- Email: [email protected]
Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience
+
MODEL
Journal of Plastic, Reconstructive & Aesthetic Surgery (2016) xx, 1e7
Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience* Clifford C. Sheckter, Afaaf Shakir, Hong Vo, Jennifer Tsai, Rahim Nazerali, Gordon K. Lee* Division of Plastic Surgery, Stanford University, 770 Welch Road Suite #400, Palo Alto, CA 94304, USA Received 6 September 2015; accepted 23 June 2016
KEYWORDS Abdominoperineal resection; Colo-rectal cancer; Gluteal flap; Rectus flap; VeY advancement
Summary Background: Abdominoperineal resection (APR) is the surgical treatment of lowlying rectal cancers and other pelvic malignancies. Plastic surgery offers a means to close these complicated defects through obliterating dead space, providing tension-free closure, and introducing vascularized tissue into a radiated field. The indications for reconstructive surgery and choice of reconstruction are debatable. This study aims to identify when and which reconstruction is preferred. Methods: A retrospective comparative analysis was performed on all patients undergoing APR at Stanford Hospital between 2007 and 2013. Data points included demographics, disease, operative positioning, and postoperative complications. Univariate analysis and multivariate logistic regression analysis were performed to identify markers of flap reconstruction and complications. Results: A total of 178 APRs were performed, of which 51 underwent flap reconstruction. The odds ratio of all complications between flap and primary closure was not significant at 1.36 (0.69e2.66). Independent predictors for flap reconstruction included prone positioning, anal squamous cell carcinoma (SCC), prior smoking, and neoadjuvant chemoradiation therapy. Univariate predictors of flap reconstruction included female gender and combined vaginectomy. Independent predictors of complications included current and prior smoking. Muscle flap closure had lower recipient site complications than V-to-Y advancement closure (20% vs. 50%, p Z 0.039). Conclusion: Flap reconstruction following APR is associated with prone positioning, neoadjuvant chemoradiation, female gender, prior smoking, and anal SCC resections. Pedicled muscle flaps had a significantly lower rate of recipient site complications than V-to-Y advancement
*
Presented at the 65th CSPS Annual Meeting, Monterrey, CA, USA, May 23, 2015. * Corresponding author. Stanford Plastic Surgery, 770 Welch Road, Suite 400, Palo Alto, CA 94304-5715, USA. Fax: þ1 650 725 6605. E-mail address: [email protected] (G.K. Lee).
http://dx.doi.org/10.1016/j.bjps.2016.06.024 1748-6815/ª 2016 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
2
C.C. Sheckter et al. flaps and therefore should be the flap reconstruction of choice. The vertical rectus abdominis myocutaneous flap was superior to the gracilis flap in terms of the overall reduction of complications. ª 2016 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
Introduction Abdominoperineal resection (APR) is the standard surgical treatment of low-lying rectal cancer, anal cancer, severe inflammatory bowel disease, and other perineal malignancies. Extensive resections often leave large pelvic defects that present a major wound closure challenge to the plastic surgeon. Immediate flap reconstruction has previously shown fewer complications than primary closure.1e6 Several options exist for local flap reconstruction of perineal wound closure,7,8 including pedicled vertical rectus abdominis myocutaneous (VRAM) flaps, local V-to-Y advancement flaps, and pedicled gracilis muscle flaps9,10; each has advantages and disadvantages. The pedicled VRAM flap has been used in multiple applications for difficult periabdominal wounds11 and results in fewer perineal complications than primary wound closure.12e14 It is also associated with lower perineal morbidity15 with excellent long-term survival results, indicating success out to 10 years postoperatively.16,17 It is also superior to gracilis flap reconstruction in terms of complications.18 However, the use of the VRAM flap can be limited by positioning, prior abdominal surgery, scarring, and number of ostomies required.7 V-to-Y advancement flaps avoid additional abdominal wall morbidity when compared with the VRAM flap19 and offer a less bulky option for flap coverage.20 Local tissues can heal well and prevent perineal hernia, but local tissue flaps are disadvantaged by being adjacent to the radiated surgical field and may be associated with increased wound healing complications.21 The pedicled gracilis muscle flap is another effective option.1,9,22 It offers a large arc of rotation with no restrictions on postoperative motion or ambulation,23 but may be too small for larger pelvic defects. Given the various options available for reconstruction, determining which specific flap type to use can be challenging. Complicating matters further, there has also been a paradigm shift in terms of operative positioning by the extirpative surgical oncology team. Recent literature suggests that a combined approach through the abdomen in the supine position and then finishing the resection in the prone position through the buttock cleft can offer multiple advantages over low lithotomy, including better surgical margins and lower recurrence rates.24e27 However, the prone position limits the plastic surgeon’s reconstructive options, as abdominal flaps are not readily accessible. While gluteal advancement flaps and pedicled gracilis flaps are described in the literature, there is no study evaluating outcomes between different reconstructive options in different patient positions. To better identify appropriate candidates for reconstruction and assess the optimal means of reconstruction
after APR, we evaluated the demographics and outcomes of a single institution over a 7-year period. We hypothesized the following: patients receiving neoadjuvant radiation will have a higher prevalence of reconstruction, the prevalence of complications after reconstruction will be equivalent to simple primary closure, muscle flaps will have fewer complications than V-to-Y advancement flaps, and the prone position will have a higher overall complication prevalence due to limited options for reconstruction.
Methods Institutional Review Board (IRB) approval was obtained following the Stanford University protocol for a retrospective comparative (cross-sectional) analysis. A retrospective chart review was performed on all patients undergoing APR at Stanford Hospital between January 1, 2007, and December 31, 2013, which evaluated data points including demographics, comorbidities, disease stage, radiotherapy, operative positioning, flap choice, and postoperative complications. Patients who underwent APR with reconstruction were followed in both surgical oncology and plastic surgery clinics after hospital discharge. Length of follow-up was determined as time in days between the date of surgery and the last visit with a surgeon who examined the patient. Patients lost to follow-up after their initial postoperative visit had their length of follow-up measured as the last visit. There was no investigation into the outside facility records for any patient, which was held constant for all groups. Demographics were obtained from preoperative history and physical documentation. This included information on diagnosis, age, gender, body mass index, smoking status, comorbidities, and neoadjuvant radiation. Chart reviewing was performed by three independent reviewers who were blinded to reconstruction when evaluating for complications, in an attempt to limit bias in complication outcomes. Complications were categorized as perineal incision, donor site, or other medical (e.g., small bowel obstruction, failure to thrive) complications. Perineal complications were further subdivided into specific categories based on frequency. Cellulitis was non-suppurative inflammation of incision requiring antibiotic treatment. Abscess was a purulent collection requiring drainage or manipulation of tissues for expression. Dehiscence was defined as dermal separation greater than one-third of the wound distance without presence of infection. Flap loss was defined as necrosis of at least one-third of the tissue. Prolonged healing was defined as absence of dermal opposition or persistent non-supportive drainage at 3 weeks from the day of surgery.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
Reconstruction following abdominoperineal resection (APR) Univariate analysis was performed with t-test for continuous variables and Fisher’s exact test for categorical variables, utilizing GraphPad (GraphPad Software Inc., La Jolla, CA). Multivariate analysis and odds ratio (OR) calculations were performed using R (The R Foundation, Vienna, Austria). For the multivariate analysis, candidate predictors included operative position, disease, procedure, body mass index, gender, current tobacco use, prior tobacco use, and neoadjuvant chemoradiation therapy. The multivariate logistic analysis was performed to evaluate the statistical significance of predictors on the response variables, namely complication and plastic surgery closure. This was performed using the generalized linear model function of R with the binomial family. The disease predictor variable had four levels, with rectal cancer as the baseline variable. The procedure variable had three levels, with laparoscopicassisted APR as the baseline. The baseline category for smoking was “never smoking”. The body mass index (BMI) variable was numeric. The significance for all statistical analysis was determined at p < 0.05.
