MGUS indicates monoclonal gammopathy of undetermined significance; MM, multiple myeloma.
eTable 1. Overall survival of MM patients with and without prior knowledge of MGUS
eTable 2. Hazard ratios, P values, and 95% confidence intervals with regard to M-protein concentration at the time of MGUS diagnosis
eTable 3. Overall survival of MM patients with and without prior knowledge of MGUS, restricted to patients diagnosed with MM between 1996 and 2005
Sigurdardottir EE, Turesson I, Lund SH, Lindqvist EK, Mailankody S, Korde N, Björkholm M, Landgren O, Kristinsson SY. The Role of Diagnosis and Clinical Follow-up of Monoclonal Gammopathy of Undetermined Significance on Survival in Multiple Myeloma. JAMA Oncol. 2015;1(2):168-174. doi:10.1001/jamaoncol.2015.23
Multiple myeloma (MM) is consistently preceded by the precursor state, monoclonal gammopathy of undetermined significance (MGUS). The average annual risk of progression from MGUS to multiple myeloma is 0.5% to 1.0%. Current guidelines suggest life-long clinical follow-up of individuals diagnosed as having MGUS depending on risk stratification. The impact of diagnosing and conducting clinical follow-up of MGUS on MM survival is unclear.
To estimate the impact of prior knowledge of MGUS diagnosis and comorbidities on MM survival.
Design, Setting, and Participants
We conducted a population-based study including all patients with MM (MM patients) diagnosed in Sweden (n = 14 798) from 1976 to 2005 (with follow-up until 2007); 394 (2.7%) had previously been diagnosed as having MGUS. Information on comorbidities was gathered for all patients. We calculated survival rates from the time of MM diagnosis, comparing patients with vs those without prior knowledge of MGUS. Using Cox proportional hazards models, we calculated hazard ratios (HRs) and 95% CIs for risk factors for death. χ2 Tests were used to evaluate differences in comorbidities.
Prior knowledge of MGUS among MM patients. In a subanalysis, monoclonal (M)-protein concentration and type were used as exposure.
Main Outcomes and Measures
Risk of death and comorbidities.
Patients with MM with prior knowledge of MGUS had significantly (HR, 0.86; 95% CI, 0.77-0.96; P < .01) better overall survival (median survival, 2.8 years) than MM patients without prior knowledge of MGUS (median survival, 2.1 years), although MM patients with (vs without) prior knowledge of MGUS had more comorbidities (P < .001). Among MM patients with prior knowledge of MGUS, low M-protein concentration (<0.5 g/dL) at MGUS diagnosis was associated with poorer MM survival (HR, 1.86; 95% CI, 1.13-3.04; P = .01).
Conclusions and Relevance
Patients with MM with prior knowledge of MGUS had better MM survival, suggesting that earlier treatment of MM leads to better survival. The observation that a low M-protein concentration at MGUS diagnosis was associated with poorer MM survival may reflect less frequent clinical follow-up. Our observations stress the importance of clinical follow-up in patients with MGUS, regardless of risk stratification.
