A Review of Perforator Flaps for Burn Scar Contractures of Joints

Abstract

Objective:

Perforator flaps are one possible surgical treatment for burn scar contractures; however, a review of evidence on this topic is lacking.

Methods:

MEDLINE was searched for articles related to perforator flaps for burn contractures. Following title and abstract screen, full texts were searched to identify articles describing perforator flaps for burn scar joint contractures. Data were extracted and summarized descriptively. Only articles that contained ≥10 patients with burn scar contracture were considered.

Results:

Two hundred forty-eight articles were identified, of which 17 met criteria for review. Of these, 16 were low-quality case series, while 1 was an open randomized controlled trial. In total, perforator flaps were performed on 339 patients (age range: 3-75 years), with the most common contracture locations being cervical (n = 218) and knee (n = 41). Nine of the 17 articles described a rehabilitation strategy. In general, functional outcomes were excellent, with the majority of patients experiencing return of normal joint range of motion and no recontracture. Compared to full-thickness skin grafts, perforator flaps showed greater improvements in joint range of motion. Cosmetically, perforator flaps were shown to have good color match with surrounding tissue, good contour around anatomical landmarks, and improved overall patient appearance. The most common complications were marginal flap necrosis (n = 26 patients) and venous congestion (n = 17 patients).

Conclusions:

Preliminary evidence from low-quality case series and 1 high-quality trial suggests perforator flaps may be successful for resurfacing released burn scar contractures; however, there is a need for additional trials comparing perforator flaps to other approaches.

Introduction

Burns represent a major cause of morbidity worldwide.¹ While resuscitation and management of burns has greatly improved over the past many years, many patients still experience ongoing functional and cosmetic disability from burn scar contracture. Burn scar contractures of the skin are more likely to occur following deep partial thickness or full-thickness burns that do not receive primary excision and skin grafting but may also occur after excision and skin grafting. The shortened and abnormally tight scar tissue may be a consequence of abnormal scar remodeling, inappropriate initial management, or lack of physical rehabilitation following burn injury. Recent studies have suggested that 23% to 54% of individuals admitted to burn units and requiring autograft eventually develop a burn scar contracture associated with some disability.^2–5

Burn scar contractures can cause abnormal or limited motion.⁶ Epidemiologically, the most common joint to experience contracture is the shoulder; however, any joint or tissue surface that experiences a major burn is susceptible.^3,4 The incidence of burn scar contractures may be reduced by implementing early physical rehabilitation intervention following burn injury.^7,8 On the other hand, it has been shown that while helpful, early joint stretching does not completely eliminate the occurrence of burn scar contractures.^9,10 Recent mathematical models of wound healing have suggested that a threshold exists, beyond which even early and aggressive physical rehabilitation may not influence burn scar contracture formation.¹¹ Once burn scar contractures have become established, surgical intervention is often necessary to release the contracture, followed by physical rehabilitation to maintain the increased range of motion and function of the affected joint.

When managing burn scar contracture, there is surprisingly little scientific evidence to support one surgical approach over another.¹² Our experience suggests the most common approach involves contracture release, followed by resurfacing of the resulting defect with a skin graft—a procedure first introduced over 100 years ago.¹³ Flaps are used when the resulting defect has exposed bone, tendon, joint capsule, or any other non-graftable surface. It may be that even in cases where the wound base is graftable, the thicker, more pliable coverage afforded by a flap is desirable; a perforator flap represents 1 flap option. Perforator flaps, designed to contain a source vessel, theoretically import skin and soft tissue with robust perfusion and greater pliability than a skin graft. However, evidence on the clinical outcomes of perforator flaps for burn scar contracture reconstruction is limited. This is a problem for the practicing plastic surgeon, as there is currently no summary of the best available evidence for reference when managing a patient who may benefit from such a procedure.

Given the scarce evidence on the topic of perforator flaps for burn scar joint contractures, we sought to conduct a literature review on the topic and summarize the best available evidence. The purpose of this article is to review reported cases where perforator flaps were used for reconstruction following release of cutaneous burn scar joint contractures and to summarize the rehabilitation, functional, cosmetic, and safety outcomes of this procedure. It is hoped that this review may serve as a resource for surgeons considering reconstruction of burn scar contractures with perforator flaps and highlight potential benefits and concerns for those involved in the management of patients with burn scar contracture.

