Hyperprogression during immunotherapy: do we really want to know?

Abstract

With the development of immune checkpoint targeted therapeutics, hyperprogressive disease (HPD) has been reported as a new pattern of progression observed with anti-programmed cell death 1 (PD-1) and anti-programmed death ligand 1 (PD-L1) therapies [1.Champiat S. Dercle L. Ammari S. et al.Hyperprogressive disease is a new pattern of progression in cancer patients treated by anti-PD-1/PD-L1.Clin Cancer Res. 2017; 23: 1920-1928Crossref PubMed Scopus (802) Google Scholar, 2.Kato S. Goodman A. Walavalkar V. et al.Hyperprogressors after immunotherapy: analysis of genomic alterations associated with accelerated growth rate.Clin Cancer Res. 2017; 23: 4242-4250Crossref PubMed Scopus (574) Google Scholar, 3.Saâda-Bouzid E. Defaucheux C. Karabajakian A. et al.Hyperprogression during anti-PD-1/PD-L1 therapy in patients with recurrent and/or metastatic head and neck squamous cell carcinoma.Ann Oncol. 2017; 28: 1605-1611Abstract Full Text Full Text PDF PubMed Scopus (369) Google Scholar, 4.Ferrara R. Mezquita L. Texier M. et al.Hyperprogressive disease in patients with advanced non-small cell lung cancer treated with PD-1/PD-L1 inhibitors or with single-agent chemotherapy.JAMA Oncol. 2018; 4: 1543-1552Crossref PubMed Scopus (458) Google Scholar]. HPD is clinically defined by the unexpected acceleration of the cancer evolution when patients start such immunotherapy and it has been associated with a poorer patient outcome [5.Champiat S. Ferrara R. Massard C. et al.Hyperprogressive disease: recognizing a novel pattern to improve patient management.Nat Rev Clin Oncol. 2018; 15: 748-762Crossref PubMed Scopus (242) Google Scholar]. It is, in fact, a paradoxical phenomenon of treatment-induced tumor flare-up that can be observed in different histologies and with diverse types of cancer therapies [5.Champiat S. Ferrara R. Massard C. et al.Hyperprogressive disease: recognizing a novel pattern to improve patient management.Nat Rev Clin Oncol. 2018; 15: 748-762Crossref PubMed Scopus (242) Google Scholar]. This phenomenon has been proposed to account for the early crossing over of progression-free and overall survival curves in some randomized phase III trials such as CheckMate 057 in non-small-cell lung cancer (NSCLC) [6.Borghaei H. Paz-Ares L. Horn L. et al.Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer.N Engl J Med. 2015; 373: 1627-1639Crossref PubMed Scopus (6769) Google Scholar], CheckMate 141 in head and neck squamous cell carcinoma [7.Ferris R.L. Blumenschein G. Fayette J. et al.Nivolumab for recurrent squamous-cell carcinoma of the head and neck.N Engl J Med. 2016; 375: 1856-1867Crossref PubMed Scopus (2889) Google Scholar] or Keynote 045 [8.Bellmunt J. de Wit R. Vaughn D.J. et al.Pembrolizumab as second-line therapy for advanced urothelial carcinoma.N Engl J Med. 2017; 376: 1015-1026Crossref PubMed Scopus (2072) Google Scholar] and IMvigor211 [9.Powles T. Durán I. van der Heijden M.S. et al.Atezolizumab versus chemotherapy in patients with platinum-treated locally advanced or metastatic urothelial carcinoma (IMvigor211): a multicentre, open-label, phase 3 randomised controlled trial.Lancet. 2018; 391: 748-757Abstract Full Text Full Text PDF PubMed Scopus (875) Google Scholar] (urothelial carcinoma) comparing anti-PD(L)1 monotherapies to standard-of-care chemotherapies. However, the relationship between immunotherapy and this detrimental phenomenon remains controversial. Some suggest that HPD could be the natural evolution of some tumors when the therapy has no effect. Indeed, there is currently no consensus on its radiological definition and several methods for HPD assessments have been proposed [5.Champiat S. Ferrara R. Massard C. et al.Hyperprogressive disease: recognizing a novel pattern to improve patient management.Nat Rev Clin Oncol. 2018; 15: 748-762Crossref PubMed Scopus (242) Google Scholar]. Also, no strong biomarker related to the phenomenon has been reported and it lacks a clear biological mechanism [5.Champiat S. Ferrara R. Massard C. et al.Hyperprogressive disease: recognizing a novel pattern to improve patient management.Nat Rev Clin Oncol. 