 51 to 72). Every additional 20% of infusions ofimmune checkpoint inhibitors received after 1630 h (among all infusions received bya patient) conferred an overall survival hazard ratio (HR) of 1·31 (95% CI 1·00 to1·71; p=0·046). A propensity score-matched analysis of patients who did (n=73) anddid not (n=73) receive at least 20% of their infusions of immune checkpoint inhibitorsafter 1630 h (54 [37%] of 146 patients were women and 92 [63%] were men, with a medianage of 58 years [IQR 48 to 68]) showed that having at least 20% of infusions in theevening was associated with shorter overall survival (median 4·8 years [95% CI 3·9to not estimable] vs not reached; HR 2·04 [1·04 to 4·00; p=0·038]). This result remained robust to multivariableproportional hazards adjustment with (HR 1·80 [1·08 to 2·98; p=0·023]) and without(2·16 [1·10 to 4·25; p=0·025]) inclusion of the complete unmatched study sample. Themost common adverse events were colitis (54 [18%] of 299 patients), hepatitis (27[9%]), and hypophysitis (15 [5%]), and there were no treatment-related deaths. Interpretation : Our findings are in line with an increasing body of evidence that adaptive immuneresponses are less robust when initially stimulated in the evening than if stimulatedin the daytime. Although prospective studies of the timing of immune checkpoint inhibitorinfusions are warranted, efforts towards scheduling infusions before mid-afternooncould be considered in the multidisciplinary management of advanced melanoma.

The Lancet Oncology 2021

Voir le bulletin