Multiple myeloma (MM) is a neoplastic disorder characterized by proliferation of plasma cells in the bone marrow, monoclonal (M)-protein in the blood and/or urine, and MM-related organ or tissue dysfunction.1 Population-based studies have shown that survival in MM has improved significantly in recent years,2- 5 mainly owing to the introduction of novel agents (eg, thalidomide, bortezomib, lenalidomide, carfilzomib, pomalidomide), high-dose melphalan with autologous stem cell transplantation in younger patients, and improved supportive care.2- 5
A recent prospective cancer screening trial including over 77 000 individuals followed for more than 10 years showed that MM is consistently preceded by a precursor state, monoclonal gammopathy of undetermined significance (MGUS).6 Most MGUS cases, however, are never diagnosed. MGUS is characterized by a detectable M protein in persons without evidence for end-organ damage or other related plasma cell or lymphoproliferative disorders.7 By screening, MGUS is detected in 2% to 3% of individuals 50 years or older and approximately 5% of persons 70 years or older.8,9 However, only a small proportion of MGUS progresses to a malignant disorder10; in fact, the annual risk of progression to MM or other related disorders is, on average, 0.5% to 1%, with varying risks according to risk groups.10- 13 Current guidelines suggest, depending on the individual patient’s clinical risk score, life-long monitoring of individuals with MGUS to detect progression to MM or related disorders.11,14
At this time, the impact of annual monitoring on the outcome of patients who eventually develop MM is unclear. Because MGUS is typically detected incidentally as part of medical workup for another cause,15 the impact of comorbidity on survival in patients with MM (hereinafter, MM patients) with prior knowledge of MGUS is also unknown. To increase our understanding of the impact of prior knowledge of MGUS diagnosis and comorbidities on MM survival, we performed a large, population-based study using data on more than 14 000 MM patients diagnosed in Sweden from 1976 to 2005, with follow-up through 2007.
•Almost 15 000 patients were diagnosed as having multiple myeloma (MM) in Sweden from 1976 to 2005.
•Three percent of patients had been clinically diagnosed as having MGUS prior to their MM diagnosis.
•Patients with prior knowledge of MGUS had significantly 14% (P < .01) better overall survival from MM (median, 2.8 years) than patients without prior knowledge of MGUS (median, 2.1 years).
•Patients with low M-protein (≤0.5 g/dL) at MGUS diagnosis had poorer survival from MM than patients who had higher M-protein concentration (0.5 to <3 g/dL) at MGUS diagnosis.
•Clinical follow-up in MGUS, regardless of risk stratification, may improve MM outcomes.
All residents of Sweden are, at birth or immigration, assigned a unique national registration number that is used in government-maintained nationwide health care and population registers, whereby medical record linkage is possible with a high degree of accuracy. Each individual’s date of death is centrally registered in the Swedish Cause of Death Registry.
Since the mid-1950s, Sweden has provided universal medical health care for the entire population (currently approximately 9 million people). In contrast to many other countries, for example, the United States (where most hematologic patients are seen and treated primarily by physicians in private practice, outside hospitals), the Swedish health care system has a geographically defined referral structure for specialist assessments. Patients with hematologic disorders are typically diagnosed, treated, and followed clinically by physicians at hospital-based hematology or oncology centers. These centers are affiliated with a few regional university hospitals, which offer inpatient hospital care to a defined primary catchment area in addition to being the hematology and oncology referral center for a larger health care region.
Since 1958, all physicians in Sweden have been required to report each case of cancer that they diagnose to the nationwide Swedish Cancer Register. The Register holds information on diagnosis, sex, date of birth, date of diagnosis, and area and hospital where the diagnosis was made, and has been reported to have more than 95% coverage for MM during the study period.16 Information was gathered on all MM patients who were reported to the Swedish Cancer Registry from 1976 to 2005, with follow-up until 2007. Information on comorbidities was gathered from the Swedish Patient Registry, which captures information on individual patient-based discharge diagnoses and discharge listings from all inpatient care, with a very high coverage.17 Information on date of death was gathered from the Cause of Death Registry.
Because MGUS is generally asymptomatic, it is usually an unexpected finding during a medical workup for another cause. In Sweden, when a clinician detects MGUS in a patient, he or she will typically consult with a hematology specialist at a regional hospital-based center, and, if needed, refer the patient for further workup, especially to rule out an underlying malignant neoplasm. These centers are affiliated with a hospital-based hematology and oncology centers.
