Introduction

The latissimus dorsi (LD) muscle and musculocutaneous flap has gained steady popularity in reconstruction fields since first introduced by Tansini in 1906 [1]. Owing to its ability to provide abundant, well-vascularized muscle tissue and a reliable vascular pedicle, the LD flap has been chosen most often for reconstruction of various defects. However, donor-site morbidity, including seroma formation, still remains a concerning issue [2–4].

Seroma, defined as a mass-like lesion or swelling resulting from the localized accumulation of fluid or serum within a tissue or organ [5], is a well recognized and not uncommon complication following LD muscle transfer. The incidence of seroma after LD muscle harvest has been reported to be high, occurring in up to 80 % of patients [2–4, 6]. Although they have yet to be documented clearly, the suggested mechanisms of seroma formation include the development of a dead space after flap harvest, potentially causing shearing effects with inflammatory reactions, and prolonged leakage into this dead space by disrupted vessels and lymphatics as a result of surgical dissection [3, 7–9]. Donor-site seroma following LD muscle transfer can lead to patient discomfort and anxiety and also may create wound problems, including wound infection due to repetitive aspiration, wound dehiscence, and overlying skin flap necrosis. Many efforts have sought to reduce the incidence of seroma formation after LD muscle harvest, including avoidance of diathermy dissection, endoscopic or robotic dissection, pressure dressings, use of quilting sutures, and instillation of fibrin sealants; [10, 11] the latter two methods have been the strategies most commonly employed in recent years [11, 12]. Several investigations have been conducted to determine their efficacy on the prevention of seroma formation following LD muscle transfer; however, these studies had contradicting results [13–18], and there remains no consensus regarding the efficacy of these techniques. As most studies have been based on a single institution and a small number of cases, their statistical power may be limited; a systematic review of the literature may be helpful to come to solid conclusions. Although Sajid et al. reviewed the efficacy of quilting sutures for the prevention of seroma after LD muscle transfer, they analyzed only a few seroma-related parameters with a limited number of articles and did not include articles on the effects of fibrin sealants alone. Given that fibrin is one of the main strategies for preventing seroma and has been frequently used with quilting sutures, analysis of the efficacy of both methods may allow a synthetic approach for reducing the seroma following LD muscle transfer. The aim of this review is to evaluate the benefits of fibrin sealants and quilting sutures for the prevention of seroma formation following LD muscle transfer by meta-analytic methodology.

Patients and Methods

In June 2014, a literature search was performed using Medline, Ovid, and Cochrane databases for articles on the efficacy of fibrin sealants or quilting sutures for the prevention of donor-site seroma after LD muscle harvest. “Latissimus dorsi” AND “seroma” AND (“fibrin sealant” OR “quilting suture”) were used as keywords. Manual screening of the reference articles was also conducted to retrieve relevant studies. After excluding the duplicates, 43 citations were identified from the first search. Then, abstracts were reviewed for further selection based on predefined inclusion and exclusion criteria. Two authors independently identified the candidate studies and resolved disagreements by consensus discussion. The primary outcome measure in this review was the incidence of seroma formation, and secondary outcomes included total drainage volume from the back, periods of drainage in situ, volume of seroma and frequency of aspiration for treatment of seroma.

Fundamentally, all studies evaluating the effects of fibrin sealants or quilting sutures on the development of seroma formation following LD muscle or musculocutaneous flap harvest were regarded as candidates. As seroma-related outcomes can vary widely according to patient characteristics, surgical technique or flap type, only two-arm studies comparing the outcome of an intervention group with that of a control group were included in the current meta-analysis. Studies comparing the outcomes of a fibrin group versus a control group, a quilting suture group versus a control group, a fibrin and quilting suture group versus a fibrin alone group, and a fibrin and quilting suture group versus a quilting suture alone group were all included in this review. Studies that included information regarding outcome measures or raw data allowing for our calculations were enrolled. Exclusion criteria were reviews or meta-analyses, case reports or series with fewer than ten cases, and letters or discussions. Single-arm studies evaluating the efficacy of fibrin instillation or quilting sutures without comparison with control groups were also excluded, as previously mentioned. Articles evaluating the efficacy of these procedures on seroma in sites other than the donor site of the LD muscle flap were also excluded. A full-text review was carried out for articles that satisfied the aforementioned selection criteria in the abstract review process. Finally, a total of 14 articles were included in the following analysis (Fig. 1).

