Doughnut mastopexy lumpectomy versus standard lumpectomy in breast cancer surgery: A prospective study

Methods 127 patients with breast cancer were offered the choice between two conservative surgical approaches: doughnut mastopexy lumpectomy (DML group, n = 39) or standard lumpectomy (SL group, n = 88). The groups were comparable for radiological tumour size, tumour location within the breast, histological size, and pT category. Comparison was performed in term of surgical variables, histological parameters, postoperative morbidity and cosmetic outcome. Results The patients undergoing DML were younger than the patients who chose SL. In the DML group, the skin incision was 3-fold longer than in the SL group but was obtained with a final scar located around the nipple areola complex without further postoperative complications. The average volume of the breast specimen was higher in the DML group compared with SL group. The clinician assessment of cosmetic outcome reported a higher rate of acceptable result in the DML group than in the SL group. However, patient's assessment did not show difference of cosmetic satisfaction between groups. Conclusion Our comparative study indicates that DML may be a useful alternative to SL not only in terms of accurate breast tissue resection but also in term of cosmetic results. Keywords Breast cancer Oncoplastic surgery Doughnut mastopexy lumpectomy Free margins Introduction Doughnut mastopexy lumpectomy (DML) is a unique breast resection in which a tissue segment is removed and the breast reshaped through a periareolar incision. 1–3 This oncoplastic technique may carry several advantages over standard lumpectomy. First, a final periareaolar scar is more discreet than a scar from direct incision, with an overall more aesthetic result. Second, glandular volume is displaced in the doughnut mastopexy lumpectomy and breast skin is reshaped. Intuitively, this suggests better cosmetic results than standard lumpectomy. Third, when skin-sparing mastectomy is indicated, the periareolar incision offers an excellent option. 1 Conversely, direct skin incisions may compromise blood supply and increase the risk of skin necrosis. 4 To our knowledge, no study to date has directly compared doughnut mastopexy lumpectomy with standard lumpectomy (MEDLINE search: English and French papers published from January 1966 to April 2006 were evaluated). The present investigation compares DML and standard lumpectomy (SL) in relation to surgical variables, histological parameters, postoperative morbidity and cosmetic outcome. Patients and methods Patient selection The study was conducted in the Oncology Unit of the Department of Obstetrics and Gynaecology at the Univers ity Hospital of Montpellier, France, between January 2004 and May 2005. 292 consecutive women with breast carcinoma were operated on. Patients with breast tumour located 2 cm or more from the areola were entered into this prospective study. A total of 127 patients who were scheduled to undergo breast lumpectomy were invited to participate and were offered the choice between the two surgical approaches: doughnut mastopexy lumpectomy (DML group, n = 39) or standard lumpectomy (SL group) without concomitant mammaplasty ( n = 88). 38 patients had a tumour located close to the areola and were operated on using a hemiareolar incision ( n = 26 patients ) or a central quadrantectomy ( n = 12 patients ). 25 patients needed planned mastectomy. The remaining 56 patients underwent an inverted-T procedure to perform the mastopexy and were excluded from the study. 46 patients were excluded because of inflammatory carcinomas ( n = 4 patients ), locally advanced tumours with gross lymph node involvement ( n = 15 patients ), local failure of previous conservative treatment ( n = 15 patients ) and metastatic disease ( n = 12 patients ). After being informed of the risks and benefits associated with each treatment alternative, the patients made their own decisions and were then assigned to either the DML or SL group. Preoperative evaluation included physical examination of the breast, mammography, breast ultrasonography and contrast-enhanced MRI. A needle core biopsy with radiology-based guidance when needed, was always performed to confirm the cancer diagnosis. Patient and tumour characteristics, details of adjuvant therapy, surgical intervention, and complications of surgery were all entered into a computerized database. Patients were examined for operative patterns and postoperative complications. Radiotherapy to the breast and lymph nodes, chemotherapy, endocrine therapy, and axillary lymph node dissection were carried out without modification to our standard protocols. Surgical techniques All patients were operated on by 2 surgeons (PLG, NEG) trained in both breast and plastic surgery. Non-palpable tumours were excised after radiologically-guided percutaneous location. Axillary lymph node dissection was performed through an axillary route in cases of invasive tumour pattern. For the DML patients, the surgeon performed a wide quadrantectomy, with no thought to the residual breast shape since it was to be remodelled by immediate mammaplasty. Preoperative markings were made with the patient in the upright position. The preoperative drawing outlined the median intermammary line, the inframammmary fold, and the sternal notch-to-nipple line. 5 The position of the new nipple was marked along the line at a distance between 19 and 21 cm from the sternal notch. The vertical inframammary distance was fixed at 50 mm and the maximum diameter of the areola, from 40 to 45 mm. Each patient was operated on in a semi-recumbent dorsal decubitus position. The area surrounding the nipple-areolar complex (NAC) was de-epithelialized. 