Diagnosis of Hyperprogressive Disease in Patients Treated with Checkpoint Inhibitors Using ¹⁸F-FDG PET/CT

TO THE EDITOR: A recent ahead-of-print publication by Castello et al. (1) provides fascinating insights on the potential prognostic role of ¹⁸F-FDG PET/CT in patients with non–small cell lung cancer treated with immune checkpoint inhibitors (ICIs). The authors unraveled the prognostic value of mainstream quantitative imaging biomarkers that can be derived from ¹⁸F-FDG PET using most clinical workstations. They confirmed that baseline tumor burden, baseline total lesion glycolysis on PET scans, and derived neutrophil-to-lymphocyte ratio (neutrophils/leukocytes minus neutrophils) were associated with overall survival in cancer patients treated with ICIs (2).

In this timely and comprehensive work written by experts in this field, the authors mined the data contained in baseline PET images from 50 non–small cell lung cancer patients treated with ICIs. They evaluated the association between overall survival and 8 candidate biomarkers, including imaging (n = 4) and biologic (n = 4) variables. They used a previously published composite criterion to diagnose hyperprogression (3), which can be simplified as patients with fast tumor growth during the first 2 mo of treatment. The important point is that these fast, progressive patients might have already been progressing rapidly before the initiation of ICIs. This fast progression during the first 2 mo of treatment was observed in 1 of 3 patients (30%, n = 14/46) treated with ICIs (1) and was more frequent in patients with higher baseline tumor burden, a higher number of metastatic sites, and proinflammatory parameters (pretreatment derived neutrophil-to-lymphocyte ratio and platelet counts). Such research should be actively pursued since it is of tremendous significance.

In this new era of immune-oncology, the treatment paradigm is shifting toward restoring tumor elimination by the immune system, hence the emergence of novel patterns of response (4) and progression, such as pseudoprogression (5) and hyperprogression (6). Although the medical community has gained experience in the management of pseudoprogression (5), the current wait-and-see strategy proposed to take into account delayed radiographic shrinkage is challenged by hyperprogression. Hyperprogression is indeed an atypical flair-up of tumor growth kinetics linked to premature death (6–8) due to a harmful effect of immunotherapy (6). Hyperprogression is a clinical phenomenon that might be underdiagnosed since it is a new concept that has only recently emerged. ICIs might harm 4% (9) to 29% (7) of patients with solid tumors through an accelerated progression profile leading to premature death. The underlying mechanism is an area of active investigation. Reported risk factors are a higher age (6) and the presence of MDM2/4 family amplification or estimated glomerular filtration rate aberrations (9).

The frequency of hyperprogression in patients with non–small cell lung cancer treated with ICIs differs widely in the literature: it was 30% in the study of Castello et al. as compared with 8% (3/38) (9) and 14% (56/406) (8) in other series. The current challenge is that distinct definitions are proposed. The same term, hyperprogression, is now used to conceptualize 2 distinct pathophysiologic phenomenon: a fast progression that may not be due to ICIs (prognostic tool) or an accelerated tumor growth after the initiation of ICIs associated with premature death (predictive tool).

In one definition, hyperprogression defines a fast progression that may be independent of ICIs. This strategy considers tumor growth rate only after the initiation of ICIs (1,3). This strategy is convenient since it requires only two response assessments. Nonetheless, this definition cannot demonstrate a causality effect: the fast progression profile cannot be attributed specifically to immunotherapy. Since cancers have exponential growth, patients with a high baseline tumor burden are, therefore, more likely to progress more quickly and to be called hyperprogressors if we only consider these two time points.

In another definition, hyperprogression defines an accelerated progression attributed to a harmful effect of immunotherapy (6,8). This strategy considers a change between pretreatment tumor growth rate and on-treatment tumor growth rate and has demonstrated a low rate of hyperprogression using PET (10). This definition aims to identify predictive biomarkers associated with a dramatic surge in tumor growth due to immunotherapy. Such a definition presupposes medical imaging before, at the start of, and during ICIs; such imaging is often available in clinical practice in patients treated with ICIs as the second line of therapy or if there is a wait time from referral to first treatment. The median time reported in the literature is 6 wk for first-line treatment of non–small cell lung cancer.

In conclusion, given the clinical and prognostic importance of hyperprogression, it is important to harmonize the criteria for its definition. These criteria could be a combination of clinical, radiologic (CT), and metabolic (PET) data. Beyond the definition, it is important to take into account and differentiate two distinct mechanisms—fast progression and accelerated progression—and harmonize how they are highlighted (through biomarkers, periodicity, or other means). Finally, it is necessary to harmonize the technical criteria for measuring target lesions, including whether new lesions are considered, and clinical data.

• © 2020 by the Society of Nuclear Medicine and Molecular Imaging.