Results
Plastic surgery closure in APR When evaluating comorbidities and treatment characteristics, data were separated into two columns e flap reconstruction and primary closure e to help identify factors leading to plastic surgery’s involvement in these cases. Univariate analysis showed the significance for flap reconstruction in women, prior and active smokers, patients receiving neoadjuvant therapy, patients with anal SCC, and those receiving vaginectomy with APR (Table 1). There were no significant differences in the prevalence of flap closure for open APRs versus laparoscopic-assisted APRs (Table 1). Multivariate analysis, considering flap closure as an outcome, found prone positioning, anal SCC, inflammatory bowel disease, use of neoadjuvant therapy, and prior smoking status as independent predictors of flap closure (Table 2). While univariate analysis showed significance for flap closure in APR with combined vaginectomy, multivariate analysis did not show a significant association.
Table 1 Univariate analysis of demographics and treatment characteristics.
General A total of 178 patients who underwent APR during the study period were included. The excluded patients from the study included one patient with a failed ileoanal pouch with multiple enterocutaneous fistulas, another patient with massive pelvic bleeding from prior sigmoidal colon surgery requiring internal iliac artery embolization, and a third patient with prior low anterior resection with pelvic abscess and fistulae. The most common indication for APR was colorectal adenocarcinoma in 128 patients (72%), with the remainder of diseases including squamous cell carcinoma (SCC) (19 patients), inflammatory bowel disease (18 patients), and other malignancies (13 patients). Plastic surgery was consulted for closure in 51 of 178 cases, with the remainder receiving primary closure by the surgical oncology team (Figure 1). The mean follow-up time for patients who underwent flap closure was 421 322 days, while the mean follow-up time for patients who underwent primary closure was 401 357 days; these were not significantly different (p Z 0.79).
Figure 1
3
Flowchart of APR reconstruction.
Characteristic
Reconstruction Primary p n Z 51 (%) Closure n Z 127 (%)
Age, years mean BMI mean Female Smoker Never Prior use Yes, smoking Comorbidities Hypertension Hyperlipidemia Diabetes Pulmonary disease Neoadjuvant CRT Positioning proneb Disease Rectal cancer Anal SCC Crohn’s and UC Otherc Procedure APR open APR lap assist APR þ vaginectomy Mean follow-up (days)
57.94 26.69 29
61.04 27.53 46
0.2147a 0.3785a 0.0183
27 14 10
69 (54) 15 43
0.8697 0.0141 0.0707
19 11 7 4 45 32
46 36 14 11 78 (61) 41
1.000 0.4525 0.6136 1.000 0.0003 0.0003
27 16 3 5
101 (80) 2 13 11
0.0008 <0.0001 0.5627 e
29 17 5 421 322
90 (70) 36 1 401 357
0.1146 0.5871 0.0078 0.7895a
p-values calculated using Fisher’s exact test unless denoted. Percentage given where n > 50. CRT, chemoradiation therapy. UC, ulcerative colitis. Abdom, abdomen. Surg, surgery. APR, abdominoperineal resection. Lap, laparoscopic. a t-test. b Compared to lithotomy. c Anal adenocarcinoma, pelvic sarcoma, familial adenomatous polyposis, traumatic perineum, pre-sacral hamartoma, melanoma.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
4
C.C. Sheckter et al. Table 2 Independent predictors of flap closure in multivariate logistic regression analysis. Factor
Odds Ratio
95% CI
p
Prone positioning Anal SCC IBD APR þ vaginectomy Current smoker Prior smoker BMI Neoadjuvant therapy
4.32 58.30 12.00 0.97 0.67 3.50 0.95 13.30
1.76e10.60 8.19e410.00 1.48e98.00 0.06e16.6 0.26e1.72 1.08e11.30 0.88e1.02 2.74e64.60
0.001 <0.001 0.020 0.098 0.417 0.036 0.091 0.001
CI, confidence interval; SCC, squamous cell carcinoma; IBD, inflammatory bowel disease; APR, abdominoperineal reconstruction; BMI, body mass index.
Complications following APR The odds ratio (OR) of all complications comparing any flap closure with primary closure did not reach significance with an OR of 1.36 (95% CI: 0.69e2.66, p Z 0.3753). Recipient site complications, notably dehiscence, occurred at a significantly high frequency when APR defects were closed with V-to-Y advancement flaps as compared to that with primary closure (Table 3). VRAM flaps had a significantly
Table 3
lower prevalence of complications than primary closure (p Z 0.04), and the prevalence of all complications for gracilis flaps was not different from that of primary closure (p Z 0.74). All V-to-Y advancement flaps were performed when the patient was in the prone position (as compared to supine or lithotomy). Positioning was determined by the surgical oncologist. When examining the prevalence of all complications in the prone position, both the gracilis flap (p Z 0.20) and the V-to-Y flap (p Z 0.12) were associated with a higher prevalence of complications than primary closure, though these differences did not reach significance. In patients who received neoadjuvant chemoradiation, the prevalence of all complications was not significantly different between any flap group and primary closure; however, there was a significantly higher prevalence of recipient site complications with V-to-Y flap closures than with primary closure (p Z 0.02). This was not true for VRAM flaps (p Z 0.34) or gracilis flaps (p Z 0.30). Although the prevalence of all complications between muscle flaps and V-to-Y flaps was not found to be significantly different (p Z 0.30), there was a significantly higher prevalence of recipient site complications with V-to-Y flaps than with muscle flaps (p Z 0.04). Multivariate analysis of complications as an outcome showed that current or prior smoking status was the only
APR complications after reconstruction compared to those after primary closure.
