Do irradiated osteo-articular recycled tumor autografts still hold promise for biological joint reconstruction? Our experience with acetabular and proximal ulna ECRT

Abstract

Background

Joint reconstruction following resection of malignant bone tumors is challenging in itself in spite of several options in hand. Ability to restore joint anatomy, function and mobility while achieving optimal oncological outcomes are the requirement of reconstructions today. While biological reconstructions (allograft or recycled tumor autografts) following tumor bone surgery are popular for intercalary resections not involving the joint, their use for osteo-articular reconstructions are associated with concerns over cartilage and joint health. We have used extracorporeal radiation therapy (ECRT) and re-implantation of the osteoarticular segment as a size matched recycled tumor autograft reconstruction after complex acetabular and proximal ulnar resections; owing to the lack of significantly superior reconstruction alternatives in these locations and also review the current literature on other biological/non-biological reconstruction options.

Questions/purposes

(1) What are the oncological, reconstruction and functional outcomes with osteo-articular reconstruction using ECRT and re-implantation of recycled tumor autograft for the acetabulum and olecranon? (2) Is there an evidence of cartilage loss, joint damage or avascular necrosis resulting from irradiation of the articular autograft?

Methods

19 patients with primary bone tumors underwent limb salvage surgery with en-bloc resection and reconstruction using the resected articular tumor bone after treating it with extra-corporeal irradiation of 50–60Gy. These included 16 acetabular and 3 proximal ulnar. While all patients were included for oncological assessment; minimum follow-up of 24 months was considered for final outcome assessment of function and joint status.

Results

MSTS scores of the 16 acetabular reconstruction patients with minimum 2 years follow-up was 87% (26/30). Neither delayed union, non-union at osteotomy sites nor was any fractures reported in the irradiated graft. There was no local recurrence within the irradiated graft and only 1 patient required graft excision for uncontrolled infection. All 3 patients of proximal ulna reconstruction achieved healing and full range of movement of the elbow. Scores of MSTS: 100% (30/30), MEPS: 100 and DASH: zero was achieved. Two patients developed osteonecrosis of the femoral head; one requiring a joint replacement and one awaiting replacement. One patient of acetabular reconstruction has joint space narrowing on radiographs with mild clinical symptoms.

Conclusions

Extracorporeal radiotherapy and re-implantation after osteo-articular resection is an oncologically safe option offering promising outcome in our small series. The availability of size-matched graft, thus avoiding inherent problems of allograft also provides a better economic option over endoprosthesis and its associated complications in select sites. The results can deteriorate over time that may require secondary reconstructive procedures like joint replacement.

Level of evidence

Level IV, Therapeutic Study.

1. Introduction

Biological reconstruction following resection of malignant tumors is popular as it provides structural stability, restoration of anatomy and function; which on successful incorporation is permanent and rarely requires revision procedures.¹ Most joint reconstructions today following resection of bone tumors require the use of metallic prostheses. Though endoprosthetic joint reconstructions continue to remain popular owing to their modular, readily available and accessible inventory, short term and long term failures remain a cause of concern.^2,3 Higher incidence of primary bone tumors in the younger population add to the challenge with prosthetic joint reconstructions. An osteoarticular bone graft would be an ideal biological reconstruction. However these reconstructions usually require allografts which requires access to a large and sophisticated bone bank for providing a proper size matched graft. Allografts also have inherent risks of infection and non-unions. Since its first report in 1968, extra-corporeal radiation therapy (ECRT)⁴ has been used to recycle tumor bone for re-implantation as radiated autograft. While these reconstructions have been of proven value in non-articular reconstructions, they have not been used for osteoarticular reconstructions owing to the presumed risk of cartilage damage with the high radiation dose delivered and subsequent failure.⁵

Considering the limited access to osteo-articular allografts and their reported failures, ECRT could offer as an attractive option of reconstruction, particularly to reconstruct the acetabulum and olecranon as has been used in our series. For the acetabulum, literature has shown excellent oncological outcomes in terms of local control and overall survival in several series^6,7; however reconstructions following pelvic resections have reported fair to superior functional outcomes after biological joint reconstructions when compared to endoprosthetic reconstructions with lower re-operation rates.⁸ To restore anatomy using ECRT of resected tumor bone and re-implantation; surgical planning considered the quality of bone affected by the tumor, the ability to repair ligaments, capsule and muscle re-attachments for maximum function without compromising stability. We used ECRT for osteoarticular reconstruction at sites where we do not have any superior conventional reconstruction option other than a complicated prosthesis (the acetabulum and the olecranon). While oncological outcomes were evaluated in all patients at follow up, functional, reconstruction and joint status (arthritis, cartilage loss, subchondral fracture/collapse) were evaluated at a minimum 24 months follow-up. From our experience^9,10 with recycled osteo-articular autografts using ECRT, considering the addition of numbers to the series of patients and incidence of newer findings in patients with longer follow up we asked:

• (1)What are the oncological, reconstruction and functional outcomes with osteo-articular reconstruction using ECRT and re-implantation of recycled tumor autograft for the acetabulum and olecranon?
• (2)Is there an evidence of cartilage loss, joint damage or avascular necrosis resulting from irradiation of the articular autograft?

