AB0553 THE NOCEBO PHENOMENON PARTLY ACCOUNTS FOR DIARRHOEA AMONG PARTICIPANTS IN THE RANDOMIZED PLACEBO-CONTROLLED TRIAL OF NINTEDANIB FOR INTERSTITIAL LUNG DISEASE ASSOCIATED WITH SYSTEMIC SCLEROSIS (SENSCIS)

Abstract

Background: The nocebo phenomenon, the opposite of placebo, defined as unfavourable changes in a patient's symptoms or condition resulting from negative anticipations to treatment and possibly leading to suboptimal outcomes and non-adherence, is more frequent than previously thought in rheumatology practice[1]. The tyrosine kinase inhibitor nintedanib has shown efficacy for the treatment of systemic sclerosis (SSc)-associated interstitial lung disease in SENSCIS, a recent randomized controlled trial (RCT)[2]. Diarrhoea was the most frequently reported adverse event in SENSCIS. Objectives: To test whether the nocebo phenomenon is involved in the prevalence of diarrhoea as an adverse event in trials with nintedanib. Methods: We compared the incidence of diarrhoea in the placebo arm between SENSCIS and all other placebo controlled RCTs involving >40 SSc patients in each arm, as well as between SENSCIS and all other nintedanib RCTs published so far. We also compared the strength of the warnings for diarrhea (ie times word is mentioned and number of lines devoted to nintedanib related “diarrhoea”) in the informed consent forms (ICFs) of different nintedanib RCTs Results: The mean percentage of patients reporting diarrhoea was 32% in the placebo arm and 76% in the active treatment arm in SENSCIS. These numbers are comparing to a prevalence of only 7% (range 2.3-9.1%) and 9% (range 5.8-14%), respectively of other RCTs in SSc (bosentan, n=2; macitentan, n=2; tocilizumab, n=1). Since the estimated point prevalence of diarrhoea in an SSc cohort similar to SENSCIS would not exceed 15% based on the literature, there was an at least 2-fold increase in the occurrence of diarrhoea in the placebo group during SENSCIS. More importantly, when looking into other nintedanib RCTs (Table 1), we found that patients reporting diarrhoea in the placebo arm were 20% and 18% in cancer and idiopathic pulmonary fibrosis (IPF) trials, respectively, which is almost half than in SENSCIS. Consistent with our hypothesis, the percentage of diarrhoea in the placebo arms of the different nintedanib RCTs increased along with the number of mentions and the number of lines devoted to “diarrhoea” in the respective ICFs. Conclusion: These results indicate that the nocebo phenomenon is partially involved in the high prevalence of diarrhoea among SSc patients participating in the SENSCIS trial. Whether patients with SSc have increased susceptibility to the nocebo phenomenon when compared to patients with IPF or cancer deserves further study.