Separated-incision versus T-incision for internal hemipelvectomy related to Enneking type II + III resection: comparative outcomes in a single-center retrospective cohort

Abstract

Background

The T-incision approach for internal hemipelvectomy necessitates extensive dissection to expose the posterior pelvic structures, leading to higher rates of wound complications. A modified separated-incision approach was developed and validated in this comparative study.

Patients and methods

The separated-incision approach used two distinct incisions: an anterior incision that combines the ilioinguinal approach with a short, straight Smith-Petersen incision; and a separated posterior incision for the posterior pelvic structures. This approach was applied to 8 patients with internal hemipelvectomy related to type II + III regions. Seventeen patients underwent a traditional T-incision approach during the same period were matched for comparison.

Results

The separated-incision approach was successfully performed in these eight patients, preserving the gluteal vessels and posterior skin-muscle flap. Lower blood loss (2375 vs. 3952 mL, p = 0.005), and similar operative time (312 vs. 398 min, p = 0.098) was observed. Postoperative haemoglobin was higher (88.1 vs. 74.8 g/L, p = 0.009), and drainage volumes were marginally reduced (1379 vs. 1917 mL, p = 0.209). Compared to 8 patients in the T-incision group who experienced wound complications including delayed wound healing (n = 4) and surgical site infection (n = 4), all patients in separated-incision group achieved primary wound healing within 3 weeks (47% vs. 0%, p = 0.026). Additionally, they had a shorter time to ambulation (39 vs. 62 days, p = 0.015) and higher MSTS scores (24.25 vs. 20.47, p = 0.006). No differences were observed in local recurrence or overall survival.

Conclusions

The separated-incision demonstrated fewer wound complications and faster mobilization, suggesting it is a promising alternative; multicentre validation is warranted.

Level of evidence

Level III, Retrospective cohort study.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-025-08944-x.

Introduction

Pelvic malignant bone tumours have inferior prognoses due to delayed diagnoses and significantly challenges in surgical management, especially for those related to Enneking type II + III regions [1–5]. They were initially managed via hindquarter amputation, at cost of severe handicap [6]. Encouragingly, internal hemipelvectomy with limb-sparing procedures - primarily in the form of en-bloc tumour resection and 3D-printed prosthetic reconstruction - emerged as optional surgical alternatives for pelvic bone tumours [7]. However, significant challenges exist in surgical exposure, bleeding, postoperative complications, and functional reconstruction [3, 4, 8].

A clear and accessible surgical approach that minimizes injury would be beneficial in addressing these challenges. Since 1970 s, the utilitarian pelvic incision and its modifications have become the most common surgical approaches [9, 10]. However, when the tumours invade a deeper area, such as the posterior structures of the acetabulum, the view of the operation field may be limited [11]. In order to gain a clear access, an auxiliary incision from the anterior superior iliac spine to greater trochanter (extended Smith-Petersen approach) was often needed, which was described as T-incision [10, 12]. Unfortunately, the routine T-incision approach required extensive blunt dissection of the iliacus to expose iliac fossa. It might lead to injury to the intrapelvic neurovascular bundles and subsequent ischemic necrosis of muscle flap. Additionally, limb functional impairment might be associated with the sacrifice of the gluteus, iliacus. Thus, a high incidence of wound complications (nearly 20–50%) and delayed recovery happened [2, 3, 8, 13].

To minimize the wound complications of T-incision approach, much attention has recently been applied, such as aortic balloon occlusion technique and negative-pressure wound therapy [14, 15]. Meanwhile, modifications of surgical approach are also considered [11, 16]. Considering the limited accessibility and observation from the sole-incision approach, separated approaches has been tried and analysed in several studies [11, 16]. At the same time, we designed and applied a modified separated-incision approach for internal hemipelvectomy since 2016 to optimize the traditional T-incision approach. The separated-incision consisted of an ilioinguinal approach with a short straight Smith-Petersen incision (a minimized T-incision) and a separated posterior incision. It avoided a direct incision from the groin to the gluteal fold, while utilizing three distinct surgical windows to achieve adequately exposure to the posterior structures of the acetabulum.

