Post-MGUS diagnosis serum monoclonal-protein velocity and the progression of monoclonal gammopathy of undetermined significance to multiple myeloma

Abstract

Background

Multiple myeloma (MM) is a common hematologic malignancy consistently preceded by monoclonal gammopathy of undetermined significance (MGUS). Little is known about post-diagnosis clinical predictors of progression of MGUS to MM to guide MGUS management. This study aimed to investigate whether the rate of rise in serum monoclonal protein concentration during the year after MGUS diagnosis – M-protein velocity – predicts progression of MGUS to MM.

Methods

Data from the U.S. Veterans Health Administration system were used. A retrospective cohort of MGUS patients who progressed to MM were matched on age at MGUS diagnosis and race in a 1:4 ratio to the MGUS patients using incidence density sampling. Kaplan-Meier curves were plotted. Univariable and multivariable conditional logistic regression analyses were fitted from the matched risk sets.

Results

A total of 128 cases and 490 matched controls were included. The case group contained a higher percentage of patients with M-protein velocity >0.1 g/dL/year than the control group (44.5% versus 28.2%, p<0.0001). M-protein velocity of >0.1 g/dL during the year following MGUS diagnosis was positively associated with progression of MGUS to MM (multivariable-adjusted odds ratio=2.15; 95% confidence interval=1.37 to 3.35).

Conclusion

Patients with a positive M-protein velocity during the year after MGUS diagnosis may be considered for more frequent monitoring for early detection and timely treatment of MM. Future prevention studies could target these patients for intervention evaluation.

Impact

Our results suggest a new clinical predictor of progression to MM following MGUS diagnosis, which has potential to identify high-risk patients for management and prevention.

INTRODUCTION

Multiple myeloma (MM) is one of the most common hematologic malignancies in the United States. In 2018, MM accounts for 12,770 deaths, and 30,770 new MM cases are expected (1). MM is characterized by clonal bone marrow plasma cells ≥10% or biopsy-proven bony or extramedullary plasmacytoma and evidence of end organ damage; or the presence of at least one of the following: ≥60% clonal plasma cells on bone marrow examination, serum involved/uninvolved free light chain ratio of ≥100, provided the absolute level of the involved light chain is ≥100 mg/L, or >1 focal bone lesion ≥5 mm in size (1).

MM is consistently preceded by monoclonal gammopathy of undetermined significance (MGUS) (2,3), a pre-malignant disorder defined by the presence of serum monoclonal protein (M-protein) of ≤3 g/dL, <10% bone marrow monoclonal plasma cell infiltrate, and the absence of end organ damage (4). The prevalence of MGUS in the population age ≥50 is ~3% (5) with a 1% annual risk of progression to more advanced diseases, including MM (6). Patients with MGUS are asymptomatic and a diagnosis of MGUS does not currently warrant treatment. Management of MGUS is restricted to monitoring for disease progression (7,8). For patients with a measurable clonal immunoglobulin (Ig), some have recommended 2–3 serum protein electrophoresis tests for the first year after diagnosis and then one test every 2–3 years for low-risk patients or annually for intermediate and high risk patients as long as there are no symptoms suggestive of progression (7,8).

Previous studies reported that serum M-protein concentration ≥1.5 g/dL at MGUS diagnosis, Ig isotype other than IgG, an abnormal serum-free light-chain ratio, proportion of bone marrow aberrant plasma cells within the bone marrow plasma cell compartment ≥95% (6,9), and reduced levels of one or two non-involved Ig isotypes (10) are associated with progression of MGUS to MM. Therefore, patients with MGUS who have one or more of these factors are classified as intermediate (1–2 risk factors) or high (≥3 risk factors) risk of progression and should be monitored annually for life (7,8). However, to date, no studies have provided clear evidence that dynamic factors measured after MGUS diagnosis can refine the prognostication, and ultimately management of MGUS. Since serum M-protein at MGUS diagnosis is used as a marker of disease burden in patients with MM (6), it is logical that increasing M-protein levels could foreshadow progression to MM. Thus, a better understanding of M-protein level changes following MGUS diagnosis could provide insights into the expected natural history of patients with MGUS.

The objective of this study is to investigate whether the rate of rise in M-protein concentration – M-protein velocity – during the year following MGUS diagnosis can predict the progression of MGUS to MM in patients diagnosed with MGUS.

