Subcutaneous mastectomy for primary breast cancer and ductal carcinoma in situ

European Journal of Surgical Oncology 1997; 23:343-347 Subcutaneous mastectomy for primary breast cancer and ductal carcinoma in situ K. L. Cheung, R. W. Blamey, J. F. R. Robertson, C. W. Elston and I. O. Ellis Professorial Unit of Surgery, Nottingham City Hospital, Hucknall Road, Nottingham NG5 IPB, UK Subcutaneous mastectomy (SCM) has been performed in 323 patients with invasive breast cancer or ductal carcinoma ill situ (DCIS) over the last 20 years. This paper assesses the oncological safety of this operation with particular reference to local recurrences (LR) in the skin flaps. LR was assessed in the 134 patients who did not receive post-operative irradiation to the flaps and were followed up for a minimum of 30 months. The rates of LR were similar for SCM (16%) and for women who underwent simple mastectomy without post-operative irradiation (SM) for invasive cancer (14%) over the same time period, both overall and after prognostic stratification. In DCIS LR was only found in women treated with SCM. Four of the five recurrences (from 33 women) were in the nipple and later in the series patients with microscopic tumour involvement in the nipple received prophylactic irradiation to the nipple only; five patients treated in this way have not suffered LR. Subcutaneous mastectomy followed by insertion of a prosthesis is a safe alternative to mastectomy, since it carries no higher risk of LR. it is the simplest form of reconstruction in those proceeding to silicone implant and has the potential of achieving an excellent cosmetic result. Failure does not compromise the chance of more complex reconstruction procedures. Key words: Subcutaneous mastectomy; invasive breast cancer; ductal carcinoma in situ; local recurrence. Introduction Treatment with breast conservation is now widely used but is inappropriate in some patients because of a high risk of local recurrence; there are recognizable tumour factors which can be used to select these patients and they are then advised to undergo mastectomyj Similarly, patients with extensive ductal carcinoma in situ (DCIS) are also advised to undergo mastectomy. A number of women who undergo mastectomy in these circumstances request reconstruction. Subcutaneous mastectomy with subsequent reconstruction by insertion of a silicone prosthesis is the simple alternative; it offers a chance of an excellent cosmetic result, which is difficult to match using other reconstructive procedures. If the procedure fails due to flap necrosis or infection, then it is still possible to use more complex reconstructive methods. Whether simple or subcutaneous mastectomy is used would not be expected to affect long-term survival, which is dependent on the presence of distant metastases at diagnosis. Local recurrence (LR) rate might be affected if the procedure used did not remove all the breast tissue. To assess oncological safety this paper compares the local recurrence rate after subcutaneous mastectomy to that after simple mastectomy. Analysis of LR was the main purpose of this paper and from the point of view of the amount of breast Correspondence to: Mr K. L. Cheung FRCSEd, Department of Surgery, The University of Hong Kong, Queen Mary Hospital, Hong Kong. tissue removed, the operation has not changed during the period of the study. An analysis of the effects and complications of SCM and of implant insertion is also given. This is less well reported than the rate of LR as the 20 years of the study represent a learning experience, and techniques important to the reconstructive part of the procedure have varied (prosthesis used, site of incision, immediate or late implant, prophylactic antibiotic usage, tissue expansion, subcutaneous or submuscular implant). Patients and methods Women aged < 70 years presenting to the breast clinic at Nottingham City Hospital with operable primary invasive breast cancer <5 cm in size or with DCIS were treated by mastectomy from 1973 and, since 1979, increasingly by wide local excision and intact breast irradiation. No further consideration is made in this paper of those women who underwent treatment with breast conservation. Since 1974 women advised to undergo mastectomy have been offered the option ofsubcutaneous mastectomy (SCM), followed by insertion of a silicone prosthesis. Up to 1994, 323 patients with a mean age of 45 years (range: 27-69) chose SCM. Patients undergoing simple mastectomy (SM) during the same period numbered 1696. Both groups of patients were managed in the same surgical unit with identical protocols for SM and SCM with respect to adjuvant systemic therapy and radiotherapy, and both 0748-7983/97/040343 +05 $12.00/0 1997 W.B. Saunders Company Limited