In this study, the modified separated-incision approach is compared with the traditional T-incision approach in terms of surgical convenience and accuracy, functional recovery status, and short-term complications, aiming to identify the superior surgical method. We hypothesized that separated-incision approach may benefit patients have malignant tumours related to II + III regions in terms of (1) reducing intraoperative injuries, (2) lowering the incidence of wound complications, and (3) promoting rapid rehabilitation.

Methods

Patients

Between January 2010 and December 2022, 87 patients with histologically periacetabular tumours were treated with different surgeries at our orthopaedic oncology institution since 2010. They were treated by hindquarter amputation, Harrington reconstruction techniques, robot-assisted tripod percutaneous reconstruction technique [17], or internal hemipelvectomy with or without reconstruction. Before 2016, we used the traditional T-shaped incision for internal hemipelvectomy; afterward, we gradually transitioned to using the new segmented incision, which is chosen based on the surgeon’s experience.

In this single-centre retrospective cohort, the inclusion criteria were: (1) patients having a periacetabular lesion that required a type II/III or I + II + III internal hemipelvectomy; (2) performed with T-incision or separated-incision approach; (3) had no bony reconstruction or the reconstruction type custom prosthesis replacement; (4) diagnosed by the pathology examination; (5) complete imaging data, and follow-up information were available. The exclusion criteria were: (1) patients died of other diseases within 3 months; (2) patients received pelvic radiotherapy before surgery; (3) active infections within one month prior to surgery; (4) previous severe fractures, joint diseases, deformities, etc., of the lower limb on the surgical side.

Informed consent was obtained from all patients, and approved by institutional review board.

Preoperative Preparation

The preoperative preparation consisted of routine laboratory tests, X-rays, bone scans, and magnetic resonance (MR) and computed tomography (CT) scans of the pelvis. Neoadjuvant chemotherapy was given to high-grade osteosarcoma, Ewing’s sarcoma, and undifferentiated mesenchymal sarcoma. All patients were at stage IIB or III (T2) according to the Enneking staging system [9].

The resection and reconstruction were performed following the operating instructions provided by Enneking et al. [9] and Malawer et al. [18]. The bony margins and soft tissue margins were defined based on Enneking’s classification, while the detailed widths of the bone and soft tissue margins were guided by Kawaguchi et al. [19]. While surgeons recommended implant reconstruction for the patient, some patients opted for pelvic resection followed by a period without reconstruction due to financial constraints.

Tranexamic acid was routinely administered intravenously 30 min prior to surgery at a dosage of 1 g, with the option to repeat the dose of 1 g every three hours. The use of autologous blood transfusion was prohibited, as it might have promoted the metastasis of these malignant tumors.

Surgical technique: the T-incision and the Stepwise separated-incision

A traditional T-incision (Fig. S1) was done for most patients, as described by Lackman [10]. Before 2016, we used the traditional T-shaped incision; afterward, we gradually transitioned to using the new segmented incision, which is chosen based on the surgeon’s experience.

The stepwise separated-incision approach was preferred to perform in patients who had large tumours invaded the posterior acetabular and the ischial tuberosity. The separated-incision approach consisted of an ilioinguinal approach with a short straight Smith-Petersen incision (a minimized T-incision), combining a separated posterior incision (Figs. 1 and 2). It avoids a continuous incision from the groin to the gluteal fold and instead utilizes three separate surgical windows.

Fig. 1.

Schematic depiction of the separated-incision approach. A An ilioinguinal incision and a short and straight Smith-Petersen incision. B There is a wide gap between the anterior incision and posterior incision to allow well-protection of muscle flap. C The posterior incision. d In the anterior incision, femoral neurovascular bundles could be protected with psaos muscle. e In the superior part of ilioinguinal incision, glutes minus, medius and part of glutes maximus are resected from iliac crest to exposure the posterior aspect of acetabular column and greater sciatic notch, the acetabular osteotomy can be performed with careful protection of gluteal vessels. f In the posterior incision, a direct visualization to ischial tubercle and its attached muscles is helpful for en bloc tumor invading ischial rami or the obturator, with a lower risk of sciatic nerve injury. (green part is marked as tumor)

Fig. 2.