2018; 15: 748-762Crossref PubMed Scopus (242) Google Scholar]. Moreover, some may question the value of differentiating HPD from regular progressive disease (PD) if the conclusion for both situations is that the patient does not benefit from anti-PD(L)1 blockade. Indeed, this issue would be solved if we had robust predictive biomarkers with strong individual positive predictive value which could select the patients who benefit from anti-PD(L)1 therapies. The first radiological definition of HPD was the tumor growth rate (TGR), which compared tumor lesion sizes on computed tomography (CT) scans before and after anti-PD(L)1 therapy [1.Champiat S. Dercle L. Ammari S. et al.Hyperprogressive disease is a new pattern of progression in cancer patients treated by anti-PD-1/PD-L1.Clin Cancer Res. 2017; 23: 1920-1928Crossref PubMed Scopus (802) Google Scholar]. This method could demonstrate retrospectively that the incidence of HPD was higher with anti-PD(L)1 therapies than with chemotherapies in NSCLC [4.Ferrara R. Mezquita L. Texier M. et al.Hyperprogressive disease in patients with advanced non-small cell lung cancer treated with PD-1/PD-L1 inhibitors or with single-agent chemotherapy.JAMA Oncol. 2018; 4: 1543-1552Crossref PubMed Scopus (458) Google Scholar]. However, pre-baseline CT scans can be difficult to collect and may be lacking in the context of the first-line therapy. Thus, fast progression (FP) has been proposed as a surrogate definition of HPD, with a ≥50% increase in the sum of long diameters of target lesions within 6 weeks of initiating treatment or death due to PD within 12 weeks [10.Gandara D. Reck M. Morris S. et al.LBA1 Fast progression in patients treated with a checkpoint inhibitor (cpi) vs chemotherapy in OAK, a phase III trial of atezolizumab (atezo) vs docetaxel (doc) in 2L+ NSCLC.Ann Oncol. 2018; 29 (doi: 10.1093/annonc/mdy511)PubMed Google Scholar]. In a post hoc analysis of the randomized phase III OAK trial, comparing atezolizumab to docetaxel in NSCLC (N = 850), a similar rate of FP (10%) was observed in both arms suggesting that FP is not specific to anti-PD-L1 therapy [10.Gandara D. Reck M. Morris S. et al.LBA1 Fast progression in patients treated with a checkpoint inhibitor (cpi) vs chemotherapy in OAK, a phase III trial of atezolizumab (atezo) vs docetaxel (doc) in 2L+ NSCLC.Ann Oncol. 2018; 29 (doi: 10.1093/annonc/mdy511)PubMed Google Scholar]. However, the proportion of fast progressors only due to ≥50% growth was numerically higher for atezolizumab (45% of FP, N = 20/44) than for docetaxel (29% of FP, N = 12/41) [10.Gandara D. Reck M. Morris S. et al.LBA1 Fast progression in patients treated with a checkpoint inhibitor (cpi) vs chemotherapy in OAK, a phase III trial of atezolizumab (atezo) vs docetaxel (doc) in 2L+ NSCLC.Ann Oncol. 2018; 29 (doi: 10.1093/annonc/mdy511)PubMed Google Scholar]. In the study by Kim et al. in this issue of Annalsof Oncology, important results about such paradoxical tumor growth acceleration during anti-PD(L)1 blockade in NSCLC are presented [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. This is the second largest size study evaluating kinetics of tumor growth in this specific context with 263 patients treated in tertiary referral centers. The authors report that 19% (45/237) of NSCLC patients treated by anti-PD(L)1 blockade presented with HPD [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. As previously reported [1.Champiat S. Dercle L. Ammari S. et al.Hyperprogressive disease is a new pattern of progression in cancer patients treated by anti-PD-1/PD-L1.Clin Cancer Res. 2017; 23: 1920-1928Crossref PubMed Scopus (802) Google Scholar, 4.Ferrara R. Mezquita L. Texier M. et al.Hyperprogressive disease in patients with advanced non-small cell lung cancer treated with PD-1/PD-L1 inhibitors or with single-agent chemotherapy.JAMA Oncol. 