Connect MM® – the Multiple Myeloma Disease Registry: incidence of second primary malignancies in patients treated with lenalidomide

Advances in treatment have extended the survival of patients with multiple myeloma (MM),[1] but this increased survival time also increases the potential for developing second primary malignancies (SPM).[2,3] According to the data from the Surveillance, Epidemiology, and End Results (SEER) Program, the 20-year risk of invasive SPM for patients with MM is 6.1%, which is similar to the overall risk of malignancy in the average patient-based population.[2]

Initial investigations in randomized clinical trials (RCTs) raised some concern over increased rates of SPM, particularly acute myeloid leukemia (AML), myelodysplastic syndromes (MDS), acute lymphocytic leukemia (ALL), and Hodgkin and non-Hodgkin lymphomas, in patients with MM treated with lenalidomide.[3–5] The MM-015 trial showed a 3-year incidence of invasive SPM in patients treated with melphalan, prednisone, and lenalidomide (MPR), MPR with lenalidomide maintenance, and melphalan and prednisone alone of 7%, 7%, and 3%, respectively.[6] In Intergroupe Francophone du Myelome (IFM) 2005-02, patients treated with lenalidomide or placebo maintenance after autologous stem cell transplant (SCT) and lenalidomide consolidation had 13% and 7% cumulative incidence of invasive SPM, respectively, at 6 years’ follow-up, with incidence rates (IRs) per 100 patient-years (PY) of 3.4 and 2.1.[7] The similar Cancer and Leukemia Group B (CALGB) 100104 trial reported increased cumulative IR of invasive SPM with lenalidomide versus placebo maintenance after SCT at 5 years’ follow-up (p = 0.005).[8] Patients ineligible for transplant treated with lenalidomide continuous and low-dose dexamethasone (Rd), 18 cycles of Rd, or melphalan, prednisone, and thalidomide in MM-020 had 3-year invasive SPM risks of 3.2%, 5.6%, and 5.0%, respectively.[9] Finally, a meta-analysis of seven clinical trials between 2000 and 2012 (which included MM-015) found invasive SPM in 3% of treated patients (105/3218).[10] Exposure to lenalidomide plus oral melphalan increased the risk of hematologic SPM overexposure to melphalan alone.[10] The meta-analysis concluded that the benefit of treatment outweighed the risk of SPM because the overall majority of patient deaths were due to progressive disease rather than SPM, and based on these findings, continued scrutiny, but no alteration to the treatment with lenalidomide of patients with MM, was recommended.[10]

We examined the occurrence of SPM at a median follow-up of approximately 3 years among treated patients in Connect MM, a US-based, multicenter, prospective observational cohort registry study designed to observe the management and outcomes of patients with newly diagnosed MM. Patients must have been diagnosed within 2 months of enrollment, and treatment was at physicians’ discretion. Full study design details have been previously described.[11] SPM were defined as new (or recurrent after > 5 years) cancers, including invasive and non-invasive (non-melanoma skin cancers), occurring after MM diagnosis and at any time during study follow-up regardless of causal relationship to MM therapy. This report focuses only on invasive SPM (categorized as a hematologic or solid tumor). Type of SPM and method of diagnosis were collected quarterly with an electronic data capture system. PY were calculated from the start of treatment until the detection of the first reported SPM, death, end of follow-up (e.g. patient lost to follow-up, withdrawn consent), or data cutoff of 30 June 2014.

Four exposure groups were analyzed: exposed to lenalidomide or not, received SCT with or without lenalidomide maintenance, exposed to melphalan or not, and exposed to oral melphalan with or without proximate lenalidomide (within 160 d of start or end of treatment). When patient-time could not be calculated for an individual due to missing dates, the average time for the remaining patients in the exposure group was used. The time to first invasive SPM analysis was estimated using adjusted cumulative incidence curves, with log-rank p values and 95% CI used to compare exposure groups. Univariate and multivariate Cox regression analyses were conducted to assess SPM risk given baseline parameters and exposure groups.

As of 30 June 2014, data were available for 1450 treated patients with a median follow-up of 33.5 months (range: 0.03–55.9 months). Invasive SPM were reported for 4.0% of patients (58/1450), with an IR per 100 PY of 1.61 (95% CI: 1.25–2.09). Of the 58 patients with SPM, 18 (1.2%) had hematologic malignancies (IR per 100 PY, 0.49 [95% CI: 0.31–0.78]), of which the most frequent were MDS (7), ALL (5), and AML (2). Forty patients (2.8%) had solid tumors (IR per 100 PY, 1.11 [95% CI: 0.81–1.51]); the most frequent were lung/bronchus (8), prostate (6), breast (6), colon/rectum (5), melanoma (5), and pancreas (2). Of 10 patients developing SPM after a prior malignancy, two were the same tumor type (one breast cancer and one melanoma). The median times to hematologic and solid tumor SPM were 24.8 and 21.2 months, respectively.

Differences were observed in age, sex, International Staging System stage, the frequency of SCT, and prior hematologic and solid tumor malignancies between patients with and without SPM; however, small patient numbers make it difficult to draw conclusions (Table 1). A higher proportion of patients with SPM had a reported history of monoclonal gammopathy of undetermined significance (MGUS; 23.2% versus 10.6%). Treatment exposure for patients with and without SPM is also shown in Table 1.

Table 1. Patient demographics, disease characteristics, and treatment exposure.

