Use of abdominal tissue is considered the best available option for autologous breast reconstruction after mastectomy. Patients with significant abdominal scars from previous surgeries frequently present for breast reconstruction and often seek or are only suited for autologous tissue reconstruction. The use of pedicled and free abdominal flaps in such patients with multiple preexisting abdominal scars has many potential problems.1 The vascularity of part or whole of the flap in these patients is potentially unreliable thus precluding its use as the first choice for many.2-4 This is compounded by the concerns regarding surgical disruption of the integrity of the abdominal wall which may lead to weakness, laxity, bulges, or frank herniation. Various operative strategies have been suggested to improve transverse rectus abdominis muscle (TRAM) flap survival and reduce donor site morbidity in patients with preexisting abdominal scars with variable success.3-9
Preexisting single abdominal scars have been shown not to increase the flap complications in patients undergoing abdominal flap breast reconstruction5,7,9 but may increase donor site morbidity.5,10,11 Previous multiple abdominal surgeries have been documented to increase postoperative abdominal wall morbidity following abdominal flap harvest but not shown to increase the risk of flap failure.3-8
There are numerous studies looking into the effect of single abdominal scars on flap and donor site outcomes,3,4,6,9,12-14 in contrast to the paucity of literature concerning the effect of previous multiple abdominal surgeries.8 In our practice, previous abdominal surgery is not considered to be a contraindication to abdominal flap breast reconstruction.9 Although patients may have a particular scar, it is more important to ascertain what surgery led to the scar because with each surgery there is increased danger to the vascularity with more scarring underneath. Depending on the locations of the scars and the type of previous surgery, the senior author (C.M.M.) makes adaptations to the reconstructive approach employed. The objective of this study was to compare the flap and donor site outcomes in patients with previous multiple abdominal surgery, who underwent an adapted approach to abdominal flap breast reconstruction, to those without previous abdominal surgery. The information gleaned could then be used in counseling patients with multiple previous abdominal surgeries requesting free flap breast reconstruction. It would also provide recommendations for successful execution of this surgery in this challenging group of patients.
PATIENTS AND METHODS
Patients with previous multiple abdominal operations who underwent abdominal flap breast reconstruction by a single surgeon (C.M.M.) from January 2000 to December 2009 were identified from the free flap database. Their case notes were reviewed for data on indications for previous abdominal surgery and locations of the previous incisions (thus current scars).
The flap type (deep inferior epigastric perforator [DIEP], muscle-sparing [MS]-I, MS-II, MS-III TRAM),10 abdominal flap design, reconstruction timing (immediate or delayed), flap outcomes, and donor site morbidity were also recorded. A note was also made of patient's age, body mass index (BMI), smoking history, and comorbidities.
They were then retrospectively compared to a group of patients who underwent the same reconstruction but with no previous abdominal surgery. We used 2 control group designs in this studyone randomized (n = 20) and the other alternate sequential allocation (n = 20). The latter group was selected by choosing the patient immediately following each patient with multiple previous abdominal surgeries. The primary reason for these 2 control groups was to avoid selection bias and nullify the surgeon's inevitable learning curve with time (time-experience bias). Only patients operated on by the same surgeon (C.M.M.) were included to avoid interoperator variability.
We used total flap loss, partial flap loss, fat necrosis (subcutaneous firmness of at least 2 cm),15 wound infection, and wound breakdown as measures of flap outcomes. Donor site outcomes included seroma, wound infections (culture-proven or cellulitis needing antibiotics), wound breakdown, skin flap necrosis, and abdominal wall laxity (bulge, hernia, or weakness). Numerical data (age, BMI) were compared using Mann Whitney U test while categorical (nominal) data were compared with Fisher exact test.
Following flap harvest, the rectus muscle was approximated with a running 2/0 Vicryl suture while the anterior rectus sheath was repaired with running looped 0 Ethilon in 2 layers. None of the patients required mesh repair, and the abdominal wound closure was achieved with interrupted 2/0 PDS (Ethicon) to Scarpa's fascia and 3/0 Monocryl (Ethicon) in 2 layers (interrupted deep dermal and continuous subcuticular). Two suction drains were used and removed prior to discharge from hospital when draining less than 30 mL per day. A supportive binder was worn day and night for 3 months.