Methods

The full study selection process is shown in Figure 1. MEDLINE was searched using the following strategy: [(perforat*.mp AND flap*.mp) OR Perforator Flap (MeSH)] AND [burn*.mp OR Burns (MeSH)]. The search was performed without any restrictions to date or language and revealed 248 articles.

Figure 1.

Process flowchart for selecting articles for review. Some articles contained patients with burn contracture and other surgical indications. In these cases, the article was only retained for review if outcomes specific to the burn contracture patients were reported.

Articles identified by the initial search were then reviewed individually by title and abstract to select articles of potential relevance to our review, based on whether the article might address burn scar contracture or perforator flaps. In total, 66 articles were identified by title and abstract screen that were of potential relevance for our review.

Finally, each of the 66 articles were reviewed in full text to select the articles that would be included in the review. Primary inclusion criteria were (1) surgical management of burn scar contracture and (2) use of perforator flaps. As the vast majority of the identified articles were isolated case studies, secondary inclusion criteria were applied to enhance the overall quality of evidence provided for this review: (1) articles must report on at least 10 cases of burn scar contracture and (2) articles must provide data specific to burn scar contracture patients. Using these criteria, articles containing reports of patients with burn scar contracture as well as other surgical indications were only considered if there were at least 10 burn scar contracture patients and with data specific to burn scar contracture patient outcomes distinguished from other patient outcomes. Together, these criteria help to ensure a minimum amount of person-time follow-up in each study and allow for greater statistical power from each study, together serving to reduce bias. This resulted in 17 articles that were eligible for inclusion in the review.

In this review, we considered a broad definition for perforator flaps to be any of the following (Figure 2): (1) a pedicled flap in which the flap site was planned to prevent transection of a known cutaneous perforator supplying the flap, (2) an island flap with a retained intact vascular perforator, or (3) a free flap with a transected vascular perforator that would be anastomosed at the defect site. Importantly, it must be acknowledged that pedicled flaps are not typically considered true perforator flaps. They were considered in this review only on the basis of an important distinction between pedicled perforator flaps and standard pedicled flaps: that the authors intentionally identified a known cutaneous perforator, usually by Doppler, and actively planned surgery around preservation of this vascular supply. This is similar to an island flap, but without full tissue dissection and elevation. This is in contrast to standard pedicled flaps where the collateral blood supply to the flap is not necessarily planned in advance, and instead assumed. While this broad definition of perforator flaps is atypical, it was considered here for completeness in summarizing all studies that used any attempt to preserve a known perforating vessel in flap design, given the paucity of summary data on this topic. Since our initial search strategy was broad and utilized standard terminology of true perforator flaps, inclusion of pedicled perforator flaps would not affect identification of true perforator flaps (ie, island and free flaps) and instead only served to broaden our review.

Figure 2.

Definitions for method of flap elevation used in this review. A, Pedicled flap, where a bridge of healthy tissue remains intact with a known cutaneous perforator supplying the elevated flap. B, Island flap, where a perforator is dissected but left intact. C, Free flap, where the perforator is dissected and then transected with the intent of anastomosing the vessel at a distant site. In each, the red lines represent the perforator source to the flap. (A) and (B) may be referred to as local flaps, as their use is restricted by the length and/or maneuverability of the tissue or vascular pedicle.

The identified articles were typically case series with heterogeneous patients, surgical methods, and outcomes descriptions. Consequently, meta-analyses of data were determined to be inappropriate. Therefore, we extracted all relevant information from each article and summarized data across articles descriptively. Data extraction was performed and confirmed by 2 researchers (R.T.L. and L.C.C.). To provide indication of study quality, we utilized the Oxford Centre for Evidence Based Medicine (OCEBM) Levels of Evidence framework for treatment studies, which provides both a level of evidence for each individual study and an overall grade of recommendation for the data supplied in this review.¹⁴ Study quality was assessed and confirmed by 2 researchers (R.T.L. and L.C.C.).

