Spine surgery for metastatic spine cancer in the era of advanced radiation therapy

Stereotactic body radiation therapy

Before the advent of SBRT, conventional external beam radiation therapy (EBRT) was considered a standard of care for MSC when surgical management was not initially indicated [24]. Although EBRT achieves high response rates for pain control (60%–80%), its efficacy is limited, with complete response rates of 0%–20% and partial response rates of approximately 60% [17]. Furthermore, approximately 20% of patients treated with low-dose EBRT require reirradiation due to pain progression within months of the first treatment course [17].

SBRT is increasingly becoming the standard treatment for spinal metastases, supported by recent trials demonstrating its superiority over EBRT [17]. The primary advantage of SBRT lies in its ability to deliver precise treatment with very high doses per fraction [25]. By accurately targeting the area of interest, SRBT enables dose-escalation while sparing surrounding organs at risk [25].

A retrospective analysis of data from the randomized Canadian Cancer Trials Group Symptom Control 24 phase 2/3 trial demonstrated improved local control rates with SBRT compared to conventional EBRT [24]. The local failure rate was significantly lower with SBRT than EBRT at both 6 months (2.8% vs. 11.2%, p<0.001) and 12 months (6.1% vs. 28.4%, p<0.001) [24]. Chang et al. [26] reported a 1-year tumor progression-free incidence of 84%, defining progression-free as no magnetic resonance imaging (MRI)-documented tumor progression. Similarly, Amdur et al. [27] reported a high local control rate of 95% at 11 months post-SBRT, with local control defined as no evidence of tumor progression on MRI at the treated site. Although SBRT yields excellent local control rates overall, variations exist based on the radiosensitivity of the tumor histotype [28]. Bernard et al. [28] found that while overall local control rates were high (82.6% at 1 year and 75.8% at 2 years), radioresistant tumors like non-small cell lung cancer or colorectal cancer had significantly higher failure rates (30.4% vs. 8.0% at 1 year, p=0.008) and poorer overall survival compared to radiosensitive tumors.

In addition to better local control rates, SBRT also offers better pain control compared to EBRT. Sahgal et al. [29] found that at 3-month post-treatment, 35% of SBRT patients experienced complete pain resolution versus 14% of EBRT patients (p=0.002). In a randomized phase 2 trial by Sprave et al. [30], SBRT led to faster pain reduction within 3 months (p=0.01) and significantly lower pain scores at 6 months post-treatment compared to EBRT (p=0.002).

SBRT has also shown efficacy in reducing epidural cord compression. Lee et al. [31] used SBRT to treat 33 patients with severe epidural compression, achieving a significant epidural tumor response in 74% of patients. Ryu et al. [32] demonstrated neurological functional recovery after SBRT in patients with metastatic epidural spinal cord compression (MESCC).

Recent studies have shown that SBRT is an effective retreatment option after failed initial radiation therapy [17]. The 1-year local control rate of SBRT after EBRT failure ranges from 66% to 90% [33]. Additionally, reirradiation with SBRT after initial SBRT achieved a local control rate of 81% [34]. Pain improvement was reported in 65%–81% of patients undergoing reirradiation with SBRT [35,36]. Finally, a systematic review reported a 12% risk of vertebral compression fracture (VCF) and a 1.2% risk of radiation myelopathy after reirradiation with SBRT [17]. Given its clinical efficacy and low risk of severe complications after retreatment with SBRT, the International Stereotactic Radiosurgery Society practice guideline recommends administering SBRT following EBRT or SBRT alone following multidisciplinary assessment [17]. The guideline also recommends consulting a spine surgeon before SBRT for patients with clinical features consistent with MESCC, mechanical instability, or baseline VCF [17].