Fig. 1
figure 1

Study attrition diagram

Full size image

Data Extraction

The following data were extracted from articles according to the guidelines of the Meta-analysis of Observational Studies in Epidemiology [19]; the first author, institution, publication year, study period, study design, number of patients and cases, indications for LD flap harvest, type of LD flaps (muscle-sparing LD, conventional LD, extended LD flaps), mean age, mean body mass index, percentage of active smokers, and patients having underlying co-morbidities, dissection methods used for LD flap harvest (sharp/diathermy dissection), procedures used for preventing seroma formation (including instillation of fibrin sealants, quilting or their combination), incidence of seroma, total drainage volume from the back, periods of drainage in situ, volume of seroma, and frequency of aspiration for treatment of seroma.

Statistical Analysis

The pooled relative risk (RR) and corresponding 95 % confidence interval (CI) of each procedure were estimated for the development of seroma using the Mantel–Haenszel test. The mean difference (MD) and 95 % CI were estimated for the continuous variables, including total drainage volume from the donor site, periods of drainage in situ, volume of seroma, and frequency of seroma aspiration using inverse variance. In the articles that did not describe the mean values for continuous outcome, median values were used instead. When the standard deviation (SD) was not described, it was calculated from the range of the values or from the p value using the guidelines of the Cochrane Collaboration. The heterogeneity of studies was explored by the Cochrane Q statistic and the I 2 test. When the I 2 value was lower than 30 %, the analyzed studies were considered relatively homogenous, and a fixed effect model was employed to pool the weighted data. Otherwise, a random effect model was applied. The findings are presented as forest plots. Statistical analysis was performed using RevMan 5.2, which was provided by the Cochrane Collaboration.

Additionally, using simple data pooling methods, the pooled incidence of seroma in patient subgroups of each procedure, including fibrin alone, quilting alone, a combination of both and a no-intervention group, was also calculated and compared with the results of this meta-analysis.

Results

Fourteen studies representing 932 cases of LD muscle or musculocutaneous flap harvest were analyzed in the present review. There were three prospective randomized controlled studies [10, 17, 20], one prospective observational study [13], two non-concurrent cohort studies [18, 21], and eight retrospective studies included [6, 14–16, 22–25] (Table 1).

Table 1 Characteristics of included studies
Full size table

Six of 14 studies investigated the efficacy of fibrin sealants for the prevention of seroma formation following LD flap harvest. Three compared the outcomes of fibrin-treated and control groups, and the other three drew comparisons between the group receiving both fibrin and quilting sutures and that receiving quilting sutures alone. Owing to the possibility of heterogeneity among studies, those two categories of comparisons were investigated separately in the following analysis. Eight studies assessed the effects of quilting sutures. All of these compared the quilting suture group with a no-intervention group as a control group except for one study that compared a combination quilting and fibrin group to fibrin alone. These two categories of comparisons were also analyzed respectively.

Fibrin Sealants and the Prevention of Seroma Formation

Fibrin Versus Lack of Intervention

Three studies [16, 17, 24] comparing the outcome of fibrin instillation with lack of intervention were analyzed; 189 cases of LD flap harvest were included (Table 2). Overall, no statistically significant differences were observed in the following three analyses for risk of seroma formation, total drainage volume, and periods of drainage in situ (Fig. 2). In estimating the pooled relative risk of seroma development, nonsignificant protective effects of fibrin sealants were observed. Fibrin instillation reduced the risk for seroma formation by approximately 23 % compared with lack of intervention, though this difference was not statistically significant (RR 0.77, 95 % CI 0.22–2.72, random effect model; I 2 = 66 %). Similar results were seen in the analysis for total drainage volume from the donor site, and there was a trend toward a decreased total drainage volume in the fibrin group, although the difference was not statistically significant either (MD −104.73, 95 % CI −461.09 to 251.62, random effect model; I 2 = 58 %). Somewhat different trends were observed in the analysis of periods of drainage in situ; fibrin instillation rather increased the period of drainage in situ compared with lack of intervention (MD 0.81, 95 % CI −0.73 to 2.36, fixed effect model; I 2 = 0 %). With regard to the volume of seroma and the frequency of seroma aspiration, only one prior study [17] has described these data, but a meta-analysis could not be performed. In the study mentioned, no statistically significant differences were observed in either the seroma volume or the aspiration frequency between the two groups.

