Outcomes and Reliability of Perforator Flaps in the Reconstruction of Hidradenitis Suppurativa Defects: A Systemic Review and Meta-Analysis

Introduction: Hidradenitis suppurativa (HS) is a common and debilitating disease, in which the only effective treatment involves a wide excision of the affected skin. Secondary wound healing and skin grafting are two well-known options for managing these defects, but perforator flaps provide a new therapeutic alternative by ensuring reconstructions of large defects, reducing donor site morbidity, and enhancing functional recovery. The aim of this study was to achieve a systematic review of perforator flaps use in HS. Patients and Methods: PubMed and Cochrane databases were searched from 1989 to 2021. The PRISMA statement was used in the study selection process and the review was registered on PROSPERO. Furthermore, patient characteristics, operative technique, complications, and recurrences were searched. Results: Thirty-six articles were selected including 286 patients and 387 flaps. Axillary localization was mostly represented (83.2%). Direct donor site closure was achieved in 99.1% of cases. In total, 15.1% of the flaps presented at least one of the following complications: wound dehiscence (5.5%), partial necrosis (2.9%), hematoma or seroma (2.1%), infection (2.1%), venous congestion (1.8%), and nerve injury (0.3%). Two cases of total necrosis were recorded. Recurrence of the disease was observed in 2.7% of the defects. Conclusions: Pedicled perforator flaps are a reliable and reproducible technique in the reconstruction of HS defects. They are associated with a low recurrence rate while ensuring an effective reconstruction with reduced morbidity and faster recovery compared to the techniques classically used in this indication.


Introduction
Hidradenitis suppurativa (HS) is a chronic, inflammatory, and recurrent skin disease, whose pathophysiology is still poorly understood [1]. The central pathogenic event is believed to be the occluded hair follicles, provoking an immune response [2] leading to local inflammation of the apocrine gland-bearing areas of the body, most commonly the axillae, inguinal, and anogenital regions. Pro-inflammatory cytokines such as TNF-α are involved [2] and the inflammatory vicious circle finally results in painful subcutaneous nodules and fluctuant draining abscesses with a constant malodorous discharge, impairing

Eligibility Criteria
Inclusion criteria were all original peer-reviewed studies describing the use of perforator flaps in the reconstruction of defects directly following wide excision in HS, regardless of their location. Exclusion criteria were: articles dealing only with secondary wound resurfacing; defects caused by another etiology than HS; reconstructions not using a perforator flap, such as musculocutaneous flaps; any free flap; duplicate studies; review articles without original data; technical descriptions; letters to the editor without case reports; communications without original data; and studies in a language other than English and French.
PubMed and Cochrane Library electronic databases from 1989 (first description of perforator flap) to 2021 were used. The following keywords were selected: the group ((hidradenitis) OR (hidradenitis suppurativa) OR (hidradenitis suppurative) OR (acne inversa) OR (verneuil) OR (verneuil's disease)) was associated successively to different

Data Collection
Data were collected independently by two researchers (CV, PR). Disagreements were resolved by the senior author (NB). Retrieved information were: author, country, date of publication, type of study, level of evidence, number of patients, age, comorbidities (high blood pressure, smoking, diabetes, body mass index), defect locations and surface, surgical margin, perioperative antibiotic therapy, number, family and type of flaps, surface, level of dissection and degree of rotation of flaps, perforator skeletonization, preoperative perforator detection, rate and type of complications, surgical revision, closure of the donor site, recurrence, and follow-up time. All data were listed in Microsoft Excel 2016 (Microsoft Inc., Redmond, WA, USA).

Assessment of Methodological Quality and Study Outcomes
Each case series and cohort study were assessed for methodological quality using a standardized critical appraisal instrument, the Joanna Briggs Institute (JBI) Critical Appraisal Checklist. Case reports and letters to the editor were not submitted to these appraisal tools. The primary endpoint was to describe the use of perforator flaps in the indication of HS defects in order to assess their safety and reliability. The secondary outcome was to identify risk factors for complications.
All statistical analyses were performed using Prism 5 software (GraphPad Software, San Diego, CA, USA).
Variables are expressed as means ± standard deviations and were compared using Fisher's exact test. A two-sided p-value < 0.05 was considered significant. Subgroup analyses were performed for the predominantly represented axillary location in order to homogenize data and identify risk factors for complications.

Results
Among 52 identified, 36 [28, studies were included, as shown in the corresponding flowchart (Figure 1). We included 286 patients with 380 defects and 387 flaps, some defects being reconstructed by two flaps. Most studies had a low level of evidence (Table 1). There were only three prospective studies and no randomized control trial. The methodological quality assessments of the studies are presented in Table 2 (case series) and  Table 3 (cohort studies). General patient and flap characteristics are presented in Table 4.
Variables are expressed as means ± standard deviations and were compared using Fisher's exact test. A two-sided p-value < 0.05 was considered significant. Subgroup analyses were performed for the predominantly represented axillary location in order to homogenize data and identify risk factors for complications.

