Stereotactic Radiosurgery

bFGF: basic fibroblast growth factor cEBRT: conventionally fractionated external beam radiation CTLA-4: cytotoxic T-lymphocyte-associated antigen 4 ESCC: epidural spinal cord compression HDR: high-dose rate ISRC: International Spine Radiosurgery Consortium MRI: magnetic resonance imaging NOMS: neurological, oncological, mechanical, and systemic NSCLC: non–small-cell lung PD-1: programmed death-1 PTV: planning target volume SINS: Spinal Instability Neoplastic Scoring system SLITT: spinal laser interstitial thermotherapy SRS: stereotactic radiosurgery 18FDG PET: 18F-fluoro-2-deoxy-d-glucose positron emission tomography SCOPE OF THE PROBLEM In 2016, 1.6 million new cancer diagnoses were made.1 Forty percent of these patients develop metastatic disease to the spine and of those 10% to 20% or roughly 64 to 128 000 develop spinal cord compression.2-4 The risk of a cancer diagnosis increases with age and the segment of the US population older than 65 yr is growing at a faster rate than the general population.5 Targeted therapies, such as biologics and immunomodulators, are being increasingly utilized, but as with cytotoxic agents, are thought to be more effective for visceral and soft tissue disease than bone metastases; thus, despite improved survivals, the expectation is that spine metastases will remain a major source of morbidity. More routine use of magnetic resonance imaging (MRI) and 18F-fluoro-2-deoxy-D-glucose positron emission tomography (18FDG PET) has led to improved detection of spinal metastatic disease. Two-year survival from the time of diagnosis of spinal metastatic disease is 10% to 20%. Taken together, the scope of this problem is greater now than ever before and will likely continue to grow in the future. The goals of treatment for spine metastases are palliative in nature and include maintenance or improvement of neurological function, local tumor control, mechanical stabilization, pain relief, and improvement in quality of life. One major advancement that has impacted, indeed revolutionized, the treatment of spinal metastases over the past 2 decades is stereotactic radiosurgery (SRS). In this article, the authors will review the NOMS (neurological, oncological, mechanical, and systemic) decision-making framework for patients with spinal metastases and the evolution of radiotherapy as a major factor in the paradigm shift driving less aggressive surgery and more radiosurgery in the treatment of this disease. The underlying radiobiology, effectiveness of SRS, and outcome data are also presented. NOMS NOMS is a decision-making framework designed to facilitate and guide therapeutic decisions for patients with spinal metastases (Figure 1, NOMS).6,7 It serves as a template for the analysis of important clinical data and as such is not wedded to any particular treatment or technology. As medical science progresses, so too will the treatment modalities utilized within NOMS. Notably, it was conceived for use in a multidisciplinary fashion in conjunction with pain physicians, interventionists, and with input from radiation, medical, and neuro-oncology. The framework is comprised of 4 major domains from which clinical information is assessed: neurological, oncological, mechanical, and systemic.FIGURE 1: The NOMS framework – Neurological, Oncological, Mechanical and Systemic. Republished with permission of AlphaMed Press, from The NOMS framework: approach to the treatment of spinal metastatic tumors, Laufer, et al,6 permission conveyed through Copyright Clearance Center, Inc.The neurological assessment involves the determination of clinical myelopathy, functional radiculopathy or evidence of cauda equina syndrome, and the degree of radiographic epidural spinal cord compression (ESCC) as has been previously described (Figure 2, ESCC Grading scale).6,7 The ESCC scale has shown good reliability and can enhance reporting and communication on the degree of spinal cord compression in clinical and research applications.8 Within the NOMS framework, the oncological assessment consists of determining the expected histology-specific tumor response to systemic therapies, conventionally fractionated external beam radiation (cEBRT; Table 1, radiotherapy sensitivity) and SRS.9,10FIGURE 2: Epidural Spinal Cord Compression Grading Scale. Reprinted from Surgical Approach to Epidural Spinal Cord Compression,7 Copyright 2006, with permission from Elsevier.TABLE 1: Summary of Expected Radiation Response by HistologyAlthough the NOMS assessments are taken together, the evaluation of mechanical instability as an independent factor prompting surgical intervention even for radiosensitive tumors should be noted.11,12 Because radiotherapy and systemic therapy do not restore mechanical stability of the spine, patients with mechanical instability require stabilization. The mechanical assessment requires discernment of tumor-associated mechanical instability and is facilitated by the Spinal Instability Neoplastic Scoring system (SINS). SINS scoring considers multiple factors important for biomechanical spinal integrity: location, pain, lesion character, radiographic spinal alignment, degree of vertebral body collapse, and the degree of posterolateral spinal element involvement (Table 2, SINS).13,14TABLE 2: SINS System—Spinal Instability Neoplastic ScoreThe last consideration in the evaluation of the patient with metastatic spinal disease is the systemic assessment, which incorporates the patient's overall burden of disease and medical comorbidities as they relate to the patient's capacity to tolerate a surgical intervention, recover adequately, and go on to receive continued systemic therapy. Part and parcel of this assessment is the realization of the complex interplay of tumor biology, molecular markers, and targeted therapies, as well as the knowledge that certain histologies such as colon, hormone-refractory prostate, nonsmall cell lung (NSCLC), and hepatocellular carcinoma are traditionally associated with poor postoperative median survivals.6,15 The recent introduction