2. Methods

The institution waived approval for the human protocol for this study, and all investigations were conducted in conformity with ethical principles of research for this retrospective analysis of patients having undergone osteoarticular reconstruction with ECRT and re-implantation of recycled tumor autograft between Feb 2010 and Dec 2018 at our institution. The retrospective analysis of patients included 16 with acetabulum involving pelvis and 3 proximal ulna resections and re-implantation after ECRT. Of these patients, 12 pelvic resections and the 3 proximal ulna resections have reported favorable outcomes on preliminary short term data.^9,10 However the incidence of newer findings at longer follow up in our acetabulum ECRT group raises questions about the long term credibility of this procedure and hence pertinent it be reported here. Assessment of outcomes included a review of radiographs and institutional tumor database for surgical and follow-up records to study oncologic (local disease recurrence), reconstruction (union of osteotomy junctions, implant failure or graft fracture), joint status and functional outcomes in patients that had a minimum 24 months follow up.

Pelvic resections were classified as per Enneking and Dunham classification. Pelvic resections [Fig. 1A] included 9 chondrosarcomas, 4 Ewings sarcomas, 2 Osteosarcomas and one Leiomyosarcoma metastasis. All 3 ulnar resections [Fig. 1B] were biopsy proven Ewings sarcoma. Osteosarcoma and Ewings sarcoma cases received neoadjuvant chemotherapy as per our institution protocol. The resection margins were planned on pre-treatment imaging (radiographs, computerised tomography (CT) and magnetic resonance imaging scans) to ensure a layer of normal tissue over soft tissue component and a 2 cm bony margin. Part of the joint was saved wherever possible without compromising the margins; thence 5 of our acetabular resections and 1 olecranon resection was partial-articular/through-the-joint. One patient received adjuvant radiotherapy owing to contaminated margins. We found the careful case selection in assessment of tumor bone integrity, ability to preserve the ligament/capsule attachments in the autograft for efficient reconstruction of soft tissue accounted for the favorable functional results in these patients, especially the hip joint capsule attachment to the acetabulum; annular ligament, capsule, triceps tendon in proximal ulna among our ECRT patients.

Fig. 1.

(A) Specimen of a pelvic type II with complete acetabular resection, which was treated with ECRT for re-implantation and post-operative radiograph showing the recycled osteo-articular graft in situ and fixation implants to hold the graft in place. (B) Proximal ulna with complete olecranon resection for Ewings sarcoma, re-implanted as an osteoarticular autograft after irradiation as seen on post-operative radiograph.

The technique of ECRT remained the same for osteoarticular and intercalary bone resections. The resected tumor bone was transported to the radiation department maintaining all aseptic precautions, where a single fraction of 50 Gy using a linear accelerator, was delivered over 20–25 min. Void if any, following ECRT was filled with bone cement.

Mean duration for autograft to reach us back in the operating room post ECRT was 50 min (Range 45–60 min). Resurfacing of hip joint using constrained liner was done in two patients with complete acetabular resections and focal joint involvement; prior to our experience with osteoarticular irradiated autografts and they are excluded from this cohort. Acetabular reconstruction patients were kept non-weight bearing until radiographic evidence of bony union and then progressed to full weight bearing mobilization.Post-operative physiotherapy after proximal ulna reconstruction allowed immediate range of movement exercises as tolerated. Loading and lifting weights were allowed once r adiological union was observed. Adjuvant chemotherapy was started as early as post-operative day 12–14 days in all indicated patients. Follow-up was at 3 months interval for local and systemic disease assessment. Thorough clinical evaluation and imaging was done for oncological and functional outcomes. Minimum follow-up of 2 years was considered for assessment of final outcomes. Radiographs were assessed for any local recurrence, union across osteotomy sites and signs of joint space narrowing or arthritic changes, (any lucencies or collapse of femoral head to look for osteonecrosis in acetabular group). Chest radiographs alternated with CT scan of chest was performed every 6 months to screen for metastatic nodules. Functional scoring for pelvis group of patients was done using Musculoskeletal Tumor Society (MSTS). Functional outcome measurement in proximal ulna group included range of movement assessment (ROM), muscle power assessment as per Lovett scale, MEPS (Mayo elbow performance score) and DASH(disability of arm, shoulder and hand score) in addition to MSTS.