344 K. L. Cheung et al. groups were followed up in the same post-mastectomy clinic. No prophylactic irradiation was given until the late 1980s. Flap irradiation (45 Gy for 15 sessions with 8 MeV electrons) was then added to cases with poor histological grade, nodal involvement and lymphovascular invasion, factors which had been shown to be associated with a high likelihood of flap recurrence.-" From that date, patients undergoing SCM in whom the nipple area was microscopically involved by tumour, were also given prophylactic irradiation by D. M. (45 Gy for 15 sessions with electrons) to the nipple and areola (a 4--6 cm 2 field), if the flaps were not otherwise to be irradiated. Surgical technique Subcutaneous mastectomy was performed in the earlier part of the series through an inframammary incision; later by an incision down the lateral side of the breast. Both incisions allowed access to the axilla and to the second intercostal space for internal mammary node biopsy; the breast was fully dissected off the chest wall with access as high as the clavicle. The skin flaps were of the same thickness as in SM. All breast tissue was excised and the area underlying the nipple was marked with a stitch to aid histological examination. The operation differed from SM only in that all the skin, including the nipple, was left intact. The weight of the excised breast tissue was recorded for the purpose of later selecting a prosthesis. Following SCM, the patients were offered reconstruction with insertion of an implant. Sometimes this was carried out at the initial operation but usually around 6 months later. Silicone implants were the commonly used prostheses, together with the Becker expander (Mentor, CA) in a few cases. Most implants were placed subcutaneously. An analysis of local recurrence Local recurrence (LR) was defined as a histologically proven recurrent tumour in the mastectomy skin flap. Regional recurrence is not studied here, since the nodal biopsy procedures and the management of the axilla were the same in and after SCM and SM, at any one time during this study. Patients were included in the assessment of LR whether or not they received a silicone prosthesis. Previous experience had shown that over 70% of LR occurred within the first 2 years. 3'4 In this study, therefore, only patients with a minimum follow-up of 30 months are included. The rate of LR is greatly reduced by irradiation, 5 and for comparison of the operations the analysis is thus restricted to those cases without post-operative irradiation; 134 cases. Separate analyses have been performed for patients with invasive cancer and for those with DCIS (101 invasive, 33 DCIS). The 101 patients undergoing SCM for invasive cancer had a mean age of 46 years (range: 27-69) and a median follow-up of 122 months (range: 30-240). During the same time period 535 patients with the same age range, also with no irradiation and with a minimum follow-up of 30 months, underwent SM for invasive cancer, with a median followup of 93 months (range: 30-244). The LR rates in the two groups were compared overall and after stratification for prognostic factors according to the Nottingham Prognostic Index (NPI, a combination of tumour size, grade and stage), 6 and for the presence of lymphovascular invasion,a powerful predictor of LR. 2 Data on lymphovascular invasion were available in 94 and 476 patients with SCM and SM, respectively. SCM was also performed for DCIS in 33 patients with a mean age of 48 years (range: 31-64). The median followup was 78 months (range: 36-228). During the same period 57 patients with DCIS of identical age range underwent SM with a median follow-up of 75 months (range: 33-192). As the aforementioned prognostic factors are not relevant in DCIS, comparison has been made for DCIS on the overall LR rates between the two operations. Statistical methods Data regarding the primary tumour status and LR were prospectively collected. A retrospective analysis was carried out with the help of standardized biomedical computer programmes (SPSSPC). Differences in the groups of patients were compared using survival tables according to Lee--desu statistics and the chi-squared test with Yates' correction when appropriate. Complications of SCM and of #Tsertion of a silicone implant Data regarding the complications of SCM were available in 277 patients; 206 received an implant coincidentally or subsequently, and the complications are reported. Final comment on result Assessment of the cosmetic result following insertion of an implant was based on findings in the post-mastectomy clinic about 6 months after the reconstruction, as documented in the hospital records. It was classified by the surgeon as good (unclothed, breasts apparently equal size and symmetrical), moderate (appeared normal only in a brassiere), or poor. At the time of the review a simple questionnaire on patient satisfaction' was sent out to all surviving patients who had implants inserted, asking their opinion on the acceptability of the whole procedure. Results Analysis of local recurrence (Table 1) lnvasive cancer. There were 16 LRs in the SCM group and 74 in the SIVi,--giving overall LR rates after SCM and SM of 16% and 14%, respectively. There were no significant differences in the LR rates between the two groups overall or when stratified according to the NPI (Table 1), or in the presence of lymphovascular invasion. After SCM all except one of the 16 LRs were of the single spot type, 7 treatable by simple excisional surgery. Two patients had microscopic involvement by DCIS in the nipple area in the SCM operative specimens and subsequently developed LR in the nipple; they were successfully treated