MATERIALS AND METHODS

Data, study population and design

We used the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes to identify MGUS (273.1) and MM (203.0) diagnoses in the United States Veterans Health Administration (VHA) system database. To confirm MM diagnosis, we further used the International Classification of Diseases for Oncology code 9732/3 to identify patients in the Veterans Affairs Central Cancer Registry, including only those who received MM treatment within 6 months of MM diagnosis. Furthermore, patient charts were reviewed to verify MGUS and MM diagnoses following the criteria defined by the International Myeloma Working Group (1,4). During chart review, MGUS and MM diagnoses as well as the date of diagnoses were confirmed; furthermore, all available levels of M-protein concentration, Ig isotype as determined by immunofixation electrophoresis, and dates of M-protein measurements were obtained. Two reviewers (JG and TST) independently reviewed patient charts, abstracted data, and resolved disagreements by consensus.

We identified 9,287 patients with ≥2 ICD-9-CM codes for MGUS diagnosis between 10/1/1999 and 12/31/2009 in all 21 regional VHA districts throughout the United States (Fig 1). The date of the first MGUS diagnosis was obtained through data abstraction. These patients were followed through 8/6/2013. Among them, 617 patients developed MM. We excluded (i) 54 patients whose M-protein at MGUS diagnosis was unknown or measured before 10/1/1998; (ii) 52 patients whose M-protein at MGUS diagnosis was >3.0 g/dL, as these patients met criteria for smoldering multiple myeloma (SMM); (iii) 34 patients whose MGUS type was light-chain or Ig isotype other than IgG or IgA, because light-chain only disease cannot be measured by serum protein electrophoresis, IgM typically does not progress to MM (10,11), and IgD is rare (12); (iv) 194 patients, who had no M-Protein measurement within 14 months post-MGUS diagnosis except that at MGUS diagnosis and; (v) 119 patients whose MM diagnosis was <2 years following MGUS diagnosis; and (vi) 36 patients, whose last M-protein was measured <6 months after MGUS diagnosis. We applied the same exclusion criteria on patients without a MM diagnosis, except for (v), for which we excluded patients who died or were censored <2 years after MGUS diagnosis. Patients with MM were then matched on age at MGUS diagnosis (≤65, >65) and race (white, black, other, unknown) in a 1:4 ratio to the MGUS patients with or without a diagnosis of MM using incidence density sampling (13). In this sampling scheme, controls to an index case are selected with replacement from all patients at risk excluding the index case itself (including patients diagnosed with MM at later times) at the event time of the index case (13). Last, in the matching process, 9 cases were only able to match to <4 controls, resulting in 128 cases and 490 matched controls (204 patients with MM and 286 patents without MM) or 396 unique patients for the subsequent analyses.

Figure 1. Consort diagram for the matched cases and controls.

*1:4 matching on age at MGUS diagnosis and race using incidence density sampling. In this sampling scheme, controls to an index case are selected with replacement from all patients at risk (including patients diagnosed with MM at later times) at the event time of the index case, excluding the index case itself. Therefore, the selected controls can include patients who developed MM later. The analytic cohort included 396 unique patients (128 unique cases and 490 controls).

Unique patient identifiers were used to obtain data on sex, race, height, weight, and comorbidities. The Romano adaptation of the Charlson comorbidity index was calculated based on comorbidities present before MGUS diagnosis (14). We computed body mass index (BMI) at MGUS diagnosis, weight in kilograms divided by the square of height in meters, and categorized as underweight (BMI <18.5), normal-weight (BMI 18.5–24.9), overweight (BMI 25–29.9), or obese (BMI ≥30) (15). We used the most frequently measured height for each patient and the weight measured one month before or after MGUS diagnosis, favoring the value closest to the date of MGUS diagnosis (16).

Institutional Review Boards at both Washington University School of Medicine and Veteran Affairs Saint Louis Healthcare System approved the study.

Exposure: 1-y post-MGUS diagnosis M-protein velocity

We defined exposure as 1-y post-MGUS diagnosis M-protein velocity. M-protein velocity was computed as the slope between the M-protein concentration values from the first and the last measurements during the year following MGUS diagnosis (g/dL/year), as follows:

$$\text{Velocity}=\frac{\text{Value}\text{at}\text{the}\text{last}\text{measurement}-\text{Value}\text{at}\text{MGUS}\text{diagnosis}}{\text{Months}\text{from}\text{MGUS}\text{diagnosis}\text{to}\text{last}\text{measurement}}\times 12.$$

M-protein velocities were further categorized into >0.1 and ≤0.1 g/dL/year. The cutoff of 0.1 was determined using all cases and controls (17–19), because it yielded the maximum of Youden’s J statistic (i.e., the sum of sensitivity and specificity) (17).