Table 2 Comparison between Fibrin group versus no-intervention group
Full size table
Fig. 2
figure 2

Forest plots evaluating the efficacy of fibrin instillation compared with lack of intervention. (Upper) Pooled relative risk for seroma formation. (Middle) Mean difference in total drainage volume from the donor site. (Lower) Mean difference in the periods of drainage in situ. None of the groups exhibited statistically significant benefits from fibrin instillation compared with lack of intervention

Full size image

Fibrin and Quilting Suture Versus Quilting Suture Alone

A total of three studies [10, 21, 23] representing 103 LD flap cases were used to evaluate the efficacy of fibrin by comparing the fibrin and quilting group with the quilting alone group (Table 3). Fibrin instillation generated significant differences in seroma-related outcomes, in contrast with prior analysis (Fig. 3). The combination of fibrin and quilting sutures almost halved the risk of seroma formation compared to quilting suture alone, although this difference was not statistically significant (RR 0.51, 95 % CI 0.12–2.16, fixed effect model; I 2 = 0 %). With regard to total drainage volume, additional instillation of fibrin into the quilting suture reduced the drainage volume significantly compared with use of quilting suture alone (MD −320.80, 95 % CI −389.92 to −251.68, fixed effect model; I 2 = 0 %). Similar results were revealed in the analysis of periods of drainage in situ where fibrin sealants used along with quilting sutures decreased drainage periods significantly, by approximately 1.6 days (MD −1.62, 95 % CI −2.91 to −0.33, random effect model; I 2 = 55 %). Data regarding seroma volume and frequency of seroma aspiration were extractable from only one study [10]. In this prospective randomized study, seroma volume was significantly lower in the fibrin-added group than in the quilting suture alone group (30 vs. 196.7 cc). The frequency of aspiration was also lower in the fibrin and quilting group, although the difference was not statistically significant (1 vs. 2.7 days).

Table 3 Comparison between Fibrin and Quilting suture group versus Quilting suture alone group
Full size table
Fig. 3
figure 3

Forest plots comparing the efficacy of combined fibrin and quilting suture versus quilting suture alone. (Upper) Pooled relative risk for seroma formation. (Middle) Mean difference in total drainage volume from the donor site. (Lower) Mean difference in the periods of drainage in situ. Fibrin instillation combined with quilting suture approximately halved the risk for seroma formation, though this difference was not statistically significant, and significantly reduced total drainage volume and drain indwelling periods

Full size image

Quilting Suture and the Prevention of Seroma Formation

Quilting Suture Versus Lack of Intervention

Seven studies [6, 13, 14, 18, 20, 22, 25] compared the outcome of a quilting suture group with a no-intervention group to evaluate the efficacy of the quilting suture for the prevention of seroma formation after LD flap harvest. Our study included a total of 594 LD flap cases, including 248 cases with quilting sutures and 346 without intervention (Table 4). Overall, significant protective effects of quilting sutures against seroma-related morbidities were observed (Fig. 4). Quilting suture reduced the risk of seroma formation significantly compared with lack of intervention (RR 0.38, 95 % CI 0.19–0.75, random effect model; I 2 = 88 %). Similar results were observed in the analysis of total drainage volume and periods of drainage in situ. There was a significant trend toward a decreased total volume of drainage in the quilting suture group compared with the no-intervention group (MD −284.10, 95 % CI −474.61 to −93.60, fixed effect model; I 2 = 0 %). With regard to the periods of drainage in situ, the use of quilting sutures decreased the drainage period significantly by approximately 3.7 days compared with lack of intervention (MD −3.65, 95 % CI −5.43 to −1.87, fixed effect model; I 2 = 0 %). Meta-analysis of seroma volume and frequency of seroma formation was conducted limitedly due to the insufficient number of studies that contained this information. Quilting suture was found to reduce seroma volume (MD −443.0, 95 % CI −1243.31 to 357.31) and aspiration frequency (MD −3.0, 95 % CI −4.35 to −1.65).

Table 4 Comparison between Quilting suture group versus No-intervention group
Full size table
Fig. 4
figure 4

Forest plots evaluating the efficacy of quilting sutures compared with lack of intervention. (Upper) Pooled relative risk for seroma formation. (Middle) Mean difference in total drainage volume from the donor site. (Lower) Mean difference in the periods of drainage in situ. All groups showed the significant protective effects of quilting suture against seroma-related morbidities compared with lack of intervention

Full size image

Quilting Suture and Fibrin Versus Fibrin Alone

Only one study compared the outcomes of the group receiving both quilting suture and fibrin and that receiving fibrin alone. In that study, quilting sutures with fibrin instillation reduced the seroma rate, total drainage volume, seroma volume, and frequency of seroma aspiration significantly compared with fibrin alone. Periods of drainage in situ were also shorter in the combination treatment group, a difference that was marginally significant.

Pooled Analysis of Seroma Incidence

Based on 14 studies, the pooled incidence of seroma for the groups of each procedure was estimated. The group that received both procedures had the lowest incidence (16.2 %), followed by quilting suture alone (33.2 %) and fibrin sealants alone (34.8 %) groups. The group that received no intervention had the highest incidence (54.6 %) of seroma formation (Fig. 5).

