Journal of Surgical Oncology 46: 190-197 (1991)
Wound Healing Complications in Soft Tissue Sarcoma Management: Comparison of Three Treatment Protocols ROBERT S. BELL, MD, FRCS(C),IAMES MAHONEY, ,LID, FRCS(C), BRIAN O’SULLIVAN, MD, FRCP(C), CAN NGUYEN, MD, F R C S ( C ) , FRED LANCER, MD, FRCS(C), BERNARD CUMMINGS, MD, FRCP(C), CHARLES CATTON, MD, FRCP(C), ANDRE1 CZITROM, MD, FRCS(C), AND VICTOR L. FORNASIER, MD, F R C P ( C ) From the University Musculoskeletal Oncology Unit, Mount Sinai Hospital (R.S.B., F.I., A.C.), Departments of Radiation Oncology (B.O., B.C., C.C.), and Pathology (V.I.F.), the Princess Margaret Hospital, Ontario Cancer Institute, and Divisions of Plastic Surgery (1.M.) and Orthopaedic Surgery (C.N.), St. Michael’s Hospital, Toronto, Ontario
A prospective, nonrandomized comparison of three treatment protocols was undertaken in 45 patients with soft tissue sarcoma designated preoperatively as being at high risk of wound healing complications. All patients underwent complete resection of the gross tumour mass ( 5 with positive and 40 with negative microscopic margins). Fourteen patients received postoperative adjuvant irradiation (group I), 16 preoperative irradiation (group 11), and 15 preoperative irradiation and vascularized tissue transfer to the surgical bed after resection (group 111). Major wound healing complications (defined as complications requiring at least 1 further surgical procedure) were lower in group I11 patients (chi-square = 5.57, PI0 cm when prior incomplete surgery necessitates wide excision of the previous surgical field 3. Axial site of tumour origin in the axilla or inguinal regions 4. Lesions requiring composite resection of skin, muscle, and underlying bone
These criteria were validated in an interim analysis of data demonstrating that the initial patients with one or more of these attributes were indeed at high risk of wound breakdown [8]. All patients underwent preoperative clinical, radiographic, and pathological staging. The tissue diagnosis (from biopsy samples), and patient management were discussed at a weekly multidisciplinary tumour conference prior to initiating treatment. The patients were divided into 3 treatment groups in temporal fashion (Figs. 1, 2, 3). For the first 8 months of the study, no patients thought to be at risk of wound healing complications were treated with preoperative radiotherapy, although our patients in the greater study group did receive preoperative adjuvant irradiation. As
191
our experience with preoperative irradiation techniques increased, we extended our indications for preoperative treatment until, at 1 year into the study, most patients designated as being at risk of wound complications were entered into group I1 (preoperative irradiation followed by wide resection). Finally, during the last 18 months of the study, concern about wound complications led to a new treatment protocol using preoperative irradiation followed by wide resection and vascularized tissue transfer to the surgical bed (group 111). Irradiation treatment was performed using cobalt or linear accelerator equipment. Treatment plans were complex and individualized with frequent use of CT simulators. In general, preoperative irradiation fields were designed to provide a normal tissue margin of 2-5 cm surrounding the lesions. Preoperative planning always permitted sparing of a longitudinal strip of normal tissue in the treatment field to minimize late distal lymphedema. Patients treated using preoperative technique (groups I1 and 111), received SO Gy in 2.5 fractions over 5 weeks prior to surgery and a final postoperative boost of 16 Gy in 8 fractions when wound healing was completed [S]. Patients in group I (postoperative treatment), received 60-66 Gy using 2-Gy daily fractions, after healing their wounds. Irradiation treatment fields for patients in group I were planned to provide SO Gy to the entire surgical field and surrounding margin with a final boost to the site of tumour origin. All resections were performed by one surgical oncologist. The resection was planned using preoperative computed tomography (CT) and/or magnetic resonance imaging (MRI) scans to provide a margin of normal tissue measuring 1-2 cm, or a fascia1 plane surrounding all of the circumference of the tumour whenever feasible. If preoperative staging suggested that a major neurovascular structure was contaminated by microscopic tumour, the risk of local recurrence and the morbidity attendant with neurovascular sacrifice was discussed with the patient prior to deciding whether neurovascular resection would be undertaken. By contrast, if neurovascular structures were grossly involved by the tumour on preoperative staging studies, resection of the involved structure (with reconstruction of resected vessels if necessary), was invariably advised and undertaken. Pathological examination of the resected specimens was performed according to a strict protocol involving 1. Preoperative review of the diagnostic biopsy and radiographic studies by the surgeon and surgical pathologist 2. Examination of the resected specimen and the surgical wound in the operating room by the pathologist for purposes of orientation 3. Painting of resection margins using silver nitrate
192
Bell et al.
Fig. 1. A high-grade synovial sarcoma of the posterior shoulder was resected and reconstructed using the Tikhoff-Lindberg technique without preoperative irradiation (group I). Wound necrosis necessitated 2 further procedures before primary healing was achieved. The tumour measured 22 cm in greatest diameter.
