Abstract
Background: Adjuvant therapy is an essential component of treatment for many early-stage (stage I-III) breast, colon, or lung cancers after curative surgery. Patients newly diagnosed with these cancers who also have a history of diffuse large B-cell lymphoma (DLBCL) or Hodgkin lymphoma (HL) have typically undergone anthracycline-based chemotherapy to treat their lymphoma, and may be less likely to receive adjuvant chemotherapy for subsequent solid cancers because of prior toxicities or limits on cumulative doses. Some studies suggested lower use of adjuvant chemotherapy in HL survivors with breast cancer (Elkin et al., JCO 2011), and worse overall survival (OS) outcomes in lymphoma survivors with solid tumors (Sanna et al., Ann Oncol 2007). Our objective was to examine the use of adjuvant chemotherapy and OS among DLBCL/HL survivors compared with the general population of patients with three common cancers: breast, colon, and lung.
Methods: From the Surveillance, Epidemiology, and End Results (SEER) database, we identified patients aged 20-79 diagnosed with breast (stage I-III), colon (stage II-III), or lung (stage II-IIIA) cancers in 2004-2015, who underwent surgery and were thus candidates for adjuvant chemotherapy by virtue of age and stage. Among them we distinguished "cases" of DLBCL or HL survivors (diagnosed with lymphoma >1 year before the index cancer), treating other patients as "controls". We compared administration of chemotherapy (versus no or unknown administration status, as recorded by SEER) and OS in each cancer, in multivariable robust Poisson regression or Cox models, respectively, reporting relative risk (RR) or hazard ratio (HR) with 95% confidence intervals (CI). All models were adjusted for age, sex, race, year of diagnosis, stage of the index cancer, lymph node involvement, type of surgery (e.g. breast conservation or mastectomy, partial or total colectomy, lobectomy or pneumonectomy), and endocrine receptor status for breast cancer.
Results: Among patients with breast cancer (n=532,686), we identified 360 DLBCL and 349 HL survivors (with median latency between lymphoma and cancer diagnoses of 79 and 222 months, respectively). Among patients with colon cancer (n=149,993), there were 165 DLBCL and 88 HL survivors (median latency 71 and 156 months, respectively). Among patients with lung cancer (n=23,396) there were 37 DLBCL and 25 HL survivors (medial latency 60 and 147 months, respectively). In breast and colon cancers, DLBCL survivors were significantly older, and HL survivors younger than controls, but this difference was absent in lung cancer (Table). HL survivors were diagnosed with breast cancer at an earlier stage than controls (P=.006), but in all other studied scenarios we observed no significant difference in stage or extent of nodal spread between DLBCL/HL survivors and controls.
After adjustment for baseline characteristics, we observed no significant difference between patients with and without a history of lymphoma in the rates of chemotherapy administration (adjusted RR, 0.87 to 1.04), except HL survivors with colon cancer, who had a 20% lower rate than controls (RR, 0.80, 95% CI, 0.64-1.00). OS was significantly worse for patients with prior HL lymphoma in all 3 cancers, but for DLBCL survivors only in breast cancer. Consistent results were observed in the subpopulation of patients who actually received adjuvant chemotherapy.
Conclusions: In most studied scenarios, despite prior exposure to intensive chemotherapy, survivors of DLBCL or HL received adjuvant chemotherapy for common cancers at similar rates relative to patients with no history of lymphoma. Therefore, contrary to prior suggestions, differences in application of chemotherapy cannot explain survival disparities between lymphoma survivors and other patients with the same cancers. HL survivors had consistently worse OS even in the subpopulation receiving chemotherapy, suggesting that other (e.g. cardiovascular) sources of mortality related to late toxicities of HL treatment may be contributing, rather than differences in treatment. Unique survival patterns among DLBCL survivors suggest that limits on anthracycline-based chemotherapy due to cumulative lifetime dosing may impair outcomes in breast cancer, but not in cancers like colon or lung, which are treated with non-overlapping taxane- or fluorouracil-based adjuvant regimens.
Olszewski:Genentech: Research Funding; TG Therapeutics: Research Funding; Spectrum Pharmaceuticals: Consultancy, Research Funding.