Over the 10-year period, 20 patients with multiple previous major abdominal surgeries underwent unilateral (n = 18) or bilateral (n = 2) abdominal flap breast reconstruction (total = 22 flaps). There were 9 immediate and 11 delayed reconstructions. The reconstruction types included 9 free muscle-sparing TRAMs and 13 free DIEP flaps. Patients' ages ranged from 42 to 63 years (mean = 54) and the average BMI was 27.45 kg/m2 (r, 24-35). Three patients were smokers at the time of the surgery (15%) (Table 1). Fourteen patients had previous chemotherapy, and 10 had previous radiotherapy (RT). One patient had a history of a deep venous thrombosis (DVT) while another had a previous history of lymphoma. The abdominal scars encountered were Pfannenstiel, caesarean sections, midline and paramedian laparotomy, subcostal scars, nephrectomy, hernia repair, laparoscopic trocar sites, and appendicectomy (Table 2). The indications for the incisions were largely obstetric and gynecological. Most patients in the series had 2 scars but 2 patients had 5 abdominal scars each (Table 3). No previously damaged donor blood vessels were encountered in any of the patients at surgery. For the 13 DIEP flaps, 2 perforators were included in each flap; no flap was raised on a single perforator.
The breasts reconstructed were of the following cup sizes: C = 3, D = 7, and DD = 5 (unspecified = 5). The mastectomy specimens weighed an average of 675 g (r, 485-903). Five patients (38%) had contralateral balancing breast surgery comprising 3 breast reductions and 2 mastopexies.
There were no flap failures, and none of the flaps developed partial skin loss, infection, or significant fat necrosis. Five patients had donor site seroma, which needed aspiration, but none required operative drainage as they all settled with repeated aspirations (average = 3). All healing problems were minor and required no operative intervention. They comprised 3 partial wound breakdowns and 2 wound infections needing antibiotic treatment (Tables 4 and 5). One of the dehisced wounds required debridement and direct closure, as the patient was about to go on vacation and could not wait for spontaneous healing with conservative measures. After an average follow-up of 26 months (r, 10-36), none of the patients developed abdominal wall weakness, bulging, or frank herniation. One patient in the multiple scar group has noticed a localized bulge at the site of the DIEP flap harvest site but did not need treatment.
The index and 2 control groups were comparable in terms of age, BMI, previous cancer treatment, and comorbidities (Fig 1). There were no significant differences in flap and donor site outcomes between the patients with multiple previous abdominal scars and the controls groups (the results are shown for the sequential allocation control versus the index group in Tables 4 and 5, as these were identical when compared with the random controls)
Our study showed that similar results can be obtained in patients who have had previous multiple abdominal surgeries, as in patients with no scars. This has been previously documented for patients with multiple scars. However, none of these previous studies have analyzed the effect of the previous multiple abdominal surgeries with respect to the indication for those surgeries. It is important to do this analysis because clinically at each operation the vessels (pedicle, perforators or both) are put at risk by the abdominal surgery even though it is being performed via the same incision. Therefore, an analysis of flap and donor site outcomes merely based on scars is an oversimplification.
Patients with previous major abdominal surgery are often informed that they are not suitable for abdominal flap breast reconstruction and in our institution often ended up in the senior author's clinic (C.M.M.) because he does not consider the scars of previous major abdominal surgery to be an absolute contraindication for surgery. Indeed, in his practice, almost a third of all patients who undergo abdominal flap breast reconstruction have had preexisting scars.9
Reconstructing patients with previous multiple surgical scars through the abdominal wall is especially challenging because tissue availability is limited, flap vascularity is potentially unreliable, and flap harvest may weaken the abdominal wall (Fig 2). However, a number of these patients insisted on abdominal flap reconstruction despite these adverse conditions and potential problems (Figs 3-5). It is therefore important to document the outcomes of these reconstructions in comparison to patients with no previous abdominal surgery both in the immediate and delayed scenarios and formulate proposals for successful execution of this surgery.
