We present a series of 18 patients with DFSP of the breast who underwent either MMS or WLE with either reconstructive surgery or primary closure. Our study showed a difference in lesion size of 3.3 cm in MMS versus 6.4 cm in WLE as well as excised specimen size of 7.0 cm in MMS compared with 10.8 cm in WLE. However, neither of these were statistically significant likely due to the small sample size. Although gross intended margins ranged from 1 to 3 cm with WLE in our series, about 67% (6/9) of patients with WLE had at least one intraoperative re-excision following positive margins on initial intraoperative pathology, indicating that the intended margins in these cases were not wide enough. Most DFSP recurrences are detected within 3 years of primary excision. Our study, with a median follow-up of 5 years, identified one patient with recurrence in the WLE cohort, but no recurrences with MMS [1]. Compared to primary closure, there was increased morbidity with more complex reconstructions, including flaps and grafts, but no surgical complications were observed in our two patients with pedicled flap coverage.

The major difference between MMS and WLE is the extent of resection of normal tissue and margin control [9]. A retrospective review of 48 patients with DFSP demonstrated more frequent positive margins in WLE than MMS, suggesting that MMS allows more focused resection resulting in accurate margin control [10]. Pathologic analysis of WLE specimens typically utilizes a vertical “breadloafing” technique, which can result in sampling error if the intervals of the sections miss extensions of tumor especially with DFSP’s infiltrative and asymmetric growth [9]. Most studies and systematic reviews report a lower recurrence rate with MMS compared to WLE [6, 11-20]. Similarly, studies utilizing modified WLE with total peripheral margin analysis and horizontal processing were able to achieve 0–1% recurrence rates, suggesting that meticulous margin evaluation is important regardless of surgical technique [5, 17, 21]. By focused excision of margins as directed by frozen section histologic review, MMS can also limit the size of postoperative defect compared to WLE [6, 7, 19, 22, 23]. Lowe et al. found a statistically significant smaller postoperative defect size by 2 cm in MMS compared to WLE in treatment of DFSP [6]. Goldberg et al. also found no recurrence in MMS despite MMS having average margin size of 1.36 cm compared to 2.33 cm for WLE [16]. This may be a deciding factor when resecting DFSP from a cosmetically sensitive area such as the breast, head, or neck [10, 21]. Similar to our study, there is also a trend of smaller lesions being more likely to be treated by MMS and larger lesions treated by WLE as DuBay et al., for example, reported in their study that preoperative lesions averaging 5.3 cm2 were treated by MMS, while preoperative lesions averaging 14.8 cm2 were treated by WLE [21].

All of our patients had unilateral DFSP of the breast, which can cause visible asymmetry following resection. There is clearly a size limit that will allow acceptable symmetry following primary closure with MMS, and this is also dependent on tumor location and breast size. For example, Fig. 1b shows loss of inferior pole with MMS and primary closure in a patient with a 6.8-cm lesion in the lower inner quadrant near the inframammary fold. Figure 2b demonstrates an excellent cosmetic result with MMS for a 2.7-cm lesion in the upper inner quadrant where there is less breast tissue. The use of a pedicled latissimus dorsi flap can preserve lower pole fullness and symmetry as seen in Fig. 3a and b, following resection of a 3.5-cm mass. MMS may be beneficial for patients who have a smaller lesion to breast size or location with less breast tissue, allowing for an aesthetically pleasing result with primary closure without further revision [24].

Interestingly for DFSP of the breast, there are no case reports that describe use of MMS for resection, very few case reports that describe reconstruction and its complications following resection, and no studies that compare complications between primary closure and reconstruction [8]. In addition to primary closure, reconstruction techniques cited in the literature include pedicled latissimus dorsi flap, rotation flap, reverse abdominoplasty, pectoralis flap, and reduction mammoplasty to provide wound coverage and to preserve breast shape [25-30]. We found no complications in our MMS and primary closure cohort but encountered complications requiring intervention with WLE and skin grafts, implant, and rotational flap reconstructions. One recurrence in the WLE cohort was detected despite pedicled flap reconstruction and resected without complications. Despite potential asymmetry, there still may be value to primary closure after MMS in select patients as it involves fewer complications, allows for time to monitor the wound for possible recurrence, and provides the patient more time to decide whether a simple repair is sufficient or a more complex delayed revision is desired. Although our flap reconstructions were immediate due to availability of intraoperative frozen sectioning, this may not translate to other institutions where frozen section is not available. All of these considerations should be discussed in shared decision-making with the patient when selecting among WLE, MMS, and timing of reconstruction.

A limitation of our retrospective study is likely selection bias in which patients were referred for WLE or MMS. Potentially, patients with larger tumors, smaller breast size, and/or those who voiced concern over cosmetics were referred to a plastic surgeon. There can be confounding factors such as surgeon preference or if patients were offered a plastic surgery consult to discuss possible elective reconstruction. Another limitation is we did not collect patient-reported outcomes, and thus, it is unknown if patients were satisfied with the aesthetic outcomes of the different techniques.