Evolution of surgical treatment of metastatic spine tumors

The treatment of cancer has transformed over the past 40 years, with medical oncologists, radiation oncologists and surgeons working together to prolong survival times and minimize treatment related morbidity. With each advancement, the risk–benefit scale has been calibrated to provide an accurate assessment of surgical hazard. The goal of this review is to look back at how the role of surgery has evolved with each new medical advance, and to explore the role of surgeons in the future of cancer care. A literature review was conducted, highlighting the key papers guiding surgical management of spinal metastatic lesions. The roles of surgery, medical therapy, and radiation have evolved over the past 40 years, with new advances requiring complex multidisciplinary care.


Defining the role of surgery
In the 1980s, the role of surgery in the management of spinal metastases was poorly defined. While limited case series supported surgical intervention, these retrospective reports included various pathologies and non-standardized techniques [1][2][3][4][5][6][7][8]. Oncologic regimens demonstrated variable efficacy, further impacting the measured utility of surgical excision. There was debate in the medical literature regarding the advantage of surgery over radiation alone when reflecting on pooled patient outcomes [9]. In 1990, the Tokuhashi scoring system was proposed to guide surgical decision making by linking patient pathology and prognosis to treatment [10]. Even with these advances, given the nature of the studies and selection bias, the role of surgery was not fully assessed.
In 2005, a prospective, randomized clinical trial compared conventional radiation to surgery plus radiation in patients with metastatic spinal disease treated from 1992 to 2002 [11]. The Patchell trial definitively concluded that surgical decompression and stabilization preserved ambulatory status and prolonged survival for patients with epidural spinal cord compression and neurological deficits for less than 48 h. In addition, steroid and opioid pain medication use was reduced in the surgical group, and the study was terminated early due to superiority of surgical treatment [11]. Further, surgery did not significantly increase hospital length of stay. Of note, patients with greater than 48 h of paraplegia were excluded from this trial, as were patients with myeloma or lymphoma and those with cauda equina compression alone.
While the Patchell study concluded that surgery plus radiation was superior to radiation alone, this was not the end of the story. The precise type of surgical intervention was still not clear, as the trial inclusion criteria allowed the surgeon to tailor his or her approach to meet each patient's specific needs, i.e. laminectomy with or without instrumented fusion. The next step in the evolution of multidisciplinary management was, therefore, to further refine the optimal surgical intervention for patients with metastatic spinal disease.
It was felt that some of the reduced pain medication use among patients in the surgical arm of the Patchell study was likely attributable to stabilization of cancer-related vertebral column instability. The Spine Oncology Study Group (SOSG) proposed the Spinal Instability Neoplastic Score 1 3 (SINS) in 2010 following the Patchell study, as a way of identifying patients likely to respond to operative fixation in addition to decompression for metastatic lesions. The SINS system detailed the patient's pain, degree of vertebral body collapse, lytic versus blastic characteristics of the tumor, posterior element involvement, level of disease, and presence of kyphosis or translation related to the level of interest [12]. Patients were classified as stable, indeterminate or unstable, which guided the use of instrumentation. With this scoring system, spine surgeons were given a framework for how to balance the risk and expense of instrumentation with the benefit of decreased pain and improved stability.
Other surgical considerations not captured by the SINS scoring system or the Patchell study include the vascularity of the tumor and minimally invasive options. The use of preoperative embolization can be considered for highly vascular metastatic lesions to reduce surgical risks. These pathologies include renal cell carcinoma, follicular thyroid cancer, hepatocellular cancer, pheochromocytoma, melanoma, hemangiopericytoma and some sarcomas. Adjuncts such as kyphoplasty have been added to facilitate shorter segment fusions [1][2][3]. Laser interstitial thermal therapy (LITT) [13][14][15][16] and radiofrequency ablation [17] have also been studied to address metastatic lesions with less morbidity than open cytoreductive approaches. While promising technologies, implementation of LITT requires MRI compatible probes and dedicated time in the MRI suite, dedicated software and training in MRI thermometry. Future prospective trials and commercialization of this applied science will likely guide adoption of these techniques.