A separated incision for resection of periacetabular chondrosarcoma. A-B, pre-surgery design of surgical approaches; C, implantation of 3D-printed prosthesis; D-F the patient achieved primary wound healing within 2 weeks after surgery

The medial window was created by the medial part of ilioinguinal approach, it was used to expose the pubic symphysis and the obturator, facilitating pubic symphysis osteotomy and tumour dissection.

The lateral window was created by an anterior incision consisting of upper section of ilioinguinal approach and a short part of Smith-Petersen approach. As the lateral window, it provided access to the iliac crest, the anterior wall of acetabulum, the entire internal iliac fossa, and the most lateral sacral ala. Further exposure incorporated full visualization of the anterior aspect of the sacroiliac joint (SI), and completes exposure of the whole hemipelvis. The skin, femoral neurovascular bundle and iliopsoas muscle could be elevated as a bridge between the lateral window and the medial window.

The posterior window (the posterior incision) was centred around the greater tuberosity, extending proximally posteriorly and distally posteriorly along the gluteal crease, and turned to ischial tuberosity. Detachment of gluteus from crest could be performed mainly from the upper section of anterior incision, and partly from posterior incision, which provided full exposure to the dorsal side of greater sciatic notch and visualization to gluteal vessels (superior and inferior). When protecting the sciatic nerve, the posterior column and wall (the ischial spine, and the lesser sciatic notch) could be fully exposed. Adjacent structures such as sacro-tuberous ligament, sacrospinous ligament, and hip external rotator abductors could be detached from ischial tuberosity and inferior pubic ramus. Further exposure of the posterior hip joint, could also conveniently provides clear view for femoral neck osteotomy.

After completing the exposure and tumour resection for each incision, thorough haemostasis was performed, followed by gauze packing. Finally, the tumours could be en bloc resected after the superior osteotomy through the greater sciatic notch, the femoral neck osteotomy, and the medial osteotomy through pubic symphysis. Then, the tumours could be dissected from the lateral window under abdominal aortic balloon occlusion.

Details of this approach was described in Supplement Materials.

Implantation of prostheses

As described in previous article, the preoperative planning of the resection, design of prosthesis was predetermined with the software Mimics version 17.0 (Materialise, Leuven, Belgium) [20] (Fig. S2).

The hemipelvic prosthesis was from Shengshi Medical Instruments Co., Ltd., Shanghai, P. R. China was adopted. The bushing part of the metal acetabular component was inserted in the iliac component and fixed. Then, the polyethylene cup bearing an anti-dislocation shoulder was augmented to the metal acetabular component with gentamycin-embedded cement, and the extroversion and anteversion of the cup were adjusted to an angle of 45° during this procedure. A femoral component was implanted, and reduction of the hip joint was made (Fig.S2, Fig.S3).

Reconstruction of the soft tissue

During the reconstruction process, preserving the gluteus maximus and psoas major muscles was crucial for maintaining stability in the hip joint. The external rotator muscles in posterior window were carefully sutured and fixed to the sacroiliac ligament, erector spinae, and other posterior muscles. Similarly, the adductor muscles in anterior window were meticulously sutured and secured to the transversus abdominis and other muscles. To further enhance the stability and reduce the risk of infection, internal sutures were pre-embedded in gentamycin-embedded cement and passed through the pre-designed holes of the acetabular cup to secure the mesh patch, and ultimately used to suture the remaining gluteus medius, iliopsoas, etc. (Fig. S3)

For those patients received no metal reconstructions, 2–3 suture anchors were inserted to iliac, and they helped to suture and fixation the remaining gluteus medius, iliopsoas, etc.