5 The dermis was incised in the half portion of the de-epithelialized area corresponding to the segmental portion of the breast involved by the tumour ( Figs. 1 and 2 ). The remaining half portion was considered to be the vascular pedicle involved in the glandular and NAC blood supply. A segmentally-oriented excision was then performed with the aim of incorporating the tumour with at least a 1-cm macroscopic margin of normal tissue ( Fig. 3 ). Surgical clips were left in place to mark the perimeter of the tumour resection site. The breast gland was then lifted off the pectoralis muscle with preservation of the fascia over the muscle to obtain sufficient breast tissue advancement to perform the remodelling procedure ( Fig. 4 ). This was done by apposing the two glandular columns to fill in the defect and recentralization of the NAC to recreate a harmonious size and shape. A purse-string closure around the nipple completed the procedure, leaving only a periareolar closure at the end of the operation ( Fig. 5 ). 6 A contralateral breast mammaplasty procedure was performed in 32 patients to achieve symmetry. Standard lumpectomy was performed following the published guidelines for breast conservation using curvilinear incisions for tumours in the upper half of the breast and radial incisions for lesions in the lower half. 1,7 After dissection of the subcutaneous layer, a full thickness resection of the tumour and surrounding glandular tissue was undertaken down to the chest wall. By widely undermining the fibroglandular tissue at the pectoralis fascia and the subcutaneous layer, the glandular defect was eliminated by approximation sutures of the mammary tissue to recreate a harmonious shape. Each patient was operated on in a dorsal decubitus position. A direct incision was used for the 88 patients. No contralateral surgery was done in this group to achieve symmetry. Histological procedure In all cases, the breast tissue specimen was oriented in three dimensions by the surgeon, sent to the pathologis t in a fresh state and measured. Intraoperative radiography of the specimen was performed for non-palpable tumours and correlated with preoperative radiography. The specimens were inked with multiple colours to assist in identifying margins. Following adequate formalin fixation, the tumour histopathologic type was categorised according to the WHO classification. Details regarding tumour staging were recorded according to the International Union Against Cancer TNM staging system. Margin status was analyzed on the anterior side of the specimen (close to the breast skin), the posterior side of the specimen (close to the pectoralis major muscle) and on the lateral sides. Tumour size as determined by the maximal histological size and margin widths was measured by ocular micrometry. Positive margins were defined as the presence of invasive carcinoma or DCIS at an inked margin. Negative pathologic margins were defined as no invasive carcinoma or DCIS present at the inked margins. A margin involved by in situ lobular carcinoma was considered a free margin. All tumours were assessed for an extensive intraductal component (EIC). The volume of each specimen was calculated by multiplying measurements of length, width and height. When secondary surgery was necessary in both the surgeon's and histologist's opinions, the multidisciplinary staff informed and guided the patient in choosing the strategy. When the lateral margins of the specimens were focally involved (<3 mm of the inked surface involved by tumour), reexision was indicated, whereas mastectomy was chosen when the margins were extensively involved (>3 mm of the inked surface involved by tumour). Follow-up Clinical data was collected in relation to age, BMI, radiological preoperative and histological tumour size, tumour location, size of the operative incision, length of the surgical procedure, and hospital stay. The postoperative histological data was of particular importance for the study purpose and included volume of the breast specimen, width of the nearest margins obtained (lateral margins, anterior and posterior margins of the tissue specimen), ratio of clear lateral margins and the number of patients who underwent secondary surgery (reexisional surgery or radical mastectomy). Finally, overall postoperative complications (during a 30-day postoperative period) were recorded in the two groups. All patients were reviewed by the surgeon, radiotherapist, and medical oncologist every 4 months. Bilateral mammograms were performed annually. The cosmetic evaluation was performed separately by a surgeon and a medical oncologist and the mean was recorded. Cosmetic assessment was performed one year after the operation using a grading system. A score of 5 to 1 (5 = excellent; 4 = good; 3 = fair; 2 = mediocre; 1 = poor) was given after evaluation of