Reconstruction of the superficial femoral vessels with muscle flap coverage for soft tissue sarcomas of the proximal thigh

Abstract Background Surgical resection of soft tissue sarcoma with a margin of healthy tissue may necessitate resection and reconstruction of major blood vessels together with soft tissues of the proximal thigh to preserve the limb. The long‐term functional outcomes of these reconstructions remain unestablished. The aim of this report was to assess the vascular and functional outcomes of soft tissue sarcoma patients with femoral vessel reconstructions. Patients and Methods Patients who had undergone oncovascular reconstruction during the treatment of proximal thigh soft tissue sarcoma in 2014–2020 were reviewed for details of the vascular and soft tissue reconstructions, and the oncological and functional outcomes. This included eight patients of a median age 59 (range 19–77) years. All had a reconstruction of at least the superficial femoral artery and vein as well as soft tissue reconstruction with a muscle flap. All vessel reconstructions were done with either autologous vein (six grafts/four patients) or allograft (10 grafts/six patients). A microvascular latissimus dorsi flap, with a skin island, was incorporated to cover the vascular grafts in five patients. A pedicled sartorius or gracilis muscle flap was used to fill the defect in three patients. Results Graft patency was assessed in seven patients with a median follow‐up of 48 (1–76) months. The arterial graft was patent in 6/8 and the vein graft in 2/8 patients. The gait had returned to normal in five of the six patients assessed. The median MTSS was 70 (43–87)% and the TESS 90 (75–100)%. No local recurrence of the sarcoma was detected. Conclusions Vascular reconstruction combined with soft tissue reconstruction enables limb‐sparing surgery in patients with soft tissue sarcoma involving proximal femoral vessels. Although the surgeries are complex with high early morbidity, the achieved long‐term functional outcomes are worthwhile.

T A B L E 1 Patient demographics, treatment and follow-up details of lower limb soft tissue sarcoma patients treated with vascular and soft tissue reconstructions  Williard et al., 1992). In cases where the tumor immediately surrounds or invades major blood vessels, the radical excision often still involves sacrifice of the involved vessels. With tumors in the proximal thigh or inguinal region, this has traditionally required proximal amputation, hip disarticulation or hemipelvectomy. Emergence of oncovascular surgery has enabled reconstruction of the resected vessels and salvage of the involved limb (Bonardelli et al., 2000;Davis et al., 2017;Fortner et al., 1977;Fujiki et al., 2020;Ghert et al., 2005;Imparato et al., 1978;Karakousis et al., 1996;Okamoto et al., 2018;Poultsides et al., 2015;Schwarzbach et al., 2005;Song et al., 2009).
The safety and efficacy of vascular reconstruction in conjunction with sarcoma resection has been well established. However, the longterm graft survival and functional outcomes are sparsely described (Fujiki et al., 2020). In addition, the reported rates of in-hospital morbidity, wound complications and reoperations are high (Davis et al., 2017;Ghert et al., 2005;Okamoto et al., 2018;Poultsides et al., 2015;Schwarzbach et al., 2005). Incorporation of local or microvascular flaps during the primary operation to improve the soft tissue coverage of the grafts has can be postulated to reduce local complications as well as enable more extensive excisions but the reported cases remain few.
The aim of our report is to assess the early and long-term outcomes of patients with soft tissue sarcoma in the proximal thigh. In particular, we evaluate the early complications associated with the surgery, the long-term patency of the vascular reconstructions and the achieved functional outcome.

| PATIENTS AND METHODS
A retrospective analysis of prospectively collected data on patients operated on for deep inguinal or proximal thigh soft tissue sarcoma requiring major vessel reconstruction in conjunction with tumor resection between year 2014 and 2020 was performed. The patient records were reviewed for demographic details, histological diagnosis and grade, details of the surgery and perioperative period, adjuvant treatments, as well as details of any local recurrence or metastases.
Eight patients with soft tissue sarcoma were included (Table 1).
Six were male. The median age at operation was 59 (19-77) years.
Photos of the patient number 7 are shown in Figure 1.