Results

A total of 248 articles were identified, of which 17 met criteria for review (Figure 1). Table 1 shows a summary of study levels of evidence, while Tables 2 and 3 show characteristics of all articles, flaps, and outcomes, grouped by contracture location. The majority of articles originated from outside of North America and all but 1 were case series (level of evidence 4) of perforator flap surgery. As such, overall study quality in this review was low, with an overall evidence grade of recommendation of C for perforator flaps for burn scar contracture. However, 1 article was a high-quality open randomized controlled trial (level of evidence 2) comparing perforator flaps (usually pedicle flaps) to full-thickness skin grafts (FTSG).³⁰ This trial found that perforator flaps resulted in greater improvements in range of motion and surface expansion compared to FTSGs without compromising skin elasticity. Furthermore, this study found that observers rated the appearance of flaps as better than FTSGs, while patients found no difference between the two. Those with perforator flaps had less color mismatch compared to those with FTSGs. Perforator flaps also had a lower rate of necrosis compared to FTSGs.

Table 1.

Study Quality and Overall Quality of Evidence.^a

Study Author	Study Design	Quality Level (1-5)
Cervical contractures
Grishkevich et al15	Case series	4
Grishkevich et al16	Case series	4
Li et al17	Case series	4
Mun et al18	Case series	4
Tsai et al19		4
Vinh et al20	Case series	4
Wang et al21	Case series	4
Wang et al22	Case series	4
Upper extremity contractures
El-Khatib et al23	Case series	4
El-Khatib et al24	Case series	4
Er et al25	Case series	4
Uygur et al26	Case series	4
Lower extremity contractures
Gupta et al27	Case series	4
Ismail et al28	Case series	4
Kim et al29	Case series	4
Miscellaneous contractures
Stekelenburg et al30	Randomized controlled trial	2
Tucker et al31	Case series	4
Overall grade of recommendation		C

^aQuality level was based on the Oxford Centre for Evidence-Based Medicine Levels of Evidence Scale from 1 to 5, where level 1 represents the highest quality level and level 5 represents the lowest quality level. Overall quality of evidence is determined based on the quality level of all studies reviewed on a scale from “a” to “d,” where level “a” represents consistently high-quality level 1 evidence and level “d” represents consistently level 5 evidence or troublingly inconsistent or inconclusive studies of any level.

Table 2.

Study, Patient, and Flap Characteristics.