Postoperative management

Prophylactic intravenous (IV) antibiotics (1,500 mg of IV cefazolin sodium, administered every three hours) were administered 30 min prior to the procedure and continued for 72 h postoperatively or until the drainage tube was removed. Vacuum-sealing drainage tubes were removed when output flow rates measured 50 mL over 24 h [15]. Anti-thrombotic stockings and calf compression devices were routinely used for prophylaxis against deep vein thrombosis. Patients were restricted to bed rest for 6–8 weeks. Skin edge necrosis was identified early and operative debridement and re-suturing were undertaken where indicated. Increasing weight bearing was started after the bed rest period. To prevent dislocation of the hip, range of motion was limited in external rotation and flexion > 90° by hip brace for 4 weeks.

Follow-up

Patients had regularly scheduled follow-up, usually four times each year for the first 2 years, every 6 months from the third to fifth year, and yearly thereafter. The 1993 MSTS lower extremity score was recorded. Additionally, imaging examinations including X-ray, CT and MRI for primary site, and bone scan or PET-CT for whole body were also performed.

Statistical analysis

Survival analysis based on Kaplan-Meier survival curves was performed from the date of surgery to the time of death or last follow-up date. The chi-square test was used to compare two variables with discrete outcomes, Student’s test or nonparametric test were used to compare unrelated samples when appropriate. A probability value (p) of less than 0.05 was considered to be significant. Statistical analysis was performed using the SPSS software version 20.0 (SPSS Inc, Chicago, IL, USA).

Results

Patient characteristics and oncological outcomes

A total of 25 patients were enrolled in the study, with the separated-incision approach successfully performed in 8 cases. There was no significant difference between these two groups regarding to gender, age, BMI, pathological diagnosis, Enneking classification, recurrent diseases, as shown in Table 1 and summarized in Table S1. Most patients in the traditional group (15/17) and all patients in the separated-incision group (8/8) had tumors invading the posterior acetabulum, with 6 of 8 patients in the separated-incision group also having ischial tuberosity invasion, compared to 8 of 17 in the traditional group. The tumor volume in the separated-incision group was 346 mL, which was equal to that in the traditional incision group (359 mL, p = 0.874). No significant difference existed in these two groups regarding to neoadjuvant therapy, including chemotherapy or arterial chemoembolization.

Table 1.