The first population-based MGUS screening studies were initiated by Waldenström’s group in Sweden in the early 1960s.18 Indisputably, these early efforts have played an important role and facilitated an increasing awareness of MGUS among Swedish clinicians. In the present nationwide study, MGUS patients diagnosed between the late 1960s and the late 1970s were primarily diagnosed by Waldenström’s group at Malmö University Hospital. During these years, diagnostic criteria were defined by the presence of an M protein in serum in the absence of an underlying lymphoproliferative malignant neoplasm.18 From the early 1980s, efforts have been made, mainly influenced by Kyle’s group at the Mayo Clinic,19 to establish stringent criteria to distinguish MGUS from asymptomatic forms of myeloma and related disorders. MGUS is now defined by the presence of a monoclonal immunoglobulin level of less than 3 g/dL in serum; if bone marrow examination was performed, a plasma cell content of less than 10%; no evidence of other lymphoproliferative disorders; and the absence of clinical manifestations related to the monoclonal gammopathy.7 These criteria are essentially the same as those used at Swedish hospitals during the study period.
The following approaches were applied to establish a nationwide MGUS cohort: first, we retrieved information on all incident patients through our national network, which comprises all outpatient units, including all major regional hospital-based hematology/oncology centers in Sweden. For all MGUS patients, we obtained information on sex, date of birth, date of diagnosis, and region and unit where the diagnosis was made. When available, we also collected information on the MGUS isotype and concentration of the monoclonal spike at diagnosis. Second, we identified all MGUS patients who were reported in the Swedish Patient Registry, which captures information on individual patient-based inpatient and outpatients discharge diagnoses and discharge listings from all inpatient care, with a very high coverage.
Information on all MGUS patients from these 2 sources was merged into 1 master database. Using the nationwide Swedish Cancer Registry, which includes information on all incident cancers diagnosed since 1958 (including date of diagnosis and region and hospital where the diagnosis was made),16 we obtained data on all cancer diagnoses for all MGUS patients. To minimize the influence of misdiagnosis (eg, smoldering myeloma), MGUS patients with a lymphoproliferative malignant neoplasm diagnosed up to 6 months after MGUS were removed from the MGUS cohort. As an additional quality control measure, we removed any MGUS patient with a recorded preceding lymphoproliferative malignant neoplasm.17
To characterize patterns of underlying comorbidities among MM patients, using the Swedish Patient Registry,17 we obtained all inpatient and outpatient discharge diagnoses for autoimmune diseases, infections, nonhematological malignant diseases, ischemic heart disease, heart failure, cerebrovascular diseases, chronic lung diseases, and renal diseases. This study was approved by the Regional Ethics Committee of Stockholm.
Survival from time of MM diagnosis comparing patients with and without prior knowledge of non-IgM MGUS, diagnosed in the period of 1976 to 2005 with follow-up until 2007, was analyzed with Kaplan-Meier method. Risk factors for death were analyzed with a Cox proportional hazards model, wherein the hazards ratios (HRs) were estimated and 95% CIs calculated for the following variables: MM patients with prior knowledge of MGUS compared with MM patients without prior knowledge of MGUS, age at MM diagnosis, year of MM diagnosis, sex, immunologic isotype (IgG and IgA), as well as M-protein concentration. All calculations were adjusted for age of diagnosis, year of diagnosis, and sex. The cohort was divided into 3 equally large calendar periods: 1976 to 1985, 1986 to 1995, and 1996 to 2005.
We conducted extensive sensitivity analyses to exclude the effect of calendar time on the outcome. First, we conducted analyses stratified by calendar time, including only patients diagnosed during the last calendar period (1996-2005). Second, we conducted a nested case-control analysis in which all MM patients with prior knowledge of MGUS (cases) were matched (sex, year of MM of diagnosis ± 1 year, and age at MM diagnosis ± 1 year) with up to 4 MM patients without prior knowledge of MGUS (controls). All sensitivity analyses are provided in the eTables 1 through 3 in the Supplement. Third, we performed 2 analyses using cause-specific survival: one with MM as the primary cause of death and another using MM when listed as any underlying cause of death. Fourth, using Pearson product moment correlation coefficient, the correlation between the number of days from MGUS diagnosis until MM diagnosis and the age at diagnosis of MM was analyzed.