In our practice, different operative strategies have been adopted to overcome the limitations posed by abdominal scars especially midline and subcostal ones. The experience has led to formulation of an algorithmic approach to this problem9 (Fig 6). The precautions that have to be taken for successful abdominal flap breast reconstruction in patients with multiple previous abdominal surgery scars are summarized in Table 6.
Flap types multiple scars = a free flap
Our series of multiple scar patients who had previously undergone multiple abdominal surgeries was nearly equally divided between immediate (9/20) and delayed (11/20) breast reconstructions. The most notable finding was that all patients with multiple scars were reconstructed with a free tissue transfer. This is in contrast with our overall experience with abdominal scars in which some flaps were pedicled.9 It is a reflection of the fact that free tissue transfers have a more robust blood supply compared to pedicled TRAMs.16-19 Furthermore, previous scarring may impair or compromise perforator anatomy or size. In addition to disturbing the perforators, abdominal scars may compromise or reduce the perfusion across the scar. We would therefore recommend that if possible more than one perforator should be used and if the scarring precludes perforator flap harvest then one should have a low threshold for intraoperative change of plan for free TRAM flap design.
Surgical incision types
Although some authors contend that Pfannenstiel incisions should not affect the flap perfusion,7 this is by no means universally accepted. The effect of Pfannenstiel incisions on flap circulation depends on how extensive the subfascial and suprafascial dissection was during the surgery, the indication for this surgery and whether any complications (eg, hematoma or infection) occurred following surgery. The latter two may increase scarring thereby making dissection of the perforators difficult while extensive dissection could have damaged the perforators profoundly or irreversibly.
In patients with lower midline scars, only a small amount of tissue on the contralateral side can be reliably included in the flap, as this tissue is variably perfused depending on how old the scar is, its length, and its previous indication (Fig 5).
We therefore determine the perfusion of the contralateral tissue after vessel anastomoses and then discard the portions that are not well perfused. This is similar to the practice of others8 and provides a valuable increase in the horizontal dimension of the flap.
Flap and donor site complications
Although Mehrara et al20 have documented increased complications of both the donor and flap sites, most authors would agree that preexisting abdominal scars do not increase flap complications. This is confirmed by our past and present studies.9
Previous surgery through the abdominal wall weakens the abdominal fascia and/or muscles and therefore it is not surprising that they may lead to abdominal wall laxity.5,10 We did not find an increased rate of donor-site complications in patients with multiple abdominal scars. This contrasts with the findings of Parrett et al (2008), who found a significantly higher rate of abdominal donor-site complications in their DIEP flaps in patients with scars versus those with no scars (24% vs 6%). Subcostal scars are a well-documented cause of increased donor site breakdown in TRAM flaps.5,8,13 None of our open cholecystectomy patients, however, experienced wound breakdown or skin flap necrosis largely due to the algorithmic precautions we undertook.9
Perforator anatomy and vascular imaging
Although usually only one good perforator is mandatory for flap survival especially with an adequate vein (1.5-mm diameter),15,21,22 we strive to use 2 or more perforators in patients with preexisting multiple abdominal scars because of concerns that the scars might have affected perfusion within the flap.
This review study was undertaken before adopting preoperative localization of the perforators with computed tomographic (CT) angiography23,24 by our unit. This is similar to other large series on this subject.8 All patients, whether they have scars or not, had the percutaneous location of their perforators marked out using a hand-held Doppler probe preoperatively.21 We have recently adopted CT angiography for all our patients. These newer imaging techniques for perforator flap anatomy may in the future facilitate the performance of breast reconstructions in patients with multiple scars.23,25
Tips for successful abdominal flap reconstruction in patients with multiple scars (multiple previous abdominal surgery)
Abdominal scars should, however, not be approached with trepidation in patients seeking abdominal flap breast reconstruction. Patients should be counseled about the possibility of an increased risk of donor site complications even with DIEP flaps.8 The precautions we undertake are summarized in Table 6 and the algorithmic approach in Figure 6. When faced with a patient who has multiple scars on her abdomen, one has to keep an open mind as to the most feasible approach. In our opinion, pedicled flaps are contraindicated unless the patient has minor scars such as those from laparoscopy or appendectomy. We select the laterality of the flap based on the side of the lower abdomen, which has the least number of scars and is therefore likely to have the least disruption to its blood supply. In cases in which bilateral reconstruction is required, the side with the most scars should be a muscle-sparing free TRAM flap as opposed to a DIEP flap (Figs 4 and 5).
