Seven years after the approval of the first immune checkpoint inhibitor for the treatment of metastatic melanoma, immunotherapy remains one of the primary areas of focus for oncology drug development. While checkpoint inhibition, chimeric antigen receptor T-cell therapy, and other immune-related therapies hold great promise in terms of both improved treatment outcomes and reduced side-effect profiles for cancer patients, they do present special challenges in both the clinic and the clinical trial setting.
Evaluation of response to therapy in immune oncology is complicated by the atypical patterns of response seen in a small but significant portion of patients receiving these treatments. In particular, trials of immune checkpoint inhibitors have shown that between ten and twenty percent of patients may exhibit signs of pseudoprogression prior to showing response. This is a concern for both medical management of the patient, where patients who may ultimately benefit from treatment can have therapy withdrawn prematurely, and for clinical trial management, where standard response criteria (RECIST 1.1) dictate that once a patient has progressed, no further positive response can be considered.
Pseudoprogression is a well-known problem in oncology, of course. A classic example is shown in Figure 1 below. In this case, a glioma patient presents with apparent tumor regrowth six weeks after successful treatment. Close examination of the image on the right, however, shows that there is no further shift of the midline or other mass effect present, relative to what is seen at the first post-treatment image on the left. This indicates that this is likely not tumor regrowth, but rather a delayed effect of radiation therapy.
Figure 1: A glioma patient post-therapy (l) and six weeks later (r).
Pseudoprogression occurs in patients undergoing immunotherapy for two reasons. First, immune therapies often take some time to take effect. It may take weeks for the immune system to fully rally, and tumors may progress in the meantime, showing significant growth at the first post-treatment radiological exam, which can be misinterpreted as treatment failure. Additionally, once the immune system does begin to attack target lesions, one of the first observed effects can be inflammation. This can be mistaken for malignant growth, again triggering a classification of progressive disease inappropriately.
These problems were recognized during the first clinical trials of ipilimumab, where pseudoprogression was observed in roughly twenty percent of patients who eventually demonstrated response to therapy. In response, Wolchok et al. (Clin Cancer Res 2009) proposed a revised set of response criteria, which modified standard response assessment in three ways:
- New lesions no longer mandate progressive disease (PD). Instead, they are measured and added into the total target lesion burden.
- Non-target lesions are no longer assessed for progression, although they still must be completely resolved in order to declare complete response.
- Progression requires confirmation by a second scan obtained at least four weeks after the initial observation of tumor growth.
These criteria, referred to as Immune-Related Response Criteria (irRC), have been applied to data from numerous clinical trials over the past eight years. While they do address many of the concerns with pseudoprogression, however, experience has shown that they introduce new concerns of their own. In particular, the lack of ability to call progression based on non-target lesions and the potential that new lesions may not be measurable have the potential to prevent declaration of progression in cases where it is clearly warranted. Additionally, these criteria are based not on Response Evaluation Criteria in Solid Tumors (RECIST), but rather on the much older World Health Organization (WHO) criteria for response, which have not been widely used in clinical trials since the advent of RECIST in 1999.
As a result of these concerns, interest quickly grew in the development of new response criteria that would address the problem of pseudoprogression, while hewing more closely to the best practices of standard response assessment. In response to this interest, the RECIST working group developed their own immune-related response criteria, which were published in The Lancet in March, 2017. Dubbed iRECIST, these criteria are identical to standard RECIST 1.1 until the first observed progression. At this point, as with irRC, a confirmatory scan is required after a delay of at least four weeks. Progression can then be confirmed in one of two ways:
- Significant worsening in the category of response used to declare initial PD. For example, if PD was due to target lesion increase, confirmation requires additional target growth of at least 5mm in aggregate.
- New progression in a category of response other than that used to declare initial PD. For example, if PD was due to non-target lesion worsening, either new target lesion progression or the appearance of an unequivocal new lesion will confirm progression.
If progression is not confirmed and the patient is clinically stable, they may remain on therapy until such time as confirmed PD is noted, and subsequent improvement may result in declaration of either partial response (iPR) or complete response (iCR).
iRECIST is currently being deployed across a broad range of clinical trials, primarily as a secondary or exploratory endpoint. At this time, the FDA and other regulatory agencies are withholding judgment as to the suitability of these criteria for use in the drug approval process. As experience with iRECIST increases, there is hope that these criteria may ultimately prove to be a better correlation to clinical outcomes for patients undergoing immunotherapy.
Learn more about emerging immunotherapeutic approaches to treat cancer by downloading our white paper, Exploring Immunotherapies: Beyond Checkpoint Inhibitors.