χ2 Tests were used to test for differences in comorbidity patterns among MM patients with prior knowledge of MGUS vs those without prior knowledge. Statistical results were considered significant at P < .05. R 3.0.2 software was used for statistical analysis. Figures were drawn with the package ggplot2,20 and models were fitted with the survival package.21
The study cohort consisted of 14 798 patients diagnosed as having MM in Sweden between 1976 and 2005 and with follow-up until 2007. A total of 394 MM patients had previously been diagnosed as having MGUS. Details of the study cohort are presented in Table 1. In MM patients with prior knowledge of MGUS vs those without prior knowledge, the median ages at MM diagnosis were 73 and 72 years, respectively (range, 19-101 years). Among MM patients with prior knowledge of MGUS, the median M-protein concentration at MGUS diagnosis was 1.2 g/dL (range, 0.1-2.9 g/dL).
The MM patients with prior knowledge of MGUS had significantly (HR, 0.86; 95% CI, 0.77-0.96) better survival (median survival, 2.8 years; 95% CI, 2.6-3.3 years) than those without prior knowledge (median survival, 2.1 years; 95% CI, 2.1-2.2 years) (Table 2). Survival of MM patients with prior knowledge of MGUS vs those without is shown in the Figure.
Among MM patients with prior knowledge of MGUS, an M-protein concentration of less than 0.5 g/dL at MGUS diagnosis was associated with poorer survival than having a concentration of 0.5 g/dL or greater (HR, 1.86; 95% CI, 1.13-3.04; P = .01) (Table 2). There was no difference in survival when monoclonal isotypes were compared (IgG vs IgA). The median time from MGUS to MM diagnosis was shorter for those MGUS patients with a higher M-protein concentration; for patients with an M-protein concentration of 0.5 g/dL or greater, the median time was 4.8 years, whereas it was 9.1 years for patients with M-protein concentrations of less than 0.5 g/dL.
In sensitivity analyses restricted to patients diagnosed as having MM after 1996, the results were essentially the same as the main model, with MM patients with prior knowledge of MGUS having significantly better survival than MM patients without prior knowledge of MGUS (HR, 0.81; 95% CI, 0.70-0.93). Similarly, in the sensitivity analysis based on a nested case-control design (see the Methods section), the results were similar to those of the main analysis (eTables 1-3 in the Supplement). Furthermore, in analysis using cause-specific survival (instead of overall survival) as the outcome, the risk of dying was lower in MM patients with prior knowledge of MGUS compared with MM patients without prior knowledge of MGUS; the HR was 0.75 (95% CI, 0.64-0.88) for patients with MM listed as a primary cause of death, and the HR was 0.79 (95% CI, 0.69-0.91) for patients with MM listed as an underlying cause of death. The correlation between the number of days from MGUS diagnosis until MM diagnosis and the age at diagnosis of MM was low (0.003) and nonsignificant (P = .94).
In models focusing on the impact of comorbidities at diagnosis of MM in relation to prior knowledge of MGUS, we found autoimmune diseases, infections, nonhematological malignant diseases, ischemic heart disease, heart failure, cerebrovascular diseases, and renal diseases to be significantly more common in MM patients with prior knowledge of MGUS (P < .001 for all comparisons) (Table 3). Patterns of comorbidity were similar independent of M-protein concentration at MGUS diagnosis (data not shown).
