Can We Say No to the ‘Nocebo Effect’ in Inflammatory Bowel Diseases?

During recent decades, the management of inflammatory bowel diseases [IBD] underwent significant changes including the evolution of treatment strategies, monitoring practices, and continuous expansion of the therapeutic armamentarium.1 Albeit, many patients experience adverse events [AEs] associated with their therapy which can negatively affect the patient’s confidence in their current [and subsequent] treatment, leading to insufficient adherence and ultimately worsening outcomes. The patient’s perception of AEs is however a complex issue, which may present itself not only based on physiological changes attributed to the drug itself, but also through other nonspecific effects, such as the patients’ negative expectations, as a negative ‘functional’ trait. This phenomenon is supported by the historical observation that, in controlled trials, patients randomised to the group receiving inert agent also present positive outcomes arising from the treatment, and simultaneously, a certain proportion report negative consequences despite not being exposed to the active agent.2 When these changes in efficacy and tolerability cannot be explained by confounding factors other than psychological effects of patient expectations, they are referred to as ‘placebo effect’ and ‘nocebo effect,’ respectively.2,3 The impact of nocebo effects has not been thoroughly evaluated in IBD. In the present issue of this journal, Ma et al.4 report a systematic review and meta-analysis on AE rates for placebo and active substance arms of 195 randomised controlled trials [RCTs] for currently approved therapies in Crohn’s disease [CD] and ulcerative colitis [UC]. The main findings of the study show overall high placebo AE rates in both CD and UC of, respectively, (70.6% (95% confidence interval [CI]: 65.8%, 74.9%) and 54.5% [95% CI: 48.5%, 60.4%]. In pooled analysis, there was no significant risk difference in AE, serious adverse event [SAE], or AE-related withdrawal rates between CD patients receiving placebo or active drug. In UC, a small (1.6% [95% CI: 0.1%, 3.1%]) increase in AE rates was observed among patients receiving active therapy. Authors also identified general and disease-specific factors contributing to the nocebo phenomenon. Predictors of AEs in CD patients treated with placebo included intravenous/subcutaneous vs oral dosing, biological therapy vs corticosteroid/aminosalicylates as study drug, and moderate to severe disease at inclusion. In UC patients, disease severity, study design, study drug, and concomitant therapies influenced the occurrence of AEs in the placebo arms. Whereas the placebo effect is widely studied and described in present literature, there is much less information on the nocebo phenomenon and its consequences. Howick et al. evaluated systematic reviews of overall 1271 randomised clinical trials, and estimated a median prevalence of AEs in trial placebo groups of 49.1% with a median dropout rate of 5%.5 Many factors may strongly influence the psychological and psychosocial context of the patients’ negative expectations—which is the trigger of the nocebo effect—such as cumulative disease experience, anxiety and depression, previous negative experiences concerning side effects, and multiple parallel treatments for disease control, which all can lead to increased symptom misattribution.3,6 All of these factors are characteristic of chronic, disabling conditions, not unlike CD or UC, which make these patients particularly susceptible for experiencing higher nocebo effect rates. Other situational/contextual and disease-specific factors may also play significant role in influencing expectations by patients and investigators. One of the main sources of patients’ negative expectations is the negative perceptions of drug safety warnings.2,4 As an example, considerably higher rates of adverse effects were reported in the placebo arms of tricyclic antidepressant trials compared with the placebo arms of selective serotonin reuptake inhibitor trials, showing that information about possible [in this case more extensive and severe] side effects influenced the rate of reported AEs substantially.7 These problems are probably more expressed in RCTs, where the exhaustive informed consent process [with all the possible side effects listed] and the continuous close monitoring [and request for self-reported progress/AEs] of the participants may augment the perception of AEs.5,8,9 The route of drug administration [parenteral vs oral] has also been previously associated with both the placebo and the nocebo effect in pain management therapies.10 In IBD patients, Jairath et al. identified several factors that are associated with placebo response rates in RCTs, which are route of administration, drug class, disease duration, trial design [induction/maintenance, multinational/single country], and trial publication date.11 The nocebo effect sets important challenges for RCT design and interpretation. High nocebo rates may lead to inaccurate estimation of treatment-related AEs and also interfere with the estimation of treatment efficacy by increasing drug discontinuation.5 The high background rate of reported AEs limits the statistical power for the detection of actual differences in side effects between the active comparator and placebo. In conditions when high nocebo rates occur, designing an RCT for the detection of relatively small differences in true AE rates would require unachievable sample sizes.8 Of note, AEs with high specificity in manifestation are more easily distinguishable and less susceptive to nocebo effects. The present findings by Ma et al. have important implications. The high rates of AEs observed in the placebo arms of RCTs suggest a high magnitude of nocebo effect in IBD patients. Authors also identified a substantial [~20%] difference in reported AE rates between CD and UC patients. Why is that? It is well known that whereas in UC symptoms [especially bleeding, diarrhoea, and urgency] are straightforward, the symptomatics in CD patients are much more complex, and the leading symptom may vary significantly across patients. This may be related to the higher perception of non-specific symptoms associated with CD12, leading to overall higher nocebo effect in CD patients. Of note, the presence of irritable bowel syndrome [IBS]-like symptoms were, however, reported to be similar in UC and CD13. In addition, when comparing the active drug and placebo arms in these RCTs, clinically relevant differences were not observed in AE, SAE, or AE-related withdrawal rates. This result may reflect an overall good safety profile, but also highlights the fact that RCTs, although robust in design, have limitations for detecting small differences in rare drug-related adverse outcomes when high AE rates are reported. Finally, the identification of factors that may increase the nocebo effect is crucial for both RCT design and daily clinical practice. Authors found that many contextual factors, such as trial design, disease severity, or route of drug administration could have substantial effect on the nocebo phenomenon. Reducing the nocebo effect would be desirable; instruments to do so are, however, fairly limited in the clinical trial environment. Framing of information to put the benefits and risks of treatment in perspective should be improved in RCTs, but it is difficult while thoroughly satisfying the requirements of informed consent.14 Structuring the reports of AEs in RCTs could be also beneficial, as many trials use only self-reported AEs and do not collect or analyse them categorically. It is implicated that in certain patient populations or drug classes with high nocebo effects, RCTs are underpowered to accurately detect rare and serious AEs. Therefore, post-marketing registries and long-term robust pharmacovigilance programmes are needed to provide additional information on rare AEs. In conclusion, research on the nocebo phenomenon and its predictors are important for drug development to improve overall performance of clinical trials, especially with the rapid expansion of new treatment options in IBD. Of note, knowledge on the nocebo effect is equally important in everyday clinical practice, when assessing efficacy, side effects, and adverse events of new biological therapies. None. None. L.G., P.L.L. contributed to the design and writing up of the paper.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.