Article	Country	Number of Patients	Age	Contracture Location	Flap Size	Perforator	Flap Material	Flap Type	Anatomical Region of Tissue Harvest	Donor Site Tissue Quality	Method of Donor Site Closure
Cervical Contractures
Grishkevich et al15	Russia	26 (10 M, 16 F)	Range of 6-54 yo	Cervical	Not reported	Superficial cervical artery perforators	Musculocutaneous	Pedicled flap	Neck	Scarred skin	Primary
Grishkevich et al16	Russia	32 (11 M, 21 F)	Range of 10-54 yo	Anterior and lateral cervical	Not reported	Superficial cervical artery perforators through sternocleidomastoid muscle	Adipocutaneous	Pedicled flap	Cervicothoracic	Scarred skin in n = 7	Primary
Li et al17	China	15 (5 M, 10 F)	Range of 10-51 yo	Cervical	Not reported	Superficial cervical artery perforators	Fasciocutaneous with muscle cuff on perforator	Pre-expanded pedicled perforator flap	Neck	Healthy skin	Primary
Mun et al18	South Korea	12 (4 M, 8 F)	Range of 21-47 yo	Anterior cervical	Range of 11 × 7 cm to 24 x 12 cm	Thoracodorsal latissimus perforator	Adipocutaneous	Free flap	Lateral torso and back	Healthy skin	Primary (n = 11), skin graft (n = 1)
Tsai et al16	Taiwan	40 (33 M, 7 F)	Range of 8-62 yo	Anterior cervical	Not reported	Anterolateral thigh (n = 32), lateral thigh perforator (n = 2), medial thigh perforator (n = 3), tensor fascia lata perforator (n = 2), thoracodorsal perforator (n = 1)	Adipocutaneous	Free flap (6 as split-free flaps)	Anterolateral thigh (n = 32), lateral thigh (n = 2), medial thigh (n = 3), tensor fascia lata (n = 2), thoracodorsal (n = 1)	Not reported	Not reported
Vinh et al20	Vietnam	17 (12 M, 5 F)	Range of 13-48 yo	Anterior cervicofacial	Range from 17 × 14 cm to 19 × 34 cm	Circumflex scapular artery perforator plus 1 additional perforator	Adipocutaneous	Bipedicled superthin free flap	Occipito-cervico-dorsal region	Some with scarred skin (unclear number)	Split-thickness skin graft
Wang et al21	China	12 (unknown male/female)	Range of 6-52 yo	Jaw-neck (n = 5), mental-neck (n = 7)	Maximum of 10 × 35 cm	Transverse cervical perforator plus adjacent bridging perforators	Adipocutaneous	Pre-expanded superthin island flap	Occipito-cervico-dorsal region	Not reported	Primary
Wang et al22	China	54 with perforator-based flaps. Within full study, 68 patients (42 M, 26 F)	Range of 18-65 yo	Anterior cervical	Range of 30 × 22 cm to 32 × 15 cm	Dependent on flap type, but usually perforators of the posterior circumflex humeral artery. Anterior perforating branches of the internal mammary artery for proximal end of extended deltopectoral flaps	Adipocutaneous or fasciocutaneous	Island deltopectoral flap (n = 36), Pedicled deltopectoral extended flap with secondary perforator supply (n = 6), free scapular flap (n = 4), island neck-shoulder flap (n = 8), Z-plasty (n = 14)	Shoulder, chest, neck, or back	Not reported	Skin graft (n = 10), remaining closed primarily