All Complications Recipient site complications Cellulitis Abscess Dehiscence Prolonged healing Flap loss Donor site complicationa Other Abdominal abscess Abdominal incision infection ECF CVA Failure to thrive Delayed viscous injury SBO Atrial fibrillation Abdominal hematoma Hernia Urosepsis Perianal Paget’s Pancreatitis
Gracilis n Z 16
p
VRAM nZ9
p
VeY n Z 26
p
Primary Closure n Z 127 (%)
12 5 2 0 0 1 2 2
0.2797 0.7644 0.6253 0.3611 0.5986 0.2995 e e
2 0 0 0 0 0 0 0
0.0778 0.1129 1.0000 1.0000 1.0000 1.0000 e e
19 13 3 2 6 2 n/a n/a
0.1875 0.0196 0.4644 1.0000 0.00169 0.1327 e e
73 (57) 33 10 12 9 2 e n/a
3 0
0.0458 1.0000
0 0
1.0000 1.0000
0 0
0.5892 1.0000
5 3
1 1 0 0 0 0 0 0 0 0 0
1.0000 0.2120 1.0000 1.0000 0.2166 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0 0 1 1 0 0 0 0 0 0 0
1.0000 1.0000 0.1284 0.2420 0.5983 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0 0 0 1 4 1 0 0 0 0 0
1.0000 1.0000 1.0000 0.5291 0.7499 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0 1 1 3 16 0 3 5 1 2 1
Significance determined with Fisher’s exact. p-values compared to primary closure. Percentage given where n > 50. VRAM, vertical rectus abdominis myocutaneous; ECF, Enterocutaneous fistula; CVA, Cerebral vascular accident; SBO, Small bowel obstruction. a Thigh seroma.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
Reconstruction following abdominoperineal resection (APR)
5
independent predictor of complications. Neoadjuvant therapy and prone positioning were associated with decreased odds of complications, but only neoadjuvant therapy reached significance. Conversely, varied pathologic diagnoses, APR plus vaginectomy, and BMI were associated with higher odds of complications, but none were significant (Table 4).
muscle flaps may be superior to local fasciocutaneous flaps for perineal defects. For the prone position, a gracilis flap as outlined in our proposed algorithm for APR reconstruction is recommended (Figure 2). The preferred position to perform an APR may still be debatable, but surgical oncologists may increasingly consider the prone position because of recent evidence supporting superior oncologic outcomes.24e27 The literature on APR reconstruction lacks guidance of these closures; published series to date typically evaluate reconstruction only in the supine/lithotomy position. When positioned supine, APR reconstruction with a pedicled VRAM showed no perineal complications, but was not statistically significant due to our limited sample size. Nonetheless, this closure did show a significant trend in reducing overall complications and had no donor site complications compared to the pedicled gracilis. These findings are consistent with previously published studies that support the use of VRAM when technically feasible.18 Comparison with a recent systematic review by Devulapalli et al. shows a few differences.28 The review demonstrated fewer wound healing complications for flap closure than for primary closure in the study on 566 patients. Notably, flap closure only included gracilis and VRAM reconstructions. Our flap closure group also included VeY advancement flaps, which are more complicated. We also found overall fewer wound healing complications when comparing muscle flaps with primary closure. The review was unable to show superiority of one muscle flap over another, and there was no mention of positioning. In terms of demographic factors, the only independent predictor for complications was current or prior smoking history. This is logical given the well-known risks of smoking and interference with wound healing. Surprisingly, prior smoking showed greater odds of complications than current smoking. There is no clear explanation to this finding, although patient truthfulness may play a role. Likely, patients who claimed to have quit smoking were still smoking to some degree. Because there was no laboratory verification of the smoking status, the absolute truth is unknown. Further, the quantity of smoking (e.g., pack years) was not measured, which may be a better measure of the smoking effects. Neoadjuvant therapy was the only factor significantly associated with reduced odds of complications. Given the risks for wound healing with neoadjuvant therapy, this is counterintuitive. Possibly, the decreased complication odds are confounded by an increased proportion of these patients receiving flap reconstruction. Alternatively, surgeon knowledge of neoadjuvant therapy may have led to wider resections of radiated tissue and more ginger handling of tissues in these patients e ultimately improving wound healing. The former explanation is unlikely, because as demonstrated, flap closure was not superior to primary closure in wound-healing complications. The latter explanation could be studied if surgeons were blinded to neoadjuvant therapy and the surgical technique was observed and studied. Multivariate analysis of predictors for flap closure yielded many congruent results including prone positioning, neoadjuvant therapy, inflammatory bowel disease (IBD), and anal SCC. It was puzzling that prior smokers were more
Discussion Primary closure was performed in the majority of patients, and there was no statistically significant difference between wound healing and medical complications compared to the group that received flap reconstruction. This should not be interpreted to mean that reconstructive surgery is unnecessary, as the reconstructed group had baseline characteristics associated with a greater challenge in closure and healing, notably neoadjuvant therapy, more complex defects (from anal SCC resections), prone positioning, and smoking status. It is logical to assume that plastic surgery was consulted in cases where surgical oncology could not close the defect primarily, because they are more difficult closures; therefore, the statistical equivalence supports reconstruction as an effective means of closure. Accordingly, reconstruction may not provide superior results to primary closure of simpler wounds. This study validates equivalent wound healing of more complicated defects with reconstruction compared to primary closure of simpler wounds. Regarding the type of closure, V-to-Y advancement showed a relatively higher wound-healing complication prevalence than other methods of closure. We postulate that V-to-Y advancement inherently provides inferior tissue bulk and involves a larger suture line prone to breakdown. Furthermore, the use of tissue that is directly adjacent to the defect is associated with more complications in patients who received neoadjuvant radiation therapy that has been delivered to the same area. All V-to-Y advancement reconstructions occurred in the prone positiondit is not physically possible in the supine/ lithotomy position. Because the rectus muscle is not available, the local alternative (gracilis) shows better success at wound healing. In general, closure with pedicled
Table 4 Independent predictors of complications in multivariate logistic regression analysis. Factor
Odds ratio
95% CI
p
Prone positioning Anal SCC IBD APR þ vaginectomy Current smoker Prior smoker BMI Neoadjuvant therapy
0.90 1.87 2.84 1.33 2.35 4.76 1.04 0.44
0.42e1.92 0.48e7.35 0.62e12.9 0.11e15.65 1.15e4.78 1.64e13.80 0.97e1.10 0.21e0.89
0.784 0.368 0.177 0.819 0.018 0.004 0.260 0.023
CI, confidence interval; SCC, squamous cell carcinoma; IBD, inflammatory bowel disease; APR, abdominoperineal reconstruction; BMI, body mass index.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
6
C.C. Sheckter et al.
Figure 2
APR reconstruction algorithm.
likely to receive flaps than current smokers (not significant). It is unlikely that surgical oncologists selectively referred prior smokers but not current smokers. If anything, the opposite is logical. This result is seemingly perchance and likely a sampling error. The association of APR with combined vaginectomy and flap closure was significant in univariate analysis but not in multivariate logistic regression. Most cases of combined vaginectomy were a result of anal SCC, which was an independent predictor of flap closure. SCC resections are larger given the nature of visualizing disease and resulting margins. Thus, anal SCC is likely a confounding variable in the univariate association of flap closure with vaginectomy. Limitations of the study include the retrospective nature and lack of randomization. All referrals to plastic surgery for reconstruction were at the discretion of the surgical oncologists, which is an inherent bias. These referral patterns may not hold true for other health systems. Further, there is bias in the reconstructive choice of a given surgeon as there was no randomization of the flap type. The study is limited by the relatively small number of VRAM flap reconstructions compared to that of gracilis and V-to-Y advancement flaps. There was no significance shown for improved perineal healing with VRAM flaps despite having a lower complication prevalence, which could possibly be explained by underpowering. Nonetheless, these results provide guidance to plastic surgeons reconstructing APR defects, specifically addressing the challenges and solutions to prone positioning reconstruction. Further studies are warranted that address prone positioning.
