Gene Expression Profiling Predicts Relapse‐free and Overall Survival in Newly Diagnosed Myeloma Patients Treated with Novel Therapies
Elisabet E. Manasanch,David Berrios,Eric Fountain,Catherine M. Claussen,Tammy Chuang,Gregory P. Kaufman,Behrang Amini,Qaiser Bashir,Yago Nieto,Muzaffar H. Qazilbash,Krina K. Patel,Sheeba K. Thomas,Donna M. Weber,Zuzana Berkova,Gökçe Törüner,Pei Lin,Feng Liu,Hans C. Lee,Robert Z. Orlowski,Chutima Kunacheewa
DOI: https://doi.org/10.1111/bjh.17344
2021-01-01
British Journal of Haematology
Abstract:Survival in myeloma varies significantly depending on host factor biology and therapy.1, 2 An RNA microarray platform (MyPRS) was the first to identify distinct low- (LR) and high-risk (HR) myeloma subgroups.3, 4 A second platform identified additional subgroups (EMC92).5-7 Both of them predicted outcome in patients treated with conventional chemotherapy.4-6, 8 We reviewed electronic medical records in 104 consecutive newly diagnosed multiple myeloma (NDMM) patients who underwent gene expression profiling (GEP MyPRS) between April 2014 and January 2017. The primary end-point of this study is the impact of GEP on relapse-free survival (RFS) and overall survival (OS). Detailed study methods are available in the Supplementary Material. For all patients GEP was available and 99% (n = 103) had fluorescent in situ hybridization (FISH) results. Patients were HR or LR GEP [29 patients (27·9%) vs 75 patients (72·1%) respectively]. Patients’ characteristics are in Tables SI–SII. The median GEP score was higher in HR versus LR GEP/FISH patients: 47·3 vs 32·84 (P = 0·0002) and 50·1 vs 32·6 (P < 0·001) respectively. Sixteen patients with LR FISH were classified as HR GEP and nine patients with HR FISH categorized as LR by GEP. 11/16 (70%) HR GEP patients contained both amplification 1q ≥ 4 copies and monosomy 13. All patients were treated with triplet initial therapy, 72·5% underwent autologous stem cell transplant (autoSCT) and most HR patients received maintenance therapy with doublets/triplets (n = 11/16 70%) versus monotherapy (n = 3/16; 19%). The overall response rate (ORR) of 94 evaluable patients after six months of therapy was 96% (n = 90/94) and comparable in both HR and LR GEP (93% and 96·9%; P = 0·087; Table I). Patients with LR GEP and/or LR FISH deepened their response over time. In contrast, HR GEP patients had significantly higher relapse rate at 12 months 28% (7/29 patients) versus only 1·6% (1/75 patients) (P < 0·0001) irrespective of their FISH classification. In addition, the LR FISH/HR GEP had a relapse rate at 12 months of 30% (3/12 patients) versus 0% in patients with LR GEP/HR FISH (0/8 patients; Table I). All patients N (%) High-risk FISH (n = 21) N (%) Standard-risk FISH (n = 81) N (%) High-risk FISH (n = 26) N (%) Standard -risk FISH (n = 77) N (%) Response at 12 months (n = 87) All patients (n = 19/104); N (%) High-risk GEP/ High-risk FISH (n = 5/19); N (%) High-risk GEP/ low-risk FISH (n = 4/12); N (%) Low-risk GEP/ High-risk FISH (n = 3/10); N (%) Low-risk GEP/ low-risk FISH (n = 7/65); N (%) At a median follow-up of 33 months (1·51–55·75 months), 85/104 (82%) patients were alive. Patients with HR GEP had significantly poorer three-year RFS of 41% vs 60% and three-year OS of 71% vs 83% than LR GEP (P = 0·034). Similarly, patients with HR FISH had shorter three-year RFS/OS ( Fig 1). The median RFS/OS for the cohort was not reached. Patients who underwent up-front autoSCT had longer median RFS (not reached vs 18 months, P = 0·0021) but not OS (P = 0·287; Figure S1). We combined the HR and LR groups from FISH/GEP to evaluate clinical outcomes (Figure S2). Patients with HR GEP/HR FISH demonstrated reduced RFS and OS when compared to LR GEP/FISH with HR GEP being the best predictor of early RFS. For example, patients with HR GEP had a death rate due to multiple myeloma (MM) of 21% (HR GEP/HR FISH), 25% (HR GEP/LR FISH), 0% (LR GEP/HR FISH) and 6% (LR GEP/LR FISH) (P = 0·03; Table I). The addition of CKS1B amplification CKS1B ≥ 4 copies resulted in statistically significant differences in RFS but not OS ( Fig 1). Deletion 17p and t(14;16) were prognostic of shorter RFS/OS (Figures S3/S4). Adequate treatment in myeloma is dependent on the reliable identification of HR patients at diagnosis. Traditionally risk assessment in myeloma has been done through cytogenetics. In 2006, the molecular classification of myeloma was described using GEP of tumour cells.3 An analysis of 4 750 patients in clinical trials using different GEP platforms confirmed GEP as a prognostic tool.9 Several studies have also indicated that GEP is more accurate than FISH in risk stratification.4, 5, 8 To our knowledge, this is the first retrospective study in standard of care treatment in NDMM patients treated with modern therapies aimed to investigate the clinical relevance of GEP its relationship to current HR FISH criteria and the impact of modified FISH criteria by adding amplification 1q as a HR cytogenetic abnormality. In our study, an ORR of 95·7% with ≥ VGPR (very good partial response) rate of 74·7% at six months was similar to that in previous studies.10, 11 GEP was able to identify a subset of HR patients with early relapse which led to shorter RFS and OS. Additionally, the mortality in patients