Outcome measures

For Kaplan-Meier analyses, the outcome measure was time from MGUS diagnosis to MM diagnosis, when present. The date for the first MM diagnosis from data abstraction was used as the date of MM diagnosis. Patients without a MM diagnosis were censored at death or 08/06/2013, whichever came first. For univariable and multivariable conditional logistic regression analyses, the outcome measure was whether a patient had a MM diagnosis.

Statistical analyses

Summary statistics of the demographic and clinical characteristics stratified by matched cases and controls were computed. To compare patients who progressed to MM and patients who did not among the 396 unique patients, we used chi-square tests to examine differences in proportions for categorical variables; for continuous variables, we used student t tests to examine differences in means and Wilcoxon signed-rank tests to examine differences in medians.

Kaplan-Meier curves were plotted on 396 unique patients to compare the progression time to MM between patients with velocity ≤0.1 g/dL/year and patients with velocity >0.1 g/dL/year. A stratified log-rank test for the matched cases and controls was performed to detect the statistical difference between the two groups.

Univariable and multivariable conditional logistic regression analyses fitted from the matched risk sets were performed (20). We included the following covariates: 1-y post-MGUS diagnosis velocity (>0.1, ≤0.1 g/dL/year), Ig isotype (IgA, IgG), sex (male, female), BMI group (underweight, normal-weight, overweight, obese, unknown), age, Charlson comorbidity index, and M-protein concentration at MGUS diagnosis. We also used 1-y post-MGUS diagnosis velocity as a continuous covariate in both univariable and multivariable conditional logistic regression analyses.

All tests were two-sided. Statistical significance was determined by an alpha level of 0.05. All statistical analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC).

RESULTS

This study included 396 unique patients diagnosed with MGUS between 10/01/1999 and 12/31/2009 (128 cases and 490 controls, due to the inclusion of patients with MM in the selected controls). Mean age at MGUS diagnosis was 67 years in both cases and controls (Table 1). Patients were predominantly male (cases: 96.1%; controls: 96.9%). More than 60% of the patients were white (cases: 60.2% white, 29.7% black; controls: 61.4% white, 30.6% black) and either overweight or obese (cases: 31.3% overweight, 33.6% obese; controls: 39.0% overweight, 28.2% obese). Although we matched on race, proportion of white and black patients are not the same for cases and controls due to the 9 cases that were able to match fewer than 4 controls. Mean Charlson score at MGUS diagnosis was 2.7 for cases and 3.0 for controls. Mean M-protein concentration at MGUS diagnosis was 1.3 g/dL for cases and 0.8 g/dL for controls. M-protein velocity >0.1 g/dL during the first year following MGUS diagnosis was seen in 44.5% of the cases and 28.2% of the controls. A majority of patients had IgG (cases: 80.5%; controls: 87.6%). Finally, mean follow-up was shorter for cases than controls (60.2 versus 96.3 months). All of the aforementioned variables were not statistically significantly different between patients who developed MM and patients who did not, except for mean M-protein concentration at MGUS diagnosis (p <0.0001), M-protein velocity >0.1 g/dL during the first year following MGUS diagnosis (p=0.0011), Ig isotype (p=0.0092), and mean follow-up (p <0.0001).

Table 1.

Demographic and clinical characteristics by matched 128 cases and 490 controls among 396 unique U.S. veterans diagnosed with MGUS between October 1, 1999 and December 31, 2009

	Overall	Cases	Controls	p-valueb
Variable\Sample size	396a	128	490
Age at MGUS diagnosis, mean (std) years	66.9 (9.6)	66.8 (9.9)	67.1 (9.0)	0.8034†
Year of MGUS diagnosis (median)	2004	2004	2004	0.7098§
Male (%)	96.2	96.1	96.9	0.6153*
Race (%)				0.9320*
White	59.3	60.2	61.4
Black	31.3	29.7	30.6
Other	0.8	1.6	0.2
Unknown	8.6	8.6	7.8
BMI group (%)				0.4584*
Underweight	1.0	0.0	1.4
Normal-weight	17.9	17.2	16.5
Overweight	34.3	31.3	39.0
Obese	30.6	33.6	28.2
Unknown	16.2	18.0	14.9
Comorbidities, mean Charlson score (std)	3.0 (2.8)	2.7 (2.6)	3.0 (2.8)	0.1734†
Ig isotype (%)				0.0092*
A	13.1	19.5	12.5
G	86.9	80.5	87.6
M-protein velocity during the year after MGUS diagnosis (%)				<0.0001†
≤0.1 g/dL/year	66.7	55.5	71.8
>0.1 g/dL/year	33.0	44.5	28.2
Serum M-protein concentration at MGUS diagnosis, mean (std) g/dL	0.9 (0.7)	1.3 (0.7)	0.8 (0.6)	0.0011†
Follow-up, mean (std) months	83.2 (36.9)	60.2 (29.5)	96.3 (35.2)	<0.0001†