Conclusion Flap reconstruction following APR is associated with prone positioning, neoadjuvant chemoradiation, female gender, prior smoking, IBD, and anal SCC resections. Wound-healing complications are associated with prior and current smoking. Pedicled muscle flaps have a significantly lower rate of recipient site complications than V-to-Y advancement flaps and should therefore be the flap reconstruction of choice. Fasciocutaneous V-to-Y advancement flaps may be simpler in terms of operative time and surgical complexity; however, this type of closure is significantly more complicated in terms of wound healing and should be generally avoided. The use of pedicled muscle flaps offers successful healing in wounds that surgical oncologists otherwise cannot close primarily. The VRAM was superior to the gracilis in the overall reduction of complications. In the prone position, where the VRAM is not available, the gracilis is the preferred flap closure over the V-to-Y advancement flap.
Funding No funding or sponsorship was involved in this study.
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.bjps.2016.06.024.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
Reconstruction following abdominoperineal resection (APR)
7
References 1. Shibata D, Hyland W, Busse P, et al. Immediate reconstruction of the perineal wound with gracilis muscle flaps following abdominoperineal resection and intraoperative radiation therapy fro recurrent carcinoma of the rectum. Ann Surg Oncol 1999;6:33e7. 2. Chessin DB, Hartley J, Cohen AM, et al. Rectus flap reconstruction decreases perineal wound complications after pelvic chemoradiation and surgery: a cohort study. Ann Surg Oncol 2005;12:104e10. 3. Chan S, Miller M, Ng R, et al. Use of myocutaneous flaps for perineal closure following abdominoperineal excision of the rectum for adenocarcinoma. Colorectal Dis 2010;12:555e60. 4. Howell AM, Jarral OA, Faiz O, Ziprin P, Darzi A, Zacharakis E. How should perineal wounds be closed following abdominoperineal resection in patients post radiotherapyeprimary closure or flap repair? Best evidence topic (BET). Int J Surg 2013;11:514e7. 5. Wang ED, Conkling N, Xu X, et al. Perineal flap reconstruction following oncologic anorectal extirpation: an outcomes assessment. Plast Reconstr Surg 2015;135:176ee84e. 6. Khoo AK, Skibber JM, Nabawi AS, et al. Indications for immediate tissue transfer for soft tissue reconstruction in visceral pelvic surgery. Surgery 2001;130:463e9. 7. Kolehmainen M, Suominen S, Tukiainen E. Pelvic, perineal and genital reconstructions. Scand J Surg 2013;102:25e31. 8. Galandiuk S, Jorden J, Mahid S, McCafferty MH, Tobin G. The use of tissue flaps as an adjunct to pelvic surgery. Am J Surg 2005;190:186e90. 9. Persichetti P, Cogliandro A, Marangi GF, et al. Pelvic and perineal reconstruction following abdominoperineal resection: the role of gracilis flap. Ann Plast Surg 2007;59:168e72. 10. Burke T, Morris M, Roh MS, Levenback C, Gershenson DM. Perineal reconstruction using single gracilis myocutaneous flaps. Gynecol Oncol 1995;S7:221e5. 11. Carey JN, Sheckter CC, Watt AJ, Lee GK. Intra-abdominal pedicled rectus abdominis muscle flap for treatment of highoutput enterocutaneous fistulae: case reports and review of literature. J Plast, Reconst Aesthet Surg 2013;66:1145e8. 12. Touny A, Othman H, Maamoon S, Ramzy S, Elmarakby H. Perineal reconstruction using pedicled vertical rectus abdominis myocutaneous flap (VRAM). J Surg Oncol 2014;110:752e7. 13. Kroll SS, Pollock R, Jessup JM, Ota D. Transpelvic rectus abdominis flap reconstruction of defects following abdominalperineal resection. Am Surg 1989;55:632e7. 14. McAllister E, Wells K, Chaet M, Norman J, Cruse W. Perineal reconstruction after surgical extirpation of pelvic malignancies using the transpelvic transverse rectus abdominal myocutaneous flap. Ann Surg Oncol 1994;1:164e8. 15. Lefevre JH, Parc Y, Kerneis S, et al. Abdomino-perineal resection for anal cancer: impact of a vertical rectus
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
abdominis myocutaneus flap on survival, recurrence, morbidity, and wound healing. Ann Surg 2009;250:707e11. Sunesen KG, Buntzen S, Tei T, Lindegaard JC, Norgaard M, Laurberg S. Perineal healing and survival after anal cancer salvage surgery: 10-year experience with primary perineal reconstruction using the vertical rectus abdominis myocutaneous (VRAM) flap. Ann Surg Oncol 2009;16:68e77. Butler CE, Gundeslioglu AO, Rodriguez-Bigas MA. Outcomes of immediate vertical rectus abdominis myocutaneous flap reconstruction for irradiated abdominoperineal resection defects. J Am Coll Surg 2008;206:694e703. Nelson RA, Butler CE. Surgical outcomes of VRAM versus thigh flaps for immediate reconstruction of pelvic and perineal cancer resection defects. Plast Reconstr Surg 2009;123: 175e83. Pantelides NM, Davies RJ, Fearnhead NS, Malata CM. The gluteal fold flap: a versatile option for perineal reconstruction following anorectal cancer resection. J Plast Reconstr Aesthet Surg 2013;66:812e20. Windhofer C, Michlits W, Heuberger A, Papp C. Perineal reconstruction after rectal and anal disease using the local fascio-cutaneous-infragluteal flap: a new and reliable technique. Surgery 2011;149:284e90. Saleh DB, Callear JA, Basheer M, Mohammed P. The partial myocutaneous gluteal flap reconstruction of extralevator abdominoperineal defects in irradiated patients. Ann Plast Surg 2015;74:403e7. Chong TW, Balch GC, Kehoe SM, Margulis V, Saint-Cyr M. Reconstruction of large perineal and pelvic wounds using gracilis muscle flaps. Ann Surg Oncol 2015. Ducic I, Dayan JH, Attinger CE, Curry P. Complex perineal and groin wound reconstruction using the extended dissection technique of the gracilis flap. Plast Reconstr Surg 2008;122: 472e8. West NP, Finan PJ, Anderin C, Lindholm J, Holm T, Quirke P. Evidence of the oncologic superiority of cylindrical abdominoperineal excision for low rectal cancer. J Clin Oncol 2008; 26:3517e22. Han JG, Wang ZJ, Wei GH, Gao ZG, Yang Y, Zhao BC. Randomized clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. Am J Surg 2012;204:274e82. Liu P, Bao H, Zhang X, et al. Better operative outcomes achieved with the prone jackknife vs. lithotomy position during abdominoperineal resection in patients with low rectal cancer. World J Surg Oncol 2015;13:39. Hu X, Cao L, Zhang J, Liang P, Liu G. Therapeutic results of abdominoperineal resection in the prone jackknife position for T3-4 low rectal cancers. J Gastrointest Surg 2015;19:551e7. Devulapalli C, Jia Wei AT, DiBiagio JR, et al. Primary versus flap closure of perineal defects following oncologic resection: a systematic review and meta-analysis. Plast Reconstr Surg 2016;137:1602e13.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
MODEL
Journal of Plastic, Reconstructive & Aesthetic Surgery (2016) xx, 1e7
Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience* Clifford C. Sheckter, Afaaf Shakir, Hong Vo, Jennifer Tsai, Rahim Nazerali, Gordon K. Lee* Division of Plastic Surgery, Stanford University, 770 Welch Road Suite #400, Palo Alto, CA 94304, USA Received 6 September 2015; accepted 23 June 2016
KEYWORDS Abdominoperineal resection; Colo-rectal cancer; Gluteal flap; Rectus flap; VeY advancement
Summary Background: Abdominoperineal resection (APR) is the surgical treatment of lowlying rectal cancers and other pelvic malignancies. Plastic surgery offers a means to close these complicated defects through obliterating dead space, providing tension-free closure, and introducing vascularized tissue into a radiated field. The indications for reconstructive surgery and choice of reconstruction are debatable. This study aims to identify when and which reconstruction is preferred. Methods: A retrospective comparative analysis was performed on all patients undergoing APR at Stanford Hospital between 2007 and 2013. Data points included demographics, disease, operative positioning, and postoperative complications. Univariate analysis and multivariate logistic regression analysis were performed to identify markers of flap reconstruction and complications. Results: A total of 178 APRs were performed, of which 51 underwent flap reconstruction. The odds ratio of all complications between flap and primary closure was not significant at 1.36 (0.69e2.66). Independent predictors for flap reconstruction included prone positioning, anal squamous cell carcinoma (SCC), prior smoking, and neoadjuvant chemoradiation therapy. Univariate predictors of flap reconstruction included female gender and combined vaginectomy. Independent predictors of complications included current and prior smoking. Muscle flap closure had lower recipient site complications than V-to-Y advancement closure (20% vs. 50%, p Z 0.039). Conclusion: Flap reconstruction following APR is associated with prone positioning, neoadjuvant chemoradiation, female gender, prior smoking, and anal SCC resections. Pedicled muscle flaps had a significantly lower rate of recipient site complications than V-to-Y advancement