Limitations of SBRT

The spinal cord is the most critical organ at risk during radiation therapy for MSC. Exceeding the dose tolerance can lead to radiation myelopathy, a rare but catastrophic complication of spine SBRT [37]. The reported risk of radiation myelopathy is approximately 0.4% for initial SBRT and 1.2% for reirradiation SBRT [17,38]. Conventional ERBT dosing (45–50 Gy in 1.8–2 Gy fractions) carries a very low risk of radiation myelopathy (0.03%–0.2%) [39]. However, these spinal cord dose constraints may not directly apply to SBRT due to differences in fractionation and dose distributions [37]. Sahgal et al. [29] recommended limiting the thecal sac maximum dosage to 12.4 Gy in one fraction, 17 Gy in two fractions, or 20.3 Gy in three fractions to keep the risk of radiation myelopathy below 5%. Notably, the majority of radiation myelopathy cases reported to date have occurred with single-fraction SBRT, suggesting that fractionation may be an important consideration to mitigate the risk [37].

While radiation myelopathy is the most feared complication of SBRT, VCFs at the treated level or adjacent levels are the most common complications [40]. Mantel et al. [41] reported a 34.4% incidence of VCFs post-SBRT, though only 5% were symptomatic, defined as an increase of >2 points on the visual analog scale for pain. Risk factors for post-SBRT VCFs include osteolytic metastases, extent of osteolysis, and preexisting VCFs [41]. Furthermore, the rate of compression fractures following SBRT is reportedly higher than that with other types of radiation therapy, reaching up to 30%. However, fewer than 5% of these cases require intervention or surgical stabilization [41].

Shifting landscape of surgical treatment

The early surgical approach for MSC involved laminectomy alone, without instrumentation, leaving anterior or lateral lesions unaddressed [42]. However, the surgical outcomes were poor, with only 30% of patients experiencing symptom improvement. Complications, including instability, deformity, postoperative pain, and neurological decline, were also common [42].

Unsatisfactory results and frequent complications from laminectomy shifted treatment preference toward EBRT [43,44]. Although EBRT reduced treatment-related morbidity compared to laminectomy, clinical outcomes remained poor, characterized by low local control rates and frequent clinical decline [43,44].

A 2005 landmark study by Patchell et al. [45] compared EBRT alone to surgery, including decompression, tumor debulking, and instrumentation. The surgery group had a higher rate of patients able to walk (84%) compared to the EBRT-only group (57%) [45]. This data shifted the paradigm back toward surgical interventions, leading to more aggressive intralesional gross total resections and even en bloc tumor resections via anterior cavitary approaches [45–47]. However, these procedures were associated with a high morbidity [47,48]. Furthermore, due to low local control rates with EBRT, this procedure led to a high local recurrence rate with poor long-term outcomes [46].

Separation surgery

The introduction of SBRT, which enables the delivery of high ablative doses of radiation to MSC, has significantly altered the concept of managing MSC [7,49,50]. Due to high local control rates, even for radioresistant tumors, many patients with MSC were initially treated with SBRT instead of surgery [1,4,51]. Studies have shown that ambulatory patients and those without neurological deficits benefit more from SBRT than surgery due to low treatment-related morbidity and similar local control rates [51]. For patients with low-grade MESCC (Bilsky grade 0 or 1), SBRT can be used as a definitive therapy instead of an additional surgery [1,21,40]. However, surgical treatment still plays a critical role for patients with high-grade MESCC (Bilsky grade 2 or 3) (Table 1) [21,52,53]. Radiation therapy needs time for the tumor to regress, which may not provide prompt spinal cord decompression. In contrast, surgical treatment enables direct decompression of neural elements, preventing deterioration of neurological function [51].