A spreadsheet was developed for data entry (including demographic data, surgical treatment, patient outcomes, complications and other case specific information). Descriptive statistics was based on frequencies and percentages (categorical variables); means and standard deviation (continuous variables) performed using SPSS Version 25 (IBM Corp, N.Y., USA).

3. Results

Acetabulum: None of the patients developed local recurrence within the irradiated autograft bone at a mean follow up of 63.5 months (27–135 months). Three patients with high grade sarcomas died of metastatic disease 10, 19 and 36 months after surgery. Another patient developed jaw osteosarcoma, 7 years after primary surgery and succumbed to metastatic disease within a year. The patient operated for metastatic disease in the pelvis was operated later for lung metastasis. He is alive with metastatic disease in lung controlled by pazopanib (selective multi-targeted tyrosine kinase inhibitor). None of these five had any local recurrence of disease at available latest follow up. Two patients who developed local recurrence in soft tissues around the operated area were low-grade secondary chondrosarcomas who also had previous debulking surgeries elsewhere. Both are currently disease free after excision of the recurrences. MSTS scores of the 14 patients with follow-up >24 months was 87% (26/30; range 17–30). All proximal osteotomy junctions healed within 10 months. No fractures were reported in the irradiated bone segment. Proximal Ulna: With a minimum follow-up of 2 years, (46, 52 and 60 months), all 3 patients achieved full range of movement of the elbow with no restriction of prono-supination. Muscle power was 5/5. Perfect scores of MSTS: 100% (30/30), MEPS: 100 and DASH: zero were achieved. Union across junctions was noted at median 8 months (range; 6–9 months), with local recurrences, no implant or graft failures at latest follow up.

Acetabulum: Three patients developed secondary osteoarthritis of the hip joint following osteonecrosis of femoral head based on radiographs [Fig. 2] in two patients at 25 and 36 months and one patient at 24 months post-operatively showed radiographic signs of joint space narrowing with mild clinical symptoms as compared to the unaffected side. Two patients with avascular necrosis still maintained joint space in CT scans initially when the diagnosis was made and on subsequent follow up presented with subchondral collapse progressing to secondary osteoarthritis. One of these has undergone total hip arthroplasty elsewhere for the arthritic hip joint and the other is awaiting replacement. We inferred that the secondary osteoarthritis was a sequele to osteonecrosis of the femoral head resulting from surgical dislocation rather than primarily caused by acetabular cartilage damage. Proximal Ulna: No reports of joint space narrowing, laxity, subluxation or dislocation in the series at minimum 46 months of follow up.

Fig. 2.

(A) 44year/Male with biopsy proven chondrosarcoma of the pelvis underwent a type II pelvic resection for the left side, (B) intra-op image showing the surgical dislocation of femoral head to deliver the specimen. (C) The tumor segment was treated with ECRT and received on the surgical table for preparation for re-implantation, focal joint involvement was appreciated after tumor clearance and the tumor cavity was filled with bone cement. (D) Post-operative radiographs show the implanted autograft with fixation implants in-situ and preserved joint space as compared to the unaffected side. (E) At 25 months follow up, patient resented with pain and restricted hip movement worsening over a few weeks, radiograph shows complete loss of joint space, subcondral collapse with avascular necrosis of femoral head and myositis within the abductors. (F) Patient underwent conventional total hip arthroplasty for resurfacing at 28 months post ECRT reconstruction to regain painless movement at the hip.

The complexity of pelvic resections and reconstuction was associated with certain inherent complications that were seen in our series too. There were three acute wound related complications. Two settled with wound washes and retained implant at the last follow-up. However one patient underwent removal of implant along with infected ECRT bone eight months post-operatively. This patient was functionally well at the last follow-up. One patient had a delayed discharging sinus 5 years post-surgery; cultures were sterile and settled conservatively without functional disability. One patient who settled with wound wash was in which a typeI + type IV + typeII resection and reconstruction was performed, also had two episodes of hip joint subluxation; at two and nine months post-operatively. Both casualties were surgically managed and the patient was functionally well at latest follow-up. There were four major acute post-operative complications; one sciatic nerve palsy; one cardiac event which was managed medically; two patients who developed femoral artery thrombi, and were managed accordingly and both independently mobile at the time of completion of this study.