It is important not to commit to a pedicle until exploration of the deep inferior epigastric vessels (DIEVs) has shown them not only to be intact and in continuity but also to be of adequate calibre and unaffected by any previous scarring. If on one side the vessels are scarred or of poor quality, then the other side must be explored and the flap pedicle should be changed appropriately. It is therefore vital that no skin be de-epithelialized before flap transfer and successful microvascular anastomosis. The surgeon therefore needs to be versatile in the use of both the left and right DIEVs, the internal mammary, and the axillary recipient sites as well as different flap inset orientations dictated not only by the shape of the breast but also by the size, site of the flap harvested, and some of its portions that perfuse satisfactorily after microvascular anastomoses.
The next important consideration is the size of the flap. It is vital to make the flap from a multiply scarred abdomen as large as possible and discard portions of it only after perfusion has been established. If the flap size is rather small, the patient should have been prepared or counseled and consented for contralateral balancing breast reduction or mastopexy. Hence, one has to use simultaneous contralateral balancing surgery liberally. Alternatively, fat transfer may be planned for a later stage to increase the size of the flap.
The donor site morbidity can be reduced by minimizing the tension on the abdominoplasty flap, minimal undermining in patients with transverse upper abdominal scars6,9,13 and appropriate management of the residual upper abdominal scars in the abdomen. The latter includes Z-plasties or multiple stab incisions of the vertical scars. Tension on the abdominoplasty flap can be reduced by minimizing the size of the skin paddle size and undermining of the mons pubis or inguinal areas as suggested by our algorithmic approach.9
This inevitably results in a high abdominoplasty scar, and the patient must be warned about this. This situation also pertains when the suprapubic area has been multiply scarred by multilevel incisions rendering it unusable for inclusion in the flap.
Although we prefer to use DIEP flaps whenever possible to reduce donor site morbidity,22,26,27 it is sometimes necessary to undertake intraoperative conversion to a free TRAM flap to improve the vascularity.21 Taken to its logical conclusion in patients with multiple scars, especially in the lower abdomen, if the DIEP vessels are not suitable or seem inadequate, one should convert the free tissue transfer to a pedicled flap albeit with a smaller tissue volume. We did not need to carry this out in any of our patients. Interestingly, although superficial inferior epigastric artery (SIEA) flaps have the least morbidity of all abdominal flap types,2 we did not use any for multiple scars because of their proneness to disruption of the superficial vessels and the variability/unreliability of the arterial supply.
Lower vertical and transverse scars of the lower abdomen were the commonest combination of 2 scars. We approach this relatively more common scenario by using a hemi-flap either as a DIEP or as a free muscle-sparing TRAM flap (Fig 2).9,11,12
Since the present study, we have adopted CT angiographic imaging to delineate perforator anatomy in terms of location and relative sizes. We hope that this will prove useful in optimizing the execution of breast reconstruction in patients with multiple previous abdominal surgeries/scars and render them more suitable for abdominal flap breast reconstruction.
The results from this small single operator case series suggest that the presence of multiple abdominal scars following major abdominal surgery does not constitute an absolute contraindication to abdominal flap breast reconstruction. With careful preoperative planning and taking into account the appropriate precautions, it is possible to safely undertake abdominal flap breast reconstruction in this challenging group of patients. However, patients must be informed of the potential risks, especially to the donor site. It is concluded that abdominal free flap breast reconstruction is not contraindicated in patients with previous multiple abdominal surgery in which the vascular pedicle was still preserved.