Editorial: Checkpoint Inhibitor-Induced Liver Injury—Time to Abandon CTCAE? Authors' Reply

Abstract

We thank Torosian and Dave for their interest in our study [1].

In this study [2], we demonstrated that CTCAE overestimates the clinical severity of liver injury in CHeckpoint Inhibitor-Induced Liver Injury (CHILI), whereas DILI classifications are more accurate in predicting actual liver dysfunction and 3-month mortality.

Indeed, in our study, half of the patients treated with second-line immunosuppressants, primarily mycophenolate mofetil (MMF), had non-severe CHILI according to DILI classifications. This is notable since MMF may carry a pro-oncogenic risk associated with cancer and remains a topic of controversy [3, 4]. Moreover, immune checkpoint inhibitor (ICI) rechallenge is not recommended after hepatitis CTCAE grade ≥ 3, whereas patients experiencing immune-related adverse events often have favourable oncological response to ICI. As a result, using the CTCAE classification may lead clinicians to overtreat patients with second-line immunosuppressants and excessively limit ICI rechallenge. It is essential to use objective criteria to assess liver injury severity to avoid depriving patients of appropriate treatment options.

The CTCAE classification may also result in the overhospitalisation of patients, as evidenced by Li et al. [5], who demonstrated that hospitalisation for severe hepatitis based on CTCAE guidelines was not associated with faster hepatitis resolution and had no impact on mortality.

Furthermore, the recommended 3-day interval for starting a second-line immunosuppressant after corticosteroids appears to be brief [6], as the median delay between the onset of CHILI and the start of second-line immunosuppression was 27 days in our cohort [2]. Yet, it appears that a 3-day period may not be sufficient in immune-mediated diseases, as in autoimmune hepatitis, which is also treated with steroids, a 7-day period is usually required to define steroid resistance [7]. Recently, De Martin et al. developed a prognostic score to predict the response to corticosteroids in patients with severe acute autoimmune hepatitis presenting acute liver injury [8]. In this study, mortality was significantly predicted by INR value at corticosteroids initiation, INR progression and bilirubin values after 3 days of steroid therapy [8].

Thus, a prognostic score in CHILI that would consider both INR and total bilirubin values when starting steroid therapy and after 3 days, may facilitate the identification of severe patients who may potentially benefit from a second-line immunosuppression.

Moreover, patients with elevated bilirubin levels appear to be more likely admitted to hospital in our study [2]: the median total bilirubin level was 72 μmol/L in hospitalised patients and 10 μmol/L in non-hospitalised patients (p < 0.001). Identifying the most severe forms of hepatitis may also encourage the proposal of more invasive treatments, such as plasmapheresis.

Plasmapheresis is an extracorporeal blood purification method which is a validated treatment in many immune-mediated disorders, as it removes antibodies, increase T-reg cells and silent innate immune cells [9], and has already been described in CHILI [10]. In our study [2], 3 patients have been treated with plasmapheresis and improved outcomes. Further prospective studies are needed to precise its indication in severe CHILI.

Lina Hountondji: conceptualization, writing – original draft, writing – review and editing. Lucy Meunier: conceptualization, writing – review and editing, supervision.

This article is linked to Hountondji et al papers. To view these articles, visit https://doi.org/10.1111/apt.18276 and https://doi.org/10.1111/apt.18308.