|Home | About | Journals | Submit | Contact Us | Français|
SY Kristinsson, M Björkholm, LR Goldin, and O Landgren designed the study. SY Kristinsson, M Björkholm, I Turesson, C Blimark, U-H Mellqvist, A Wahlin, and O Landgren obtained data. All the authors were involved analyses and the interpretation of the results. SY Kristinsson, M Björkholm, I Turesson, and O Landgren initiated this work. SY Kristinsson and O Landgren wrote the report. All authors read, gave comments, and approved the final version of the manuscript. All the authors had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
There are emerging data to suggest a role for genetic factors in the pathogenesis of multiple myeloma (MM). Based on small numbers, certain solid tumors have been reported to occur more frequently among blood relatives of MM patients. Using population-based data, we assessed risks for hematologic malignancies, monoclonal gammopathy of undetermined significance (MGUS), and solid tumors among first-degree relatives of MM patients. We included 13,896 MM patients and 54,365 matched controls. Also we identified first-degree relatives of MM patients (n=37,838) and controls (n=151,068). Using a marginal survival model, we estimated relative risks (RRs) and 95% confidence intervals (CIs) for hematologic and solid tumors among family members of MM patients and controls as measures of familial aggregation. Compared to relatives of controls, relatives of MM patients had an increased risk of developing MM (RR=2.1; 95% CI 1.6–2.9), MGUS (2.1; 1.5–3.1), acute lymphoblastic leukemia (ALL) (2.1; 1.0–4.2), any solid tumor (1.1; 1.0–1.1), and bladder cancer (1.3; 1.0–1.5). No significantly increased risk was found for other hematologic or solid malignancies. Our findings support a role for a shared susceptibility (genetic, environmental, or both) that predisposes to MM, MGUS, ALL, and bladder cancer.
Multiple myeloma (MM) is a malignant B-cell disorder characterized by the proliferation and accumulation of malignant plasma cells in the bone marrow, which is coupled with the overproduction of monoclonal proteins in serum or urine.1
MM has an age-adjusted incidence rate of 2.5 to 8.0 per 100,000 in Western countries,2, 3 and the median age at diagnosis of around 70 years.2 Males and African Americans are known to be more frequently affected.4 A personal history of the precursor condition, monoclonal gammopathy of undetermined significance (MGUS), is associated with an increased risk of developing MM.5
Although the etiology of MM and MGUS is unknown, there is evidence to support a role for genetic factors, including studies showing familial aggregation of MM6–14 and familial aggregation of MGUS.15 In addition, racial disparities in incidence patterns for MGUS and MM support a role for germline genes in the etiology of MM.10 In a recent study, we found first-degree relatives of MGUS patients to have an increased risk of MGUS, MM, lymphoplasmacytic lymphoma (LPL)/Waldenström’s macroglobulinemia (WM), and chronic lymphocytic leukemia (CLL), supporting the operation of shared common germline susceptibility genes in these disorders.15 However, no study has defined the risk of MGUS among relatives of MM patients. Furthermore, two recent studies reported certain solid tumors (including malignant melanoma and prostate cancer) to co-aggregate among blood relatives to MM patients, suggesting an overlap in the etiology of MM and other types of cancer.11, 12
Given the implications for future studies aimed at uncovering underlying susceptibility genes, it is important to define the spectrum of tumors associated with MM. We conducted a comprehensive study designed to evaluate familial aggregation patterns of hematologic malignancies, MGUS, and solid tumors among first-degree blood relatives of MM patients. Using high-quality population-based data from Sweden, we were able to identify 37,838 first-degree relatives of 13,896 MM patients and 151,068 first-degree relatives of 54,365 matched population-based controls.
Sweden provides universal medical health care for the entire population, which is currently approximately 9 million people. In contrast to many other countries, patients with lymphoproliferative malignancies in Sweden are typically diagnosed, treated and followed clinically by physicians at hospital-based hematology or oncology centers.
Since 1958, all physicians and pathologists/cytologists in Sweden are required by law to report each case of cancer they diagnose or treat to the centralized nationwide Swedish Cancer Registry, which has a very high rate of completeness and diagnostic accuracy.16, 17
Using the Swedish Cancer Registry, we identified all MM patients diagnosed between 1958 and 2005. For each MM patient, four population-based controls (matched by sex, year of birth, and county of residence) were chosen randomly from the Swedish Population database. All controls had to be alive at the time of MM diagnosis for the corresponding case and with no previous hematologic cancer at the date of the corresponding case’s diagnosis. Using the Swedish Multigenerational Registry18, we obtained information on all first-degree relatives (parents, siblings, and offspring) of MM patients and controls. All MM patients and controls without linkable relatives were removed from the study. As a final step, we conducted record-linkages with the Swedish Cancer Registry, a nationwide MGUS cohort established from a national hospital network including MGUS patients (N=4,458) diagnosed in Sweden 1967–2005; described elsewhere 19), a nationwide LPL/WM cohort (described elsewhere 19), and a nationwide myeloproliferative disorders (MPD) cohort (described elsewhere20) to obtain information on malignancies and MGUS among first-degree relatives of MM patients and controls.