2018; 4: 1543-1552Crossref PubMed Scopus (458) Google Scholar], the authors confirmed that HPD patients have a dramatically poorer prognosis, with a median overall survival of 50 days compared with 205 days in patients displaying a PD without HPD (hazard ratio = 5.079) [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. Despite sharing the lack of benefit from immunotherapy, one can easily understand that having pre-treatment aggressive disease is different than having a detrimental effect of the therapy which accelerates the disease. It is, therefore, obvious that measuring FP without kinetics of tumor growth (with a pre-baseline CT) does not explore whether tumor growth accelerates or not on therapy. An HPD definition that misses tumor kinetics increases the risk of overlapping HPD patients with other clinical realities (Figure 1). Indeed, if oncologists just have the information of progression on the first CT scan evaluation, they cannot differentiate among progressive patients the ones that have HPD from the ones with conventional progression or even pseudo-progression. Some patients can even be misclassified as HPD when they will in fact present with a pseudo-progression pattern [4.Ferrara R. Mezquita L. Texier M. et al.Hyperprogressive disease in patients with advanced non-small cell lung cancer treated with PD-1/PD-L1 inhibitors or with single-agent chemotherapy.JAMA Oncol. 2018; 4: 1543-1552Crossref PubMed Scopus (458) Google Scholar]. Some HPD patients could also fail to meet the FP criteria if the tumor growth accelerates significantly but without a ≥50% increase in the sum of the long diameters of target lesions. Due to the absence of consensus on the accurate definition of HPD, Kim et al. did comprehensive work at comparing three different definitions of HPD: TGR [1.Champiat S. Dercle L. Ammari S. et al.Hyperprogressive disease is a new pattern of progression in cancer patients treated by anti-PD-1/PD-L1.Clin Cancer Res. 2017; 23: 1920-1928Crossref PubMed Scopus (802) Google Scholar], tumor growth kinetics (TGK) [3.Saâda-Bouzid E. Defaucheux C. Karabajakian A. et al.Hyperprogression during anti-PD-1/PD-L1 therapy in patients with recurrent and/or metastatic head and neck squamous cell carcinoma.Ann Oncol. 2017; 28: 1605-1611Abstract Full Text Full Text PDF PubMed Scopus (369) Google Scholar] and time to treatment failure (TTF) [2.Kato S. Goodman A. Walavalkar V. et al.Hyperprogressors after immunotherapy: analysis of genomic alterations associated with accelerated growth rate.Clin Cancer Res. 2017; 23: 4242-4250Crossref PubMed Scopus (574) Google Scholar]. To exclude the possibility of overestimating the incidence of HPD, they defined HPD as a more than twofold increase in TGK and in TGR of the experimental period compared with the reference period in patients determined to have PD by RECIST 1.1 at the first evaluation after PD1/PD-L1 blockade. In addition, HPD based on TTF was defined as TTF <2 months, as suggested previously [2.Kato S. Goodman A. Walavalkar V. et al.Hyperprogressors after immunotherapy: analysis of genomic alterations associated with accelerated growth rate.Clin Cancer Res. 2017; 23: 4242-4250Crossref PubMed Scopus (574) Google Scholar]. They demonstrate in a robust manner that TGR and TGK are very similar but outperform TTF at defining HPD [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. Moreover, the percentage of patients with TTF <2 months was higher (37%) than patients with HPD by the TGK (20, 9%) or TGR (20, 5%) definition. This suggests that FP could not be used as a surrogate marker for hyperprogression. Also, as previously reported [1.Champiat S. Dercle L. Ammari S. et al.Hyperprogressive disease is a new pattern of progression in cancer patients treated by anti-PD-1/PD-L1.Clin Cancer Res. 2017; 23: 1920-1928Crossref PubMed Scopus (802) Google Scholar], the authors have found an inverse correlation between treatment response to immunotherapy according to RECIST 1.1 with kinetics of tumor growth (TGR or TGK) during the reference period. This supports the idea that patients with a slowly progressive tumor may have a higher risk of presenting with HPD and indicate that HPD differs from a spontaneous acceleration of disease progression. Whether the immune system can have a pro-tumor effect in certain situations is not questionable: the anti-PD(L)1 axis is just one among many examples of how the immune system can actually favor tumor growth [12.Coussens L.M. Werb Z. Inflammation and cancer.Nature. 2002; 420: 860-867Crossref PubMed Scopus (11186) Google Scholar, 13.Dunn G.P. Bruce A.T. Ikeda H. et al.Cancer immunoediting: from immunosurveillance to tumor escape.Nat Immunol. 