A total of 35 patients (24 men, 11 women) with a median age of 42 years (range, 27-68) were enrolled in our research. These 35 pathologically confirmed malignant soft tissue tumors included 17 dermatofibrosarcoma protuberans (DFSP), 8 SCCs, 7 liposarcomas, and 3 rhabdomyosarcomas. Histologic grades 1, 2, and 3 accounted for 34.3%, 48.6%, and 17.1%, respectively. 14 (40.0%) of 35 tumors located in the posterior trunk, 9 (25.7%) of 35 tumors located in the anterior trunks, 8 (22.9%) of 35 tumors located in the lower extremities, and 4 (11.4%) of 35 tumors located in the upper extremities. All patients underwent tumor en bloc resection and PPPFs to cover the wound defect, followed by either primary closure (91.4%) or skin grafting to the donor site (8.6%). The patient demographics are summarized in table 2.
Tumor size measured on 3DR models on X, Y, and Z reference lines were compared to the relative territory in the pathologic specimen. The Bland-Altman plot for each comparison demonstrated relatively high concordance rates within the 95 % limits of agreement (Fig. 2).
The characteristics of the PPPFs for these 35 patients are delineated in table 3. In our research, the 3DR models clearly displayed the main source arteries, which contained 26 lumbar arteries, 10 superior gluteal arteries, 12 posterior intercostal arteries, 3 anterior intercostal arteries, 5 thoracodorsal arteries, 4 thoracoacromial arteries, 2 lateral thoracic arteries, 2 superficial cervical arteries, 6 dorsal scapular arteries, 1 lateral sacral artery, 2 internal thoracic arteries, 3 musculophrenic arteries, 10 inferior epigastric arteries, 4 superior epigastric arteries, 2 superficial epigastric arteries, 2 the descending branch of lateral circumflex femoral arteries (LCFA), and 1 the ascending branch of LCFA, 1 medial circumflex femoral artery (MCFA), 2 profunda femoral arteries, 4 superficial femoral arteries, 2 peroneal arteries, 1 posterior tibial artery, 1 anterior tibial artery, 3 profunda brachii arteries, 2 radical collateral arteries, 3 brachial arteries, 1 superior ulnar collateral artery, 1 radical artery and 1 ulnar artery. An average of 3 perforators (range: 1, 7) with a mean diameter of 0.32 cm (range: 0.18, 0.74 cm) were present in our current study. The average distance between the tumor boundary and the exact location of perforator piercing from the deep fascia was 2.2 cm (range: 1.2, 7.7 cm). The average length of the artery perforator coursing along the subcutaneous tissue was 5.8 cm (range: 3.3, 25.1 cm). These values were obtained from the 3DR without intraoperative identification. The average flap size was 92.2 cm2 (range: 32, 126 cm2). The mean flap harvest time was 55 minutes (range: 36, 97 min). All preferred perforators were verified during the PPPF elevations without false-positive findings, hence, we concluded that there was a 100 percent positive predictive value in terms of perforator vessels evaluation by 3DR based on thin-slice MRI sequence.
2 patients presented with postoperative complications (5.7%). One patient appeared venous congestion. The flap healed uneventfully after local subdermal heparinization. The distal partial necrosis occurred in the other flap due to mixed vascular issues. After conservative treatment, the patient underwent secondary operation with debridement and a split-thickness skin graft (2.9%).
A 47-year-old man was diagnosed with DFPS in the left lateral lumbar region (Fig. 3). MRI showed that the tumor had invaded the muscle. The 3DR showed a 5.8´4.1´3.4 cm tumor with muscle infiltration and four perforators adjacent to the tumor originating from the lumbar arteries. The diameters of these perforators were as follows: 0.28, 0.36, 0.41 and 0.44 cm. The distances between tumor boundaries and piercing sites were as follows: 3.5, 6.0, 9.2 and 9.2 cm. The PPPF was planned based on the perforator with the diameter of 0.28 cm and the distance of 3.3cm. The skin incision 2 cm distance from tumor boundary and the projection of four perforators were outlined over the skin according to the surgical guide templates preoperatively. The patient underwent tumor wide excision with an 11´8 cm wound defect. During surgery, these two left perforators were identified. After ligating the lower perforator, an 18´7 cm flap was rotated 130 degrees to resurface the defect. The donor site was closed by primary closure. The flap survived without postoperative complications.
A 63-year-old man was admitted with sSCC in the left posterior thigh (Fig. 4). MRI showed that the tumor spread extensively along the subdermal tissue and invaded the gluteal muscle. 3DR visualization showed that the tumor with the size of 15.4´15.9´13.3 cm infiltrated the underlying muscles with the extent of 9.8´8.7´4.3 cm. There were three main branches distributing around the left femur according to the 3DR. The lower perforators originating from the profunda femoral artery were chosen to supply the PPPF. The patient underwent tumor wide excision with a 20´24 cm wound defect. A PPPF with the size of 18´14 cm was rotated 160 degrees to cover the defect. The donor site was closed by free skin graft. The flap survived without postoperative complications.
A 34-year-old man was diagnosed with DFSP located in the right lumbosacral region (Fig. 5). The MRI showed that the tumor adhered to the underlying muscle and perforators around the tumor were widespread. The 3DR showed a 4.5´4.5´2.2 cm the tumor overlay the thoracolumbar fascia. The preferable perforator originated from the lateral sacral artery piercing the second anterior and posterior sacral foramina, with the diameter of 0.24 cm and the distance of 5.3 cm from the tumor margin. The patient underwent tumor wide excision with a 14´8cm wound defect. A PPPF with the size of 20´6.5 cm was rotated 160 degrees to cover the defect. The donor site was closed directly. The flap survived uneventfully.