Approval was obtained from the Karolinska Institutional Review Board (IRB) for this study. Informed consent was waived because we had no contact with study subjects. An exemption from IRB review was obtained from the National Institutes of Health Office of Human Subjects Research because we used existing data without personal identifiers.
The statistical approach is based on a model proposed by Liang21 and described in detail elsewhere.22 We classified relatives as “affected” if they had a primary cancer registration with the tumor of interest (examining up to 5 cancer registrations). Here, the age or age at onset of disease in a relative of a proband is modeled by a proportional hazards model. Familial aggregation for each condition is evaluated by testing the hazard ratio of being a relative of a case compared with being a relative to a control. The model was fitted to the data using the PHREG procedure in SAS v8.02. We use relative risk (RR) to denote the hazard ratio defined above, with 95% confidence intervals (CI). Since we have complete ascertainment of cases, every case is a “proband” and thus families with more than one case appear twice in the dataset. The robust sandwich covariance matrix accounts for dependencies among family members, including dependence due to the overlapping family clusters.22 We tested separately for increased risk for MM, MGUS, LPL/WM, non-Hodgkin lymphoma (NHL) (i.e., NHL excluding LPL/WM), chronic lymphocytic leukemia (CLL), Hodgkin lymphoma (HL), acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML)/myelodysplastic syndrome (MDS), MPD, chronic myeloid leukemia (CML), and 27 different solid tumors. Because cases and control probands were matched (see above), the relatives should be generally well matched. However, because they cannot be individually matched, we adjusted for sex in all analyses.
A total of 13,896 MM patients (mean age at diagnosis 68 years (range 19–96), 56% males), 54,365 matched controls and corresponding first-degree relatives of patients (n=37,838) and controls (n=151,068) were identified.
As shown in Table 1, first-degree relatives of MM patients had a significantly increased risk of developing MM (RR=2.1; 95% CI 1.6–2.9), MGUS (RR=2.1; 95% CI 1.5–3.1), and ALL (RR=2.1; 95% CI 1.0–4.2). However, we found no significantly increased risk of developing NHL (RR=1.2; 95% CI 0.9–1.3), CLL (RR=1.1; 95% CI 0.8–1.7), LPL/WM (RR=1.4; 95% CI 0.7–2.8), HL (RR=0.9; 95% CI 0.6–1.4), MPD (RR=1.1; 95% CI 0.7–1.6), AML/MDS (RR=0.8; 95% CI 0.5–1.2), or CML (RR=0.5; 95% CI 0.2–1.2).
When we assessed the risk of hematologic disorders in relation to type of first-degree relative (parent, sibling, offspring), age at diagnosis for the probands (above/below 65 years), and sex of the first-degree relative, the estimates for first-degree relatives of MM patients were very similar (data not shown).
First-degree relatives of MM patients had an increased risk of developing “any solid tumor” (RR=1.1; 95% CI 1.0–1.1) and bladder cancer (RR=1.3; 95% CI 1.0–1.5) (Table 2). We also found borderline increased risks for prostate cancer (RR=1.1; 95% CI 1.0–1.2; P=0.06). No significantly increased risk was found for any other solid tumor.
In this population-based case-control study including over 13,000 MM patients and their almost 40,000 first-degree relatives, we made several important and novel observations. Given the large sample size, we were able to conduct the most comprehensive evaluation of MGUS, hematologic and solid tumors among first-degree relatives of MM patients and controls to date.
Compared to first-degree relatives of controls, we found that first-degree relatives of MM patients had a 2-fold higher risk of developing MM. This pattern is consistent with prior studies.9–14 We also found, that first-degree relatives of MM patients had a 2-fold increased risk of MGUS. This finding is consistent with our recent investigations showing that first-degree relatives of MGUS patients are at increased risk of developing both MGUS and MM.15 Also a recent study from the Mayo Clinic found relatives to MM and MGUS patients to have 2- and 3.3-fold increased risks of MGUS.23 Compared to first-degree relatives of controls, we also found first-degree relatives of MM patients to have a 2-fold increased risk of developing ALL. A positive familial aggregation for ALL and MM has not been reported previously in any large study; however, it has been observed in a prior single high-risk MM family.24 Because we evaluated a large number of malignancies, it cannot be ruled out that this finding is due to chance. In contrast to our previous studies showing overlapping excess familial risks for various types of lymphomas19, 25–27, with the exception of ALL, we found no increased risk for other hematologic malignancies among MM relatives.