Patients characteristics

Patient No.	Age	Gender	BMI	Pathologic diagnosis	Enneking staging	Tumor location	Posterior acetabular involvement	Ischial tuberosity involvement	Tumor size (cm)	New adjuvant therapy	Follow-up (Months)	Status
1	42	F	22	Chondrosarcoma	IIB	I + II + III	Yes	no	9*8*6	no	170	NED
2	39	F	21	Chondrosarcoma	IIB	I + II + III	Yes	Yes	12*9*8	no	146	NED
3	38	F	24	Malignant GCTB	IIB	I + II + III	Yes	Yes	12*8*7	no	124	NED
4	44	F	28	Recurrent Chondrosarcoma	IIB	II + III	Yes	no	8*7*6	no	181	NED
5	43	M	21	Recurrent Chondrosarcoma	IIB	I + II + III	Yes	Yes	11*8*6	no	78	NED
6	20	F	19	Ewing Sarcoma	IIB	I + II + III	Yes	no	17*6*6	Chemotherapy	177	NED
7	42	M	23	Recurrent Chondrosarcoma	IIIB	II + III	Yes	Yes	11*8*7	Arterial embolism	48	DOD
8	23	F	19	Ewing sarcoma	IIIB	I + II + III	Yes	no	9*8*5	Chemotherapy	42	DOD
9	47	M	23	Recurrent Chondrosarcoma	IIB	II + III	Yes	no	14*11*7	no	150	NED
10	48	M	22	Metastatic Thyroid cancer	IIIB	II + III	no	no	8*6*5	Arterial embolism	25	DOD
11	56	F	24	Metastatic Breast cancer	IIIB	I + II + III	Yes	no	7*8*6	Chemotherapy	46	DOD
12	54	F	28	Recurrent Chondrosarcoma	IIIB	I + II + III	Yes	Yes	14*11*7	Arterial embolism	26	AWD
13	70	F	18	Metastatic Chondrosarcoma	IIIB	II + III	no	no	10*8*6	no	29	DOD
14	69	M	21	Fibrosarcoma	IIIB	I + II + III	Yes	Yes	16*12*8	no	4	DOD
15	33	M	24	Chondrosarcoma	IIB	II + III	Yes	no	10*7*7	0	170	NED
16	33	F	17	Recurrent MSFT	IIIB	I + II + III	Yes	Yes	13*12*10	no	31	DOD
17	55	F	25	Metastatic Leiomyosarcoma	IIIB	II + III	Yes	Yes	11*8*6	no	34	DOD
18	47	F	26	Leiomyosarcoma	IIB	I + II + III	Yes	Yes	14*10*8	Arterial embolism	52	NED
19	37	F	23	Malignant GCTB	IIB	II + III	Yes	Yes	13*9*6	Denosumab	44	NED
20	56	M	21	Chondrosarcoma	IIB	II + III	Yes	Yes	14*8*6	no	36	NED
21	56	M	27	Osteosarcoma	IIB	II + III	Yes	Yes	12*7*7	Chemotherapy	35	AWD
22	46	M	21	Chondrosarcoma	IIB	II + III	Yes	Yes	12*8*6	no	42	NED
23	47	M	29	Osteosarcoma	IIIB	II + III	Yes	No	9*8*8	Chemotherapy	38	DOD
24	50	M	26	Chondrosarcoma	IIB	I + II + III	Yes	Yes	13*8*6	no	110	NED
25	37	F	26	Malignant GCTB	IIB	II + III	Yes	no	9*7*6	no	100	NED

AWD alive with disease, DOD died of disease, GCTB giant cell tumor of bone, MSFT malignant solitary fibrous tumor, NED no evidence of disease

After a median duration of follow-up of 46 months (range: 4 to 181 months), 1 patient experienced local recurrence in the traditional group, and none in the separated-incision group. The 5-year overall survival in the whole group was 56.5 ± 4.4%, 50.4 ± 5.9% in the traditional group, and 75.0 ± 15.0% in the separated-incision group, and there was no significant difference (p = 0.278) (Fig. S4).

Intraoperative injuries

Most patients in the traditional incision group (15/17) underwent prosthesis reconstruction, compared to 5 patients (5/8) in the separated-incision group. The separated-incision group exhibited significantly less blood loss (2,375 ± 680 ml vs. 3,952 ± 910 ml, p = 0.039) and a similar operative time (312 ± 39 min vs. 398 ± 64 min, p = 0.098) than did the traditional group. Twelve patients (12/17) in the traditional group experienced intraoperative abdominal aortic balloon embolism, and the average time was 56 ± 19 min. All of the patients in separated-incision group underwent intraoperative abdominal aortic balloon embolism, and the average time was 59 ± 11 min; there was no significant difference compared to the traditional group.

The total volume of packed red blood cells that was transfused during and after the operation was 1,606 ± 474 mL in the separated-incision group, which was lower than that in the traditional group (2,094 ± 448 mL, p = 0.204). Postoperative hemoglobin levels were greater in the separated-incision group (88.1 ± 6.8 g/L vs. 74.8 ± 5.4 g/L, p = 0.009), and drainage volumes were marginally reduced (1379 ± 315 mL vs. 1917 ± 534 mL, p = 0.209) (Table 2).

Table 2.