Notes: MGUS, monoclonal gammopathy of undetermined significance; BMI, body mass index; Ig, immunoglobulin.

^aIncidence density sampling selects controls with replacement from all people (including patients with a diagnosis of MM) at risk at the time of case occurrence, excluding the index case itself, so controls include patients with MM. Therefore, the analytic cohort included only 396 unique patients (128 unique cases and 490 controls). Also see Fig 1 for the consort diagram.

^bStatistical tests (^†student-t-test; ^§Willcoxon-rank-sum test; *Ch-square test) were performed to compare differences between patients who progressed to MM and patients who did not among the 396 unique patients.

The Kaplan-Meier curves (Fig 2) show a significantly higher proportion of progression among patients with M-protein velocity >0.1 g/dL/year (dashed curve) with median time to MM of 119 months, compared to patients with M-protein velocity ≤0.1 g/dL/year (solid curve) with median time not reached. Among patients who progressed to MM, the median time to MM for patients with M-protein velocity >0.1 g/dL/year was 44 months, and the median time to MM for patients with M-protein velocity ≤0.1 g/dL/year was 59 months. The stratified log-rank test, p<0.0001, for the matched cases and controls showed a statistically significant difference between the two groups.

Figure 2. Kaplan-Meier curves for progression to MM on 396 unique patients with MGUS by 1-y Post-MGUS diagnosis M-protein velocity.

Horizontal axis label: Months from MGUS diagnosis

Vertical axis label: Proportion without progression to MM

+ Censored

––– Post-MGUS diagnosis M-protein velocity ≤0.1 g/dL/year

----- Post-MGUS diagnosis M-protein velocity >0.1 g/dL/year

Stratified log-rank test: p<0.0001

Table 2 presents crude odds ratios (ORs) and multivariable-adjusted ORs (aORs) from the univariable and multivariable conditional logistic regression, respectively. For the multivariable analyses, M-protein velocity was dichotomized (left) or included as a continuous variable (right). In the multivariable analysis using dichotomized M-protein velocity (concordance index=0.8475), M-protein velocity of >0.1 g/dL increase in the first year following MGUS diagnosis was positively associated with progression of MGUS to MM (aOR=2.15; 95% confidence interval, CI=1.37 to 3.35; p=0.0008). A higher level of M-protein concentration at MGUS diagnosis (aOR=5.03; 95% CI=3.46 to 7.33; p<0.0001 per 1 g/dL increase) and IgA relative to IgG isotype (aOR=4.11; 95% CI=2.25 to 7.52; p<0.0001) were positively associated with progression to MM. BMI group, age, and Charlson comorbidity index at MGUS diagnosis were not statistically significantly associated with progression to MM. In the multivariable analysis using M-protein velocity as a continuous variable (concordance index=0.8455), each 1 g/dL increase in M-protein velocity during the year following MGUS diagnosis was associated with a 115% increase in the odds of progression to MM (aOR=2.15; 95% CI=1.24 to 3.72; p=0.0066 per 1 g/dL increase).

Table 2.

Univariable and multivariable-adjusted odds ratios for developing multiple myeloma among the matched 128 cases and 490 controls.