The hypothesis that detection and follow-up of MGUS may influence survival in MM patients is unlikely to ever be tested in a prospective clinical study owing to the large sample size required, long follow-up time, and consequent extreme costs. Indeed, the current population-based study, including all patients diagnosed as having MM in Sweden from 1976 to 2005, linked with a nationwide clinical MGUS database, provides a unique opportunity to address this clinically important question. Based on a nationwide population-based cohort study including almost 15 000 MM patients, we found MM patients with prior knowledge of MGUS diagnosis to have better survival than patients diagnosed as having MM in the absence of prior MGUS knowledge. These findings were true even though MM patients with prior knowledge of MGUS (typically being detected incidentally as part of medical workup for another cause) had significantly higher prevalence of comorbidities than other patients. Quite counterintuitively, MM patients with prior knowledge of MGUS with low M-protein concentration (<0.5 g/dL) had significantly worse survival than those with higher M-protein concentrations (≥0.5 g/dL). We speculate that this may be reflective of current guidelines11 suggesting less frequent monitoring of MGUS patients with lower M-protein concentrations. Our findings are striking and of clinical relevance because they demonstrate that, despite the lack of treatment options for patients with MGUS, the diagnosis and long-term follow-up of MGUS prolongs survival of patients with MM.
Current clinical guidelines suggest life-long monitoring of individuals diagnosed as having MGUS,11,14 but how it influences outcomes remains largely unknown.22 Interestingly, a recent study by a Spanish group23 showed that treating high-risk smoldering myeloma resulted in improved progression-free and overall survival. We speculate that the reason for the prolonged survival observed in our study most likely reflects the fact that MGUS patients are evaluated more often for signs of MM progression and may be diagnosed and started on antimyeloma therapy at an earlier stage. This argues for early treatment approaches in MM23 and raises the question of whether systematic screening for MGUS should be initiated. Alternatively, those with more aggressive disease may have a shorter duration of MGUS and are less likely to be diagnosed and identified during the MGUS state. Although, on a biological note, MM is consistently preceded by MGUS, most individuals are unaware of their MGUS diagnosis. This is due to the fact that MGUS typically is asymptomatic. Consequently, in the absence of population screening, most MM patients will be diagnosed owing to their symptoms rather than biochemical progression from MGUS to myeloma. Our results need to be confirmed by prospective studies.
Based on prior studies, the annual risk of progression from MGUS to MM is, on average, 0.5% to 1%.10 However, recent risk models show a varying degree based on 4 risk factors: isotype, M-protein concentration, free light chain, and immunoparesis.10,12 Interestingly, in our study, we found that among MM patients with a prior knowledge of MGUS, those individuals with an M-protein concentration of less than 0.5 g/dL at MGUS diagnosis had worse MM survival. Based on the fact that current guidelines suggest that low-risk (ie, a low concentration of M protein on serum protein electrophoresis test) MGUS require less frequent monitoring.11 The reasons for this are unclear and need to be studied further. We have speculated that low-risk MGUS patients might be lost to follow-up or followed less intensely. Most low-risk patients will never develop MM, but according to our findings those who do develop MM seem to fare worse than high-risk MGUS patients who develop MM. Our findings are consistent with those of a prior smaller study22 of 116 MM patients seen at the Mayo Clinic between 1973 and 2004. That study showed that low-risk MGUS patients are more often diagnosed as having serious complications compared with high-risk MGUS patients. Although the risk of progression to MM is very low in individual patients, our findings suggest that life-long surveillance is important in MGUS patients independent of risk score.
As expected, we found that MM patients with prior knowledge of MGUS had significantly greater prevalence of comorbidities than other MM patients. These results most likely reflect the fact that MGUS is most often diagnosed during workup for an unrelated disease.24 Previous studies have shown that comorbidity is a poor prognostic factor in the general population, with increasing mortality as number of comorbidities rises.25,26 Despite our observation of a heavier burden of comorbidities, MM patients with previous knowledge of MGUS had significantly better survival than MM patients without prior knowledge of MGUS. Given this paradoxical difference, it can be argued that our study underestimates the true survival benefit of diagnosing and conducting clinical follow-up of MGUS. The impact of comorbidity in patients with MM and MGUS needs to be studied further.