with HR GEP was almost all due to progressive myeloma, whereas in HR FISH most mortality was due to infection, other types of cancer or unknown. Overall survival outcomes for our cohort are comparable or even favourable in terms of RFS to those published in the recent ENDURANCE study which included mostly LR patients.12 There the median RFS was 34·4 and 34·6 months with bortezomib, lenalidomide and dexamethasone (VRd) and carfilzomib, lenalidomide and dexamethasone (KRd) respectively. In our cohort, including HR patients, the median PFS and OS was not reached at a median follow-up of 33 months. The difference in outcome with regard to RFS may be due to the higher number of patients undergoing up-front ASCT in our cohort versus the ENDURANCE study (72·5% vs 25%). In our study GEP identified more HR patients than FISH; 70% (n = 11/16) of patients who were reclassified from standard-risk FISH to HR GEP presented with amplification 1q ≥ 4 copies. This supports previous studies that showed that over 30% of HR myeloma genes mapped to chromosome 1. In particular, the presence of ≥ 4 copies of amplification CKS1B/1q21 led to worse clinical outcomes.13, 14 In our cohort, modified FISH ≥ 4 copies CKS1B impacted RFS significantly. Out study has limitations including its retrospective nature, lack of uniform treatment and unknown cause of death in some patients. Despite this, our study shows the predictive value of GEP in a large cohort of patients treated with modern therapies. It also shows that HR GEP patients continue to have worse clinical outcomes despite triplet initial therapy and triplet maintenance therapy. Thus, new therapies need to be urgently incorporated in these patients’ frontline treatment. GEP is a useful tool that can reliably predict patients that will have worse outcomes with current standard therapies. In this cohort of patients, it also identified more accurately patients that died of progressive myeloma than HR FISH. Larger prospective studies may help further clarify the incremental value of GEP over FISH. EM and CK designed the research, gathered data, analyzed the data and wrote the manuscript. DB, EF, CC, TC, PL, GT and ZB gathered and analyzed data. HL, KP, GK, BA, QB, YN, MQ, ST and DW contributed patients. LF analyzed data. RZO contributed patients and provided manuscript comments. EM has received research support from Sanofi, Quest Diagnostics, Novartis, JW Pharma, Merck; consultant fees from Takeda, Celgene, Sanofi, GSK and Adaptive Biotechnologies. HL has received consulting fees from Adaptive Biotechnologies, Celgene, Pimera and Takeda and research support from Amgen, Daiichi Sankyo, Janssen and Takeda. ST has received consulting fees from Amgen and research support from Acerta Pharma, Amgen, Array BioPharma, Bristol-Myers-Squibb, Celgene and Idera. RO has received consulting fees from Amgen, Bristol-Myers-Squibb, Celgene, GSK Biologicals, Ionis Pharmaceuticals, Janssen, Juno Therapeutics, Kite Pharma, Legend Biotech, Molecular Partners, Sanofi, Servier, and Takeda and research support from BioTheryX. EF, CC, TC, KP, GK, BA, SS, QB, YN, MQ, DW, GT, LF and CK report no conflicts of interest. This work was supported in part by The MD Anderson Cancer Center Support Grant (P30 CA016672), the Leukemia and Lymphoma Society Specialized Center of Research (LLS SCOR), the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation, the Multiple Myeloma Research Foundation, the Chapman-Perelman Foundation, and the University of Texas MD Anderson Moon Shot Program. RZO, the Florence Maude Thomas Cancer Research Professor, would like to acknowledge support from the National Cancer Institute (R01s CA184464 and 194264, and U10 CA032102), the Leukemia &Lymphoma Society (SCOR-12206-17), the Adelson Medical Research Foundation, the Brock Family Myeloma Research Fund, and the Jean Clarke HR Myeloma Research Fund. We thank participating patients and their families. Table SI. Patients’ baseline characteristics and treatment according to gene expression profiling (GEP), fluorescent in situ hybridization (FISH) and modified FISH criteria. Table SII. Patients’ baseline characteristics and treatment, according to gene expression profiling (GEP) in combination with modified fluorescent in situ hybridization (FISH) criteria. Table SIII. A. Response to treatment by gene expression profiling (GEP) in combination with modified fluorescent in situ hybridization (FISH) criteria; B. Cause of death. Table SIV. Relapse-free survival (RFS: A, C, E, G) and s (OS: B, D, F, H) by fluorescent in situ hybridization (FISH), FISH (with amplification CKS1B), FISH (with amplification CKS1B ≥ 4 copies) and gene expression profiling (GEP). Fig S1. Relapse-free survival (RFS; A, C) and overall survival (OS: B, D) of the entire patient cohort and according to transplant status. Fig S2. Relapse-free survival (RFS; A, B, C, D, E) and overall survival (OS: F, G, H, I, J) of HR cytogenetics t(4;14), t(14;16), deletion 17p and amplification CKS1B. Fig S3. Relapse-free survival (RFS) and overall survival (OS) by current fluorescent in situ hybridization (FISH; Fig 4A), current FISH plus amplification CKS1B any copies (Fig 4B) and modified FISH with amplification CKS1B ≥ 4 copies (Fig 4C). Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.