| Surgical technique
A radical resection extending to a natural tissue barrier or at least 5 cm away from the tumor was done, including the affected vessels.
The superficial femoral artery and vein were reconstructed in all patients. The reconstruction included the external iliac vessels in three patients, one of whom had also the deep femoral vessels reconstructed (Table 1) The median operative time was 7.9 (4.3-9.1) h ( Table 2). The median blood loss during surgery was 1565 (92-5000) ml, and 1 (0-14) units of red blood cells were transfused during the day of the surgery. Five of the patients were treated in the high dependency unit, most for 1 day. The median hospital stay was 10 (8-54) days.
The thromboprophylaxis regime included perioperative intravenous heparin at a dose of 100 IU/kg and long-term acetylsalicylic acid 100 mg once daily. Low molecular weight heparin was used postoperatively for up to 3 months. A double platelet inhibition with the addition of clopidogrel was introduced for 3 months after balloon angioplasty for graft stenosis.
The patients were recalled for clinical evaluation at the discretion of the senior author, with a special focus on the long-term vessel patency and limb function. Duplex ultrasound was used to assess graft patency. The examination was done by an experienced vascular nurse specially trained in vascular graft surveillance.
The presence of limb edema was assessed at time of the last vascular follow-up by a physiotherapist trained in the procedure with the truncated cone method with serial circumferential measurements at 4 cm intervals (Brorson et al., 2015).    Assessment of graft patency with duplex ultrasound was available for seven patients with a follow up of 48 (2-76) months (Table 3) (Fujiki et al., 2020). The aim to achieve adequate disease control without above the knee amputation or hip disarticulation reflects the perceived functional advantage a reconstructed limb provides. Studies comparing the health-related quality of life in sarcoma patients treated with limb-sparing surgery and amputation have suggested that limb-salvage surgery is associated with less functional and psychological handicap (Davis et al., 1999(Davis et al., , 2017Malek et al., 2012;Mei et al., 2014;Reijers et al., 2022). However, no overall quality of life benefit has been consistently observed (Mei et al., 2014). In our view, preserving the ability to perform tasks of daily living independently is an important goal even in patients with limited long-term prognosis. Therefore, we do not consider high-grade tumors or metastatic disease as an indication to favor more mutilating options over limb-sparing surgery but emphasize case by case assessment.
The median MSTS score of the six patients assessed in our cohort was 70%, median 48 months after the operation. This is lower than previously reported in two small series for patients with lower limb sarcomas requiring oncovascular surgery (Davis et al., 2017;Emori et al., 2012). The median TESS score was 90%, slightly higher than reported elsewhere (Ghert et al., 2005). As the TESS is a self-reported instrument measuring the ability to perform tasks of daily living, it appears our patients did not feel restricted by the physical limitations Autologous vein grafts were used whenever possible, saphenous vein when available and cephalic vein as the second option. Allografts were used only when no suitable vein was available for grafting. This reflects our preference to avoid the use of synthetic material at iliac vessels and distally, as well as in areas where the likelihood of delayed wound healing or postoperative infection is perceived high. In addition, the long-term patency of venous grafts for arterial bypass in the femoral region supersedes that for PTFE grafts (Klinkert et al., 2004). A similar preference for autologous grafts has been reported by other centers (Baxter et al., 2007;Davis et al., 2017;Miyamoto et al., 2017;Nishinari et al., 2015;Okamoto et al., 2018;Poultsides et al., 2015;Umezawa et al., 2013). However, some surgeons routinely use PTFE graft as the first choice (Emori et al., 2012;Ghert et al., 2005;Schwarzbach et al., 2005). The use of allografts, although common in reconstruction of retroperitoneal vessels, appears rare (Homsy et al., 2020;Poultsides et al., 2015).
Duplex ultrasound assessment of the vascular graft patency was done for seven patients at median follow-up of 48 months. One of the patients had an obstructed arterial graft while another required repeat angioplasties. The obstruction rate was higher for the vein grafts with five of the vein grafts thrombosed during the follow-up and only two of the grafts unobstructed at 24 months of follow-up.
The arterial graft patency rates we observed were similar to, but the venous graft rates worse than, those reported elsewhere for sarcoma patients (Adelani et al., 2007;Fujiki et al., 2020;Nishinari et al., 2015;Okamoto et al., 2018;Poultsides et al., 2015). The risk factors for graft occlusion have not, to our knowledge, been studied in the context of sarcoma reconstruction. With arterial bypass grafts for arteriosclerosis, the graft patency is dependent on the distal vasculature, not properties of the graft (Albäck et al., 1998).
