Abstract
8 min readMultiple myeloma (MM) is a plasma cell malignancy that arises from 2 precursor conditions: an early stage, called monoclonal gammopathy of undetermined significance (MGUS), and a later one, called smoldering multiple myeloma (SMM).1, 2 Both stages are asymptomatic, but SMM is distinguished by a higher tumor burden, as denoted by bone marrow (BM) infiltration of ≥10% plasma cells or M-protein >3 g/dL, which corresponds to higher risk of progression to overt MM.3 While the progression rate of MGUS is 1%-2% per year, SMM progresses at a rate of 10% per year.4 Nevertheless, not all patients with MGUS or SMM progress to MM, whereby certain SMM cases could behave like MGUS. For these reasons, management of patients with MGUS and SMM is based on progression risk stratification systems, which distinguish patient groups with varying risk of progression to overt MM. As such, SMM can be low, intermediate, or high risk when one, two, or three of the following risk factors, respectively, are present: BM infiltration with ≥10% plasma cells, M-protein >3 g/dL, and free light chain (FLC) ratio <0.125 or >8.5 Similarly, MGUS is divided into low, intermediate, or high risk, if none, one/two, or three of the following risk factors, respectively, are present: non-IgG type of M-protein, M-protein level >1.5 g/dL, and abnormal FLC ratio (<0.26 or >1.65).6 While for SMM the degree of BM infiltration is an established measure of tumor burden that correlates with progression risk,3 it is currently not being used in the stratification of patients with MGUS. In fact, patients with low-risk MGUS are not recommended to undergo a BM biopsy as part of their initial evaluation.7 Defined by the concurrent presence of low M-protein concentration and a normal FLC ratio, low-risk MGUS has a low progression risk of 5% at 20 years,6 which explains why deferral of BM biopsy is currently recommended for this group. These recommendations were based on a prospective cohort of 1384 patients, out of which only 12% had had a BM biopsy performed at diagnosis.8 It is thus possible that MGUS-like SMM cases could be hiding within the un-biopsied low-risk MGUS group. To address this hypothesis, we studied a total of 239 asymptomatic patients with IgG M-protein <1.5 g/dL, who underwent a BM biopsy as part of their initial evaluation at diagnosis. We identified 239 patients with IgG M-protein <1.5 g/dL at the Center for Prevention of Progression of Blood Cancers at Dana-Farber Cancer Institute, USA and the Department of Clinical Therapeutics, Greece, in the period between 1998 and 2016. This cohort is part of prospective tissue banking studies at these institutions. Patients included in this cohort underwent a BM biopsy immediately following the detection of M-protein in their serum and were asymptomatic with absence of CRAB criteria. Patients with IgA and IgM-type disease were excluded, given their established higher risk of progression.8 M-protein was measured by serum protein electrophoresis (SPEP) and confirmed by immunofixation. Time to Progression (TTP) was defined as time elapsed between time of diagnosis and progression to overt MM, as defined by the Revised IMWG criteria.7 Cox proportional hazards regression was used to model TTP, while logistic regression was used to model the presence of an abnormal BM biopsy (≥10% plasma cells). Receiver operating characteristic (ROC) analysis was performed for optimal threshold selection. The median follow-up was 25 months (range: 3-227 months). Since all of our patients had IgG-type disease with an M-spike <1.5 g/dL, their staging depended entirely on the presence or absence of an abnormal FLC ratio (<0.26 or >1.65). Twenty-nine patients did not have FLC ratio assessed at diagnosis and were excluded from the analysis. One hundred seven patients (45%) had a normal FLC ratio, placing them in the low-risk MGUS (LRMGUS) category, while 104 patients (43.3%) had an abnormal FLC ratio. According to the recommendations referenced above, the latter patients would be considered intermediate-risk MGUS (IRMGUS) and a BM biopsy would be ordered for evaluation anyway, allowing for recognition of SMM cases. Based on the BM biopsy findings in this group, 69 (66.4%) of those patients actually had SMM, and of those, 40 (58%) were low-risk (LRSMM) and 29 (42%) were intermediate-risk (IRSMM). More importantly, though, 48 (45%) patients with LRMGUS actually had Smoldering-level BM tumor burden (Supporting Information Figure 1A), which would have been missed had a biopsy not been performed. Although 75% of those patients had infiltration in the range of 10%-20% of cellularity, the distribution was right-skewed, with 4 cases (8%) exhibiting burden that would place them in the intermediate-risk SMM category (Supporting Information Figure 1B). Next, we sought to determine whether the degree of BM infiltration was truly a significant