Surgery details and outcomes

Patients No.	Approach	Reconstruction with prosthesis	Operation durations (Min)	Blood loss (mL)	Hb after surgery (g/L)	Drainage volume (mL)	Major complications	Second surgery	Time to ambulate (days)	MSTS score
1	T-incision	Yes	420	3800	78	1860	no	no	42	23
2	T-incision	Yes	480	4200	68	2050	no	no	45	24
3	T-incision	Yes	660	5600	65	3800	DWH DL Superficial infections Implant breakage	Debridement and open reduction	75	17
4	T-incision	Yes	540	4500	64	3100	DWH Deep infections	Hindquarter amputation	48	14
5	T-incision	Yes	660	6000	65	4600	DWH DL Deep infections	Implant removal	87	13
6	T-incision	Yes	340	4800	89	2080	no	no	45	22
7	T-incision	Yes	320	3000	78	1560	no	no	47	24
8	T-incision	Yes	470	4500	81	1120	DWH	Debridement	85	21
9	T-incision	Yes	330	7000	65	1965	no	no	48	24
10	T-incision	Yes	320	6000	86	870	no	no	42	24
11	T-incision	Yes	440	3000	82	560	no	no	56	21
12	T-incision	Yes	460	6500	64	2340	DWH Deep infection	Implant removal	86	16
13	T-incision	Yes	210	1600	63	1245	DWH	Debridement	67	18
14	T-incision	Yes	290	1400	84	2305	DWH		120	15
15	T-incision	No	230	600	102	750			42	27
16	T-incision	Yes	270	2500	70	465	Local recurrence	Hindquarter amputation	42	15
17	T-incision	No	340	2200	68	1915	DWH	Debridement	78	21
18	Separated-incision	Yes	290	2800	110	885	no	no	25	29
19	Separated-incision	Yes	280	2000	85	1230	no	no	42	24
20	Separated-incision	Yes	240	800	85	1210	Implant breakage	Rechange	36	24
21	Separated-incision	Yes	270	3000	92	1960	DL	Open reduction	26	21
22	Separated-incision	Yes	320	1600	87	1320	no	no	47	25
23	Separated-incision	No	420	2000	87	2200	no	no	56	21
24	Separated-incision	No	360	2800	86	1080	no	no	42	24
25	Separated-incision	No	320	4000	76	1150	no	no	42	24

ABBE abdominal aortic balloon embolism, DL dislocation, PRBC packed red blood cells, DWH delayed wound healing

Complications

All of the patients underwent separated-incision approach achieved primary wound healing within 3 weeks, whereas 47.06% (8/17) of patients in the traditional group experienced wound complications (0% vs. 47.06%, p = 0.026). Regarding major complications in the traditional group, 8 (8/17) patients experienced wound healing disorders, 4 of whom had subsequent infections, 7 of whom needed a second operation. Two patients (2/17) experienced dislocation and subsequent open reduction, 1 patient experienced implant breakage, and 1 patient experienced local recurrence and needed hindquarter amputation. Meanwhile, one patient in the separated-incision group experienced dislocation due to a fall at 4 months after surgery, and 1 patient experienced implant breakage.

Functional recovery

The time to ambulation post-surgery was shorter for the separated-incision group than for the traditional group (39 ± 7 days vs. 62 ± 11 days, respectively; p = 0.015). Additionally, the mean MSTS score at 12 months was significantly greater (24.25 ± 1.80 vs. 20.47 ± 1.98, p = 0.006). There was no difference between patients who underwent bony reconstruction and those who did not in terms of either the time to ambulate or the MSTS (55 vs. 52 days, p = 0.757; 21.15 vs. 23.8, p = 0.204).

Discussion

Internal hemipelvectomy is inclined to result in prolonged anaesthesia and operation time, massive blood loss, neurovascular damage, urogenital/abdominal damage, and other risks [3]. These risks may lead to postoperative wound hematoma, poor wound healing, and infection [7, 21]. According to our results, internal hemipelvectomy using traditional T-incision might lead to a high volume of blood loss (3,952 mL), more wound complications (47%), and inferior functional recovery (MSTS 68%) (Table S1). These results are similar to previous reports in Table 3, that internal hemipelvectomy is associated with high rates of morbidity, especially wound complications (30–50%) [3, 4, 10, 11, 22–25].