2002; 3: 991-998Crossref PubMed Scopus (3628) Google Scholar, 14.Donnell J.S. Long G.V. Scolyer R.A. et al.Resistance to PD1/PDL1 checkpoint inhibition.Cancer Treat Rev. 2017; 52: 71-81Abstract Full Text Full Text PDF PubMed Scopus (342) Google Scholar]. However, whether this unexpected acceleration of tumor growth during anti-checkpoint therapy is directly related to immunotherapy remains discussable. The study by Kim et al. brings an additional retrospective dataset supporting the biological reality of the phenomenon. First, they nicely show that classical anti-PD-(L)1 predictive biomarkers such as tumor PD-L1 positivity, as previously shown [1.Champiat S. Dercle L. Ammari S. et al.Hyperprogressive disease is a new pattern of progression in cancer patients treated by anti-PD-1/PD-L1.Clin Cancer Res. 2017; 23: 1920-1928Crossref PubMed Scopus (802) Google Scholar, 3.Saâda-Bouzid E. Defaucheux C. Karabajakian A. et al.Hyperprogression during anti-PD-1/PD-L1 therapy in patients with recurrent and/or metastatic head and neck squamous cell carcinoma.Ann Oncol. 2017; 28: 1605-1611Abstract Full Text Full Text PDF PubMed Scopus (369) Google Scholar, 4.Ferrara R. Mezquita L. Texier M. et al.Hyperprogressive disease in patients with advanced non-small cell lung cancer treated with PD-1/PD-L1 inhibitors or with single-agent chemotherapy.JAMA Oncol. 2018; 4: 1543-1552Crossref PubMed Scopus (458) Google Scholar], but also EGFR mutations, tumor mutational burden or even MHC-I tumor expression are not associated with HPD and therefore cannot help to identify HPD patients pre-treatment [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. Second, by looking at the phenotype of circulating CD8 lymphocytes by flow cytometry at baseline they have found that a lower frequency of effector/memory subsets (CD45RA-CCR7-) and higher frequency of exhausted TIGIT+ in PD1+ cells were associated with HPD. A recent publication by Kamada et al. has also shown that PD-1 blockade may facilitate the proliferation of highly suppressive intratumoral PD-1+ effector Treg cells (eTreg,) with a FoxP3high CD45RA- CD4+ phenotype, resulting in HPD [15.Kamada T. Togashi Y. Tay C. et al.PD-1+ regulatory T cells amplified by PD-1 blockade promote hyperprogression of cancer.Proc Natl Acad Sci USA. 2019; 116 (201822001)Crossref Scopus (436) Google Scholar]. These suspected mechanisms are now added to other reports that support the role of therapeutic antibody activation of M2-like CD163+ CD33+ PD-L1+ macrophages through the Fc γ receptor [16.Russo G. Moro M. Sommariva M. et al.Antibody-Fc/FcR interaction on macrophages as a mechanism for hyperprogressive disease in non-small cell lung cancer subsequent to PD-1/PD-L1 blockade.Clin Cancer Res. 2018; 25: 1-12Google Scholar], senescent CD4 T cells with a highly differentiated (CD28- CD27-) profile [17.Zuazo-Ibarra M, Arasanz H, Fernández-Hinojal G, Highly differentiated CD4 T cells unequivocally identify primary resistance and risk of hyperprogression to PD-L1/PD-1 immune checkpoint blockade in lung Cancer. bioRxiv 320176; doi: https://doi.org/10.1101/320176.Google Scholar], granulocytic myeloid-derived suppressor cells (Gr-MDSC, CD33+ HLADR- CD15+ CD14-) [18.Faure M. Rochigneux P. Olive D. et al.Hyperprogressive disease in anorectal melanoma treated by PD-1 inhibitors.Front Immunol. 2018; 9: 797Crossref PubMed Scopus (32) Google Scholar], or tumor cells themselves [19.Wartewig T. Kurgyis Z. Keppler S. et al.PD-1 is a haploinsufficient suppressor of T cell lymphomagenesis.Nature. 2017; 552: 121-125Crossref PubMed Scopus (155) Google Scholar, 20.Du S. McCall N. Park K. et al.Blockade of tumor-expressed PD-1 promotes lung cancer growth.Oncoimmunology. 