4. Discussion

The best method of joint reconstruction of the defect after tumor resection is controversial. While we restricted our osteoarticular ECRT technique for the pelvis and proximal ulna, other joints were eliminated due to superior reconstruction options available; namely endoprostheses around the knee, arthrodesis across the wrist or ankle and prosthetic or bone-prosthesis composite for the proximal humerus to allow deltoid reattachment. Tumor prosthesis have several obvious advantages, like convenience, availability, immediate mobilization and loading capabilities,¹¹ however it comes with its own drawbacks, including infection, soft tissue failure, mechanical failure, aseptic loosening and an increased probability of revision surgery within 10 years.11, 12, 13 Osteoarticular allografts, being a biological option for reconstruction of bone defect offer hope but requires the availability and maintenance of a bone bank and other problems associated, such as graft rejection and risk of viral transmission.¹⁴ Need for complex and custom made endoprosthesis for the acetabulum and proximal ulna, 3-dimensional [3-D] printed prostheses and newer ways for osseointegration have been described, though concerns still remain about the cost, functional outcomes and complications associated.15, 16, 17 This allowed us to explore osteorticular ECRT at these sites as a biological scaffold for integration as a cost-effective, readily available and size matched graft and yet leaving the option for prosthetic reconstruction in the advertent event of graft failure.

Drawbacks of using ECRT treated autografts is the lack of material available for histopathological examination; for knowing the effect of neoadjuvant chemo/radiotherapy on tumor biology; and assessing the tumor margins. The main drawback of our study is the limited number of cases with relatively short follow-up. Even though long term data from patients who underwent ECRT treated autograft reconstruction after extremity tumor resection have not raised any concerns,^5,6 we found that three patients developed secondary osteoarthritic changes in the hip, owing to avascular necrosis of the hip in two and not particularly due to acetabular cartilage damage. That said, we limited our assessment of joints to plain radiographs, CT scans and comparing joint space with contralateral normal joint, and did not perform sophisticated investigations to assess the cartilage status like subchondral bone scan uptake¹⁸ or SPECT-CT imaging.¹⁹ There is also mention of a novel micro computerised tomography (micro-CT) imaging which was used for quantitative evaluation of articular cartilage damage and the associated changes in subchondral bone by Gatenholm et al.²⁰ in the literature. We opine that long-term studies with incorporation of similar quantitative tests are warranted in future to alleviate some of these concerns.

On comparing with other options for reconstruction, such as megaprosthesis and allograft reconstruction, there was no increased rate of local recurrence and no recurrence within the irradiated graft for the follow-up in our small series or other similar studies.^21,22 Infection is a potentially devastating complication mentioned with major reconstructions^23,24; however in our series we required graft excision in only one patient with uncontrolled infection. In our small series of ECRT and re-implantation for the acetabulum and proximal ulna, we have promising results of this technique as a reconstruction option.

In our series we did not find any signs of cartilage degeneration such as subchondral sclerosis or joint space narrowing limited to the irradiated autograft. The two patients that resulted in avascular necrosis of the femoral head were postulated to have resulted from surgical dislocation of the femoral head or intra-articular disease extension into the acetabulum, the cavity of which was filled with bone cement post ECRT in one patient which presented with joint space narrowing. Uyttendaele et al.^25,26 anticipated the risk of avascular necrosis and resurfaced the acetabulum, however an objective measurement of factors determining this risk need to be ascertained to define which joints need to be surfaced and which can be salvaged since the technique of surgical dislocation or cementing the tumor void was not different to these three cases when the techniques were retrospectively analyzed from surgical records. While the opinion on joint resurfacing is divided, we chose not to resurface where there was no tumor invasion into the joints and as we gained experience from follow ups, we reserved resurfacing for all patients - as a procedure for the future in the event of joint degeneration. With the ability to salvage the recycled autograft and allow resurfacing with conventional arthroplasty in the one patient with joint arthritis, we believe that ECRT as an osteo-articular biological option may be explored for other joints as well which may be easily converted to a bone-prosthesis composite with a conventional joint arthroplasty procedure while achieving biological soft tissue attachments (eg: patellar tendon over proximal tibia, deltoid insertion over proximal humerus). In conclusion, we found that ECRT of osteoarticular resections to reconstruct joints (acetabulum and proximal ulna reconstructions) can be an excellent option in selected cases with joint stability and good functional outcomes. We recommend long term data analysis with addition of joint health and cartilage status following ECRT. Even with the potential risks, reconstruction with an anatomically size-matched irradiated autograft eliminates donor morbidity and associated logistics issues, offers an affordable alternative to endoprosthesis and provides good functional outcomes. In the concerns of joint health, the ability to retain the graft and resurface the joint offers an additional option of reconstruction for other joints as well.

Author statement

Investigation performed at P.D. Hinduja National Hospital and Medical Research Centre.