Our study has several strengths. This is a register-based cohort study, which ensured a population-based setting and generalizability of the results. According to prior validation study, the Swedish Cancer Registry has a very high rate of completeness and diagnostic accuracy.16 All information was registered before and independent of this study. In addition, all information was recorded prospectively in the database. As described in the Methods section, in Sweden, most individuals diagnosed as having an M protein are referred for further evaluation at specialized hematology centers. Based on the study design, we believe that the MGUS cohort in this investigation is quite representative of all individuals diagnosed as having MGUS and who subsequently developed MM in Sweden during the study period. Still, we cannot exclude the possibility that some general practitioners or other specialists diagnose MGUS without any further evaluation. Given the difference in the date of MGUS diagnosis, with most patients diagnosed in the last calendar period, we wanted to rule out the possibility that the findings were affected by superior survival in MM during that calendar period. Consequently, we performed extensive sensitivity analysis, including cause-specific survival, and the results were essentially the same. We therefore feel confident that the observed findings are robust. Our study also has some limitations. We did not have information from individual medical records. Therefore, we did not retrieve any information concerning treatment or other factors known to influence survival. Type and concentration of M proteins were available only for a proportion of the study cohort. In our study, individuals were identified as MM patients whenever they met the diagnostic criteria for MM. The proportion of patients diagnosed as having smoldering MM is unknown, and, because MGUS patients are typically diagnosed during workup for an unrelated condition, it is possible that they are more thoroughly investigated, and thus lead-time bias is a concern. However, median age at diagnosis is similar in both cohorts, which suggests minimal lead time bias. In addition, owing to lack of information on actual costs, we were unable to conduct cost-analyses designed to evaluate the financial burden of MGUS follow-up.
We found that despite a higher prevalence of comorbid conditions, MM patients with prior knowledge of MGUS have a survival superior to that of MM patients without prior knowledge. Furthermore, among MM patients with prior knowledge of MGUS, we found that a low M-protein concentration (<0.5 g/dL) was associated with shorter MM survival. This may be due to less intense monitoring in low-risk MGUS and needs to be studied further. Our results reflect the importance of life-long follow-up for individuals diagnosed as having MGUS, independent of risk score, and highlight the need for better risk models based on the biology of the disease. Patients should receive balanced information stressing not only the overall very low risk of progression to malignant neoplasm but also the symptoms that could signal such development and the need to consult their physician. The higher prevalence of comorbid conditions in MM patients with prior knowledge of MGUS supports the finding that MGUS most often is diagnosed during follow-up for unrelated conditions. Our findings raise the question whether screening for MGUS in the general population could translate into earlier detection and treatment of MM and lead to better MM survival.
Corresponding Author: Sigurdur Y. Kristinsson, MD, PhD, Faculty of Medicine, University of Iceland, Stapi V/Hringbraut, 101 Reykjavik, Iceland (firstname.lastname@example.org).
Accepted for Publication: January 6, 2015.
Published Online: March 5, 2015. doi:10.1001/jamaoncol.2015.23.
Author Contributions: Dr Kristinsson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Sigurdardottir, Lund, Korde, Björkholm, Landgren, Kristinsson.
Acquisition, analysis, or interpretation of data: Sigurdardottir, Turesson, Lund, Lindqvist, Mailankody, Landgren, Kristinsson.
Drafting of the manuscript: Sigurdardottir, Landgren, Kristinsson.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Sigurdardottir, Lund, Lindqvist.
Obtained funding: Landgren, Kristinsson.
Administrative, technical, or material support: Lindqvist, Korde, Kristinsson.
Study supervision: Björkholm, Kristinsson.
Conflict of Interest Disclosures: None reported.
Funding/Support: This research was supported by grants from the Swedish Cancer Society, the Karolinska Institutet Foundations, Marie-Curie CIG, the University of Iceland Research Fund, Icelandic Centre for Research (RANNIS), and Landspitali University Hospital Research Fund.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Information: Drs Landgren and Kristinsson contributed equally to this study.