This article is more than 5 years old. Information may no longer be current.
Marrow neoplasm risk increased after adjuvant breast cancer therapy
Click Here to Manage Email Alerts
Outcomes for early-stage breast cancer have improved, but incomplete examination and follow-up may have underestimated subsequent risks for treatment-associated marrow neoplasm in patients who undergo adjuvant therapy, according to study results.
Prior first-generation adjuvant chemotherapy trials reported a 0.27% cumulative incidence of myelodysplastic syndrome/acute myelogenous leukemia after 8 years.
Antonio C. Wolff
In the current study — conducted by Antonio C. Wolff, MD, professor of oncology at Johns Hopkins Sidney Kimmel Comprehensive Cancer Center, and colleagues — the risk for marrow neoplasm among patients who underwent radiation and/or adjuvant chemotherapy for early-stage breast cancer remained low, but was higher than previously described.
Wolff and colleagues used the National Comprehensive Cancer network Breast Cancer Outcomes Database to evaluate 20,063 patients with stage I to stage III breast cancer treated at US academic centers between 1998 and 2007.
After median follow-up of 5.1 years, 50 of these patients developed marrow neoplasm (42 myeloid, 8 lymphoid). All 50 patients had similar breast cancer stage distribution, race and chemotherapy exposure, but these patients were older compared with those who did not develop marrow neoplasm (median age, 59.1 years vs. 53.9 years; P=.03).
Data about marrow cytogenics and family history of cancer were available for 41 of the 50 patients who developed marrow neoplasm. Twenty-six of those 41 had abnormal cytogenics, including 24 (67%) of 36 patients with myeloid disorders and two (40%) of five patients who had lymphoid disorders.
Researchers determined risk for marrow neoplasm increased substantially among patients who underwent surgery and chemotherapy (HR=6.8; 95% CI, 1.3-36.1), as well as among those who underwent surgery, chemotherapy and radiation (HR=7.6; 95% CI, 1.6-35.8), compared with patients who did not receive chemotherapy.
Wolff and colleagues calculated the following marrow neoplasm rates per 1,000 person-years: 0.16 for surgery alone; 0.43 for surgery plus radiation; 0.46 for surgery plus chemotherapy, and 0.54 for surgery, chemotherapy and radiation.
Cumulative incidence of marrow neoplasm doubled between years 5 and 10 (0.24% to 0.48%), and 9% of patients were alive at 10 years.
The results demonstrate the individual risk for marrow neoplasm must be balanced against the absolute survival benefit conferred by adjuvant chemotherapy, researchers concluded.
“This cohort is likely to be representative of patients treated in most large cancer centers,” Wolff and colleagues wrote. “Although the risk of marrow neoplasm after breast cancer therapy we report is small, it is not zero, and it is not short-lived. Our findings highlight the challenges of studying infrequent but important clinical events and demonstrate the potential benefits of prospective, well-annotated, large, longitudinal databases.”
Disclosure: One researcher reports remuneration from UpToDate.
Perspective
Back to Top
Charles L. Shapiro, MD
Acute leukemia and myelodysplasia are well-described but rare complications of adjuvant chemotherapy for early-stage breast cancer (Morton L, et al. Blood. 2013;doi:10.1182/blood-2012-08-448068). In particular, treatment-associated marrow neoplasms are related to exposures to cyclophosphamide and doxorubicin.With this mind, investigators from the National Comprehensive Cancer Network used the Breast Cancer Outcomes Database to assess the incidence of treatment-associated marrow neoplasms in more than 20,000 women with early-stage breast cancer treated between 1998 and 2007. Their findings show that the 10-year cumulative frequency was 0.48%; that, relative to surgery, the addition of radiation, chemotherapy or both increased the risks for developing marrow neoplasms; and that the incidence was about twofold higher than reported by the SEER population registry.
How might we put these risk estimates in perspective? Most individuals with cancer receive treatment in the community, not in cancer centers. The original NCCN was comprised of 13 of the largest cancer centers, and the Breast Cancer Outcomes Database is a registry of sorts that captures demographic, pathologic, treatment and clinical outcomes that are updated periodically. As a general rule, registries are prone to under or over-reporting. For example, women who choose to be treated at cancer centers may be subject to more intensive follow-up and diagnostic testing, hence raising the possibility of case ascertainment bias. This will lead to apparent higher rate per 1,000 patient-years than the SEER rate, which is population-based.
Perhaps more importantly, the rate of treatment-associated marrow neoplasms is very low. In addition, since 2007 — when ASCO’s updated recommendations for the use of tumor markers in breast cancer included a recommendation for the Oncotype DX breast cancer assay (Genomic Health) —many women who otherwise would have been treated were spared chemotherapy (Harris L, et al. J Clin Oncol. 2007;doi:10.1200/JCO.2007.14.2364). Likewise, as we develop improved methods to identify those who do and do not benefit from adjuvant chemotherapy, the therapeutic ratio of benefit to risk of treatment-related marrow neoplasms and other rare side effects will favor treatment even more.
Click Here to Manage Email Alerts