*
Presented at the 65th CSPS Annual Meeting, Monterrey, CA, USA, May 23, 2015. * Corresponding author. Stanford Plastic Surgery, 770 Welch Road, Suite 400, Palo Alto, CA 94304-5715, USA. Fax: þ1 650 725 6605. E-mail address: [email protected] (G.K. Lee).
http://dx.doi.org/10.1016/j.bjps.2016.06.024 1748-6815/ª 2016 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
2
C.C. Sheckter et al. flaps and therefore should be the flap reconstruction of choice. The vertical rectus abdominis myocutaneous flap was superior to the gracilis flap in terms of the overall reduction of complications. ª 2016 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
Introduction Abdominoperineal resection (APR) is the standard surgical treatment of low-lying rectal cancer, anal cancer, severe inflammatory bowel disease, and other perineal malignancies. Extensive resections often leave large pelvic defects that present a major wound closure challenge to the plastic surgeon. Immediate flap reconstruction has previously shown fewer complications than primary closure.1e6 Several options exist for local flap reconstruction of perineal wound closure,7,8 including pedicled vertical rectus abdominis myocutaneous (VRAM) flaps, local V-to-Y advancement flaps, and pedicled gracilis muscle flaps9,10; each has advantages and disadvantages. The pedicled VRAM flap has been used in multiple applications for difficult periabdominal wounds11 and results in fewer perineal complications than primary wound closure.12e14 It is also associated with lower perineal morbidity15 with excellent long-term survival results, indicating success out to 10 years postoperatively.16,17 It is also superior to gracilis flap reconstruction in terms of complications.18 However, the use of the VRAM flap can be limited by positioning, prior abdominal surgery, scarring, and number of ostomies required.7 V-to-Y advancement flaps avoid additional abdominal wall morbidity when compared with the VRAM flap19 and offer a less bulky option for flap coverage.20 Local tissues can heal well and prevent perineal hernia, but local tissue flaps are disadvantaged by being adjacent to the radiated surgical field and may be associated with increased wound healing complications.21 The pedicled gracilis muscle flap is another effective option.1,9,22 It offers a large arc of rotation with no restrictions on postoperative motion or ambulation,23 but may be too small for larger pelvic defects. Given the various options available for reconstruction, determining which specific flap type to use can be challenging. Complicating matters further, there has also been a paradigm shift in terms of operative positioning by the extirpative surgical oncology team. Recent literature suggests that a combined approach through the abdomen in the supine position and then finishing the resection in the prone position through the buttock cleft can offer multiple advantages over low lithotomy, including better surgical margins and lower recurrence rates.24e27 However, the prone position limits the plastic surgeon’s reconstructive options, as abdominal flaps are not readily accessible. While gluteal advancement flaps and pedicled gracilis flaps are described in the literature, there is no study evaluating outcomes between different reconstructive options in different patient positions. To better identify appropriate candidates for reconstruction and assess the optimal means of reconstruction
after APR, we evaluated the demographics and outcomes of a single institution over a 7-year period. We hypothesized the following: patients receiving neoadjuvant radiation will have a higher prevalence of reconstruction, the prevalence of complications after reconstruction will be equivalent to simple primary closure, muscle flaps will have fewer complications than V-to-Y advancement flaps, and the prone position will have a higher overall complication prevalence due to limited options for reconstruction.
Methods Institutional Review Board (IRB) approval was obtained following the Stanford University protocol for a retrospective comparative (cross-sectional) analysis. A retrospective chart review was performed on all patients undergoing APR at Stanford Hospital between January 1, 2007, and December 31, 2013, which evaluated data points including demographics, comorbidities, disease stage, radiotherapy, operative positioning, flap choice, and postoperative complications. Patients who underwent APR with reconstruction were followed in both surgical oncology and plastic surgery clinics after hospital discharge. Length of follow-up was determined as time in days between the date of surgery and the last visit with a surgeon who examined the patient. Patients lost to follow-up after their initial postoperative visit had their length of follow-up measured as the last visit. There was no investigation into the outside facility records for any patient, which was held constant for all groups. Demographics were obtained from preoperative history and physical documentation. This included information on diagnosis, age, gender, body mass index, smoking status, comorbidities, and neoadjuvant radiation. Chart reviewing was performed by three independent reviewers who were blinded to reconstruction when evaluating for complications, in an attempt to limit bias in complication outcomes. Complications were categorized as perineal incision, donor site, or other medical (e.g., small bowel obstruction, failure to thrive) complications. Perineal complications were further subdivided into specific categories based on frequency. Cellulitis was non-suppurative inflammation of incision requiring antibiotic treatment. Abscess was a purulent collection requiring drainage or manipulation of tissues for expression. Dehiscence was defined as dermal separation greater than one-third of the wound distance without presence of infection. Flap loss was defined as necrosis of at least one-third of the tissue. Prolonged healing was defined as absence of dermal opposition or persistent non-supportive drainage at 3 weeks from the day of surgery.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
Reconstruction following abdominoperineal resection (APR) Univariate analysis was performed with t-test for continuous variables and Fisher’s exact test for categorical variables, utilizing GraphPad (GraphPad Software Inc., La Jolla, CA). Multivariate analysis and odds ratio (OR) calculations were performed using R (The R Foundation, Vienna, Austria). For the multivariate analysis, candidate predictors included operative position, disease, procedure, body mass index, gender, current tobacco use, prior tobacco use, and neoadjuvant chemoradiation therapy. The multivariate logistic analysis was performed to evaluate the statistical significance of predictors on the response variables, namely complication and plastic surgery closure. This was performed using the generalized linear model function of R with the binomial family. The disease predictor variable had four levels, with rectal cancer as the baseline variable. The procedure variable had three levels, with laparoscopicassisted APR as the baseline. The baseline category for smoking was “never smoking”. The body mass index (BMI) variable was numeric. The significance for all statistical analysis was determined at p < 0.05.