To our knowledge, there are only limited data available on familial aggregation of solid tumors and MM. Recently, one case report and one registry-based study reported evidence of an excess occurrence of malignant melanoma12 and prostate cancer11, 12 among relatives of MM patients. Given the implications for future studies aimed to uncover underlying susceptibility genes, we were motivated to evaluate a broad range of a solid tumors (n=27) among first-degree relatives of MM patients and controls. We found a significantly increased risk for any solid tumor and bladder cancer among relatives of MM patients. Although this could be a chance finding due to screening a large number of tumors, there is one prior study showing evidence of co-aggregation of MM and bladder cancer.28 In the two previously mentioned studies11, 12, a significantly increased risk of prostate cancer among MM relatives was found. In our study we found only a borderline increased risk, suggesting there might be shared susceptibility in MM and prostate cancer. The reason for the discrepancy between our results and previous studies might, at least in part, be due to selection mechanisms. In the present study, we applied a population-based study design based on all MM patients diagnosed in Sweden over the past four decades. For relatives of MM patients and their matched controls, we obtained information on incident cancers through record linkage with the Swedish Cancer registry. In contrast, the study by Lynch et al.11 was based on a single high-risk MM family (including 5 MM cases) seen at a research center specialized in familial disease. The study by Camp et al.12 was based on 2.5 million persons with genealogic data linked to the Utah Surveillance, Epidemiology, and End Results (SEER) cancer registry including a total of 1354 MM patients, and their 13,288 first-degree relatives, 45,575 second-degree relatives, and 118,363 third-degree relatives.
Although long-term follow-up studies from the Mayo Clinic have shown that MGUS patients have an increased risk of developing MM5, until recently, it has been unknown whether all MM patients are preceded by MGUS, or if MM can occur de novo. In a recent prospective nationwide U.S. cancer screening trial enrolling 77,469 healthy adults, a total of 71 individuals were identified who developed MM during the course of the study in whom serially-collected (up to 6) pre-diagnostic serum samples obtained (up to 10 years) prior to MM diagnosis were available. In that study, asymptomatic MGUS stage preceded the diagnosis of MM in almost all MM patients.29 Importantly, the findings from our present familial MM study, as well as the observations from the above-mentioned prospective screening study, link MM and MGUS closely together and establish a key role for MGUS in the pathway to MM. This is of relevance for investigations designed to uncover pathogenetic mechanisms in MM, for the identification of novel molecular targets, and ultimately for future preventive trials.30, 31 In particular, the observed familial aggregation patterns in the present study support a role for shared common susceptibility genes in MM and MGUS.11 Consequently, our results support the application of gene mapping and candidate gene approaches in high risk families and case-control studies.
Our study has several strengths, including its large size as well as the application of high-quality data, including a large MGUS cohort, from Sweden in a stable population with access to standardized universal medical health care during the entire study period. The use of the nationwide register-based case-control design ruled out recall-bias, ensured a population-based setting, and the generalizability of our findings. Limitations include lack of information on potential confounders (although the matched design and analyses ensured adjustment for sex, age, and geography), and lack of detailed clinical data. Also, because of the nature of this hypothesis-generating study, one has to interpret our findings with some caution due to the large number of tested malignancies.
In summary, compared to first-degree relatives of matched controls, we found an increased risk of developing MM, MGUS, ALL, and bladder cancer among first-degree relatives of MM patients. These results support a role for some shared susceptibility (genetic, environmental, or both) that predisposes to these malignancies. Our study provides novel information supporting the application of gene mapping and candidate gene approaches in high risk families and case-control studies.
This research was supported by grants from the Swedish Cancer Society, Stockholm County Council, the Karolinska Institutet Foundations, and the Intramural Research Program of the NIH, NCI. The authors thank Ms. Shiva Ayobi, The National Board of Health and Welfare, Stockholm, Sweden; Ms. Susanne Dahllöf, Statistics Sweden, Orebro, Sweden; Ms. Charlotta Ekstrand, Ms. Molly Collin and Ms. Lisa Camner, Karolinska Institute, Stockholm, Sweden, for invaluable ascertainment of MGUS data; and Ms. Emily Steplowski, Information Management Services, Silver Spring, MD, for important efforts in the development of this database. Charlotta Ekstrand. The authors have no conflict of interest to declare.