Table 3.

Comparison of various internal hemipelvectomy studies. (3DP, 3D-printed)

	Patients (N)	Type (I + II + III or II + III)	Reconstruction Methods	Approaches	Operation Time (Minutes)	Blood Loss (mL)	Wound Healing Disorders	Infections	Dislocations	MSTS Score
Lackman et al. (2009) [20]	30	16/30	Most had no reconstruction	T-incision	300	1500	27%	13%	NA	69%
Ji et al. (2013) [14]	100	59/100	Modular prosthesis	T-incision	NA	2762.6	9%	15%	9%	57.2%
Bus et al. (2017) [3]	47	26/47	LUMiC prosthesis	NA	NA	NA	38%	28%	13%	70%
Ogura et al. (2018) [23]	80	46/80	C-THA prosthesis	T-incision	562	2887	31%	39%	5%	43%
Ji et al. (2020) [15]	80	56/80	Modular prosthesis	T-incision	276	1899.5	10%	6.3%	2.5%	83.9%
Ogura et al. (2020) [22]	123	25/100	Bony reconstruction	NA	654	3750	34.1%	NA	NA	NA
Hu X et al. (2022) [11]	25	13/13	3DP Prosthesis	Two-incision (MGMII)	250	1700	0	0	0	96.7%
		11/12	3DP Prosthesis	Single incision	315	2400	25%	8.3%	0	80%
Zhu et al. (2022) [33]	40	12/15	3DP Prosthesis	T-incision	324	3413	6.7%	6.7%	0	81.5%
		20/25	Screw-rod-cage system	T-incision	394	3782	12%	16%	12%	73.3%
Current study	25	17/17	3DP Prosthesis (15/17)	T-incision	398	3952	23.5%	23.5%	12%	68%
		8/8	3DP Prosthesis (5/8)	Separated-incision	312	2375	0	0	12.5%	81%

Wound complications from the traditional T-incision approach may be due to the skin necrosis and subsequent wound healing disorders at the ‘‘T’’ intersection [9, 10]. Potential reasons may be intraoperative skin ischemia caused by constant retraction, and impaired blood supply (gluteal vessels) resulting from longitudinal incisions across the groin area [11, 26]. Significant intraoperative blood loss and extensive postoperative wound exudate led to increased drainage, further impacting wound healing. What is worse, extensive dissection and resection of the iliopsoas and gluteus maximus, among other structures, may impair functional recovery and hip joint stability [27–29].

Regarding the history and development of internal hemipelvectomy, we have summarized and compared various studies since 2009, focusing particularly on surgical approach design, prosthetic reconstruction methods, and their associated complications (Table 3). It is evident that the traditional T-shaped incision approach has been adopted by the majority of researchers, but it has notable shortcomings in terms of blood loss, surgical duration, and complications. Only within certain high-volume case studies, such as those conducted by the Bone Tumour Team at Peking University People’s Hospital [14, 15], have there been improvements and controls on wound complications after years of refinement.

Additionally, the Gibson and ilioinguinal separated approaches proposed by Hu et al. [11] are somewhat similar to our technique. Their separated dual incisions demonstrated better damage control regarding surgical time, blood loss, and wound complications. However, our incision has notable differences compared to Hu et al. For the anterior pelvic incision, we utilize a shortened T-shaped incision that ensures two windows, while Xu’s described ilioinguinal approaches employ a single window, resulting in a longer incision that extends to the perineum and the pubic symphysis, completely exposing the inferior pubic ramus. For the posterior pelvic incision, our approach is similar to the Gibson approach but is also modified; specifically, the distal part of the incision is adjusted to follow the gluteal fold. This adjustment allows the anterolateral incision to adequately address the proximal femur without the need for the Gibson approach to expose the proximal femur.