Table 1

Bilsky grade for the assessment of epidural spinal cord compression

Nevertheless, SBRT helped change the surgical strategy even when initial surgical intervention is indicated [6,54]. The goals of surgery shifted from aggressive surgical resections to creating a safe space between the metastatic tumor and adjacent neural structures [20,21]. Separation surgery refers to a posterior decompressive procedure that restores the cerebrospinal fluid space around the spinal cord by resecting epidural tumor tissue without attempting gross total tumor removal [21,55,56]. The goal is to create a safe margin between the tumor and the thecal sac, enabling the delivery of high-dose SBRT with reduced risk of radiation-induced myelopathy [21]. Surgery is followed by high-dose SBRT to provide further tumor regression and local control [2,9]. While separation surgery may increase the efficacy of radiation therapy by allowing treatment with higher dosage without increased risk of complications, current literature does not support that separation surgery affects the radiosensitivity of MSC [54,57].

Surgical procedures for separation surgery include (1) instrumentation of levels above and below the level that needs separation surgery to address iatrogenic instability; (2) decompression of thecal sac by removing the lamina; (3) removal of facet joints and pedicles of levels that need tumor resection to enable safe access to ventral epidural space and vertebral body; and (4) resection of tumor tissues located lateral and anterior to dura to achieve separation of thecal sac from MSC lesions [55]. Separation surgery typically involves a single posterior approach, eliminating the need for additional transcavitary procedures [18]. Furthermore, it has a minimal risk of bleeding-related complications since aggressive anterior debulking is not performed with separation surgery [6]. When paired with postoperative SBRT, separation surgery achieves a reported 1-year local control rate of 84% [56,58]. Significant improvements in local control were seen when patients with high-grade preoperative epidural disease were surgically downgraded to Bilsky grade 0–1 [7,59]. Amelink et al. [6] demonstrated the safety and efficacy of separation surgery by directly comparing it to the corpectomy group. The study found that separation surgery resulted in lower blood loss and shorter operating times compared to corpectomy combined with anterior reconstruction. Local control, survival, and reoperation rates were similar between the two groups [6].

Optimizing patient selection

Optimizing surgical indications is crucial for patients with MSC who are prone to complications. Patients presenting with neurological compromise and severe mechanical pain that prevents mobilization are considered candidates for surgery at initial presentation (Fig. 1) [2]. However, for ambulatory patients without neurologic deficits, radiation therapy is often chosen as the initial therapy, given the morbidity associated with surgery [4]. A systematic review found no additional benefit in ambulatory status or local control rates with surgery for patients who remain ambulatory despite MESCC [4]. Although SBRT offers high local control rates, there remains a risk of clinical deterioration due to disease progression [60].

Fig. 1

Illustrative case 1. A 69-year-old male previously treated for a neuroectodermal tumor presented with bilateral lower leg weakness for 5 days. His initial lower leg motor grade was 2 throughout. (A) Sagittal magnetic resonance imaging (MRI) revealed a metastatic tumor at the T10 level. (B) MRI showed grade 3 epidural cord compression. (C) Separation surgery with posterior fusion from T8–L2 level was performed. After surgery he gradually recovered strength of lower extremity and was able to ambulate within a week. (D) Postoperative stereotactic radiation therapy was administered. (E) Postoperative MRI taken 2 years after surgery demonstrates that the spinal cord is free from compression by metastatic mass. The patient was able to ambulate using a cane.

Radiosensitivity is a critical factor in clinical decision-making [19,28,60]. Radiosensitive tumors, such as hematologic malignancies (multiple myeloma, lymphoma), prostate cancer, and breast cancer, often respond well to SBRT alone, even in cases of moderate to high-grade epidural compression [28]. In contrast, radioresistant tumors, such as renal cell carcinoma, colorectal cancer, and non-small cell lung cancer, exhibit significantly lower local control rates with SBRT alone [28,60]. Bernard et al. [28] found that the 1-year local failure rate after SBRT was significantly higher in patients with radioresistant histologies (30.4%) compared to radiosensitive ones (8.0%) (p=0.008). For radioresistant tumors, adding surgical decompression via separation surgery may allow for higher radiation doses with reduced toxicity risk, potentially improving local control, even in previously irradiated or refractory tumors [57,59]. Additionally, certain tumor types, such as renal cell carcinoma, are often hypervascular and prone to hemorrhage, complicating radiation planning and favoring pre-radiation surgical intervention [61–63]. Moreover, the extent of epidural spinal cord compression and neurological deficit is closely linked to tumor biology [19,60]. In particular, neurologic deficits caused by high-grade compression in the context of a radioresistant tumor often require prompt surgical intervention to prevent irreversible functional decline [21,59,64]. Therefore, the tumor histology should be considered alongside mechanical instability, neurologic status, and prior treatments when deciding between surgery and radiation therapy alone [60,65,66].