	SPM	No SPM
	(n = 58)	(n = 1392)
Patient demographics
Median age (range), years	69 (49–90)	66 (24–94)
Male, n (%)	31 (53.4)	803 (57.7)
Race, n (%)^a
White	49 (84.5)	1145 (84.8)
Black	9 (15.5)	178 (13.2)
Asian	0	4 (0.3)
Other	0	23 (1.7)
Disease characteristics at enrollment
ISS stage, n (%)^b
I	8 (21.6)	209 (25.8)
II	21 (56.8)	281 (34.7)
III	8 (21.6)	321 (39.6)
Prior hematologic malignancy, n (%)	0	18 (1.3)
Prior solid tumor, n (%)	10 (17.2)	111 (8.0)
History of MGUS, n (%)^c	13 (23.2)	138 (10.6)
Treatment exposure (all courses), n (%)
SCT	23 (39.7)	499 (35.8)
Triplet treatment	19 (32.8)	622 (44.7)
Steroids	50 (86.2)	1257 (90.3)
Mean exposure, months (SD)	7.2 (7.8)	7.7 (8.5)
Proteasome inhibitors	48 (82.8)	1183 (85.0)
Mean exposure, months (SD)	6.5 (7.4)	6.5 (7.2)
Immunomodulatory agents	44 (75.9)	1030 (74.0)
Mean exposure, months (SD)	9.3 (6.4)	11.0 (10.1)
Alkylating agents	36 (62.1)	729 (52.4)
Mean exposure, months (SD)	4.7 (9.2)	3.1 (5.7)
Anthracycline agents	7 (12.1)	154 (11.1)
Mean exposure, months (SD)	1.9 (1.1)	3.6 (5.6)

a Number of patients with race information available: n = 58 (SPM); n = 1350 (no SPM).

b Number of patients with ISS assessment available: n = 37 (SPM); n = 811 (no SPM).

c Number of patients with information on MGUS history available: n = 56 (SPM); n = 1307 (no SPM).

SCT, stem cell transplant; ISS, International Staging System; MGUS, monoclonal gammopathy of undetermined significance; SPM, second primary malignancies.

No significant differences in the 3-year cumulative probability of hematologic or solid tumor SPM were reported between any of the four exposure comparison groups (Figure 1A–D). These results are broadly consistent with a meta-analysis of RCTs, which reported 3-year cumulative incidence with lenalidomide versus no lenalidomide exposure of hematologic SPM (1.4% versus 0.4%) and solid tumor SPM (2.6% versus 2.9%).[10] However, the meta-analysis found lenalidomide exposure to significantly increase the incidence of hematologic SPM,[10] whereas Connect MM showed no effect.

Figure 1. Cumulative incidence of SPM. Time to hematologic (A and C) and solid tumor (B and D) second primary malignancies (SPM) in patients exposed or not exposed to lenalidomide (LEN) and exposed or not exposed to melphalan (MEL; A and B) and patients who received stem cell transplant (SCT) with or without LEN maintenance and oral MEL with or without proximate LEN (C and D). Adjusted for prior history of invasive malignancy and prior history of monoclonal gammopathy of undetermined significance (MGUS). HR, hazard ratio.

SPM incidence was similar regardless of melphalan exposure (Figure 1A and B). The overall IR of invasive SPM for patients receiving SCT (IR per 100 PY, 1.62; 3-year cumulative incidence of 2.6%) was similar to that found in a prospective, observational cohort study performed by the Center for International Blood and Marrow Transplant Research (CIBMTR; IR per 100 PY, 1.2; 3-year cumulative incidence of 2.6%).[12] Lenalidomide maintenance following SCT did not significantly affect rates of hematologic or solid tumor SPM (Figure 1C and D), unlike the increased risk of SPM with lenalidomide maintenance reported in CALGB 100104 [8] and IFM 2005-02.[7] The risk of SPM likewise was not affected by lenalidomide proximate to oral melphalan (Figure 1C and D), counter to the RCT meta-analysis, which found this treatment combination to significantly increase the risk of hematologic SPM.[10]

Preliminary multivariate analysis found that significant risk factors for the occurrence of SPM were prior history of invasive malignancy (hazard ratio [HR], 2.01 [95% CI: 1.01–3.99]; p = 0.046) or MGUS (HR, 2.24 [95% CI: 1.20–4.18]; p = 0.011). This supports the major finding of a Swedish population-based report that patients with prior MGUS had an elevated risk of developing MDS or AML.[13] Multivariate analysis performed in the CIBMTR study found age (≥ 70 years), male sex, and obesity to significantly increase the risk of SPM in patients with autologous SCT,[12] but these associations were not seen in this Connect MM analysis.

Limitations of this preliminary analysis of SPM in Connect MM are those inherent to the nature of an observational study. Particularly, the lack of randomized, protocol-defined treatments leads to substantial variations in treatment duration and intensity, and data capture is less structured compared with prospective trials. Additionally, the 3-year follow-up period is relatively short, especially for a registry (cf. 20 years for SEER).[2] However, this large patient population provides valuable information regarding patterns of SPM in diverse clinical settings, including the preliminary finding that lenalidomide exposure was not associated with an increased risk of SPM. Continued follow-up and a supplemental patient cohort will be investigated to confirm these results.

Potential conflict of interest

Disclosure forms provided by the authors are available with the full text of this article at http://dx.doi.org/10.3109/10428194.2015.1132419.