Subcutaneous mastectomy for prhnary breast cancel" and DCIS 345 Table 1. Local recurrence (LR) in patients undergoing subcutaneous mastectomy (SCM) and simple mastectomy (SM) Total SCM SM (all cases) n LR % n LR % n LR % DCIS 33 5 15 Invasive cancer 101 16 16 lnvasive cancer Good NPI 51 7 7 Moderate NPI 43 8 18 Poor NPl 7 1 14 lnvasive cancer Lymphovascular invasion absent 75 10 13 Lymphovascular invasion present 19 5 26 57 0 0 -- -- -- 535 74 14 -- -- -- 210 26 12 261 33 13 298 40 13 341 48 14 27 8 30 34 9 26 394 52 13 469 62 13 82 15 18 101 20 20 NPI =Nottingham Prognostic Index. by excision of the nipple and areola. Thirteen patients had LR in the mastectomy skin flap and were treated by either excision alone (n = 11) or excision followed by radiotherapy. The LR in the remaining patient was situated along the border of the sternum and was probably, in fact, a regional recurrence in an internal mammary nodes and showed complete response to radiotherapy. All except two had no further LR--one had re-excision once and one twice, after which no further LR occurred. No patient who underwent SCM developed uncontrolled LR. Seventy-four LR occurred in the SM group; these were of the single spot (n=64) and multiple spot 7 (n= 10) types. Over half of the single-spot LR were successfully treated by excision alone while most of the multiple-spot LRs received radiotherapy. Ductal carcinoma in situ. Local recurrences, all of the single spot type, were found in five (15%) of the 33 patients undergoing SCM, while no LR was encountered in their counterparts with SM. Four of the five patients' LRs occurred in the nipple area. Two were treated by excision alone and two by excision followed by radiotherapy; one developed further LR requiring re-excision. The remaining patient developed LR away from the nipple and was successfully treated by simple excision and post-operative flap irradiation. Of the four recurrences at the nipple, three had DCIS in the nipple ducts on histology of the SCM specimen. As a result, since 1986, patients with microscopic tumour involvement in the nipple area at SCM have received prophylactic irradiation to the nipple only; five patients (four CIS, one invasive cancer) treated in this way have not developed LR after a median follow-up of 46 months (range: 18-131). Complications of subcutaneous mastecton O, and of insertion of a silicone hnplant Subcutaneous mastectomy Data are available on 277 of the 323 patients who underwent SCM (both those who did and did not receive flap irradiation). Eighty-nine of the 277 patients had undergone a recent open biopsy in the same breast prior to SCM. The mean operation duration of SCM has been entered on the theatre computer over the last 2 years and during that time it has been a mean of 60min (range: 20-170). Skin flap necrosis was the major complication, occurring in 26 patients. Serious necrosis requiring surgical debridement occurred in 18 patients (6.5%): none of these 18 proceeded to subsequent insertion of an implant. Significantly more necrosis was found in patients with immediate implant insertion, but there was no significant association with recent open biopsy nor with the operation duration (Table 2). Implant insertion. Seventy-one patients decided not to proceed to delayed implant insertion, 18 patients had suffered flap necrosis at SCM and were unable to proceed to implant insertion. Insertion of an implant was carried out as part of the SCM in 24 patients and delayed in 182 patients after a mean duration of 8 months (range: 2-37) following SCM. All implant insertions were completed within an operating time of 30 rain. The great majority of implants were placed subcutaneously. One hundred and eighty-four patients had silicone implants, whilst 22 underwent insertion of a tissue expander, which in three was subsequently followed by insertion of a definitive silicone implant. No patient undergoing a delayed implant lost their prosthesis in the immediate post-operative period. Of the 206 patients undergoing immediate or delayed implant insertion, 60 patients had the implants removed for Table 2. Flap necrosis in subcutaneous mastectomy with implant inserted immediately and after a delay No necrosis Necrosis Immediate implant 17 7 No immediate implant 230 23