	Univariable matched analyses			Multivariable matched analyses
Covariate	OR (95% CI)	p-value	aOR (95% CI)	p-value	aOR (95% CI)	p-value
1-y Post-MGUS diagnosis M-protein velocity (g/dL/year)					2.15 (1.24 to 3.72)	0.0066
≤0.1 g/dL/year	1.00 (referent)	--	1.00 (referent)	--	--	--
>0.1 g/dL/year	2.48 (1.71 to 3.60)	<0.0001	2.15 (1.37 to 3.35)	0.0008	--	--
Baseline serum M-protein (1 g/dL)	4.20 (3.01 to 5.83)	<0.0001	5.03 (3.46 to 7.33)	<0.0001	5.09 (3.50 to 7.40)	<0.0001
Ig isotype
G	1.00 (referent)	--	1.00 (referent)	--	1.00 (referent)	--
A	1.85 (1.15 to 2.98)	0.0119	4.11 (2.25 to 7.52)	<0.0001	3.88 (2.14 to 7.03)	<0.0001
BMI group
Normal-weight	1.00 (referent)	--	1.00 (referent)	--	1.00 (referent)	--
Overweight	0.92 (0.54 to 1.56)	0.7606	0.79 (0.42 to 1.47)	0.4524	0.80 (0.43 to 1.50)	0.4910
Obese	1.36 (0.80 to 2.32)	0.2591	1.49 (0.79 to 2.84)	0.2217	1.51 (0.80 to 2.85)	0.2088
Sex
Male	1.00 (referent)	--	1.00 (referent)	--	1.00 (referent)	--
Female	0.69 (0.24 to 1.97)	0.4915	0.34 (0.08 to 1.33)	0.1202	0.38 (0.10 to 1.45)	0.1555
Comorbidity score	0.96 (0.90 to 1.03)	0.2094	1.02 (0.94 to 1.11)	0.6414	1.02 (0.94 to 1.11)	0.5799
Age	0.99 (0.92 to 1.06)	0.7454	1.00 (0.92 to 1.09)	0.9925	1.00 (0.92 to 1.08)	0.9464
C-index			0.8472		0.8455

Notes: MGUS, monoclonal gammopathy of undetermined significance; OR, odds ratio; aOR, multivariable-adjusted odds ratio; CI, confidence interval; --, reference group, estimate not available

DISCUSSION

To our knowledge, this study is the first to examine the association between M-protein velocity during the first year following MGUS diagnosis and progression to MM. We observed that independent of the known risk factors, a positive M-protein velocity during the year after MGUS diagnosis was associated with a >2 fold increase in the odds of progression to MM. This finding is important in that it identifies a new clinical predictor of progression to MM following MGUS diagnosis, in addition to the existing ones.

Our finding is partially consistent with that of a previous study, which observed that the evolutionary pattern of M-protein during the first three years of follow-up is the most important risk factor for disease progression in 359 MGUS patients in a single institution (21). Although similar in the conclusion that rising serum M-protein after MGUS diagnosis is an important predictor of the disease progression, our study improves upon their study in several ways. First, they categorized MGUS into non-evolving MGUS versus evolving MGUS, defined as a progressive increase in the M-protein concentration in each of the annual consecutive measurements during the first three years after MGUS diagnosis, while we computed velocity based on M-protein concentration measured in the first year following diagnosis to provide a more clinically practical and meaningful period to observe an indicative pattern of M-protein. Second, their outcome, although not explicitly defined, appeared to be time from 4^th year post-diagnosis to MM diagnosis/censoring to any transformation to a symptomatic disease, including MM and Waldenström macroglobulinemia, while the outcome of our study focused on progression of MM only. As transformation increases with age/time and may increase exponentially for patients of the evolving type, they found that evolving type is “the most important” risk factor for progression; while we found a positive velocity in the first year following diagnosis as “a risk factor among several risk factors”. Interestingly, a much lower hazard ratio (HR=5.1; 95% CI=3.4 to 7.6 (22), relative to previously reported 12.1; 95% CI=5.8 to 25.4 (21)) for evolving type (defined as a progressive increase of ≥10% in the M-protein during the year following SMM diagnosis when M-protein at MGUS diagnosis was ≥3 g/dL or a progressive increase in M-protein in each of the annual measurements when M-protein at MGUS diagnosis was <3 g/dL) was found in a recent study of 206 patients diagnosed with SMM in a single institution in Spain (22). Despite this, several studies pointed out that “evolving MGUS” potentially could be a marker for an early MM with a slow rate of progression, although this has not been confirmed (3,23). Our study supports this. Furthermore, comparable to the finding of a study that evaluated stored biospecimens in patients later diagnosed with MM, which found that ~half of the MM patients had year-by-year increase in M-protein before MM diagnosis, while the other half had a stable pattern (3), our study finds that ~45% of the cases had a positive velocity during the first year of MGUS diagnosis, while 55% did not.