Advancements in radiation therapy
As surgical decision making has become increasingly protocol-driven, the role of radiation therapy has expanded in parallel. Conventional external beam radiation therapy (EBRT), the standard therapy used in the Patchell clinical trial, has been supplanted by newer technologies. Historically, cancer pathologies were divided into very radiation sensitive, moderately sensitive and radioresistant groups, relative to EBRT. Very sensitive lesions included lymphoma, myeloma, and small cell lung cancer (SCLC). Sensitive lesions were breast, prostate, and thyroid pathologies, while radioresistant lesions were colon, renal (RCC), melanoma, sarcoma, and non-small cell lung cancers (NSCLC; squamous and adenocarcinoma). The invention of stereotactic radiosurgery (SRS) challenged the sensitive versus resistant classification system, as higher doses of targeted radiation were delivered to index levels of disease in fewer treatments, effectively treating tumors that were previously deemed radioresistant [18,19]. Using SRS limited the damage to surrounding tissues and decreased the length of the treatment time [19,20]. Further, SRS provided durable local control rates and improved patient reported quality outcomes post treatment [21][22][23][24]. Protocols were developed to administer EBRT and SRS together for the same patient, with careful monitoring of spinal cord toxicity [25].
Given this expansion of radiation protocols, minimally invasive techniques for surgical decompression gained wider adoption. Prior to SRS, the only effective means to address epidural cord compression was direct surgical excision in patients with radioresistant pathologies. Ventral epidural compression was particularly challenging, as partial and complete corpectomies increased the complexity of the surgical approach. SRS gave spine surgeons a new tool to treat ventral epidural compression for an expanded list of tumor types, and the role of surgery expanded to facilitate a safe radiation treatment field and distance between the tumor and the spinal cord [21-23, 26, 27]. This hybrid approach provided durable local control rates for a variety of pathologies [22,26,27]. An accompanying grading scale, the Bilsky cord compression scale, was developed to create a uniform system for assessing the degree of epidural extension of a spinal cord lesion, and to determine eligibility for separation surgery [26,28].
With wider use of separation surgery, the limitations of this approach were discovered. Radiation induced myelopathy to the spinal cord can occur when maximal safe doses are exceeded, and patients with recurrence at an index level cannot have unlimited, repeated radiation treatment [29]. This has complicated multidisciplinary decision making, particularly when a local recurrence occurs. Surgeons and radiation oncologists alike have struggled with when to offer limited surgical resection with separation surgery for projected long-term survivors versus a more aggressive, cytoreductive surgical approach. Furthermore, while studies are limited and of low-quality evidence regarding postoperative wound healing, preoperative radiation is known to compromise skin integrity and disrupt natural tissue planes [30]. Radiation can also impact rates of postoperative fusion in patients with instrumentation [31]. Data is limited, however retrospective series suggest that targeted SRS poses lower risk than EBRT [31]. Radiation can also accelerate the development of vertebral compression fractures (VCF) in patients with lytic lesions or prior EBRT, and the risks may be higher when higher SRS doses are used at a single level [32][33][34]. Kyphoplasty may be used as an adjunct to restore vertebral body height in patients with spinal metastases at risk of VCF, further highlighting the necessity of complex, multidisciplinary treatment planning [35][36][37].
In an effort synthesize the rapidly evolving treatment recommendations, the NOMS framework was proposed to incorporate the patient's neurologic, oncologic, metastatic and systemic disease status plus radiation considerations into the presurgical evaluation process [38,39]. The NOMS framework suggests when to offer separation surgery for metastatic lesions, and incorporates the SINS scoring system to consider when instrumented fusion will be required. This scoring system adds complexity to earlier protocols by considering the nuances of radioresistant versus radiosensitive pathologies when EBRT is offered as well as the safety profile of radiation treatment fields.
Prognostically, pathology-specific actuarial scales are widely available to help surgeons determine estimated survival [40-46]. As patients are surviving longer with metastatic disease, these scoring systems will continue to evolve to reflect treatment options and outcomes data for cancer patients. The NOMS framework has been modified recently to include a molecular typing category, reflecting the impact of targeted therapeutics on prognosis and treatment planning [18]. Figure 1 provides a case example that illustrates these concepts. The patient is a long-term survivor of metastatic RCC and has undergone SRS, separation surgery and fusion over 20 years battling the disease.