346 K.L. Cheung et al. Reason Table 3. Reasons for removal of implant Number of patients Early Necrosis/sepsis 26 Haematoma 5 Late Capsule fonnation l0 Rupture I Zumour recurrence 4 Patient request 14 Total 60 Table 4. Cosmetic result assessed by surgeon of implant insertion following subcutaneous mastectomy Result Number of patients Good 83 Moderate 49 Poor 6 Table 5. Results of questionnaires sent to patients with implant insertion following subcutaneous mastectomy Would choose SCM again (n) Would not choose SCM and implant again (n) Pleased 42 3 Not pleased 7 10 various reasons (Table 3) and 19 developed severe capsule formation requiring open capsulotomy. Final cosmetic result. After excluding the 60 patients with implants removed, the surgeon's assessment of cosmetic result was available in 138 of the remaining 146 patients who had implants (Table 4). Those still alive were sent simple questionnaires on patient satisfaction and 62 were returned: 42 patients were pleased with their results; seven not happy with the result said that they would still choose to try to obtain a better cosmetic result by SCM if faced with the same situation again (Table 5). Discussion The main intention of this paper was to assess LR rates after SCM. Distant recurrence and survival of a patient with breast cancer depend on the presence of metastatic disease at the time of presentation and are not influenced by the type of local operation performed. In our centre there has been no difference in axillary prophylaxis between patients undergoing SM and SCM, therefore regional recurrence is unlikely to be affected. Long-term LR rates might be dependent on the type of operation used and have been evaluated to assess the oncological safety of SCM. A minimum follow-up period of 30 months has been used because over 70% of local recurrences occurred within 2 years.'comparison of LR rates was made between patients treated by SCM with those treated by SM. To allow for prognostic factors which we know to influence the rate of LR., stratifications according to NPI, age and the presence of lymphovascular invasion, have been used. Prophylactic flap irradiation is known to reduce LR markedly; ~ for the comparison in this paper only women who did not receive flap irradiation have been analysed. There were no differences in protocol for selection for flap irradiation between patients with SCM and with SM. Patients with DCIS alone have been considered separately. These strict criteria have made a fair comparison of subcutaneous mastectomy with conventional simple mastectomy. The issue of selection for prophylactic nipple irradiation is also addressed. We confirm our earlier result s that there is no significant difference in the LR rates between well-matched patients undergoing SCM or SM for invasive primary breast cancer. The overall rates of LR are also comparable to other series studying LR, following SCM or SM without irradiation. 5"9'~ With one exception all LR was of single-spot type which was amenable to simple excision,-' and no patients undergoing SCM developed uncontrollable local disease, Our results are consistent with those in other series, showing that SCM is an oncologically safe option. ~ '" Subcutaneous mastectomy differs from simple mastectomy in that the nipple and the skin are retained; histological involvement of the nipple was found in 10% of cases at SCM. In the patients with DC1S the LR which occurred were, with one exception, in the nipple and in patients with microscopic nipple involvement found in the operative histology. Treatment of the nipple with radiotherapy seems a logical addition in those with this finding and has to date eliminated nipple recurrence. Subcutaneous mastectomy is simple and relatively cheap and relatively quick compared with the more complex reconstructive procedures such as latissimus dorsi, TRAM or free flaps. The main immediate post-operative complications are skin flap and areolar necrosis and/or sepsis and haematoma formation, leading to implant extrusion. The assessment of cosmesis in this paper is of less relevance than the rate of LR, since the reconstructive element represents a learning experience over 20 years. Insertion of an implant at the time of SCM led to a high rate of flap necrosis and we suggest that this should be avoided if the implant is to be placed subcutaneously. The main long-term complication is capsule formation and our rate" (14%) of capsule formation necessitating open capsulotomy or implant removal is comparable to other series, t-''t3 Again, placing a tissue expander submuscularly may overcome this problem. The final cosmetic result is similar to that we found earlier in our series. Certainly a number of'perfect' cosmetic results are obtained from this technique, when it is difficult to see that any operation has been performed, and these results are better than any achieved using flap reconstruction.