Past studies concluded that M-protein concentration ≥1.5 g/dL at MGUS diagnosis is the most important risk factor of progression of MGUS to MM (8,11). However, no predictor has been found beyond the diagnosis of MGUS to further guide the management of MGUS. Our study found that a positive M-protein velocity during the year after MGUS diagnosis significantly impacted progression of MGUS to MM, in addition to M-protein concentration at diagnosis. Our study enhances the current risk stratification and has two clinical implications. First, MGUS patients with a positive M-protein velocity during the first year following MGUS diagnosis should consider a more frequent monitoring plan, regardless of the level of M-protein concentration at diagnosis. This could lead to a more timely diagnosis and earlier treatment of MM, potentially resulting in better MM survival (24). Second, with advancement in the potential treatments in high-risk SMM patients (25,26) and possibly in high-risk MGUS patients (12,27), M-protein velocity could help identify the high-risk MGUS patients appropriate for intervention. Finally, if a preventive treatment were available for MGUS patients, it is conceivable that reduction in M-spike velocity (or reduction of M-spike concentration in general) could serve as a surrogate marker to measure the efficacy of such a therapy.

M-protein level is currently used to measure response and progression in patients with active MM, as it is highly correlated with disease burden. Similarly, M-protein velocity is associated with progression of MGUS to MM, a process generally associated with expansion of clonal plasma cells. Currently, prospective studies of MGUS or low-risk SMM are not viable due to the low event rate, which requires large numbers of patients followed for a long period of time to achieve adequate statistical power. Thus, there is significant value in the identification of an easily measured biomarker that could be considered a surrogate for efficacy in the prevention setting.

Our study is somewhat different from studies in prostate cancer that used prostate-specific antigen (PSA) velocity during the year before a prostate cancer diagnosis as a prognostic marker (28,29). They found that men with increasing PSA levels of >2.0 ng/mm in the year before diagnosis had a higher risk of death from prostate cancer despite undergoing radical prostatectomy (29) or external beam radiation therapy (28). While there is value in using biomarkers to improve prognostication, predictive biomarkers offer greater value as there is potential for intervention before the devastating disease.

Our study has several strengths. Through patient chart review, we verified MGUS and MM diagnoses and the dates of diagnoses by reviewing patients’ charts, instead of only relying on ICD-9-CM codes and diagnosis dates in the administrative database to ensure accuracy of the data. Moreover, through chart review, we were able to determine MGUS subtype and levels of serum M-protein concentration. Furthermore, to avoid predicting MM diagnosis using levels of M-protein measured too close to the development of MM, we only used those measured ≤1y post-MGUS diagnosis and excluded patients whose MM diagnosis was <2 years following MGUS diagnosis. Last, despite the relatively uncommon outcome of MM, the large number of patients in the national VHA database provided sufficient statistical power to form cases and controls to study this association.

Our study also has limitations. First, patients served by the VHA are frequently from older age, predominately male sex, and lower socioeconomic background (30–32). If these characteristics of the population influence the diagnoses of MGUS or MM, then our conclusion may not be generalizable to broader populations with MGUS. Second, we noticed a lower mean comorbidity score at MGUS diagnosis in the cases than in the controls, although not statistically significant. Our estimates could overestimate the magnitude of the true association between M-protein velocity and transformation to MM, because more comorbidities among controls could result in a higher death rate, and thus a higher likelihood of censoring due to death before MM diagnosis. For this, we would expect the odds ratio for comorbidity scores to be <1. Nonetheless, the results do not show any evidence of this. Last, although we designed our study to minimize potential biases, some sources of bias from unmeasured confounders could remain. For example, incidence-prevalence bias may exist. Including patients whose MGUS diagnosis was not truly incident (due to the fact that patients with MGUS are asymptomatic) may overestimate the true association of velocity. This happens because the computed velocity is based on the year following MGUS diagnosis rather than the year following MGUS incidence. If MGUS had already progressed to MM during the year after MGUS diagnosis, then the M-protein velocity would be higher due to MM. falsely increasing the estimated association would be stronger. Conversely, excluding patients with a diagnosis of MM ≤2 years after MGUS diagnosis may underestimate the true association, because these patients may represent the patient subpopulation with highest risk of disease progression.

CONCLUSION

Our study demonstrated that an M-protein velocity >0.1 g/dL during the year following MGUS diagnosis is associated with a higher risk of progression to MM. More frequent monitoring could be considered in these patients to allow for early detection and timely diagnosis of MM. Future studies could target these patients for evaluation of preventative interventions.