the following parameters: volumetric symmetry of breasts, shape of breast mounds, symmetry of NAC placement, ipsilateral and contralateral scars. In addition, patient satisfaction was assessed by subjective questionnaire using the same grading scale. Those with an average score of 3 or more were considered to have an acceptable result. Statistical analysis Statistical analysis was performed with Stat View software (Stat View 512, Brain Power, Inc., Calabasas, CA). The data analyst and the pathologist were blinded to the surgery groups. Baseline preoperative variables were compared using χ 2 analysis for categorical data or, when appropriate, the Fisher's exact test; the Mann–Whitney U -test was used to compare medians of non-parametric variables. Spearman's correlation test and regression analysis were used to analyze the correlation between two different variables. P < 0.05 was considered statistically significant. Results Patient and tumour characteristics ( Table 1 ) The patients undergoing DML were younger than the patients who chose SL. However, mean (SD) body mass index did not differ between groups. The two groups were comparable for radiological tumour size, preoperative histology, tumour location within the breast and rate of axillary dissection. In the DML group, patient breast biometries and tumour location were recorded. The average notch to areola distance was 23.4 (2.1) and the average nipple-areola complex diameter was 4.7 (0.65). The average distance of the tumour from the areola was 4.72 (1.9) cm for tumours of the superior external quadrant, 3.9 (1.7) for superior internal quadrant tumours, 4.5 (1.9) for inferior external quadrant tumours and 3.7 (1.6) for inferior internal quadrant tumours. The diameter of the de-epithelialized surface was 9.4 (1.3) cm, and the length of the skin incision to perform the breast resection was 14.5 (1.9) cm. In the SL group the average length of the skin incision to perform breast resection was 5.2 (1.2 cm). This was significantly less than the incision length obtained with the DML technique (14.5 ± 1.9 cm; p = 0.0007). The average operative time in the DML group was longer than in the SL group [mean (SD): 107.1 (43.5) minutes for the DML patients versus 80.5 (30.9) minutes for the SL patients; p = 0.002]. However, no difference was found for the duration of postoperative hospital stay [mean (SD): 5.2 (4.5) days for the DML patients versus 5.1 (3.6) days for SL; p = 0.86]. In the DML group, the average operative time was 113 (44.3) minutes when a contralateral procedure was performed, versus 80.8 (30.3) minutes when no contralateral surgery was done ( p = 0.05). Finally, operative time was not different between DML patients without contralateral mastopexy and SL patients ( p = 0.91). There were early complications in patients in each group. Delayed wound healing was recorded in 2 and 3 patients, respectively. Postoperative hematoma was reported in one case in each group, whereas one case of partial areola epidermolisis was seen in the DML group ( p = 0.2). No extra complications were recorded for the opposite breast of the 32 patients of the DML group who underwent contralateral mammaplasty. Histological data ( Table 2 ) The data on postoperative histological factors are shown on Table 2 . The pT category, histological size and tumour histology did not differ between groups. A larger volume of breast tissue was excised during DML compared with SL but the rate of free lateral margins was not higher in the DML than in the SL group. However, Spearman's correlation test showed a significant correlation between the volume of glandular resection and the anterior margin width ( R 2 = 0.39; p < 0.0001), and between the lateral margin widths ( R 2 = 0.21; p = 0.03). Five patients of the DML group and 13 of the SL group had positive lateral margins ( p = 0.18). The patients were comparable for mean (SD) body mass index ( p = 0.76), radiological tumour size ( p = 0.27), histological tumour size ( p = 0.43) and tumour histology ( p = 0.76). All the patients with positive margins underwent secondary surgery. Four patients of the DML group and 11 of the SL group underwent mastectomy. Cosmetic evaluation ( Table 3 ) At the time of the cosmetic evaluation, all the patients had completed their postoperative radiation therapy. The results of the cosmetic evaluation are reported on Table 3 . The clinician assessment of cosmetic outcome reported a higher rate of acceptable result in the DML group than in the SL group ( p = 0.006). However, patient's assessment did not show difference of cosmetic satisfaction between patients groups ( p = 0.23). Discussion Although oncoplastic procedures are a relatively novel approach to breast cancer treatment, several reports have been published demonstrating the widespread appeal of these techniques. 5,12–14 Our comparative study showed the clear advantage of DML vs SL in terms of the length of the skin incision, volume of the resected breast spec imen and cosmetic results. A poor cosmetic result after breast-conserving therapy is a very undesirable outcome, especially for younger women who have high expectations and active social lives. DML and other oncoplastic techniques, which usually use optimal one-stage glandular reconstruction and plastic reshaping of the healthy breast, are well accepted by younger women. This can explain the younger age of the patients undergoing DML in our study (a mean of 46 years) compared with the patients who chose SL. Similar data have been reported previously. 