Decision-making frameworks like neurologic, oncologic, mechanical, systemic and LMNOP (location, mechanical, neurologic, oncologic, response to previous treatment) can guide clinical decisions [64,65,67]. Mechanical instability, considered a relative indication for surgery, can be objectively assessed using the Spine Instability Neoplastic Score (SINS) system (Table 2, Fig. 2) [68,69]. While high-grade MESCC (Bilsky grades 2 and 3) carries a high risk of neurologic compromise, further investigation is required to determine whether prognosis differs between grades 2 and 3 [52]. Furthermore, although the introduction of SBRT has reduced the importance of histology and radiosensitivity of primary tumor in MSC treatment, radiosensitivity remains a consideration for surgical decisions [19]. Highly radiosensitive tumors, such as hematologic malignancies, may still benefit from radiation despite initial high-grade cord compression [19]. Di Perna et al. [54] proposed a novel neurology-stability-epidural compression assessment (NSE) scoring system to address the need for surgical treatment. The system considers neurologic status (score 0–5), performance status, stability as assessed by SINS (score 0–5), and epidural compression assessed by Bilsky grade (score 0–3) [54]. Surgery is recommended for patients with an NSE score exceeding 5 (Fig. 3) [54]. The study suggests that the NSE system can reliably predict the need for surgery and identify patients who would benefit most from surgical treatment [54]. Given that the factors included in the NSE system are critical for assessing the need for initial surgical treatment, its adoption in clinical practice appears both practical and reasonable (Table 3) [70].

Table 2

Spinal instability neoplastic score

Fig. 2

A 50-year-old female with a history of hepatocellular carcinoma presented with neck pain. (A) Initial magnetic resonance imaging (MRI)revealed a metastatic lesion at the C2 level. Shortly after MRI evaluation, the patient complained of severe neck pain which prevented her from sitting or standing. (B) A pathologic fracture of C2 was noted on lateral cervical spine radiograph. Her Spine Instability Neoplastic Score was 14 (location: 3, pain: 3, bone lesion: 2, alignment:4, collapse, 2, and posterolateral involvement: 0), indicating spinal instability. (C) Due to intolerable pain, reduction was performed via traction using a Gardner-Wells device. (D) Occiput–C4 fusion was performed for stabilization.

Fig. 3

Illustrative case 3. A 73-year-old male with a history of renal cell carcinoma presented with severe back pain with subjective gait disturbance. (A) Preoperative sagittal magnetic resonance imaging (MRI) demonstrated a pathologic fracture at the T10 level. (B) Axial MRI revealed grade 3 epidural cord compression. His Neurology-Stability-Epidural score was 8 (neurology: 0, stability: 3, epidural: 5), indicating surgical intervention. (C) Separation surgery with posterior fixation of T8–L2 levels was performed. (D) Postoperative MRI taken 1 month after surgery demonstrates complete separation of tumor from the spinal cord. (E) Postoperative stereotactic body radiation therapy was administered. (F) A follow-up MRI 4 years after surgery showed no compression at the treated level. The patient remained ambulatory with minimal back pain.