Fig. 2 . A high-grade malignant fibrous histiocytoma of the medial thigh was resected following 50 Gy preoperative irradiation treatment (group IT). Wound dehiscence necessitated 2 surgical debridements, and eventual split thickness skin grafting to achieve wound healing.
Tumour grade was defined as high or low based on 4. Examination of the serially sectioned specimen by the surgeon and pathologist in the surgical pathol- Enneking’s staging criteria [ 111. All designations of ogy suite (The closest margins to the tumour were tumour grade were performed by a single experienced recorded. If the margin of resection was I5 cm: 7 Prior Incomplete Resection: 4 Location: 3 Composite resection: 1 8
7 9
5 9 1 5.2 hr
0 4.9 hr 2
0
High grade: 15 Low grade: 1 0
High grade: 14 Low grade: 1 1
7
6
9 7 0 7
6 5 3
16 4
19 5
I 5
31
42
15
In group 111, vascularized tissue transfer was provided using rectus abdominis myoplasty or myocutaneous flaps in 8 patients, gastrocnemius rotationplasties in 3, rectus femoris rotations in 2, and microvascular flaps in 2 patients. In group I, 8 of 14 patients (about 50%), suffered major complications necessitating 16 secondary operative procedures. In this treatment group, 4 patients required eventual skin grafting for major wound complications, 3 required vascularized tissue transfer, and 1 healed by secondary intention. For group 11, 7 of 16 patients (about 50%), were returned to the operating room for a total or 19 secondary procedures. In group 11, 3 patients required eventual skin grafting, 2 required secondary vascularized tissue transfer, and 2 healed by secondary intention. By contrast, only 1 patient in group 111 required 1 secondary operative procedure. This patient, treated for a massive proximal thigh sarcoma by resection and myocutaneous rectus abdominis transfer, suffered partial necrosis of the distal skin paddle, requiring skin debridement and immediate split-thickness skin grafting. This wound then healed without further problems. Excluding patients with major complications, 4 of 6
1
patients in group I, 5 of 9 patients in group 11, and 5 of 14 patients in group 111 suffered minor complications requiring dressings for more than 2 weeks after surgery. Using the chi-square test to evaluate the incidence of major complications in the 3 groups, significantly fewer complications were found in group I11 (x2 = 5.57, K 0 . 0 3 ) . Using analysis of variance to evaluate the mean number of secondary procedures in each group, group 111 was again significantly better (KO.01). There were no significant differences in the incidence of minor complications among the 3 groups. Mean postoperative hospital stay was 15 days for group 111, 31 days for group I, and 42 days for group 11. This difference in mean postoperative hospital days was statistically significant (P
Wound Healing Complications in Soft Tissue Sarcoma Management: Comparison of Three Treatment Protocols ROBERT S. BELL, MD, FRCS(C),IAMES MAHONEY, ,LID, FRCS(C), BRIAN O’SULLIVAN, MD, FRCP(C), CAN NGUYEN, MD, F R C S ( C ) , FRED LANCER, MD, FRCS(C), BERNARD CUMMINGS, MD, FRCP(C), CHARLES CATTON, MD, FRCP(C), ANDRE1 CZITROM, MD, FRCS(C), AND VICTOR L. FORNASIER, MD, F R C P ( C ) From the University Musculoskeletal Oncology Unit, Mount Sinai Hospital (R.S.B., F.I., A.C.), Departments of Radiation Oncology (B.O., B.C., C.C.), and Pathology (V.I.F.), the Princess Margaret Hospital, Ontario Cancer Institute, and Divisions of Plastic Surgery (1.M.) and Orthopaedic Surgery (C.N.), St. Michael’s Hospital, Toronto, Ontario
A prospective, nonrandomized comparison of three treatment protocols was undertaken in 45 patients with soft tissue sarcoma designated preoperatively as being at high risk of wound healing complications. All patients underwent complete resection of the gross tumour mass ( 5 with positive and 40 with negative microscopic margins). Fourteen patients received postoperative adjuvant irradiation (group I), 16 preoperative irradiation (group 11), and 15 preoperative irradiation and vascularized tissue transfer to the surgical bed after resection (group 111). Major wound healing complications (defined as complications requiring at least 1 further surgical procedure) were lower in group I11 patients (chi-square = 5.57, PI0 cm when prior incomplete surgery necessitates wide excision of the previous surgical field 3. Axial site of tumour origin in the axilla or inguinal regions 4. Lesions requiring composite resection of skin, muscle, and underlying bone
These criteria were validated in an interim analysis of data demonstrating that the initial patients with one or more of these attributes were indeed at high risk of wound breakdown [8]. All patients underwent preoperative clinical, radiographic, and pathological staging. The tissue diagnosis (from biopsy samples), and patient management were discussed at a weekly multidisciplinary tumour conference prior to initiating treatment. The patients were divided into 3 treatment groups in temporal fashion (Figs. 1, 2, 3). For the first 8 months of the study, no patients thought to be at risk of wound healing complications were treated with preoperative radiotherapy, although our patients in the greater study group did receive preoperative adjuvant irradiation. As
191
