FormalPara What does this study add to the clinical work

This analysis confirms latissimus dorsi flap-based breast surgery as a robust technique with a low complication rate. The most common complication is the development of a seroma and flap loss is a rare occurrence.

Introduction

The use of autologous tissues is considered gold standard for patients undergoing breast reconstruction, and is the preferred method in the post-radiation setting [1, 2]. Several procedures are available including pedicled flaps such as latissimus dorsi flap (LDF) and transverse rectus abdominis myocutaneous flap (TRAM), as well as free flap requiring microvascular anastomoses like the deep inferior perforator flap (DIEP). Some of these, like LDF, may be combined with implants to provide adequate volume and symmetry. In patients with limited soft tissue following mastectomy, breast reconstruction may involve several steps, e.g., tissue expander placement before positioning a permanent implant, surgery of the contralateral breast, or nipple and areola reconstruction.

While LDF is one of the oldest described muscle flaps, it did not gain wide recognition until the 1970s [3]. In the following years, numerous variations were introduced, such as de-epithelialized flap for volume replacement and extended flap including lumbar fat to maximize tissue amount [4, 5]. Although LDF has been replaced by abdominal flaps as autologous tissue of choice for breast reconstruction in most countries, it remains a reliable alternative and a valuable option in several specific clinical situations [6]. Its advantages are the robust vascular supply making flap necrosis a rare event, and, in case of combination approach, lower rates of capsular contracture compared to implant-only based reconstruction [7,8,9,10].

Although the use of LDF is regaining popularity in recent years [7, 8], evidence on complication rates and outcomes is limited. The 2015 German S3 guideline on breast reconstruction with autologous tissue identified only 22 articles on LDF-based reconstruction that reported on complication rates, the largest including 78 flaps [11]. In this work, we present an 18-year retrospective analysis of a single-institution single-surgeon experience with LDF-based reconstruction with a focus on early complications and reconstructive failures.

Material and methods

In this retrospective study, clinical records of all patients undergoing breast surgery for any reason in the Certified Breast Cancer Center, Regio Klinikum Pinneberg, Germany between 01.04.2005 and 30.10.2022 were reviewed. 142 latissimus dorsi flap (LDF) reconstructive procedures were identified. The hospital electronic record was then reviewed for patient-, tumor- and treatment-related information. The following information were included in the database: age, body mass index (BMI), smoking habits, concomitant diseases and medication, timing of reconstruction, previous breast surgery, type of breast and axillary surgery, surgery duration, simultaneous use of an implant, complications, and secondary surgeries were included in the database. The protocol version 1.1 of this retrospective analysis has been reviewed by the Ethical Committee of the Medical Board Schleswig–Holstein (27/02/19).

Statistical analysis

Chi-squared test and Fisher’s exact test were used to evaluate the relationship between complication rates and surgical parameters. All reported p-values are two-sided. Statistical analysis was performed by SPSS, version 18 (SPSS Inc., Chicago, IL, USA).

Results

Patient characteristics, diagnosis, surgical setting, and previous and subsequent oncological treatments for the 142 LDF procedures performed in 140 patients are presented in Table 1. All procedures were performed by the same surgeon (P.P.). All procedures were performed as open surgery and neither endoscopic nor robotic approach was used. The humeral insertion of the latissimus dorsi muscle was transected in all cases [12]. All patients received single-shot antibiotic prophylaxis. Extended antibiotic prophylaxis was not used. Mean age was 61 and the median follow-up was 38.4 months. 25% of patients were smokers at time of surgery. Invasive breast cancer was the main leading diagnosis accounting for 78% of the cases (Table 1). 139 patients were female and only 1 patient was male. 138 patients received a unilateral LDF surgery and in two cases, LDF was used for both sides, albeit not simultaneously: (1) 38-year-old patient undergoing a mastectomy with immediate LDF- and implant reconstruction for locally advanced breast cancer was diagnosed with a contralateral multicentric breast cancer 8 years later and received skin-sparing mastectomy with LDF and implant; (2) 83-year-old patient undergoing a radical mastectomy for an ulcerated 19 cm large tumor with LDF coverage was diagnosed with extended contralateral cutaneous metastases 4 months later and underwent a mastectomy with wide-excision of all skin manifestations combined with LDF defect coverage. The number of procedures per year decreased during the period analyzed from 41 surgeries performed between 2005 and 2007 to 12 between 2020 and 2022 (Fig. 1, Supp. Table 1).