Results
Among 52 identified, 36 [28,] studies were included, as shown in the corresponding flowchart ( Figure 1). We included 286 patients with 380 defects and 387 flaps, some defects being reconstructed by two flaps. Most studies had a low level of evidence (Table 1). There were only three prospective studies and no randomized control trial. The methodological quality assessments of the studies are presented in Table 2 (case series) and Table 3 (cohort studies). General patient and flap characteristics are presented in Table 4.      The information was found for all flaps ( Table 5). Most of them were in the axillary region, with 322 flaps (83.2%); 45 (11.6%) flaps in the anogenital region; 19 (4.9%) in the inguinal region; and 1 (0.3%) in the cervical area.

. Defect and Flap Surfaces
Data regarding the defect and flap surfaces were found for 99 and 149 cases, respectively. The mean surface area of the defects was 135.7 ± 82.5 cm 2 (minimum: 15; maximum: 920). The mean surface area of the flaps was 128.3 ± 70.1 cm 2 (minimum: 20; maximum: 374). This difference in the extreme values is related to the fact that some articles specified only the surface area of the defect and not those of the flaps used and vice versa.

Flap Types
Data are presented in Table 6. The most frequently used flap type was the propeller perforator flaps in 224 cases (62.9%). We also found 43 advancement flaps (12.1%), 38 keystones (

Flap Procurement and Donor Site Closure
Preoperative perforator detection was mentioned in 26 articles (238 flaps). Doppler mapping was performed in 197 cases (82.8%). Perforator skeletonization only referred to propeller flaps, advancement flaps, peninsular flaps, and interpolation flaps. It was found in all of them, and skeletonization was performed in 100% of cases until satisfactory mobilization of the flap was obtained. Flap harvesting was specified for 206 flaps (propeller flaps, advancement flaps, peninsular flaps, and interpolation flaps), i.e., 61.5%. For the flaps involved, the dissection was subfascial in 172 cases (83.5%) and suprafascial in 34 cases (16.5%). For donor site closure, data were found for 322 flaps (83.2%). In total, 319 donor sites were managed with primary closure (99.1%). Three donor sites required coverage using skin grafting (0.9%).

Based on the Flap Type
The complication rate based on axillary location is presented in Table 8. The information was found for 83.

Based on the Type of Perforator
The complication rate according to the type of perforator for the axillary location is presented in Table 9. This data was specified for 80.1% of the axillary flaps (n = 258). No significant difference was found depending on the type of perforator ( Figure 3): TDAP flaps were no more complicated than PAP flaps (=0.3201), CSAP flaps (p = 0.3665), LTAP flaps (p = 0.5865), or IAP flaps (p > 0.9999). Table 9. Complication rate based on the perforator type for axillary location.  The complication rate based on the type of perforator among the axillary propeller flaps is presented in Table 10. The information was found for 81% of the axillary propeller flaps (n = 158). No significant difference was shown according to the type of perforator in this subpopulation (Figure 4). The complication rate found for TDAP flaps was not

Based on the Type of Perforator among the Propeller Flaps
The complication rate based on the type of perforator among the axillary propeller flaps is presented in Table 10. The information was found for 81% of the axillary propeller flaps (n = 158). No significant difference was shown according to the type of perforator in this subpopulation (Figure 4). The complication rate found for TDAP flaps was not significantly higher than for CSAP flaps (=0.3586), PAP flaps (p = 0.5935), IAP flaps (p > 0.9999), or LTAP flaps (p > 0.9999). Table 10. Complication rate based on the perforator type among the propeller flaps for axillary location.

Discussion
HS-related defects represent a true challenge for the plastic surgeon, as they require a large resurfacing with subsequent additional morbidity to the original disease. Conventional techniques represented by secondary healing and skin grafting have several limits with a slow recovery time which impacts the resumption of professional activities. Our review showed an overall recurrence rate of 2.7% after wide excision of the lesions and coverage with a perforator flap, for a mean follow-up time of 16.6 months. Second-