Results
Plastic surgery closure in APR When evaluating comorbidities and treatment characteristics, data were separated into two columns e flap reconstruction and primary closure e to help identify factors leading to plastic surgery’s involvement in these cases. Univariate analysis showed the significance for flap reconstruction in women, prior and active smokers, patients receiving neoadjuvant therapy, patients with anal SCC, and those receiving vaginectomy with APR (Table 1). There were no significant differences in the prevalence of flap closure for open APRs versus laparoscopic-assisted APRs (Table 1). Multivariate analysis, considering flap closure as an outcome, found prone positioning, anal SCC, inflammatory bowel disease, use of neoadjuvant therapy, and prior smoking status as independent predictors of flap closure (Table 2). While univariate analysis showed significance for flap closure in APR with combined vaginectomy, multivariate analysis did not show a significant association.
Table 1 Univariate analysis of demographics and treatment characteristics.
General A total of 178 patients who underwent APR during the study period were included. The excluded patients from the study included one patient with a failed ileoanal pouch with multiple enterocutaneous fistulas, another patient with massive pelvic bleeding from prior sigmoidal colon surgery requiring internal iliac artery embolization, and a third patient with prior low anterior resection with pelvic abscess and fistulae. The most common indication for APR was colorectal adenocarcinoma in 128 patients (72%), with the remainder of diseases including squamous cell carcinoma (SCC) (19 patients), inflammatory bowel disease (18 patients), and other malignancies (13 patients). Plastic surgery was consulted for closure in 51 of 178 cases, with the remainder receiving primary closure by the surgical oncology team (Figure 1). The mean follow-up time for patients who underwent flap closure was 421 322 days, while the mean follow-up time for patients who underwent primary closure was 401 357 days; these were not significantly different (p Z 0.79).
Figure 1
3
Flowchart of APR reconstruction.
Characteristic
Reconstruction Primary p n Z 51 (%) Closure n Z 127 (%)
Age, years mean BMI mean Female Smoker Never Prior use Yes, smoking Comorbidities Hypertension Hyperlipidemia Diabetes Pulmonary disease Neoadjuvant CRT Positioning proneb Disease Rectal cancer Anal SCC Crohn’s and UC Otherc Procedure APR open APR lap assist APR þ vaginectomy Mean follow-up (days)
57.94 26.69 29
61.04 27.53 46
0.2147a 0.3785a 0.0183
27 14 10
69 (54) 15 43
0.8697 0.0141 0.0707
19 11 7 4 45 32
46 36 14 11 78 (61) 41
1.000 0.4525 0.6136 1.000 0.0003 0.0003
27 16 3 5
101 (80) 2 13 11
0.0008 <0.0001 0.5627 e
29 17 5 421 322
90 (70) 36 1 401 357
0.1146 0.5871 0.0078 0.7895a
p-values calculated using Fisher’s exact test unless denoted. Percentage given where n > 50. CRT, chemoradiation therapy. UC, ulcerative colitis. Abdom, abdomen. Surg, surgery. APR, abdominoperineal resection. Lap, laparoscopic. a t-test. b Compared to lithotomy. c Anal adenocarcinoma, pelvic sarcoma, familial adenomatous polyposis, traumatic perineum, pre-sacral hamartoma, melanoma.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
4
C.C. Sheckter et al. Table 2 Independent predictors of flap closure in multivariate logistic regression analysis. Factor
Odds Ratio
95% CI
p
Prone positioning Anal SCC IBD APR þ vaginectomy Current smoker Prior smoker BMI Neoadjuvant therapy
4.32 58.30 12.00 0.97 0.67 3.50 0.95 13.30
1.76e10.60 8.19e410.00 1.48e98.00 0.06e16.6 0.26e1.72 1.08e11.30 0.88e1.02 2.74e64.60
0.001 <0.001 0.020 0.098 0.417 0.036 0.091 0.001
CI, confidence interval; SCC, squamous cell carcinoma; IBD, inflammatory bowel disease; APR, abdominoperineal reconstruction; BMI, body mass index.
Complications following APR The odds ratio (OR) of all complications comparing any flap closure with primary closure did not reach significance with an OR of 1.36 (95% CI: 0.69e2.66, p Z 0.3753). Recipient site complications, notably dehiscence, occurred at a significantly high frequency when APR defects were closed with V-to-Y advancement flaps as compared to that with primary closure (Table 3). VRAM flaps had a significantly
Table 3
lower prevalence of complications than primary closure (p Z 0.04), and the prevalence of all complications for gracilis flaps was not different from that of primary closure (p Z 0.74). All V-to-Y advancement flaps were performed when the patient was in the prone position (as compared to supine or lithotomy). Positioning was determined by the surgical oncologist. When examining the prevalence of all complications in the prone position, both the gracilis flap (p Z 0.20) and the V-to-Y flap (p Z 0.12) were associated with a higher prevalence of complications than primary closure, though these differences did not reach significance. In patients who received neoadjuvant chemoradiation, the prevalence of all complications was not significantly different between any flap group and primary closure; however, there was a significantly higher prevalence of recipient site complications with V-to-Y flap closures than with primary closure (p Z 0.02). This was not true for VRAM flaps (p Z 0.34) or gracilis flaps (p Z 0.30). Although the prevalence of all complications between muscle flaps and V-to-Y flaps was not found to be significantly different (p Z 0.30), there was a significantly higher prevalence of recipient site complications with V-to-Y flaps than with muscle flaps (p Z 0.04). Multivariate analysis of complications as an outcome showed that current or prior smoking status was the only
APR complications after reconstruction compared to those after primary closure.
All Complications Recipient site complications Cellulitis Abscess Dehiscence Prolonged healing Flap loss Donor site complicationa Other Abdominal abscess Abdominal incision infection ECF CVA Failure to thrive Delayed viscous injury SBO Atrial fibrillation Abdominal hematoma Hernia Urosepsis Perianal Paget’s Pancreatitis
Gracilis n Z 16
p
VRAM nZ9
p
VeY n Z 26
p
Primary Closure n Z 127 (%)
12 5 2 0 0 1 2 2
0.2797 0.7644 0.6253 0.3611 0.5986 0.2995 e e
2 0 0 0 0 0 0 0
0.0778 0.1129 1.0000 1.0000 1.0000 1.0000 e e
19 13 3 2 6 2 n/a n/a
0.1875 0.0196 0.4644 1.0000 0.00169 0.1327 e e
73 (57) 33 10 12 9 2 e n/a
3 0
0.0458 1.0000
0 0
1.0000 1.0000
0 0
0.5892 1.0000
5 3
1 1 0 0 0 0 0 0 0 0 0
1.0000 0.2120 1.0000 1.0000 0.2166 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0 0 1 1 0 0 0 0 0 0 0
1.0000 1.0000 0.1284 0.2420 0.5983 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0 0 0 1 4 1 0 0 0 0 0
1.0000 1.0000 1.0000 0.5291 0.7499 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0 1 1 3 16 0 3 5 1 2 1
Significance determined with Fisher’s exact. p-values compared to primary closure. Percentage given where n > 50. VRAM, vertical rectus abdominis myocutaneous; ECF, Enterocutaneous fistula; CVA, Cerebral vascular accident; SBO, Small bowel obstruction. a Thigh seroma.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
Reconstruction following abdominoperineal resection (APR)
5
independent predictor of complications. Neoadjuvant therapy and prone positioning were associated with decreased odds of complications, but only neoadjuvant therapy reached significance. Conversely, varied pathologic diagnoses, APR plus vaginectomy, and BMI were associated with higher odds of complications, but none were significant (Table 4).
