Comparison of RECIST 1.1 and iRECIST in Patients Treated with Immune Checkpoint Inhibitors: A Systematic Review and Meta-Analysis

Abstract

Despite wide recognition of iRECIST, evidence regarding the impact of iRECIST over RECIST 1.1 is lacking. We aimed to evaluate the impact of iRECIST on assessing treatment efficacy of immune checkpoint inhibitors (ICIs) over RECIST 1.1. Articles that evaluated the treatment response and outcome based on both RECIST 1.1 and iRECIST were eligible. Data regarding overall response rates (ORR) and disease control rate (DCR) based on RECIST 1.1 and iRECIST, and data required to estimate individual patient data of progression-free survival (PFS) were extracted. Estimates were compared using meta-regression and pooled incidence rate ratios. The pooled difference of restricted mean survival time (RMST) of PFS between two criteria were calculated. Eleven studies with 6210 patients were analyzed. The application of iRECIST had no impact on the response-related endpoint by showing no significantly different ORR and DCR from RECIST 1.1 (pooled ORR, 23.6% and 24.7% [p = 0.72]; pooled DCR, 45.3% and 48.7% [p = 0.56] for iRECIST and RECIST 1.1, respectively) and had a minor impact on a survival endpoint by showing longer RMST of PFS than RECIST 1.1 (pooled difference, 0.46 months; 95% CI, 0.10–0.82 months; p = 0.01). Such a modest benefit of iRECIST should be considered when we design a clinical trial for immune checkpoint inhibitors.

Response criteria for immunotherapy and the radiologic patterns of immune-related adverse events

“…The major difference between the various immune response criteria is the definition of pseudoprogression: while irRC and irRECIST only define PD after showing progression in two consecutive studies at least 4 weeks apart, iRECIST introduces a new concept of unconfirmed PD (iUPD) for new lesions, which can become later confirmed PD (iCPD) if there is a continued progression in the next imaging study in 4-8 weeks [18,22]. Some studies reported that iRECIST can capture atypical immune response and might show a modest benefit in the survival endpoint compared to RECIST that might underestimate the benefit of immunotherapy in some patients [19,[32][33][34][35], for instance, an underestimation rate of 12.8-15% was reported in advanced melanoma patients receiving pembrolizumab using RECIST 1.1 [36,37]. Further studies are necessary to compare between various immune response criteria and RECIST 1.1 and also to evaluate if the benefits of iRECIST can outweigh the increased burden in imaging interpretation, data management, and cost [33,34].…”

“…Some studies reported that iRECIST can capture atypical immune response and might show a modest benefit in the survival endpoint compared to RECIST that might underestimate the benefit of immunotherapy in some patients [19,[32][33][34][35], for instance, an underestimation rate of 12.8-15% was reported in advanced melanoma patients receiving pembrolizumab using RECIST 1.1 [36,37]. Further studies are necessary to compare between various immune response criteria and RECIST 1.1 and also to evaluate if the benefits of iRECIST can outweigh the increased burden in imaging interpretation, data management, and cost [33,34].…”

Evolution of Radiological Treatment Response Assessments for Cancer Immunotherapy: From iRECIST to Radiomics and Artificial Intelligence

“…Another challenge in making an accurate response assessment is the emergence of atypical response patterns, such as pseudoprogression or hyperprogression, following immunotherapy. Therefore, radiologists should also be familiar with the incidence, criteria, and imaging features associated with these atypical response patterns [ 5 6 7 ]. Current immune response assessment criteria, such as the Immune Response Evaluation Criteria in Solid Tumors (iRECIST) or LYmphoma Response to Immunomodulatory therapy Criteria (LYRIC), may also have limitations in terms of providing accurate response assessment following immunotherapy [ 5 ].…”

“…Therefore, radiologists should also be familiar with the incidence, criteria, and imaging features associated with these atypical response patterns [ 5 6 7 ]. Current immune response assessment criteria, such as the Immune Response Evaluation Criteria in Solid Tumors (iRECIST) or LYmphoma Response to Immunomodulatory therapy Criteria (LYRIC), may also have limitations in terms of providing accurate response assessment following immunotherapy [ 5 ]. Hence, significant efforts have been made to apply radiomics or artificial intelligence (AI) techniques to these evaluations.…”

Radiomics in immuno-oncology

“…14 The iRE-CIST criteria were undoubtedly steps in the right direction, especially for patients who would be classified as progressive disease under previous RECIST criteria. 15 However, considering that only 20%-40% of patients respond to immunotherapy, there is a pressing need for reliable predictive biomarkers. 16 In comparison with other diagnostic modalities (e.g.…”