Upper Extremity Contractures
El-Khatib et al23	Egypt	18 (11 M, 7 F)	Range of 3-12 yo	Antecubital fossa	Range of 20 to 63 cm2	Proximal septal perforators of the radial artery	Fasciocutaneous	Island flap	Forearm	Not reported	Split-thickness skin graft
El-Khatib et al24	Egypt	13 (8 M, 5 F)	Range of 5-17 yo	Antecubital fossa	Range of 28 to 84 cm2	Proximal septal perforators of the ulnar artery	Adipofascial + skin graft	Island flap	Forearm	Not reported	Primary
Er et al25	Turkey	15 (15 M)	Range of 21-29 yo	Axillary	Range of 10 × 6 cm to 27 × 15 cm	Thoracodorsal latissimus perforator	Fasciocutaneous	Island flap	Lateral torso and back	Not reported	Primary (n = 14), skin graft (n = 1)
Uygur et al26	Turkey	11 (11 M)	Range of 19-22 yo	Palm	Range of 2 × 3 cm to 4 × 5.5 cm	Medial plantar artery	Fasciocutaneous	Free flap	Medial foot	Not reported	Primary (n = 1), skin graft (n = 10)
Lower Extremity Contractures
Gupta et al27	India	19 (8 M, 11 F)	Range of 5-56 yo	Knee	Range of 11.5 × 6 cm to 19 × 11 cm	Peroneal artery (n = 7), superior medial genicular artery (n = 5), superior lateral genicular artery (n = 3), lateral sural artery (n = 1), medial sural artery (n = 1), saphenous/inferior medial genicular artery (n = 3)	Fasciocutaneous	Pedicled perforator plus flap	Thigh or leg	Not reported	Split-thickness skin graft
Ismail et al28	Egypt	10 (10 M)	Range of 15-47 yo	Knee	Range of 8 × 16 cm to 12 × 26 cm	Lateral circumflex femoral artery perforators	Adipocutaneous with muscle cuff on perforator	Island flap	Thigh	Not reported	Primary (n = 1), split-thickness skin graft (n = 9)
Kim et al29	South Korea	12 (11 M, 1 F)	Range of 16-75 yo	Knee	Range of 7 × 9 cm to 15 × 23 cm	Medial sural perforator alone or in combination with adjacent perforators	Adipocutaneous	Island flap or island perforator plus flap	Calf	Scarred skin (2nd degree)	Primary (n = 1), split-thickness skin graft (n = 11)
Miscellaneous Contractures
Stekelenburg et al30	The Netherlands	30 (perforator group: 16 (8 M, 8 F); FTSG group: 14 (7 M, 7 F)	Perforator group mean age 49 (19), FTSG group mean age 39 (21)	Head/neck, arms, legs, back/thorax, abdomen	Not reported	Various, depending on contracture location	Adipocutaneous	Pedicled flap preferred (n = 14), converted to island flap if vascularization of flap appeared to be compromised (n = 2)	Various	Inclusion of scarred skin (n = 9 in flap group, n = 3 in FTSG group)	Primary
Tucker et al31	Nepal	17 patients, 22 cases.	Not reported	Neck (n = 10), axilla (n = 5), elbow (n = 5), hand (n = 1), back (n = 1)	Not reported	Circumflex scapular artery perforators, thoracodorsal artery perforators, or transverse cervical artery perforator, or ad hoc unnamed perforators	Fasciocutaneous	Pedicled flap	Area adjacent to defect	Healthy skin	Primary