muscle flaps may be superior to local fasciocutaneous flaps for perineal defects. For the prone position, a gracilis flap as outlined in our proposed algorithm for APR reconstruction is recommended (Figure 2). The preferred position to perform an APR may still be debatable, but surgical oncologists may increasingly consider the prone position because of recent evidence supporting superior oncologic outcomes.24e27 The literature on APR reconstruction lacks guidance of these closures; published series to date typically evaluate reconstruction only in the supine/lithotomy position. When positioned supine, APR reconstruction with a pedicled VRAM showed no perineal complications, but was not statistically significant due to our limited sample size. Nonetheless, this closure did show a significant trend in reducing overall complications and had no donor site complications compared to the pedicled gracilis. These findings are consistent with previously published studies that support the use of VRAM when technically feasible.18 Comparison with a recent systematic review by Devulapalli et al. shows a few differences.28 The review demonstrated fewer wound healing complications for flap closure than for primary closure in the study on 566 patients. Notably, flap closure only included gracilis and VRAM reconstructions. Our flap closure group also included VeY advancement flaps, which are more complicated. We also found overall fewer wound healing complications when comparing muscle flaps with primary closure. The review was unable to show superiority of one muscle flap over another, and there was no mention of positioning. In terms of demographic factors, the only independent predictor for complications was current or prior smoking history. This is logical given the well-known risks of smoking and interference with wound healing. Surprisingly, prior smoking showed greater odds of complications than current smoking. There is no clear explanation to this finding, although patient truthfulness may play a role. Likely, patients who claimed to have quit smoking were still smoking to some degree. Because there was no laboratory verification of the smoking status, the absolute truth is unknown. Further, the quantity of smoking (e.g., pack years) was not measured, which may be a better measure of the smoking effects. Neoadjuvant therapy was the only factor significantly associated with reduced odds of complications. Given the risks for wound healing with neoadjuvant therapy, this is counterintuitive. Possibly, the decreased complication odds are confounded by an increased proportion of these patients receiving flap reconstruction. Alternatively, surgeon knowledge of neoadjuvant therapy may have led to wider resections of radiated tissue and more ginger handling of tissues in these patients e ultimately improving wound healing. The former explanation is unlikely, because as demonstrated, flap closure was not superior to primary closure in wound-healing complications. The latter explanation could be studied if surgeons were blinded to neoadjuvant therapy and the surgical technique was observed and studied. Multivariate analysis of predictors for flap closure yielded many congruent results including prone positioning, neoadjuvant therapy, inflammatory bowel disease (IBD), and anal SCC. It was puzzling that prior smokers were more
Discussion Primary closure was performed in the majority of patients, and there was no statistically significant difference between wound healing and medical complications compared to the group that received flap reconstruction. This should not be interpreted to mean that reconstructive surgery is unnecessary, as the reconstructed group had baseline characteristics associated with a greater challenge in closure and healing, notably neoadjuvant therapy, more complex defects (from anal SCC resections), prone positioning, and smoking status. It is logical to assume that plastic surgery was consulted in cases where surgical oncology could not close the defect primarily, because they are more difficult closures; therefore, the statistical equivalence supports reconstruction as an effective means of closure. Accordingly, reconstruction may not provide superior results to primary closure of simpler wounds. This study validates equivalent wound healing of more complicated defects with reconstruction compared to primary closure of simpler wounds. Regarding the type of closure, V-to-Y advancement showed a relatively higher wound-healing complication prevalence than other methods of closure. We postulate that V-to-Y advancement inherently provides inferior tissue bulk and involves a larger suture line prone to breakdown. Furthermore, the use of tissue that is directly adjacent to the defect is associated with more complications in patients who received neoadjuvant radiation therapy that has been delivered to the same area. All V-to-Y advancement reconstructions occurred in the prone positiondit is not physically possible in the supine/ lithotomy position. Because the rectus muscle is not available, the local alternative (gracilis) shows better success at wound healing. In general, closure with pedicled
Table 4 Independent predictors of complications in multivariate logistic regression analysis. Factor
Odds ratio
95% CI
p
Prone positioning Anal SCC IBD APR þ vaginectomy Current smoker Prior smoker BMI Neoadjuvant therapy
0.90 1.87 2.84 1.33 2.35 4.76 1.04 0.44
0.42e1.92 0.48e7.35 0.62e12.9 0.11e15.65 1.15e4.78 1.64e13.80 0.97e1.10 0.21e0.89
0.784 0.368 0.177 0.819 0.018 0.004 0.260 0.023
CI, confidence interval; SCC, squamous cell carcinoma; IBD, inflammatory bowel disease; APR, abdominoperineal reconstruction; BMI, body mass index.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
6
C.C. Sheckter et al.
Figure 2
APR reconstruction algorithm.
likely to receive flaps than current smokers (not significant). It is unlikely that surgical oncologists selectively referred prior smokers but not current smokers. If anything, the opposite is logical. This result is seemingly perchance and likely a sampling error. The association of APR with combined vaginectomy and flap closure was significant in univariate analysis but not in multivariate logistic regression. Most cases of combined vaginectomy were a result of anal SCC, which was an independent predictor of flap closure. SCC resections are larger given the nature of visualizing disease and resulting margins. Thus, anal SCC is likely a confounding variable in the univariate association of flap closure with vaginectomy. Limitations of the study include the retrospective nature and lack of randomization. All referrals to plastic surgery for reconstruction were at the discretion of the surgical oncologists, which is an inherent bias. These referral patterns may not hold true for other health systems. Further, there is bias in the reconstructive choice of a given surgeon as there was no randomization of the flap type. The study is limited by the relatively small number of VRAM flap reconstructions compared to that of gracilis and V-to-Y advancement flaps. There was no significance shown for improved perineal healing with VRAM flaps despite having a lower complication prevalence, which could possibly be explained by underpowering. Nonetheless, these results provide guidance to plastic surgeons reconstructing APR defects, specifically addressing the challenges and solutions to prone positioning reconstruction. Further studies are warranted that address prone positioning.