Molecular therapies
The status of molecular therapeutics for melanoma, NSCLC and RCC is of particular interest for spine surgeons. Patients with these primary tumor types have historically demonstrated poor survival due to poorly controlled systemic disease and radio resistance. For hormone sensitive tumors, specifically breast and prostate cancer, selective estrogen receptor modulators and androgen deprivation, respectively, are utilized to treat a subset of patients. Hormonal therapies have been a key driver of long-term survival in these diseases. Discussion of all chemotherapy options for these pathologies is beyond the scope of this review, however targeted therapy for breast cancer will be included, as recent molecular profiling has uncovered new targets for therapy [47][48][49][50][51][52][53][54]. While this review will describe the status of targeted therapies for common metastatic pathologies, it is critical to recall that primary and metastatic tumors from the same patient may not share the same genetic mutations. For example, epidermal growth factor receptor (EGFR) mutations seen on biopsy in patients with non-small cell lung cancer may not correlate with an increased response of spinal metastatic lesions to EGFR-tyrosine kinase inhibitors (EGFR-TKI) [55]. Further, some targeted therapies have offtarget effects, the most serious of which is treatment-related death (TRD) [56].

Breast cancer
Breast cancer can be characterized by estrogen (ER) and progesterone (PR) receptor, human epidermal growth factor receptor 2 (HER2) and androgen receptor positivity. Patients are also routinely tested for the presence of breast cancer genes (BRCA1/2) to determine the risk of developing ovarian cancer. Targeted therapies have been explored in clinical trials for patients stratified by genetic markers and receptor positivity [47][48][49][50][51][52][53][54]. Patients with ER/PR/HER2 negative (triple negative) metastatic breast cancer patients are more likely to have recurrent or metastatic disease [57].

Melanoma
Historically, melanoma metastatic to the spine was associated with a < 4 month average survival [58]. This was due to the resistance of metastatic melanoma to chemotherapy, and the ability of this cancer to phenotype switch, acquiring novel genetic mutations to facilitate survival in the face of systemic threats such as hypoxia. In recent years, interleukin-2 as well as drugs targeting PD-1, PD-L1, CTLA-4 and BRAF mutations have demonstrated efficacy in local control rates and overall survival [59-63]. These drugs have been investigated in small groups of patients with metastatic melanoma in the spine, and have demonstrable efficacy with CNS disease [63] as well as 40-50% response rates in patients with metastatic melanoma [62,64]. For a subset of patients with BRAF mutations within their cancer, BRAF inhibitors such as vemurafenib reduce the risk of death compared with traditional chemotherapy, and improve overall survival [65,66]. These survival advantages are reported for visceral metastatic melanoma and may not be directly applicable to patients with spinal lesions. Further studies are ongoing comparing combinations of these medications to develop the ideal treatment regimen for patients with metastatic melanoma.
When these therapies are combined with SRS, the response rates in spinal disease are even higher. Specifically, IL-2 therapy coupled with SRS yielded a 62.5% response rate in a small group of patients with metastatic melanoma in the spine [60]. In a larger cohort study of patients with metastatic spinal lesions undergoing separation surgery, nine patients with metastatic melanoma demonstrated local recurrence rates under 10% at follow-up [22]. Further, a phenomenon known as the abscopal effect may play a role in improved survival and regression of metastatic disease following radiotherapy [67]. This phenomenon has been seen in patients receiving anti CTLA-4 drugs (ipilimumab) with metastatic melanoma, and may be related to immunemediated tumor regression at sites distal to the radiation treatment field [67].

