Postmastectomy Radiation Therapy and Breast Reconstruction--Marching Forward Together or Apart?

Abstract

Postmastectomyradiationtherapy (PMRT)and immediate reconstruction eachplay important roles in the treatment of patientswithbreast cancer.GivingPMRT increases cure rates for subgroupswith substantial risks of local-regional recurrence.1 Many individuals wish to have immediatebreast reconstruction to reduce the psychological and social consequences of mastectomy. However, PMRT increases the risks of complications andpoor cosmetic results after breast reconstruction. Immediate reconstruction limits the technical approaches available for delivering PMRT, potentially resulting in increased heart and lung doses. Improvements in surgical and radiation techniquesover timehavedecreasedbutnot entirely eliminated these problems.2,3 The studybyFrasier and colleagues4 in this issueof JAMA Oncology showed that use of both PMRT and breast reconstruction increasedduring the recentpast in theUnitedStates, despite these deleterious interactions. The increase in PMRT rates was almost entirely confined to patients with tumors 5 cm or smaller with 1 to 3 involved axillary nodes at “intermediate” risk of local regional recurrence, for whom the 2009 guidelines of the National Comprehensive Cancer Network (NCCN) said PMRT should be “strongly considered.” (This increase seems to have begun in earnest in 2007 [see their Figure4].This likely reflects the impactofearlierguidelinesand editorials that encouragedPMRT tobeusedmoreoften in this group.)Asexpected, absolute ratesof reconstruction throughout the study period remained higher for patientswith lowor intermediate risks of local-regional recurrence than for individuals at high risk, for whom the guidelines unequivocally recommended PMRT. These findings are quite plausible and consistent with other studies, which they ably discuss. The authors also cogently point out the many limitations of using the Surveillance, Epidemiology, and End Results (SEER) database to determine patterns of care. One major problem is that SEER does not capture PMRT or reconstructive surgery performed more than 4 months after the initiation of therapy. Many patients now receive postoperative chemotherapy programs lasting 3 to 6 months before starting PMRT; and, particularly but not exclusively in community practices, consultation with a radiation oncologist is often not scheduled until near the end of or after chemotherapy, delaying the final decision well beyond this 4-month window. (One way to overcome this analytic problem is to use health insurance databases that capture care over longer periods.5) Themost critical questions regarding theuseofPMRTand breast reconstructionare, however, difficult to answer in such a study as the present one. First, did patients who would potentially benefit from PMRT not receive it because they underwent reconstruction? Second, didpatientswhowanted reconstructionnot receive itbecause theirphysiciansanticipated that they might need PMRT preoperatively? Treating physicians’ opinions are important to patients trying to decide whether they will undergo PMRT.6 That seems even more likely when guidelines do not explicitly mandate whether treatment be given. Such ambiguity (reflecting a lack of consensus in the profession) leaves room for potentially relevant but also extraneous medical and nonmedical factors to affect decision making. The authors found that tumor size (T1 vs T2) and grade (1 vs 2 or 3) affected the frequency of PMRT use in the intermediate-risk subgroup. However, other pathologic findings often used to select patients for PMRT are not included in the SEER database (eg, the presence of lymphovascular invasion and margin involvement). Patient socioeconomic status and comorbidities are not available in SEER. Hence, biases against using PMRT in patients who already underwent reconstruction could be hard to tease out even with matched-pair and propensity analysis. Still, it might be helpful to examine whether PMRT was used more or less often for the subgroups of patients who did or did not undergo reconstruction, both for the population as a whole and for the NCCNdefined subgroups. This might shed light on the disturbing fact that only two-thirds of patients aged 64 years or younger in the “recommend radiotherapy” group received it at the end of the study period. Was the increased use of reconstruction in this group partly to blame for the rate not being higher? For almost all patients, recommendations for PMRT are made after surgery, based on the pathologic findings. (The exception is patients with clinical stage III, or “locally advanced,” breast cancer.) The decision whether to perform reconstruction is made without this knowledge, as the authors note. Nonetheless, in practice the well-known though imperfect relationship between clinical and pathologic stages results inapreoperative“impression”ofwhetherPMRTwill likely be needed thatmay influence physician and patient behavior. For example, do physicians equally often recommend immediate reconstruction topatientswith4-cmpalpablemassesand severalpalpable ipsilateralaxillarynodesastopatientswithnonpalpable 1-cm cancers and nonpalpable nodes? These recommendations again are likely tohave an effect becausepatients’ concernsabout increasedrisksofcomplications fromPMRTare often decisive in their decisions whether to undergo immediate reconstruction.7Perhaps therewasamodest increase in the useof reconstruction inpatientswhohadPMRTbut a substantial increase in reconstruction in patients who did not have PMRT. Or perhaps the use of reconstruction increased by the Related article page 95 Research Original Investigation Trends in Postmastectomy Radiation Therapy and Breast Reconstruction