Abbreviations: F, female; M, male; yo, years old.

Table 3.

Study Outcomes.

Article	Follow-Up	Rehabilitation Strategy	Functional Outcome	Cosmetic Outcome	Complications
Cervical Contractures
Grishkevich et al15	Range of 6 months to 9 years	No neck splinting was used	Full release and return of normal ROM in 24/26 patients. Two patients experienced mild recontracture	Normal cervicomental angle and scars became invisible	None
Grishkevich et al16	Range 6 months to 12 years	No neck splinting was used	Complete contracture release and maintenance of normal skin elasticity in all cases	Only operational scars visible after primary procedure in 25 patients. No displacement of breast tissue. Normal neck contours and overall improvement in patient appearance	Marginal flap necrosis (n = 8), venous congestion (n = 12)
Li et al17	At least 1 year after reconstruction	Not reported	All patients were able to extend beyond 110°, representing nearly normal ROM. No patients reported restricted ROM in neck flexion or rotation	Improved esthetic contour noted in 2 cases. Cosmesis not reported in other cases	Infection during tissue expansion (n = 2), expander deflation (n = 1), marginal necrosis (n = 1)
Mun et al18	Range of 6 to 34 months	Not reported	Improved neck extension by 19°, on average	Average cervicomental angle improvement of 48°	Marginal flap necrosis (n = 2)
Tsai et al19	Functional evaluation performed an average of 6.3 months postoperatively, follow-up period mean = 11 months	ROM exercises starting 1 week postoperatively to prevent neck stiffness. No rehabilitation garment was required. Patients were evaluated at monthly intervals. Rehabilitation contributed to increased flap width.	Improved neck ROM. Preoperative extension 83-110.2°, postop extension 114.5-120.1°. Preoperative lateral flexion 23.1°-30°, postop lateral flexion 34.4°-43.2°	Not reported	Venous thrombosis (n = 2) but healed after thrombectomy
Vinh et al20	6 months	Not reported	15/17 patients were satisfied with the functional outcomes	15/17 patients were satisfied with the cosmetic outcomes	Partial necrosis (n = 2), hypertrophic scar (n = 2)
Wang et al21	Range of 6 months to 3 years	Not reported	All patients had improved range of motion of their neck	All patients had improved appearance	Flap necrosis (n = 1) requiring skin grafting, superficial necrosis (n = 2)
Wang et al22	Range of 1 to 10 years	Not reported	No recontractures noted. Neck extension improved to 110° (angle of the chin and neck), preop was <85° in 17 cases, 85-95° in 37 cases	Patient’s profiles were more natural postoperatively and all patients were pleased with the resulting appearance	None
Upper Extremity Contractures
El-Khatib et al23	Not reported	Immobilization of elbow in semiflexed position for 10 days. Following this, rehabilitation therapy was encouraged	Full motion of the elbow in all cases	The flap was thin giving a normal appearance	Marginal flap necrosis (n = 2)
El-Khatib et al24	Not reported	Immobilization of elbow for 10 days in a semiflexed position. Following this, patients underwent alternating passive and active exercise of the elbow. At night, patients slept with a splint until full extension was achieved	Not reported	Not reported	None
Er et al25	Not reported	Not reported	Abduction to 46.6° (SD 19.3°) preoperation, 159.0° (SD 12.4°) postoperation	Satisfactory from patients’ perspective	None
Uygur et al26	Patients were discharged on average, 12 days postoperatively. Estimated follow-up range: 10 to 22 months	Not reported	Light touch, 2 point discrimination, pinprick, cold/warm sensation all found normal. MCP joint extension improvement ranged from 10° to 70°. MCP joint flexion improvement ranged from 0° to 40°. Grasp function of the hand was greatly improved	Satisfactory	Venous congestion (n = 2) resulting in marginal flap necrosis (n = 1) or debulking procedure (n = 1)
Lower Extremity Contractures
Gupta et al27	6 months	No splinting needed. Joints were mobilized after 1 week.	52.8° preoperation. Six months following operation, all patients could maintain an unassisted symmetrical upright posture and a bipedal locomotion, and they could squat without difficulty. 16 could sit cross-legged on the floor. All were satisfied with their function	Good outcome in 8, average outcome in 11 according to physician assessment. All patients were satisfied	Wound infection and dehiscence (n = 2), marginal flap necrosis (n = 1), anterolateral thigh paresthesia (n = 1), hypertrophic scar (n = 8)
Ismail et al28	Range from 6 months to 27 months (mean = 16.2 months)	Physiotherapy was started 10-14 days postoperatively, consisting of passive and active range of motion for the knee, strengthening of all muscle groups around the knee, and proprioception. Goal of rehabilitation was for the patient to regain ROM and power, comparable to the contralateral side	Functional outcomes were measured using a goniometer. Eight of 10 participants demonstrated improved knee ROM. Two patients did not regain full ROM but did improve to a point that no longer disrupted their usual motion	Participants did not complain about the donor site scar as it could easily be hidden	Hypertrophic scar (n = 3), superficial necrosis (n = 2), venous congestion (n = 2), partial flap loss from hematoma (n = 1), skin hyperpigmentation (n = 1), secondary debulking needed (n = 2)
Kim et al29	Range from 7 months to 32 months	Affected knees were immobilized for less than 2 weeks and then passive and active ROM exercises were started. All patients were asked to apply silicone gel under a pressure garment on donor site for at least 6 months	All patients were completely satisfied with the results. ROM specifically only reported for 1 selected patient: full range of motion at the knee by 15 months	All patients were completely satisfied with the results	None reported
Miscellaneous Contractures
Stekelenburg et al30	Mean follow-up of 3.1 ± 0.7 months and again at 12.5 ± −0.9 months	Not reported	Statistically significant increase in surface area of the flaps to 123% at 3 months and 142% at 12 months, versus 87% and 92% for full-thickness skin grafting (P < .001). Greater improvements in ROM were observed in the flap group versus FTSG group. No difference in elasticity between flap and graft group	Observers rated the scar assessment as significantly lower among those with flaps, although the patients did not report statistically significant differences in scar appearance between the flap versus graft group. The flap group reported less color difference between flap and normal skin when compared to the graft group	Lower rate of necrosis among flaps versus full-thickness skin grafts (6% compared to 17%). One full-thickness skin graft required reoperation due to complete necrosis
Tucker et al31	Not reported	Splinted with plaster cast until wound healing	Improved function and ROM. Only 1 case reported specific details, where shoulder abduction improved from 45° to 120°	Aesthetic outcomes better than expected with split skin grafting	Superficial flap necrosis (n = 3), venous congestion (n = 1). Three cases of delayed wound healing from minor dehiscence of the donor site reported

Abbreviations: F, female; FTSG, full-thickness skin graft; M, male; ROM, range of motion; yo, years old.