Table 3

Neurology-Stability-Epidural score

Enhancing safety of separation surgery

Separation surgery reduces the risk of bleeding-related complications by avoiding aggressive tumor debulking, but hypervascular tumors still pose a risk [66]. Preoperative embolization can effectively decrease intraoperative bleeding during separation surgery [61,66]. The use of preoperative embolization for MSC is increasingly being adopted, especially among patients with renal cancer [62]. Additionally, preoperative intravenous tranexamic acid has been shown to reduce intraoperative bleeding [5], but its potential to increase thromboembolic risk is a concern, especially in high-risk MSC patients [5]. However, Pennington et al. [5] found that limiting tranexamic acid dose to <20 mg/kg does not increase the risk of thromboembolism. Zhang et al. [63] developed a predictive model for postoperative thromboembolism in patients with MSC, which can help identify patients suitable for tranexamic acid use. Cell-saver salvaged blood can reduce transfusion requirements during MSC surgery [71]. Although concerns about reinfusing malignant cells exist, recent literature suggests that salvaged blood transfusion does not increase the risk of clinically detectable metastasis [71]. Evidence supports the clinical safety of salvaged blood in oncologic surgeries, with post-filtration samples showing only fragmented cytoplasmic debris [71]. However, potential complications of intraoperative cell-saver, such as electrolyte disturbances, air or fluid embolism, and dilutional coagulopathy, which can lead to multiorgan failure, should be carefully considered [71].

Minimally invasive separation surgery is gaining popularity due to its similar clinical efficacy and lower complication rates compared to open surgery [8,23]. This approach involves inserting pedicle screws through small incisions [9,72], and achieving decompression or separation via a small midline posterior incision or tubular retractor system [9,72]. Benefits include reduced wound complications, blood loss, length of hospital stay, and transfusion rates [23]. Echt et al. [8] suggested that faster wound healing with this approach can help shorten the interval between surgery and SBRT. However, limitations of the minimally invasive approach include potential incomplete thecal sac decompression due to limited visibility and its unsuitability for long-level surgeries [8].

Stereotactic navigation is increasingly used in minimally invasive separation surgery, providing valuable intraoperative guidance [9]. It enhances anatomical understanding through limited incisions and exposure, minimizes radiation exposure [2,9,20,73], and allows verification of complete ventral decompression via intraoperative CT scanning [9,74]. Navigation also enables accurate placement of instruments [12]. While pedicle screw insertion is generally straightforward for experienced surgeons, anatomical distortions caused by osteolysis and compression fractures may complicate spinal instrumentation [9].

Instrumentation failure can occur after separation surgery due to poor bone quality caused by tumor-related osteolysis and osteoporosis [75,76]. Fenestrated screws with poly-methylmethacrylate cement augmentation improve fixation in poor quality bone by allowing cement to interdigitate with the screw and vertebral body, enhancing pullout strength and reinforcing vertebral body architecture [21,75,76]. Newman et al. [75] reported low failure rates for cement-augmented constructs: 2.8% for long-segment fixations and 2.2% for short-segment fixations.

Wound complications are a common postoperative issue in MSC surgeries, with varying incidence rates [2,4,77]. Perioperative radiation therapy is a risk factor for wound complications, but the optimal interval between radiation therapy and surgery is debated [55]. However, a 2–3 week interval between surgery and postoperative radiation is generally recommended [78]. Minimally invasive surgery has been shown to reduce the risk of wound complications compared to open surgery [75]. Poor nutritional status, diabetes mellitus, smoking, adjuvant radiation therapy/chemotherapy are known risk factors for surgical site infection [79]. Given the significantly higher rate of surgical site infection in patients undergoing MSC surgery, compared to other spine surgeries, identifying and managing modifiable factors preoperatively is crucial to reduce the risks [79]. Combined treatment with plastic surgery can reduce the risk of wound complications. Franck et al. [80] reported that routine use of local muscle flap closure significantly decreased the wound complication rate after MSC surgery. Hersh et al. [81] also found that the use of local paraspinous advancement flap after spine surgery for MSC significantly decreased wound complications.