Table 1 Clinical characteristics of all LDF procedures included in the analysis
Full size table
Fig. 1
figure 1

Changes in the number of LDF-based surgeries performed per year in the period analyzed (2005–2022)

Full size image

118 patients (83%) received immediate reconstruction, either with (n = 82) or without (n = 36) implant placement (Fig. 2, Table 1). The median volume of the inserted implant was 230 ml (range, 95–560 ml). In five patients, LDF surgery was performed for delayed reconstruction after mastectomy. Ten patients required LDF for defect coverage after removal of an advanced tumor (Fig. 3), and in nine cases LDF surgery was performed as revision procedure due to complications after previous surgery (e.g., implant-to-flap conversion in case of capsular fibrosis or wound healing disorders). Median duration of surgery, defined as time between first incision and final skin closure, was 117 min, including all additional procedures such as full axillary lymph node dissection (ALND, performed in 45 cases), sentinel lymph node biopsy (SLNB, 43 cases), implant removal and/or placement, removal of the port system, contralateral surgery, nipple reconstruction etc. The shortest duration of surgery was 56 min (83-year-old patient with breast cancer metastasizing to the bone receiving a palliative radical mastectomy for a large ulcerated breast cancer with defect coverage using an LDF flap). Comparing different clinical settings, the shortest median duration of surgery was observed in patients receiving LDF surgery for defect coverage (median 79 min). Patients remained in the hospital for a median of 7 days (range: 2–23 days). Most patients received at least one surgery of the ipsilateral breast prior to LDF procedure (range: 1–6). 12% underwent ipsilateral radiation therapy before and 37% after the surgery.

Fig. 2
figure 2

Postoperative result 12 years after a nipple-sparing mastectomy with complete removal of the skin of both lower quadrants and LDF-based direct-to-implant reconstruction of the left breast

Full size image
Fig. 3
figure 3

Pre- and postoperative view of a male patient with a large breast cancer receiving wide excision and LDF defect coverage

Full size image

Complications following LDF surgery

Observed complications are summarized in Table 2. The most common complication was a seroma occurring in 26% of procedures. 19% of patients received at least one needle aspiration for seroma (range: 1–6) and in one case revision surgery with an easy flow drain placement was necessary. Total flap loss due to complications was observed in three cases (2.1%). Revision surgery, defined as a second surgery for any reason within 30 days of LDF procedure, was performed in 13 out of 142 (9.1%) patients. Details on these patients are shown in Table 3. Previous application of chemotherapy did not affect complication rate, but previous ipsilateral radiation therapy was associated with significantly higher rates of infection (24% vs. 5%, p = 0.005). Lower rates of infection and wound dehiscence were observed in patients receiving primary reconstruction, compared to delayed reconstruction and defect coverage. Higher number of previous breast surgeries was associated with higher rates of nipple and/or skin necrosis (p = 0.010) and hematoma (p = 0.020).

Table 2 Complications following LDF surgery and their association with other factors
Full size table
Table 3 Details on patients receiving revision surgery
Full size table

Follow up

The mean follow up was 38.4 months (range 0–200 months). None of the patients reported twitching or animation deformity of the flap. Out of 93 patients receiving an implant simultaneously with LDF surgery, 24 (26%) underwent implant removal. The median time of implant remaining in situ was 151.2 months (95% CI 90.8–211.6). The implants were mostly exchanged due to capsular fibrosis. In three cases the implant was permanently removed without replacement: in two cases the implant was removed due to invasive recurrence of the thoracic wall after 19 and 96 months, respectively, and in one case it was removed due to capsular contraction following radiation therapy without placement of a new implant (patient’s wish). The rate of implant removal and/or exchange in the long-term follow up was numerically higher in patients receiving pre- or postoperative radiation therapy but the difference was not statistically significant (32.4% vs. 21.8%, p = 0.256). The time point of radiation therapy was not significantly associated with implant removal and/or exchange (42.9% in case of radiation therapy before LDF surgery vs. 29% in case of postoperative radiation therapy).