Fig. 5
figure 5

Pooled incidence of seroma formation in the groups of each procedure. The group receiving combination fibrin and quilting suture had the lowest incidence of seroma (16.2 %), followed by the quilting suture alone (33.2 %), the fibrin alone (34.8 %) and the no-intervention (54.6 %) groups

Full size image

Discussion

Fibrin sealants seal transected capillaries and lymphatics and promote hemostasis and tissue adherence through a fibrin-mediated cascade, leading to obliteration of dead space and consequently preventing seroma formation. The effects of fibrin and their prevention of seroma formation have also been proven in animal models [11]. However, according to this meta-analysis, the instillation of fibrin sealants alone failed to reduce the incidence of seroma and drainage volume following LD flap harvest. No statistical significance was observed in the risk for seroma formation and total drainage volume between the fibrin group and the no-intervention group. Furthermore, the periods of drainage in situ were even longer in the fibrin group, although this difference was not significant. These results were consistent with those of other clinical studies conducted on seroma that developed in other sites, which also failed to prove the efficacy of fibrin. Fibrin sealants alone may not have a significant protective effect against seroma formation following LD muscle harvest.

Investigations seeking to identify the optimal conditions to increase the efficacy of fibrin sealants have been conducted, but no recommendations have been firmly established. In three studies included in this meta-analysis, external compression was applied for 3–5 min to ensure the obliteration of any dead space after fibrin instillation. Given that fibrin failed to demonstrate any significant benefit, it is likely that sustained pressure following fibrin instillation is suboptimal for maximizing the adhering effects of fibrin.

In this study, the results of a meta-analysis of studies comparing combination fibrin and quilting suture group and a quilting suture alone group were noteworthy. In contrast with prior analysis, the application of fibrin almost halved the risk of seroma formation and made a significant difference on total drainage volume and drain indwelling period when performed in patients receiving quilting suture. Therefore, the efficacy of fibrin sealants may be greatly magnified when they are applied with a combination of quilting sutures, and the fixation effects of quilting suture may maximize the adherence effects of fibrin, eventually preventing seroma formation. The results of Shin et al.’s study [15], which demonstrated that the combination of fibrin and quilting suture was significantly superior to fibrin alone for reducing seroma rate and drainage volume, can be interpreted in a similar context.

The efficacy of quilting suture has been well established in animal models [5] and by other clinical investigations for seroma developed in other sites [11, 26, 27]. The present review demonstrated analogous results, in that quilting suture significantly reduced the incidence of seroma, total drainage volume, and periods of drainage in situ regardless of flap type and dissection method. Tissue fixation with quilting suture can successfully obliterate any dead space and minimize the shearing effects of skin flaps against the underlying structures, resulting in effective reduction of seroma development, as suggested in previous literature [18, 26].

Above all, the combination of both quilting suture and fibrin sealants showed significantly greater protective effect, reducing the risk of seroma formation, total drainage volume and periods of drainage in situ to a greater extent than either fibrin sealants alone or quilting suture alone. These results suggest that the combination of those two procedures may generate synergistic effects in the prevention of seroma-related morbidity following LD muscle transfer. Although further well-controlled studies will be required to obtain more solid conclusions and to investigate the optimal protocols for their use, the application of both procedures together can be recommended as a standard procedure.

The results of a pooled analysis also corroborated those of this meta-analysis. Although data from each study were simply combined without being weighted according to the total number of cases or the incidence of events, similar trends were observed in that the combination treatment had the lowest pooled incidence of seroma, followed by quilting suture, fibrin instillation, and lack of intervention, which also supported the results of this meta-analysis.

Despite a robust review of the literature focusing on two main strategies for seroma prevention, some inherent limitations could not be avoided in the present meta-analysis. More than half of the included studies had a retrospective and non-randomized design; therefore, the variables that could influence outcome, such as flap type, recipient site, or dissection method, could not be fully controlled, which may have reduced the strength of this review. Furthermore, the definition and diagnostic criteria for seroma formation and protocols of drainage removal were different among the studies, which can result in heterogeneity of the analyses. To overcome these shortcomings, further large-scale, well-designed studies are needed.

Conclusions

According to the present review, fibrin instillation alone did not significantly influence the development of seroma; however, it produced significant protective effects when used with quilting suture. Quilting suture significantly reduced the risk of seroma-related donor complications following LD muscle transfer. Above all, the combination of both procedures showed the best efficacy for preventing seroma-related morbidities compared with fibrin or quilting suture alone, which was supported by both meta-analysis and pooled analysis. Although further research involving large-scale, randomized studies is warranted, a combination of quilting suture and fibrin sealants may be recommended to reduce seroma-related donor morbidity after LD muscle transfer.