A separated-incision approach proposed by us, avoiding a direct incision from the groin to the gluteal fold, may minimize skin necrosis and subsequent wound healing disorders at the ‘‘T’’ intersection. Furthermore, the separated posterior incision can thoroughly expose the deep of pelvis, including the posterior sacroiliac joint, foramen magnum, posterior column/wall of the acetabulum, and sciatic tubercle. It also can protect the sciatic and gluteal neurovascular bundle under direct vision (Fig. S5). This may compensate for the shortcomings of the insufficiency of the T-incision approach [11, 16]. As shown in the current study, in the absence of significant differences in baseline characteristics and prosthetic reconstruction rates, the separated-incision approach exhibited significantly less mean blood loss with an equal operative time. Without extending the surgical time, less bleeding is benefit for minimize wound healing disorders and infections [3, 21, 30].

What is more, the modified separated-incision approach involves two surgical incisions, and has three observation windows as described in Supplement Materials. The incisions can connect to each other through submerged skin-muscle channels. This will facilitate the preservation of the skin and muscles, especially the gluteus maximus and iliopsoas, and reduce intraoperative surgical injury to skin-muscle flaps. The preservation of the gluteus maximus with the flap guarantees its viability and successful primary healing [31]. Preservation of the gluteus maximus and iliopsoas also promotes postoperative motor rehabilitation of the hip joint [32, 33]. In the current study, patients underwent separated-incision approach had fewer wound complications, and better MSTS scores. This could be substantiated by other reports on separated incisions for internal hemipelvectomy [11, 16, 34].

Limitations

This study has several limitations. The small sample size results from the low incidence of Enneking II + III pelvic tumors, which affects the general of the findings. Additionally, not being a randomized controlled trial introduces potential selection bias, indicating a need for multicenter studies to create a larger and more diverse participant pool. Furthermore, the retrospective nature of the study may lead to inconsistencies and limit the establishment of cause-and-effect relationships. However, together with the modified previous surgical approaches and innovated novel approaches in Table 3, we believe that the separated-incision approach is a novel complement to the surgical strategy for internal hemipelvectomy.

Conclusions

We proposed a separated-incision approach for internal hemipelvectomy related to type II + III regions, avoiding a direct incision from the groin to the gluteal fold, while utilizing three distinct surgical windows. It has the potential to decrease intraoperative injuries and wound complications, as well as improve functional recovery. It shows potential to be an alternative to the traditional T-incision approach for internal hemipelvectomy, further studies with larger cohorts are needed to validate these results.

Authors’ contributions

All authors contributed to the study conception and design. Study Design: Dr. Zhang CL was responsible for the overall design of the study, including formulating the research questions and hypotheses to ensure scientific rigor and feasibility. Data Collection: Dr. Hu JP and Dr. Cai T led the data collection efforts, participating in case screening and data entry to ensure the accuracy and completeness of the dataset. Data Analysis: Dr. Hu JP, Dr. Wang YJ, Dr. Xu EJ and Dr. Zhu KP conducted the statistical analysis of the data, employing relevant software for data processing and writing the results section. Manuscript Writing: Dr. Hu JP, Dr. Ma XL contributed to the initial draft of the manuscript, compiling the literature review and ensuring the accuracy of the citations. Review and Revisions: All authors participated in the manuscript review and revision process, conducting a comprehensive evaluation of all sections and providing valuable feedback for improvement.

Funding

This project was supported by grant from National Natural Science Foundation of China (No.82072963, 82103513, 82303898), Program of Shanghai Sailing Program (No.20YF1437700), Climbing Talents Program of Shanghai Tenth People’s Hospital (2021SYPDRC021) and Clinical Research Program of Shanghai Tenth People’s Hospital (YNCR2B002, YNCR2C012).

Data availability

Data and materials for this study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Each author certifies this study was approved by the Institutional Review Board (IRB) of Shanghai Tenth People’s Hospital (SHSY-IEC-5.0/22K14/Y01) and that it complies with the Helsinki Declaration. Informed consent to participate was obtained from all participants involved in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.