Revision surgery for instrumentation failure after total en bloc spondylectomy: a retrospective case series

Total en bloc spondylectomy (TES) was developed for complete surgical resection for spinal tumors [1, 2]. TES was reported to produce less local recurrence and better survival than piecemeal resection for primary tumors [3‐7], aggressive benign tumors [1, 8], or metastatic spinal tumors [1, 9]. TES involves complete resection of the affected vertebra(e) and surrounding musculoligamentous supportive tissues. The surgery presents a challenge for spinal reconstruction owing to the complete discontinuity in the spinal column. To restore spinal stability after resection, robust instrumentation and bone grafting are needed, with anterior column support [10‐12]. Because TES is performed in patients with a relatively good prognosis [13], long-term stabilization of the spine is important. Failure of bone fusion can cause instrumentation failure, causing back pain, neurological deterioration, and decreased performance of activities of daily living (ADL) [13‐18]. In most cases, revision surgery is required to improve spinal stability and symptoms.

Since 2010, instead of harvesting autografts from the ilium or fibula, the resected lamina and vertebral body from the TES are frozen in liquid nitrogen and used as grafted bone for spinal reconstruction. This technique has the following benefits: no pain at the bone harvest site, shortened operative time, decreased blood loss, and additional antitumor immune response [19]. In this procedure, the strategy of revision surgery is more important than that in the conventional procedure because the instrumentation failure rate may increase with a decreased bone fusion rate.

Despite the fact that several studies have examined instrumentation failure after TES [11‐14], there are few published reports on revision surgery for instrumentation failure after TES. Therefore, the aims of this study were to determine the rate of instrumentation failure and to document clinical outcomes of revision surgery for instrumentation failure after TES with reconstruction using frozen autografts treated with liquid nitrogen.

Of the 61 patients, 26 (42.6%) experienced instrumentation failure at an average of 32 (range, 11–92) months after TES. The mean body mass index of the patients who experienced instrumentation failure was 23.4 (range, 16.4–32.3). Of the 26 patients who experienced instrumentation failure, 4 (15.3%) underwent radiation therapy for spinal tumors before TES. Ten patients had primary tumors, comprising giant cell tumor in 3 patients, chondrosarcoma in 1, epithelioid sarcoma in 1, liposarcoma in 1, Ewing’s sarcoma in 1, synovial sarcoma in 1, chordoma in 1 and aggressive hemangioma in 1. Sixteen patients had metastatic tumors from renal cell carcinoma in 4 patients, breast cancer in 3, thyroid cancer in 2, lung cancer in 2, hepatocellular carcinoma in 1, esophageal cancer in 1, parotid cancer in 1, leiomyosarcoma in 1, unknown primary squamous cell carcinoma in 1. Rod breakage was observed in 20 patients, screw breakage in two patients, rod and cage breakage in two patients, cage breakage in one patient, and screw and cage breakage in one patient. However, no evidence of local recurrence was found at the surgical site throughout the follow-up period in any patient. Of the 26 patients, 19 (73.1%) experienced back pain, 8 (30.8%) had lower-extremity neurological deterioration, and 6 (23.1%) were unable to walk because of these symptoms (Table 1). Of the 26 patients, 23 underwent revision surgery and three did not undergo revision surgery. In patients without revision surgery, one patient (patient 26; Table 1) with an asymptomatic rod breakage died from cancer 1 year after instrumentation failure. Patients with asymptomatic cage and screw breakage (patients 24 and 25; Table 1) survived for 1 year and 3 years, respectively. Neither experienced advanced spinal instability nor related symptoms. Revision surgery was performed with posterior instrumentation replacement in all 23 patients, posterior bone grafting in 21 patients, and additional bone grafting around the cage in three patients. To increase stability, the procedure was extended by three segmental fixations above and below in three patients, two segmental fixations above and below in one patient, and two segmental fixations above and four segmental fixations below in two patients. All of the six patients with extended fixation had screw loosening. In the remaining cases without screw loosening, instrumentation was not extended. Two rods were used in four patients, three rods were used in one patient, and four rods were used in 18 patients. Titanium alloy rods (diameter: 5.5 mm) and cobalt chromium rods (diameter: 6.0 mm) were used in 4 and 20 patients, respectively. The average operative time and intraoperative blood loss were 207 ± 45 (range, 121–287) min and 93 ± 81 (range, 20–270) ml, respectively. Intraoperative findings via a posterior approach did not reveal any obvious pseudoarthroses in the patients because the scar tissue covered the anterior column where the pseudoarthroses existed, but mild metallosis at the rod fracture site was observed in nine patients. All patients, including the six patients who were unable to walk due to symptomatic instrumentation failure, maintained or recovered their ambulation function after revision surgery. Perioperative complications of revision surgery were superficial surgical site infection in two patients and delayed wound healing in one patient. They were successfully treated with conservative therapies using antibiotics and basic fibroblast growth factor spray.

Eighteen patients could be followed up for > 2 years after revision surgery (average follow-up period: 41 months) by radiological evaluation. At the last follow-up, bone fusion was achieved within the anterior cage and at the posterior bone graft in 11 patients, within the anterior cage in two patients, and at the posterior bone graft in five patients. There was no re-instrumentation failure after revision surgery (Table 2).