Discussion

To the best of our knowledge, this is the largest single-surgeon report on clinical characteristics as well as complications in patients undergoing latissimus dorsi flap-based surgery of the breast. In contrast to other publications, which mostly reported on delayed LDF reconstructions, the majority of patients in our study received LDF in an immediate reconstruction setting directly following a mastectomy. [7, 13].

The most common early complication was a seroma, occurring in 26% of patients. This is in line with previous reports and some authors suggested that asymptomatic seroma should not be considered a complication but rather an inevitable side effect of LDF surgery [11, 14]. The current practice of electrocautery dissection may further contribute to development of symptomatic or recurrent seromas, and it is not uncommon to leave drains for several days or even weeks at the donor site. Tomita et al. analyzed potential risk factors for seroma development in 174 patients with LDF surgery [15]. Seroma occured in 28% of patients receiving mastectomy and LDF, while the incidence was much lower (11%) in patients undergoing a breast conserving surgery with LDF-based volume replacement. Higher age and BMI significantly increased the risk of seroma formation. In our study, age (p = 0.045) but not overweight were correlated to seroma development. Different approaches to prevent seroma formation have been explored in previous studies. Lee et al. performed a meta-analysis of 14 studies, including three randomized controlled trials, on quilting sutures and fibrin sealants [16]. Both interventions contributed in varying degrees to reducing seroma-related morbidity following LDF transfer, and their combination might have a synergistic effect. However, more than half of the included studies had a retrospective and non-randomized design and the definition of seroma varied. Further, operative time required for either intervention, additional costs, and the potential for allergic reaction to sealants need to be considered. Therefore, neither intervention is accepted as standard-of-care and the decision to use them depends on surgeon’s preference. In the present study, the prevalence of seroma was comparable to that reported in previous studies, even though no intervention for seroma prevention was used. While drain placement is commonly used in LDF surgery, recommendations regarding the timepoint of drain removal vary. Some authors recommend removing the last drain when drainage volume decreased under a specific amount (e.g., 20, 30 or 40 ml) per 24 h, while others prefer to leave the drains for a defined time, irrespective of the output [17, 18]. In the present analysis, drains were removed once the drainage volume decreased to < 30 ml in 24 h.

The most serious complication in the setting of flap-based reconstructive procedures is a partial or total flap loss, usually occurring due to disrupted perfusion and subsequent necrosis. While total flap loss has been observed in 1–2% of patients across studies, partial flap necrosis may occur in up to 9% of cases [11] (Table 4). In our study, we observed a total flap loss rate of 2%. No patient suffered from partial flap ross. Thus, these results confirm LDF surgery as a safe and robust method with a low failure risk.

Table 4 Overview of complication rates following latissimus dorsi flap-based surgery reported in the literature (only publications on ≥ 100 procedures were considered)
Full size table

Arm and shoulder morbidity and—in case of a simultaneous implant placement—capsular contracture, are considered typical late complications of LDF based surgery. In the setting of immediate or delayed reconstruction, LDF usually requires a combination with an implant to achieve desired breast size, due to the limited volume of the flap itself. In the present study, 26% of patients who received an implant required implant removal during follow-up. In most cases, this was followed by an immediate placement of a new implant (implant exchange). The reported implant removal or exchange rate in studies with longer follow-up may be as high as 50% [22,23,24]. Rates on the prevalence of capsular fibrosis following combined LDF and implant reconstruction vary widely among studies. Hardwicke et al. examined 277 consecutive LDF procedures combined with a placement of a textured cohesive gel implant, focusing on late complications [7]. After a mean follow-up of 47 months, higher-grade capsular contracture, defined as Baker III or IV, occurred in 3.6% of patients, resulting in a capsulotomy in all patients. Interestingly, administration of a chemotherapy was associated with a lower rate of capsular contracture, while no association with radiation therapy was seen. Similarly, in our analysis the administration of radiation therapy was not associated with implant removal rate in the long-term follow up. The mean time to first implant exchange was 22.4 months and thus shorter than in the present study.