our experience with preoperative irradiation techniques increased, we extended our indications for preoperative treatment until, at 1 year into the study, most patients designated as being at risk of wound complications were entered into group I1 (preoperative irradiation followed by wide resection). Finally, during the last 18 months of the study, concern about wound complications led to a new treatment protocol using preoperative irradiation followed by wide resection and vascularized tissue transfer to the surgical bed (group 111). Irradiation treatment was performed using cobalt or linear accelerator equipment. Treatment plans were complex and individualized with frequent use of CT simulators. In general, preoperative irradiation fields were designed to provide a normal tissue margin of 2-5 cm surrounding the lesions. Preoperative planning always permitted sparing of a longitudinal strip of normal tissue in the treatment field to minimize late distal lymphedema. Patients treated using preoperative technique (groups I1 and 111), received SO Gy in 2.5 fractions over 5 weeks prior to surgery and a final postoperative boost of 16 Gy in 8 fractions when wound healing was completed [S]. Patients in group I (postoperative treatment), received 60-66 Gy using 2-Gy daily fractions, after healing their wounds. Irradiation treatment fields for patients in group I were planned to provide SO Gy to the entire surgical field and surrounding margin with a final boost to the site of tumour origin. All resections were performed by one surgical oncologist. The resection was planned using preoperative computed tomography (CT) and/or magnetic resonance imaging (MRI) scans to provide a margin of normal tissue measuring 1-2 cm, or a fascia1 plane surrounding all of the circumference of the tumour whenever feasible. If preoperative staging suggested that a major neurovascular structure was contaminated by microscopic tumour, the risk of local recurrence and the morbidity attendant with neurovascular sacrifice was discussed with the patient prior to deciding whether neurovascular resection would be undertaken. By contrast, if neurovascular structures were grossly involved by the tumour on preoperative staging studies, resection of the involved structure (with reconstruction of resected vessels if necessary), was invariably advised and undertaken. Pathological examination of the resected specimens was performed according to a strict protocol involving 1. Preoperative review of the diagnostic biopsy and radiographic studies by the surgeon and surgical pathologist 2. Examination of the resected specimen and the surgical wound in the operating room by the pathologist for purposes of orientation 3. Painting of resection margins using silver nitrate
192
Bell et al.
Fig. 1. A high-grade synovial sarcoma of the posterior shoulder was resected and reconstructed using the Tikhoff-Lindberg technique without preoperative irradiation (group I). Wound necrosis necessitated 2 further procedures before primary healing was achieved. The tumour measured 22 cm in greatest diameter.
Fig. 2 . A high-grade malignant fibrous histiocytoma of the medial thigh was resected following 50 Gy preoperative irradiation treatment (group IT). Wound dehiscence necessitated 2 surgical debridements, and eventual split thickness skin grafting to achieve wound healing.
Tumour grade was defined as high or low based on 4. Examination of the serially sectioned specimen by the surgeon and pathologist in the surgical pathol- Enneking’s staging criteria [ 111. All designations of ogy suite (The closest margins to the tumour were tumour grade were performed by a single experienced recorded. If the margin of resection was I5 cm: 7 Prior Incomplete Resection: 4 Location: 3 Composite resection: 1 8
7 9
5 9 1 5.2 hr
0 4.9 hr 2
0
High grade: 15 Low grade: 1 0
High grade: 14 Low grade: 1 1
7
6
9 7 0 7
6 5 3
16 4
19 5
I 5
31
42
15
In group 111, vascularized tissue transfer was provided using rectus abdominis myoplasty or myocutaneous flaps in 8 patients, gastrocnemius rotationplasties in 3, rectus femoris rotations in 2, and microvascular flaps in 2 patients. In group I, 8 of 14 patients (about 50%), suffered major complications necessitating 16 secondary operative procedures. In this treatment group, 4 patients required eventual skin grafting for major wound complications, 3 required vascularized tissue transfer, and 1 healed by secondary intention. For group 11, 7 of 16 patients (about 50%), were returned to the operating room for a total or 19 secondary procedures. In group 11, 3 patients required eventual skin grafting, 2 required secondary vascularized tissue transfer, and 2 healed by secondary intention. By contrast, only 1 patient in group 111 required 1 secondary operative procedure. This patient, treated for a massive proximal thigh sarcoma by resection and myocutaneous rectus abdominis transfer, suffered partial necrosis of the distal skin paddle, requiring skin debridement and immediate split-thickness skin grafting. This wound then healed without further problems. Excluding patients with major complications, 4 of 6
1
patients in group I, 5 of 9 patients in group 11, and 5 of 14 patients in group 111 suffered minor complications requiring dressings for more than 2 weeks after surgery. Using the chi-square test to evaluate the incidence of major complications in the 3 groups, significantly fewer complications were found in group I11 (x2 = 5.57, K 0 . 0 3 ) . Using analysis of variance to evaluate the mean number of secondary procedures in each group, group 111 was again significantly better (KO.01). There were no significant differences in the incidence of minor complications among the 3 groups. Mean postoperative hospital stay was 15 days for group 111, 31 days for group I, and 42 days for group 11. This difference in mean postoperative hospital days was statistically significant (P