11,12 Localisation, length and orientation of the scar are among the surgical factors that influence the cosmetic outcome and may explain the higher rate of acceptable result reported by clinicians in the DML group than in the SL group. In the same way, we found no differences in cosmetic score reported by the patients. This feature was previously reported by others. This lack of difference may reflect it being more convenient for the patient to accept and respect the cosmetic outcome as it is. 13,14 The average radiological and histological tumour sizes, distribution of pT categories, tumour locations within the breas t and histologic al patterns of the tumours were comparable between groups. Interestingly, the average specimen height was not different between groups, which confirmed that breast resection was performed down to the chest wall with both DML and SL. Conversely, the average specimen width, length and volume were greater in the DML group because of the large glandular exposure created by the wide skin-sparing dissection. In the DML group, the skin incision was nearly three-fold longer than in the SL group. This surgical exposure was obtained with a final scar located around the nipple areola complex without further postoperative complications. DML allows larger superficial dissection and at the end result is greater volume tissue resection. DML has other major advantage over direct incision regarding length and location of final scar. Periareolar de-epithelialization provides a large operative area and a final aesthetically acceptable periareolar scar. In addition, when a radical mastectomy is needed, the periareolar location of the scar enables the surgeon to perform an adequate skin incision incorporating the nipple areola complex, which may facilitate future breast reconstruction. 1 Conversely, transverse incisions in the upper half of the breast may necessitate a high mastectomy scar to incorporate the biopsy scar, and vertical incisions in the lower half may involve transgressing this scar when mastectomy is needed. 1 In addition, when skin-sparing mastectomy is indicated, the periareolar incision centered on the nipple areola complex offers an excellent option. Unfavorable scar locations (curvilinear incision close to the areola, direct radial scar) may result in skin flaps with compromised blood flow. 4 The most rigorous method to determine the differences between the two types of surgical management would have been a randomized trial. We nevertheless think we were able to limit some of the bias by using narrowly-defined inclusion criteria. This study design therefore may have sufficient statistical power to encourage serious consideration of our results and — we hope — to generate further comparative studies with longer follow-up. Conclusion This study reports on a small number of patients in self-selected comparison groups. Despite these shortcomings, the data strongly indicates that DML surgery of breas t tumours may be a useful alternative to SL not only in terms of accurate breast tissue resection but also in term of cosmetic results. Obviously, a large controlled trial with longer follow-up that rigorously analyzes both oncological and economic aspects is needed to confirm what we already suspect: oncoplastic management will one day be the gold standard of breast-conserving surgery. References 1 B.O. Anderson R. Masetti M.J. Silverstein Oncoplastic approaches to partial mastectomy: an overview of volume-displacement techniques Lancet Oncol 6 2005 145 157 2 R. Masetti P.G. Pirulli S. Magno Oncoplastic techniques in the conservative surgical treatment of breast cancer Breast cancer 7 2000 276 280 3 C. Amanti A. Moscaroli M. Lo Russo Periareolar subcutaneous quadrantectomy: a new approach in breast cancer surgery G Chir 23 2002 445 449 4 D.A. Hidalgo P.J. Borgen J.A. Petrek A.H. Heerdt H.S. Cody J.J. Disa Immediate reconstruction after complete skin-sparing mastectomy with autologus tissue J Am Coll Surg 187 1998 17 21 5 N. Bricout Chirurgie du sein 1992 Springer-Verlag Paris 130–42 6 L. Benelli A new periareolar mammaplasty: the “round block” technique Aesthetic Plast Surg 14 1990 93 100 7 E.J.T. Rudgers Guidelines to assure quality in breast cancer surgery Eur J Surg Oncol 31 2005 568 576 11 N. Kaur J.Y. Petit M. Rietjens Comparative study of surgical margins in oncoplastic surgery and quadrantectomy in breast cancer Ann Surg Oncol 12 2005 539 545 12 Giacalone PL, Roger P, Dubon O, et al., Comparative study of the accuracy of breast resection in oncoplastic surgery and quadrantectomy in breast cancer. Ann Surg Oncol , in press. 13 L.A. Woerdeman J.J. Hage E.A. Thio F.A. Zoetmulder E.J. Rutgers Breast-conserving therapy in patients with a relatively large (T2 or T3) breast cancer: long-term local control and cosmetic outcome of a feasibility study Plast Reconstr Surg 113 2004 1607 1616 14 H. Kaija S. Rauni I. Jorma H. Matti Consistency of patient- and doctor-assessed cosmetic outcome after conservative treatment of breast cancer Breast Cancer Res Treat 45 1997 225 228