predictor of progression in the LRMGUS group. Indeed, by means of Cox regression, the percentage of BM infiltration was the only significant predictor of progression in this risk group (hazard ratio 1.1, P-value <.001). Therefore, not only did 45% of the LRMGUS population in our cohort actually have Smoldering-level tumor burden in the BM, but that burden was significantly associated with a higher risk of progression, too. This finding demonstrates the significance of performing a BM biopsy in the LRMGUS population. Provided that BM biopsies are not risk-free and cause considerable discomfort to patients, we then searched for a biomarker that could predict an abnormal biopsy (≥10% monoclonal plasma cells) in this population with reasonable sensitivity and specificity to assist in clinical triage. Using logistic regression, we modeled the presence of an abnormal biopsy in the LRMGUS group and determined M-protein level to be a statistically significant predictor (P-value = .008). Based on the distribution of M-spike values among patients with and without an abnormal biopsy, we tested the predictive power of 2 potential cutoffs for M-protein, 0.5 and 0.6 g/dL, and found both to be statistically significant predictors of an abnormal biopsy (Supporting Information Figure 1C). We performed ROC analysis to determine the optimal threshold for our biomarker that would give a sensitivity and specificity of at least 75%. With an area under curve (AUC) of 0.86, we found that a cutoff of 0.6 g/dL for M-protein corresponded to ∼80% sensitivity and ∼76% specificity, with a positive predictive value of ∼77% and a negative predictive value of ∼79.2%; all metrics were superior to those of the 0.5 g/dL threshold (Supporting Information Figure 1D). Based on that, we thus recommend the use of 0.6 g/dL as the M-protein threshold for performing a BM biopsy in the LRMGUS population. In an effort to better understand the cumulative risk of progression of LRMGUS patients with ≥10% BM plasma cells, we attempted to compare it to that of the LRMGUS, the IRMGUS, and the SMM patients in our cohort (Figure 1). Although LRMGUS patients with higher degree of BM infiltration have a higher risk of progression as compared to truly LRMGUS, it appears that their risk level is similar to that of IRMGUS patients in our cohort and remains relatively stable throughout follow-up, in contrast to that of LRSMM which demonstrates a significant increase around the 100th month of follow-up. As such, these patients seem to partly retain the relatively low risk associated with low M-protein concentrations and a normal FLC ratio. Thus, even in the presence of significant tumor burden, which places patients at the SMM stage of disease, an M-protein of <1.5 g/dL and a normal FLC ratio seem to predict an MGUS-like behavior. Although our study's follow-up is not long enough to allow for definitive conclusions, our findings identify a novel risk group, which is distinct from LRMGUS and comprises MGUS-like SMM patients with a ∼20% probability of progression at 10 years. Larger studies with longer follow-up are required to validate the identity of this risk group and comprehensively describe its progression risk. In conclusion, our results suggest that BM biopsy should be performed in patients suspected for LRMGUS, so long as their M-protein is higher than 0.6 g/dL, as it can reveal an as of yet undescribed MGUS-like SMM group of patients with significantly higher probability of progression, while identifying truly low-risk MGUS patients who appear to remain stable throughout follow-up. If these results are confirmed, it could mean that this revised low-risk MGUS group almost exclusively comprises nonprogressors, with important implications for risk stratification and management. This study was supported in part by National Institutes of Health grant (NIH R01 CA 205954 and F32 CA220859) and Multiple Myeloma Research Foundation (MMRF)—Perelman Family Foundation grant for early detection and prevention of Multiple Myeloma. M.B. has advisory role and received honoraria from Takeda. E.K. has received honoraria and research funding from Amgen, Genesis Pharma, Janssen, Takeda, Prothena, M.A.D. has received honoraria from Amgen, Celgene, Janssen, Takeda. E.T. has received honoraria and research funding from Amgen, Genesis Pharma, Janssen, Takeda. K.C.A. has advisory role in Celgene, Takeda, Bristol-Myers Squibb (BMS), Janssen, Sanofi, and Gilead. K.C.A. is a scientific founder of OncoPep and C4 therapeutics. P.G.R. has a consulting role and has received research funding from Karyopharm, Oncopeptides, Celgene, Takeda, Amgen, and Janssen. I.M.G. has a consulting/advisory role and has received research funding from Celgene, Takeda, BMS, Janssen Pharmaceuticals, and Amgen. 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.
Discussion(0)
No comments yet. Be the first to comment.