We reported the instrumentation failure rate of patients with instrumentation failure after TES with reconstruction using frozen autografts treated with liquid nitrogen. Revision surgery was performed using the posterior approach alone. Bone fusion was achieved, and there was no re-instrumentation failure in any patient at the follow-up period > 2 years.

With continuing advances in cancer therapy, acceptable long-term prognosis can be expected even in patients with metastatic spinal tumors [21‐24]. In TES, en bloc resection of a tumor-bearing vertebra can be curative, leading to longer-term survival, and achieving bone fusion of the reconstructed vertebral body is essential [25]. However, instrumentation failure caused by unsuccessful bone fusion is not a rare complication. Park et al. [10] reported that 12 (37.5%) of 32 patients experienced rod breakage at an average of 29.2 (range, 8–93) months after TES. Sciubba et al. [14] reported instrumentation failure in 9 (39.1%) of 23 patients who underwent lumbar-spine TES. Matsumoto et al. [12] reported instrumentation failure in 6 (40%) of 15 patients who underwent TES. In our study, instrumentation failure following the TES procedure was identified in 26 (42.6%) of 61 patients at an average of 32 (range, 11–92) months after TES, which was comparable to that of other studies.

The previously reported incidence rate of instrumentation failure after TES using the same reconstruction method as ours (except using fresh autologous bone for bone grafting) was 17.0% (8/47) [11]; in the present study, the instrumentation failure rate after TES using frozen bone was 42.6% (26/61). It was reported that bone formation tended to be delayed when frozen bone autografts were used compared with fresh bone autografts [25]; therefore, the instrumentation failure rate in the present study was higher than that previously reported [11]. Although bone fusion was delayed, complete bone fusion within the cage was obtained in the TES model canine using frozen bone [25]. The instrumentation failure rate following the first procedure was higher in the present study; however, stability was maintained for a long time in 35 (57.4%) of 61 patients. Considering the advantages of using liquid nitrogen-treated bone, we continue using frozen bone autografts in spinal reconstruction during TES. To decrease the incidence of instrumentation failure, we recently began using a more robust cage and cobalt chrome rods to create a stiffer construct of the operated spine and having additional bone graft around the cage to facilitate bone fusion.

In the present study, back pain and neurological deterioration caused by instrumentation failure developed in 19 (76.1%) and 8 (30.8%) patients, respectively. Matsumoto et al. [12] reported that all 6 (100%) patients experienced back pain and 1 (16.7%) experienced neurological deterioration at the time of instrumentation failure. Park et al. [10] reported that back pain developed in 7 (58.3%) patients, and no patients had neurological deterioration at the time of instrumentation failure. These findings suggest that most patients with instrumentation failure experienced significant clinical symptoms. Revision surgery is necessary to prevent decreased ADL performance among symptomatic patients with instrumentation failure.

Instrumentation failure is caused by delayed union between the cage and the vertebral body [11], and revision surgery is performed to achieve robust restabilization and bone fusion. In the present study, we performed robust restabilization in most patients by replacing titanium rods with cobalt chromium rods and by increasing the number of rods. We also performed bone grafting at the posterior aspect of the spine. During the primary surgery, bone grafting at the posterior element was difficult because there was no bed for bone grafting at the level of the resected vertebra. Moreover, because the space was covered with scar tissue, bone grafting was straightforward and secured during the revision surgery.

In the present study, 13 (72.2%) of 18 patients with > 2 years of follow-up after revision surgery achieved bone fusion within the cage, but the remaining 5 (27.8%) did not achieve bone fusion; nevertheless, bone resorption within the cage improved. We believe that robust restabilization of the posterior instrumentation increased the stability of the spine, which facilitated bone fusion within the cage. In addition, bone fusion at the posterior aspect of the spine, which could not be applied in the primary surgery without the scar tissue of the resected vertebral area, was achieved earlier than that within the cage in cases where posterior bone grafting was performed. We believe that attaining bone fusion at the posterior aspect further increased the stability of the spine and favored bone fusion within the cage. This finding indicates that replacement of the cage using the anterior approach is unnecessary when the cage is restabilized using a stiffer construction by exchange and supplement of posterior instruments. Our recommended revision procedure is to perform posterior restabilization with three or four cobalt chromium rods and posterior bone grafting. In cases with a severely damaged cage (not observed in this study cohort), cage replacement using the anterior approach should be considered.

This study has some limitations. The small and heterogenous cohort with several tumor histologies and adjuvant therapies and the retrospective manner of data collection can introduce bias and errors. This retrospective study included various reconstruction procedures (e,g, the material and number of rods in revision surgery) differed depending on the time of surgery. The follow-up time was limited as well. Longer follow-up is required to determine the accurate incidence of instrumentation failure after revision surgery. Although all patients were diagnosed as having oligometastatic cancer in the present study, the indications for TES remain controversial because less invasive surgeries (e.g. separation surgery) have shown significant results recently. Despite these limitations, this study demonstrated the rate of instrumentation failure after TES with reconstruction using frozen autografts treated by liquid nitrogen and described a relatively simple and effective strategy for revision surgery and its favorable outcomes. The results obtained from this study will contribute to revision surgery for instrumentation failure after TES.

In patients with instrumentation failure after TES, posterior instrumentation replacement and posterior bone grafting were performed during revision surgery. In this study, bone fusion was observed in all patients with > 2 years of follow-up after revision surgery, and no re-instrumentation failure was observed during follow-up. Our findings suggest that bone fusion can be achieved by revision surgery using the posterior approach.