Improving the efficacy of separation surgery

While separation surgery followed by postoperative SBRT is effective for MSC, not all patients undergo these procedures routinely [16,82]. Real-world studies show that only 21%–48% of patients receive postoperative radiation therapy [16,60,83]. Delayed radiation therapy increases the risk of local progression and decreased overall survival [84]. Achieving timely postoperative radiation therapy can be challenging due to patients’ poor general health status [20]. Despite SBRT’s superior local control, radiation oncologists often opt for EBRT due to challenges with SBRT [16], including the complexity of radiation planning due to spine instrumentation, increased risk of radiation myelopathy from preoperative radiation [25], and verifying adequate tumor-thecal sac separation [16]. Postoperative MRI and intraoperative sonography can help assess the quality of separation, but further verification may be required [16]. Close communication between the surgeon and radiation oncologist, starting preoperatively, is crucial to optimize postoperative radiation therapy and modality [60].

ERAS protocol for patients with MSC can lead to improved short-term patient recovery [22,85–87]. ERAS programs involve an evidence-based multidisciplinary approach to improve perioperative outcomes [86]. Given the frailty and comorbidities common in MSC patients, ERAS protocols would be particularly beneficial in facilitating postoperative recovery and optimizing outcomes [86]. A study by Chakravarthy et al. [22] found that implementing an ERAS protocol resulted in decreased intraoperative bleeding, less perioperative opioid use, earlier ambulation, shorter urinary catheter use, and shorter hospital stays. Their ERAS protocol included multimodal preemptive pain medication, intraoperative liposomal bupivacaine injection, goal-directed fluid therapy, early initiation of diet, and early ambulation [22].

Comparative efficacy of radiation therapy alone versus surgery combined with radiation

While SBRT has revolutionized local control of MSC, growing evidence shows its efficacy varies significantly with tumor histology, degree of cord compression, and neurologic symptoms. In particular, SBRT achieves excellent local control in patients with low-grade epidural spinal cord compression (Bilsky grade 0–1) and radiosensitive tumors, potentially serving as a standalone treatment [88]. However, for patients with radioresistant histologies or high-grade compression (Bilsky grade 2–3), combining surgery with SBRT yields superior outcomes [88].

Chang et al. [35] reported that SBRT alone achieved a 1-year tumor control rate of 84% in a cohort with varied tumor types, primarily effective in patients with preserved neurological function and minimal epidural involvement. In contrast, Bernard et al. [28] found that SBRT’s efficacy is significantly reduced in radioresistant tumors such as renal cell carcinoma and colorectal cancer, with 1-year failure rates of 30.4% compared to 8.0% in radiosensitive tumors (p=0.008). These findings underscore the need for careful histologic stratification when selecting SBRT monotherapy.

For patients with high-grade MESCC, separation surgery followed by SBRT provides effective local control and neurologic preservation. Laufer et al. [56] reported 1-year local control rates of approximately 84% with separation surgery combined with high-dose SBRT, even in patients with prior radiation or radioresistant disease. Alghamdi et al. [59] found that surgically downgrading Bilsky grade 2 or 3 compression to grade 0 or 1 before SBRT significantly improved local control and reduced retreatment rates.

The landmark trial by Patchell et al. [45] randomized MESCC patients with neurologic deficits to radiation therapy alone or surgery followed by radiation. Surgical patients showed better ambulation preservation (84% vs. 57%), delayed loss of walking ability, and improved functional outcomes [45]. Despite preceding the SBRT era, the study highlighted the importance of mechanical decompression in symptomatic patients [45].

A consolidated comparison of outcomes by treatment modality is summarized in Table 4. Collectively, these data suggest that SBRT alone may be sufficient for select patients with radiosensitive tumors and minimal compression. However, surgery combined with SBRT should be considered in patients with radioresistant tumors, significant neurologic deficits, or high-grade cord compression.

Table 4

Outcomes of SBRT alone and combined surgery with SBRT for spinal metastasis