2018; 7: 1-10Crossref Scopus (81) Google Scholar]. Health authorities such as Food and Drug Administration and European Medicines Agency and clinical trial sponsors could request the collection of pre-baseline CT scans for the measurement of tumor flares upon treatment of cancer patients. As discussed during the recent American Association for Cancer Research Annual Meeting 2019 in Atlanta, GA, in a dedicated HPD session, tumor flare-up should be monitored and considered as any other adverse event [21.Poh A. Understanding hyperprogression in cancer. Cancer Discov 2019. doi: 10.1158/2159-8290.cd-nd2019-003.Google Scholar]. In fact, patients’ advocacy representatives present at this 2019 AACR dedicated session clearly mentioned that whether HPD exists or not, patients need to be informed of the frequency of tumor flare-ups upon treatment when considering their decision to receive therapy [21.Poh A. Understanding hyperprogression in cancer. Cancer Discov 2019. doi: 10.1158/2159-8290.cd-nd2019-003.Google Scholar]. As clinicians, we could probably implement a simple practice change to address this potential major problem in our daily practice. Indeed, pseudo-progressions have already modified the way we deal with tumor progression during anti-checkpoint therapies [22.Seymour L. Bogaerts J. Perrone A. et al.iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics.Lancet Oncol. 2017; 18: e143-e152Abstract Full Text Full Text PDF PubMed Scopus (1171) Google Scholar]. For most progressive patients at first CT evaluation (after 2 months), we prolong therapy until the assessment of a confirmatory CT scan (1 month later). However, the 6%–29% reported frequency for HPD [5.Champiat S. Ferrara R. Massard C. et al.Hyperprogressive disease: recognizing a novel pattern to improve patient management.Nat Rev Clin Oncol. 2018; 15: 748-762Crossref PubMed Scopus (242) Google Scholar] is a little more concerning than the 1%–9% reported frequency for pseudo-progression [23.Borcoman E. Kanjanapan Y. Champiat S. et al.Novel patterns of response under immunotherapy.Ann Oncol. 2019; 30: 385-396Abstract Full Text Full Text PDF PubMed Scopus (252) Google Scholar]. Also, we should be mostly concerned with cancer histologies suspected to be associated with higher frequency of HPD such as NSCLC, head and neck squamous cell carcinoma or urothelial cancers (where excess of early death and crossing survival curves in phase III trials have been observed). HPD is a phenomenon happening at the onset of anti-PD(L)1 monotherapies. The habit of obtaining the first CT scan tumor assessment routinely between 6 and 12 weeks after the initiation of immunotherapy is just an inheritance of cytotoxic chemotherapies clinical trials and could be changed. A simple CT scan assessment (e.g. after 3 weeks of anti-PD(L)1) could prevent us from exposing patients to a potential ineffective or even deleterious therapy. In the case of mild or non-threatening PD at this early time point, patients could have a confirmatory CT scan 1 month later, reducing the time for treatment discontinuation decision of more than 1 month and therefore offer maybe extra time to initiate a rescue chemotherapy. This change in practice is actually strongly supported by the data presented by Kim et al., as the median PFS for HPD was 19 days compared with 48 days in PD without HPD patients (hazard ratio = 4.619) [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. As Kim et al. report in their study, 70% of HPD patients do not retain the possibility to receive subsequent treatment as compared with 39% for conventional progressive patients (Figure 2) [11.Kim C. Kim K. Pyo K.-H. et al.Hyperprogressive disease during PD-1/PD-L1 blockade in patients with non-small-cell lung cancer.Ann Oncol. 2019; 30: 1104-1113Abstract Full Text Full Text PDF PubMed Scopus (157) Google Scholar]. Despite revolutionizing systemic therapy for cancer in different tumor types with prolonged survival in subsets of patients with metastatic cancers, the majority of patients do not respond to anti-PD(L)1 blockade monotherapies [24.Hirsch L. Zitvogel L. Eggermont A. Marabelle A. PD-Loma: a cancer entity with a shared sensitivity to the PD-1/PD-L1 pathway blockade.Br J Cancer. 2019; 120: 3-5Crossref PubMed Scopus (64) Google Scholar]. Also, as observed in the reported studies, current biomarkers are not able to predict HPD and therefore could not be used to exclude patients with a risk of anti-PD(L)1 deleterious effect. We believe that kinetics of tumor growth should be part of clinical trials evaluation to proper identify and evaluate the phenomenon of HPD. Beyond confirming the reality of HPD, the identification of immune-related mechanisms behind such dramatic tumor growth acceleration could be key for future anti-cancer therapy development. So, who really doesn’t want to know if hyperprogression exists?