Subcutaneous mastectomy for prhnary breast cancer and DCIS 347 Analysis of the questionnaire indicates that women are generally happy with their result from this technique. As patient numbers available for assessment differed, as some had immediate implants, some were delayed and some decided not to proceed to implant--a summary of the success rate of the operation is that SCM is technically successful in 85% but 15% suffered skin loss and could not proceed to implant; 10% of those proceeding to implant suffered delayed implant loss and 15% had a poor cosmetic result. Overall, 20% of those who embarked on the choice of SCM and implant had a near perfect visual result and 49% an acceptable result when dressed. Conclusion This series has demonstrated that the incidence of LR after SCM is no higher than after SM. No cases of uncontrolled LR have resulted. The operation is as oncologically sound as SM, although the subnipple area must be examined histologically and radiotherapy given to the nipple if the ducts are involved. Skin flaps must be cut to the same thickness as in SM and the extent of dissection at the level of the muscle must be as wide as in SM. Subcutaneous mastectomy with silicone implant insertion is much simpler and quicker and a great deal less expensive than reconstructive procedures based on myocutaneous flaps or free flaps. In this series it gave a 20% chance of a near perfect appearance with a normal nipple and with just one inconspicuous scar, a better appearance than is achieved by the more complex reconstructive operations. Even when the appearance was less satisfactory, many women appeared pleased with their result. However, this series represented a learning experience in one of the few units carrying out this operation for many years. Subcutaneous mastectomy is often dismissed by those with little experience of the procedure, on the spurious grounds of higher LR or poor cosmesis. We suggest that SCM should be the first choice when advising patients with invasive breast cancer or ductal carcinoma in situ who wish for good cosmesis but are unsuitable for treatment with breast conservation. If the procedure fails to produce an acceptable result little is lost, since the use of more complex reconstructive procedures still remains an option. References 1. Sibbering DM, Galea MH, Morgan DAL, Elston CW, Ellis IO, Robertson JFR, Blarney RW. Selection criteria for breast conservation in primary operable invasive breast cancer. Eur J Cancer 1995; 3IA: 2191-5. 2. O'Rourke S, Galea MH, Morgan DAL, et al. Local recurrence after simple mastectomy. Br J Surg 1994; 81: 386--9. 3. Weichseibaum R, Marck A, Hellman S. The role of postoperative irradiation in carcinoma of the breast. Cancer 1976; 37: 2682-90. 4. Zimmerman KW, Montague ED, Fletcher GH. Frequency of anatomical distribution and management of local recurrence" after a definitive therapy for breast cancer. Cancer 1966; 27: 67-74. 5. Berstock DA, Houghton J, Haybittle J. Baum M. The role of radiotherapy following total mastectomy for patients with early breast cancer. Worm J Surg 1985: 9: 667-70. 6. Galea MH, Blamey RW, Elston CW, Ellis IO. The Nottingham Prognostic Index in primary breast cancer. Br Cancer Res Treatment 1992; 22: 207-19. 7. Blacklay PF, Campbell FC, Hinton CE Blarney RW. Morgan DAL, Elston CW. Haybittle JL. Patterns of flap recurrence following mastectomy. Br J Surg 1985; 72: 719-20. 8. Hinton CE Doyle PJ, Blamey RW, Davies CJ, Holliday HW, Elston CW. Subcutaneous mastectomy for primary operable breast cancer. Br J Surg 1984; 71: 469-72. 9. Andry G, Sucio S, Vico E Faverly D, Andry-t'Hooft M, Verhest A, Nogaret J-M, Mattheiem W. Locoregional recurrences after 649 modified radical mastectomies; incidence and significance. Eur J Surg Oncol 1989; 15: 476-85. 10. Palmer BV, Mannur KR, Ross WB. Subcutaneous mastectomy with immediate reconstruction as treatment for early breast cancer. Br J Surg 1992; 79: 1309-11. I1. lino, Ishikita T, Takeo T, et al. Subcutaneous mastectomy with axillary dissection for early breast cancer. Anticancer Res 1993; 13: 1183-6. 12. Woods JE, Verheyden CN. Pitfalls and problems with subcutaneous mastectomy. Mayo Clin Proc 1980; 55: 687-93. 13. Holzgreve W, Belier FK. Surgical complications and followup evaluation of 163 patients with subcutaneous mastectomy. Aesthetic Plast Surg 1987; 11: 45-8. Accepted for p,tblication 3 April 1997