Giacalone et al. reported on the long-term clinical follow-up of 104 patients in two cohorts: 26 patients received an immediate reconstruction using a LDF and prosthesis (either a tissue expander or a permanent implant) after neoadjuvant chemotherapy and 78 a delayed reconstruction after chemo- and radiation therapy [21]. Early implant loss occurred more frequently in the delayed reconstruction setting (12% vs. 0%), whereas late complications were more common after immediate reconstruction (30% vs. 21%). Capsular contracture rates were similar in both cohorts (15.3% after immediate and 11.5% after delayed reconstruction, respectively). The prevalence of capsular contracture increases with time [24].

In the present work no arm and shoulder-related problems were documented after a median follow-up of 38 months. However, data on shoulder impairment were not prospectively collected and no measurements of shoulder strength performed. Therefore, underreporting cannot be excluded. Altogether, data on arm and shoulder morbidity after LDF surgery vary across studies [11, 24, 25]. According to a meta-analysis of 26 studies, there is a tendency that the LDF transfer may affect shoulder function, but this limitation seems to be minimal, and few patients experience a major impact on shoulder function [26].

Regarding the surgical technique, two aspects need to be considered. First, it has been shown that transection of the tendinous insertion of the latissimus dorsi muscle on the humerus may improve aesthetic results [27]. Specifically, it helps to avoid the development of the displeasing bulge that patients often refer to as “carrying a book” under the armpit and may increase the mobility of the flap [27]. In the present study, the humeral insertion of the muscle was transected in all cases. Second, transecting the thoracodorsal nerve has been controversially discussed in the literature [12, 28,29,30,31]. In the present study, a simple transection of the thoracodorsal nerve was performed in all patients, which reflects the standard approach followed by most surgeons in Germany [27, 28, 32]. While leaving the nerve intact may lead to unintentional muscle twitching and breast animation, sometimes referred to as “jumping breast” phenomenon, some authors suggested that cutting the nerve may result in a decrease of flap volume and proposed an individual approach depending on the setting and the complexity of nerve identification [29, 30]. Interestingly, several studies reported that reinnervation may occur after simple nerve transection and that reinnervation rates may be reduced when larger nerve segments are resected [30].

The present study shows a slow decrease in the use of LDF-based breast reconstruction. A similar trend is currently being observed world-wide. Leff et al. examined immediate breast recontructions performed in the United Kingdom [33]. After an initial increase in the use of the technique observed between 1996 and 2008, the number of procedures decreased rapidly. At the same time, the use of other techniques of autologous tissue transfer increased from 0.44% in 1996 to 2.76% in 2012. A similar trend was noted in the United States as well [34,35,36]. Various reasons have been proposed to explain the decreased use of LDF procedures. First, the relatively low volume of the harvested flap makes a simultaneous use of a breast prosthesis necessary in most cases, which makes the technique unsuitable for patients opting for an implant-free reconstruction. Second, a worldwide trend towards bilateral mastectomy is observed [37, 38]. While LDF-based reconstruction of both breasts is possible, arm and shoulder morbidity after a bilateral procedure remains a concern. Finally, the improved access to abdominal flap transfer with microvascular anastomosis at high-volume centers makes pedicled flaps less attractive.

Conclusions

The present study confirms latissimus dorsi flap to be a safe and robust method for reconstruction of the breast and skin replacement of chest wall. As in previous studies, the leading early complication was a seroma, while serious complications such as total flap loss were a rare occurrence. Future studies should focus on the long-term evaluation of arm and shoulder morbidity as well as patient satisfaction with the aesthetic outcome.