In total, this review summarizes the results of perforator flaps in 339 patients (age range: 3-75 years) with burn scar contractures across varying joints. There were 157 males, 99 females, and 83 in which sex was not reported or sex distribution could not be determined. The most common locations for contracture were cervical (n = 218), knee (n = 41), antecubital fossa (n = 36), axilla (n = 20), hand (n = 12), back (n = 1), and a mixture of head/neck, arm, leg, back/torso, abdomen (n = 16). Flap size ranged from 2 × 3 cm (6 cm² area) up to 19 × 34 cm (646 cm² area). Flap type was described primarily based on the vascular perforator and secondarily by the method of elevation (Figure 2). Among these flaps, 84 were performed as free flaps (most commonly for cervical contractures), 126 as island flaps, and 129 as pedicled flaps. The most common free flap used was the anterolateral thigh free flap. Flap composition was most commonly adipocutaneous, followed by fasciocutaneous and rarely musculocutaneous or adipofascial. Follow-up time ranged from 3 months to 12 years, although follow-up time was not reported in 4 studies. In the case of pedicled perforator flaps, donor sites were most often closed primarily. For island and free flaps, donor sites were generally closed primarily when the donor site was on the torso or back and generally closed with a skin graft when the donor site was on the extremities. At least 54 patients received a flap that contained scarred skin, usually as pedicled flaps. For repair of cervical contractures, superficial cervical artery pedicled flaps were most common (n = 73), as musculocutaneous, adipocutaneous, or fasciocutaneous. For knee contractures, the medial sural perforator adipocutaneous island flap was most common (n = 12). For antecubital fossa contractures, fasciocutaneous island flaps with proximal septal perforators of the radial artery were most common (n = 18). For axilla contractures, thoracodorsal latissimus perforator fasciocutaneous island flaps were the most common (n = 15). For hand contractures, medial plantar artery fasciocutaneous free flaps were most common (n = 11).

Nine of the 17 articles described a postoperative rehabilitation strategy. This generally involved some form of immobilization for 7 to 10 days after which mobilization (with or without formal physiotherapy) was encouraged. It is unclear how long physical rehabilitation was continued.

In general, functional outcomes were excellent, with most patients experiencing return of normal joint range of motion. Among the 17 articles, 8 gave objective data regarding range of motion of the contracted joint following surgery. The remaining articles simply reported functional outcomes in subjective terminology such as “excellent,” “all patients were satisfied,” or “improved range of motion.” Two articles provided information concerning performance of tasks in a functional capacity such as walking or sitting. Function was usually surgeon defined and not based on patient perception. Two articles commented on the neurologic or mechanical features of the flap such as sensation and elasticity.

Cosmetically, perforator flaps were believed by the authors to have good color match with surrounding tissue, good contour around anatomical landmarks, and improved overall appearance of the patient. There was no standardized approach for assessing cosmetic outcomes across studies, and most simply reported that patients were satisfied.

The most serious complication was flap necrosis (1 flap; pre-expanded transverse cervical perforator island flap) and partial flap loss (1 flap), but these were rare outcomes. The most common complications were venous congestion (17 flaps) and marginal/superficial flap necrosis (26 flaps). In total, 73 complications occurred at the recipient site, representing an overall complication rate of approximately 20%. Interestingly, 45 of 73 complications occurred with pedicled flaps, representing a complication rate of 35% with these flaps. Among island/free flaps, the complication rate was 13%.

Discussion

This review summarizes evidence on the use of perforator flaps for management of burn scar joint contracture. By providing levels of evidence for each study, it is hoped that the practicing physician may be able to utilize this review as a resource when considering applying a flap to their own patients with burn scar joint contractures. While overall quality of evidence identified in this review was low, with an overall OCEBM grade of recommendation of “C” based on most studies being case series in design, there was consensus in the findings across articles that perforator flaps led to favourable outcomes in terms of function, cosmesis, and low rates of severe adverse events. The evidence from these case series was further supported by a high-quality randomized controlled trial, where it was found that perforator flaps were superior to FTSG in terms of function, with fewer side effects, and perhaps superior in terms of cosmesis as well. Consequently, there is preliminary evidence from several case series and 1 high-quality randomized controlled trial to suggest that perforator flaps may be an effective option for management of burn scar contractures overlying joints. Importantly, this trial predominantly used pedicled perforator flaps, and no trial was identified assessing predominantly island or free flaps. A variety of different perforator flaps were successfully used across studies, suggesting that the specific flap used may depend on specific patient factors and surgeon preference, but it seems critical that further higher quality studies be undertaken to convincingly document the benefit of these procedures.