Discussion
HS-related defects represent a true challenge for the plastic surgeon, as they require a large resurfacing with subsequent additional morbidity to the original disease. Conventional techniques represented by secondary healing and skin grafting have several limits with a slow recovery time which impacts the resumption of professional activities. Our review showed an overall recurrence rate of 2.7% after wide excision of the lesions and coverage with a perforator flap, for a mean follow-up time of 16.6 months. Secondary healing has the advantage of not increasing the operative time, but is associated with problematic skin retraction in the peri-articular regions [26], as well as a considerable healing time, up 2-3 months [24,25]: in the active population of patients, this represents a delayed return to work and significant healthcare costs. Moreover, the recurrence rates observed in the literature after secondary wound healing are not negligible: Ovadja et al. [23] observed an 11% recurrence rate; Deckers et al. [73] reported a recurrence rate of 37.6% in a total of 253 cases. The median time to recurrence in this last retrospective study was 6 months (IQR: 3.0-13.0). The use of skin grafts after wide HS excision is also widely adopted: nevertheless, the uncertainty of the graft take can lead to a period of secondary healing [25], immobilization is necessary to promote success [74], and there is non-negligible morbidity of the donor site. Moreover, several studies report a failure rate of over 40% in this indication [27,28,75], and the time to return to work activity was significantly longer than after reconstruction with a perforator flap [28]. Regarding recurrence after skin grafting, the meta-analyses of Ovadja et al. [23] and Mehdizadeh et al. [20] found rates of 2% and 6%, respectively, even up to 20% in some studies [24,76]. In our review, with the use of perforator flaps, we found an overall complication rate of 15.1%, i.e., 58 complications for 383 flaps, with only 2 total necroses. In total, 12 cases of partial necrosis were observed, the most frequent complication being wound dehiscence. These complications led to revision surgery in 3.6% of cases. When comparing the use of perforator flaps in HS to other indications, a recent literature review reported an overall complication rate of 12.3% for 432 thoracic pedicled perforator flaps used for reconstruction after breast cancer [77]. In the axillary region, Jiang et al. [78] described a complication rate of 12.5% in a series of 32 TDAPs used to treat axillary scar contractures. In a retrospective study, Brunetti et al. [79] reported a complication rate of 23.1% in a total of 130 consecutive flaps for reconstructions of various locations. The majority of these defects were related to carcinologic resection (117 cases). Finally, in the lower limbs, the complication rate found after coverage of burns and traumatic, tumoral, or infectious lesions with perforator flaps was 23% [80]. In the axillary region, our review of 183 perforator propeller flaps showed a complication rate of 20.2%, with 37 complications. A meta-analysis by Florczak et al. [81] showed an average complication rate of 9.9% for 182 flaps in the thoracic region. In the review by Lazzeri et al. [82] on 288 propeller perforator flaps in the head and neck, trunk, and upper limb regions, the complication rate was 13.8%. Complications were more frequent when the defects were located at the extremities: regarding the upper limb, 23.9% of the flaps were complicated in the review by Vitse et al. [83] (on 117 defects): for the lower limb, 25.2% of flap complications were reported in the meta-analysis by Bekara et al. [84], on a total of 428 propeller flaps. The origin of these complications was mainly traumatic. The complication rate of perforator flaps in HS thus appears to be comparable to the complication rates of the same type of flaps encountered for other surgical indications and locations. In this review, we analyzed the complication rates for the different families and types of perforator flaps in the axilla. Similarly, the type of perforator flap did not appear to be a risk factor for complication, both in the total population of axillary reconstructions and in the subgroup of propeller flaps.
The surgical procedure is therefore not trivial, and the complication rate is significant, especially in case of highly active disease. An interesting approach is the combination of a surgical procedure with an adjuvant or even neo-adjuvant treatment. For moderate to severe HS stages, immunomodulatory drugs seem to have certain efficiency. If Adalimumab is currently the only validated biotherapy, Janus kinase and C5a inhibitors and antagonists are under investigation [85].
The SHARPS randomized clinical trial [86] showed an improved clinical response with adalimumab vs. placebo perioperatively for all treated body regions. However, the results were poorly significant (p-value of 0.049), the follow-up period was relatively short (12 weeks), and the trial highlighted the potential complications of this biotherapy. Further studies should assess better the effects of combined treatment with surgery and adalimumab or future approved molecules. A phase IV trial is currently underway that may provide guidance on the preoperative use of adalimumab to improve local status at the time of surgery [87].
We recognize several limitations to our study. First of all, as this was a systematic review, we had to include articles with a low level of scientific evidence given that they were based on retrospective studies. Our team is currently involved in a randomized multicentric trial comparing the outcomes of secondary wound healing versus pedicled perforator flaps after radical HS excision [88] to bring better evidence in this field. Secondly, the complication rate may have been underestimated because of a subjective assessment depending on the surgeon. The lack and heterogeneity of data concerning surgical techniques were recurrent problems, which may have compromised the identification of risk factors for complications according to the flap's characteristics. As our review was observational, a randomized controlled trial with standardized data collection would be needed to establish a management decision algorithm based on risk factors for complications.
In conclusion, our study highlights both the safety and reliability of perforator flaps for covering defects following the excision of HS. Perforator flaps offer a suitable option to cover wide defects after HS surgical treatment. It is a method that ensures a good quality reconstruction, is custom-made, reliable, and reproducible, as has already been demonstrated for various etiologies. Its low morbidity, due to the respect for the underlying muscle and the direct closure of the donor site, makes it a technique of choice for the coverage of these defects in patients whose quality of life is already significantly impacted by the disease.