Lung
Lung cancers are broadly grouped into two categories: small cell lung cancer (SCLC) and non small lung cancer (NSCLC) due to the clinically distinct natures of these two categories. When taken as a monolithic group, the survival for patients with lung cancer metastases to the spine following surgical intervention is quoted in contemporary literature at 2-7.5 months, depending on the type of chemoradiation used to treat the disease [55, 68,69]. However, grouping these pathologies together often complicates the interpretation of outcomes. Modern studies have attempted to delineate both pathological and genotypic subtypes to better understand disease course and prognosis [70]. More modern studies suggest SCLC with spinal metastases tend to have worse median survival than NSCLC (2 vs 5mo) [71]. With respect to treatment modalities, SCLC has historically been considered chemoradiation sensitive, while NSCLC is more resistant to conventional radiation therapy. Even with the limited life expectancy, there is strong data to support surgical intervention to maintain patient quality of life and ambulation [1] Targeted therapies are available for NSCLC, with new trials emerging in recent years. Pembrolizumab is a novel PD-1 inhibitor that has demonstrable efficacy against metastatic NSCLC [72]. EGFR mutations have also been noted within a subset of lung cancers, and EGFR-TKI have also been used to treat patients with metastatic spinal lesions in lung cancer, with mixed results. In a study of 65 patients who underwent surgery for NSCLC metastases to the spine, 35 patients were given the TKI gefitinib or erlotinib in addition to radiation therapy [55]. Their outcomes were compared to patients receiving traditional chemotherapy, surgery and radiation therapy. No significant differences were seen in survival between these groups, however the authors did report evidence of body regeneration postoperatively in both groups, indicative of successful treatment to the vertebral bodies [55].

Renal
Over the past 40 years, drug development and clinical trial and error have led to a sustained improvement in outcomes for patients with metastatic renal cell carcinoma [73,74]. 20-60% of patients with locally advanced renal cell carcinoma (RCC) develop metastases, and patients are grouped into good, intermediate and high risk pools to determine their prognosis and likely response to molecular therapies [75,76]. Patients with RCC spinal metastases have wide range in reported survival that reflects patient Tokuhashi scores, from 5 to 32.9 months, according to one recent review [77]. Currently, nivolumab (PD-1 inhibitor) with or without ipilimumab (CTLA-4 targeted therapy) is the most commonly studied treatment regimen [78][79][80], with vascular endothelial growth factor receptor (VEGFR) TKIs and mammalian target of rapamycin (mTOR) inhibitors for additional targeted therapy [76,81]. Future studies are needed to determine the optimal combination of drugs and their timing in order to minimize side effects [82,83]. Targeted therapies like VEGFR-TKIs and mTORis can prolong patient survival, but the prevalence of treatment-related adverse effects is high; up to 98% in one study including 51% of patients with grade 3-4 events [81].

Conclusion
Scientific and technological advances have improved the management of spinal metastatic lesions over the past 40 years. After the publication of the Patchell trial, surgery was established as a crucial part of metastatic cancer care. In the past 10 years, the scope of surgery has been further refined to reflect a patient's symptoms, pathology, and prognosis. Throughout this evolution, surgery has remained the most reliable option for emergent decompression. Despite advances in radiation therapy, SRS with separation surgery requires surgical intervention to be effective in cases of highgrade epidural compression. Molecular targeted therapeutics are emerging for systemic disease, and may supplant invasive surgical procedures, but the optimal combinatorial regimens remain unknown. For this reason, surgeons must remain engaged in medical advancements to guide decisionmaking and to take the best possible care of patients.
Author contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by PZS and ZLG. The first draft of the manuscript was written by PZS and SS and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
Data availability Data sharing not applicable to this article as no datasets were generated or analysed during the current study.