An important observation across studies was the significant degree of heterogeneity in outcomes reported. More objective reporting of outcomes is necessary to compare this surgical strategy to others. Specifically, many studies did not describe a rehabilitation plan or used vague terminology to describe functional and cosmetic results. Rehabilitation strategies may be important in ensuring optimal outcomes, and thus, future studies on this topic should provide detailed rehabilitation plans. For functional outcomes, we recommend detailed joint range of motion recording with a goniometer in the preoperative and postoperative state to quantify the benefit achieved from surgery. For cosmetic outcomes, we propose that a standardized scale be utilized that captures information such as normal contour/bulk, normal color, scarring, and patient satisfaction. Together, a more detailed approach to documenting rehabilitation, functional, and cosmetic outcomes will allow for improved knowledge translation regarding the benefits, risks, and indications for perforator flap surgery in patients with burn scar contractures.

Perforator flaps have been performed using a variety of modes of elevation, each with general success. However, complication rates appeared to be somewhat higher for pedicled flaps despite a seemingly more simple procedure. Since pedicled flaps were more likely to contain scarred tissue in flaps in this review, it is possible this damaged tissue may affect flap success. Nonetheless, it must be emphasized that these additional complications were generally minor, such as tip necrosis or transient venous congestion. Additionally, since the flaps contained scarred skin, it is possible these patients were more challenging cases, and thus, it remains unknown whether perhaps there would have been more severe complications for these patients if an island or free flap were to have been used.

This review is limited by substantial heterogeneity across studies, making any form of meta-analysis unfeasible. We attempted to reduce bias that could be introduced from low person-years follow-up (and therefore less time exposed to show benefits and adverse events) by selecting reports with a minimum sample size of 10 patients. However, in so doing, bias could also be introduced by failure to capture all existing data on the topic, and so other potential benefits and risks of perforator flap surgery for burn scar contracture may not be documented in this review. Importantly, we considered a broad definition for perforator flaps in this review, which included pedicled flaps with active attempt to preserve a known perforator. This broad definition did not likely affect our ability to identify true perforator flaps (ie, island and free flaps) in this review due to the use of recognized and broad terminology for perforator flaps in the original search strategy. However, it is possible we may have missed some studies that used pedicled flaps that did not use the terminology of “perforator” in their article since it is atypical to define these flaps as such. Thus, this review should be considered as a comprehensive review of the best quality evidence for island and free perforator flaps for burn scar contractures, with supplemental data from identified studies using pedicled flaps with a known perforator. Based on the overall low quality of evidence, as assessed by the OCEBM framework, and identification of just one high-quality trial, it is recommended that randomized controlled trials be undertaken to better establish the benefits of perforator flap surgery for burn scar contracture compared to other treatment modalities, especially since the only trial we identified predominantly considered pedicled flaps. Additionally, more detailed surgical procedure reporting would allow for improved clinical translation. Future investigations should explore the preferred type of perforator flap given the contracture location, whether it is preferable to use local tissue or a free flap, whether it is necessary to use skin grafting to close the donor site, what factors predict improved success with a perforator flap versus FTSG, and the recontracture rate between the 2 procedures.

In conclusion, this review has provided a summary of evidence available on the topic of perforator flap surgery for management of burn scar contractures overlying joints. Studies generally suggested favourable functional and cosmetic outcomes of perforator flaps with low complication rates; however, since overall study quality was low, with just one high-quality trial, more research is required to substantiate these findings and identify which patients and clinical situations are most likely to benefit from this surgical approach compared to more conventional approaches such as FTSG.