Conclusion Flap reconstruction following APR is associated with prone positioning, neoadjuvant chemoradiation, female gender, prior smoking, IBD, and anal SCC resections. Wound-healing complications are associated with prior and current smoking. Pedicled muscle flaps have a significantly lower rate of recipient site complications than V-to-Y advancement flaps and should therefore be the flap reconstruction of choice. Fasciocutaneous V-to-Y advancement flaps may be simpler in terms of operative time and surgical complexity; however, this type of closure is significantly more complicated in terms of wound healing and should be generally avoided. The use of pedicled muscle flaps offers successful healing in wounds that surgical oncologists otherwise cannot close primarily. The VRAM was superior to the gracilis in the overall reduction of complications. In the prone position, where the VRAM is not available, the gracilis is the preferred flap closure over the V-to-Y advancement flap.
Funding No funding or sponsorship was involved in this study.
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.bjps.2016.06.024.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024
+
MODEL
Reconstruction following abdominoperineal resection (APR)
7
References 1. Shibata D, Hyland W, Busse P, et al. Immediate reconstruction of the perineal wound with gracilis muscle flaps following abdominoperineal resection and intraoperative radiation therapy fro recurrent carcinoma of the rectum. Ann Surg Oncol 1999;6:33e7. 2. Chessin DB, Hartley J, Cohen AM, et al. Rectus flap reconstruction decreases perineal wound complications after pelvic chemoradiation and surgery: a cohort study. Ann Surg Oncol 2005;12:104e10. 3. Chan S, Miller M, Ng R, et al. Use of myocutaneous flaps for perineal closure following abdominoperineal excision of the rectum for adenocarcinoma. Colorectal Dis 2010;12:555e60. 4. Howell AM, Jarral OA, Faiz O, Ziprin P, Darzi A, Zacharakis E. How should perineal wounds be closed following abdominoperineal resection in patients post radiotherapyeprimary closure or flap repair? Best evidence topic (BET). Int J Surg 2013;11:514e7. 5. Wang ED, Conkling N, Xu X, et al. Perineal flap reconstruction following oncologic anorectal extirpation: an outcomes assessment. Plast Reconstr Surg 2015;135:176ee84e. 6. Khoo AK, Skibber JM, Nabawi AS, et al. Indications for immediate tissue transfer for soft tissue reconstruction in visceral pelvic surgery. Surgery 2001;130:463e9. 7. Kolehmainen M, Suominen S, Tukiainen E. Pelvic, perineal and genital reconstructions. Scand J Surg 2013;102:25e31. 8. Galandiuk S, Jorden J, Mahid S, McCafferty MH, Tobin G. The use of tissue flaps as an adjunct to pelvic surgery. Am J Surg 2005;190:186e90. 9. Persichetti P, Cogliandro A, Marangi GF, et al. Pelvic and perineal reconstruction following abdominoperineal resection: the role of gracilis flap. Ann Plast Surg 2007;59:168e72. 10. Burke T, Morris M, Roh MS, Levenback C, Gershenson DM. Perineal reconstruction using single gracilis myocutaneous flaps. Gynecol Oncol 1995;S7:221e5. 11. Carey JN, Sheckter CC, Watt AJ, Lee GK. Intra-abdominal pedicled rectus abdominis muscle flap for treatment of highoutput enterocutaneous fistulae: case reports and review of literature. J Plast, Reconst Aesthet Surg 2013;66:1145e8. 12. Touny A, Othman H, Maamoon S, Ramzy S, Elmarakby H. Perineal reconstruction using pedicled vertical rectus abdominis myocutaneous flap (VRAM). J Surg Oncol 2014;110:752e7. 13. Kroll SS, Pollock R, Jessup JM, Ota D. Transpelvic rectus abdominis flap reconstruction of defects following abdominalperineal resection. Am Surg 1989;55:632e7. 14. McAllister E, Wells K, Chaet M, Norman J, Cruse W. Perineal reconstruction after surgical extirpation of pelvic malignancies using the transpelvic transverse rectus abdominal myocutaneous flap. Ann Surg Oncol 1994;1:164e8. 15. Lefevre JH, Parc Y, Kerneis S, et al. Abdomino-perineal resection for anal cancer: impact of a vertical rectus
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
abdominis myocutaneus flap on survival, recurrence, morbidity, and wound healing. Ann Surg 2009;250:707e11. Sunesen KG, Buntzen S, Tei T, Lindegaard JC, Norgaard M, Laurberg S. Perineal healing and survival after anal cancer salvage surgery: 10-year experience with primary perineal reconstruction using the vertical rectus abdominis myocutaneous (VRAM) flap. Ann Surg Oncol 2009;16:68e77. Butler CE, Gundeslioglu AO, Rodriguez-Bigas MA. Outcomes of immediate vertical rectus abdominis myocutaneous flap reconstruction for irradiated abdominoperineal resection defects. J Am Coll Surg 2008;206:694e703. Nelson RA, Butler CE. Surgical outcomes of VRAM versus thigh flaps for immediate reconstruction of pelvic and perineal cancer resection defects. Plast Reconstr Surg 2009;123: 175e83. Pantelides NM, Davies RJ, Fearnhead NS, Malata CM. The gluteal fold flap: a versatile option for perineal reconstruction following anorectal cancer resection. J Plast Reconstr Aesthet Surg 2013;66:812e20. Windhofer C, Michlits W, Heuberger A, Papp C. Perineal reconstruction after rectal and anal disease using the local fascio-cutaneous-infragluteal flap: a new and reliable technique. Surgery 2011;149:284e90. Saleh DB, Callear JA, Basheer M, Mohammed P. The partial myocutaneous gluteal flap reconstruction of extralevator abdominoperineal defects in irradiated patients. Ann Plast Surg 2015;74:403e7. Chong TW, Balch GC, Kehoe SM, Margulis V, Saint-Cyr M. Reconstruction of large perineal and pelvic wounds using gracilis muscle flaps. Ann Surg Oncol 2015. Ducic I, Dayan JH, Attinger CE, Curry P. Complex perineal and groin wound reconstruction using the extended dissection technique of the gracilis flap. Plast Reconstr Surg 2008;122: 472e8. West NP, Finan PJ, Anderin C, Lindholm J, Holm T, Quirke P. Evidence of the oncologic superiority of cylindrical abdominoperineal excision for low rectal cancer. J Clin Oncol 2008; 26:3517e22. Han JG, Wang ZJ, Wei GH, Gao ZG, Yang Y, Zhao BC. Randomized clinical trial of conventional versus cylindrical abdominoperineal resection for locally advanced lower rectal cancer. Am J Surg 2012;204:274e82. Liu P, Bao H, Zhang X, et al. Better operative outcomes achieved with the prone jackknife vs. lithotomy position during abdominoperineal resection in patients with low rectal cancer. World J Surg Oncol 2015;13:39. Hu X, Cao L, Zhang J, Liang P, Liu G. Therapeutic results of abdominoperineal resection in the prone jackknife position for T3-4 low rectal cancers. J Gastrointest Surg 2015;19:551e7. Devulapalli C, Jia Wei AT, DiBiagio JR, et al. Primary versus flap closure of perineal defects following oncologic resection: a systematic review and meta-analysis. Plast Reconstr Surg 2016;137:1602e13.
Please cite this article in press as: Sheckter CC, et al., Reconstruction following abdominoperineal resection (APR): Indications and complications from a single institution experience, Journal of Plastic, Reconstructive & Aesthetic Surgery (2016), http://dx.doi.org/ 10.1016/j.bjps.2016.06.024