Notably, all resected femoral veins were reconstructed in our patients. This reflects our approach of reconstructing the superficial femoral vein, even when the deep femoral vein is left intact by the sarcoma resection. While most studies suggest a benefit from venous reconstruction at least in cases where two major veins of the lower limb have been resected and no collateral venous circulation is detected, some challenge the need for the reconstruction of femoral veins altogether (Adelani et al., 2007;Fujiki et al., 2020;Tsukushi et al., 2008). In our view, it is conceivable that the role of the vein reconstructions in reducing limb edema is most pertinent in the early postoperative period, while the later formation of new collaterals may mitigate the importance of the grafts on venous return. Thus, although graft occlusion is more frequently observed in the reconstructed veins than in the arteries, it seems better tolerated in the long run.
Edema is one of the major complications of proximal thigh resections in lower limb sarcoma patients. With major lymph vessels coursing along the femoral vessels and superficial vein, resection of these vascular structures severs also the lymphatic pathways. Thus, the edema results from disruption to both the lymphatics and the venous circulation and can be further exacerbated by adjuvant radiotherapy.
Methods proposed to enhance lymph flow recovery following tissue reconstruction have the placement of the flap in an orientation where pre-mapped lymphatic channels align axially with the lymphatic flow present at the defect site prior to resection, either with or without a lymphaticovenous anastomosis (Pereira et al., 2021;Scaglioni et al., 2021Scaglioni et al., , 2022Yamamoto et al., 2018Yamamoto et al., , 2021.  (Baxter et al., 2007;Ghert et al., 2005;Nishinari et al., 2015;Umezawa et al., 2013).
The perioperative morbidity associated with the oncovascular sarcoma surgery described here was high and the employed multidisciplinary resources significant. The median duration of the operation was almost 8 h and the blood loss during surgery 1565 ml.
While the perioperative mortality was zero, five of the patients required high dependency unit treatment postoperatively. Elsewhere, in-hospital mortality of up to 12.5% has been reported (Okamoto et al., 2018).
Significant perioperative morbidity was observed with six of the patients having a Clavien-Dindo class 3 complication within 30 days of surgery (Dindo et al., 2004). Three of the re-operations were done for surgical site infection, two of these in the context of an infected seroma. While none of the flaps were lost during follow-up, additional flap reconstructions to improve soft tissue coverage were done in three patients. Notably, only one patient required interventions for graft occlusion during the perioperative period.
Our early experiences and reports of prolonged lymphatic collection despite meticulous ligation of the lymphatic vessels during the surgery and the resulting delayed wound closure have shaped our practice to favor immediate flap reconstruction, with a pedicled or a microvascular flap (Davis et al., 2017;Ghert et al., 2005). In this series, the patients number 2 and 5 illustrated how a seroma formation around the vessel reconstructions led to a later requirement for a flap despite the soft tissue coverage of the grafts having been judged sufficient in the primary operation. In addition, no patients with soft tissue sarcoma resection and vascular reconstruction without the need for a soft tissue reconstruction were identified. These observations support the incorporation of a muscular flap in the primary operation to cover the vessel grafts and to fill the created cavity to prevent wound healing complications.
The use of flaps is our present practice unless the grafts can be covered directly with surrounding muscles. When no pedicled flap of sufficient volume is available in the thigh, or preoperative radiotherapy renders the local flap less reliable, musculocutaneous LD is our preferred choice. Although the LD could be harvested without a skin component, the additional volume created by the muscle flap would often render a direct closure of the wound edges too tight. Thus, we frequently include a skin island in the flap, harvested in an axial orientation on the anterior border of the muscle, where a skin island of at least 9 cm wide can easily be harvested without difficulty in primary closure of the wound. The skin island also enables easy clinical monitoring of the flap perfusion. The arterial inflow to the flap is frequently connected to the arterial graft while a native vein is preferred for the venous outflow to minimize the circulatory compromise an obstructed venous graft would inflict on the microvascular flap. The microvascular surgery involved is relatively simple as the vessel diameter is large and the anastomosis site is well-exposed.
Our report is limited by the small number of patients and the short follow-up for some of them. In addition, no postoperative functional outcomes assessment was available for two of the eight patients, further limiting the population. The absence of a control group of either patients treated with amputation or those that did not require vascular reconstructions as part of their sarcoma removal precludes any comparative analysis.

| CONCLUSIONS
In conclusion, oncovascular surgery combined with soft tissue reconstruction through local or microvascular flaps enables limb-sparing surgery in patients with soft tissue sarcoma necessitating resection of major blood vessels in the inguinal region or the proximal thigh.
Despite the high early morbidity, the long-term functional outcomes are generally good.

FUNDING INFORMATION
This work was supported by funding from Helsinki University Musculoskeletal and Plastic Surgery Research Centre.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.