of targeted therapies has dramatically improved expected survivals for many tumor types, which has necessitated a re-evaluation by both oncologists and spine surgeons in offering aggressive therapy for spine metastases. Tumor histology alone no longer predicts survival requiring an adjustment in who we consider to be appropriate candidates for SRS or surgery. The Tomita and Tokuhashi scoring systems both rely on histology to predict survival and the type of therapy offered for spine metastases.16,17 With the dramatic evolution of systemic therapy, these scoring systems will need to be adjusted to reflect significant changes in survival. For example, in the treatment of melanoma, a single PD-1 inhibitor, Nivolumab, has demonstrated a greater than 2-fold improvement in survival compared to standard chemotherapy.18 cEBRT AND SRS Due to the limited success of laminectomy as first line treatment for malignant cord compression throughout the 1960 to 1980s, cEBRT was adopted as the standard of care for patients with spinal metastases of all histologies.19 With surgical advancement and improved instrumentation, stronger evidence in support of surgical decompression combined with cEBRT became available.20-22 In a recent systematic literature review, the role of cEBRT as compared to SRS in the treatment of spinal metastases was evaluated and a greater understanding of the limitations of cEBRT and the potential benefits of SRS have emerged.23-25 In an analysis of 7 studies exploring the histology-specific treatment response with cEBRT-lymphoma, seminoma, myeloma, breast, and prostate histologies were considered largely radiosensitive while sarcoma, melanoma, gastrointestinal, NSCLC, and renal histologies were considered radioresistant (Table 1, radiotherapy sensitivity).24 Radiosensitive histologies exhibited a median response duration of 11 mo and a 2-yr local control rate of 86%, which compared favorably with radioresistant histologies, which demonstrated a 3-mo median response duration and 2-yr local control rates of 30%.26,27 A major limitation of cEBRT is its inability to deliver high-dose radiation to tumor in a conformal fashion. Rather, it is delivered in 1 or 2 beams of larger field size, typically encompassing vertebral bodies beyond the involved level and incorporating the adjacent spinal cord and other organs at risk in its path. Consequently, neither the conformality of treatment nor the tumor dose are maximized. This is reflected clinically in the lower median response duration consistently seen in radioresistant histologies in the literature.6,24,28 In contrast, radiosurgery can provide conformal high-dose radiation contoured to optimize the gross tumor volume whilst minimizing injury to neural elements. Inherent in the technology, high-dose single fraction conformal radiosurgery can increase the biological effective tumor dose compared to conventional radiotherapy. Accordingly, SRS yields a clinical benefit regardless of histology, providing a more durable symptomatic response and higher local control rate.20,24,25,29,30 Fittingly, cEBRT is reserved for patients with radiosensitive histologies regardless of the degree of ESCC, the rare exception being the patient with high-grade ESCC, rapidly progressive neurological deficit and a favorable systemic assessment (surgical decompression and instrumentation) or mechanical instability (vertebroplasty/kyphoplasty or percutaneous instrumentation). The advent of high-dose per fraction conformal radiation for use in the treatment of patients with spinal metastases is changing the way surgeons and radiation oncologists treat patients with this disease. Advances in SRS have ushered in a new era of treatment. Typical radiosurgical dose schedules are shorter and allow greater cytotoxic tumor dose delivery than cEBRT. Technological advancements including noninvasive immobilization, isocenter verification using cone beam CT, image guidance, and conformal beam shaping have increased the availability, accuracy, and safety of radiotherapy and have implications both for the indications for and the type of surgery offered by surgeons. Dose Matters SRS has the ability to deliver an ablative dose while sparing normal tissue tolerance for metastatic spine tumors in the absence of spinal cord compression, ie, ESCC 0–1c. In a multi-institutional review published in 2014, Guckenberger et al20 reviewed 387 cases comprised of various solid tumor histologies with multiple dose fractionation schemes. The median dose was 8 Gy × 3 fractions and the mean follow-up was 11.8 mo. Local control at 24 and 48 mo was 84% and 62%, respectively.20 In a large dataset based on a single-center institutional experience from Memorial Sloan Kettering Cancer Center,31 811 spinal tumors treated with single-fraction SRS were evaluated. The cohort was comprised of 82% radioresistant histologies and median follow-up was 26.9 mo. Contouring was in accordance with the International Spine Radiosurgery Consortium (ISRC) guidelines.31 The prescription dose ranged from 18 to 26 Gy in a single fraction and for analysis dosing was considered as a continuous variable. The median planning target volume (PTV) D95, that is the median dose that covers 95% of the PTV, in the low-dose group was 1644 cGy compared to 2240cGy in the high-dose group. Local failure rates for the low-dose and high-dose groups, respectively, were 5% vs 0.41% at 12 mo, 15% vs 1.6% at 24 mo, and 20% vs 2.1% at 48 mo. This study demonstrates the significant local control advantage imparted by high-dose single-fraction SRS.29 Importantly, local control rates at 2 yr approached 98% independent of tumor histology, underscoring further the importance of high-dose conformal radiation.29 Currently, most centers do not use SRS alone to treat high-grade ESCC (ie, ESCC 2 or 3) based on a number of factors. Lovelock et al32 performed a dose-failure analysis of patients undergoing SRS. This paper identified that all failures occurred in patients receiving less than 15 Gy to even a small percentage of the PTV. Given a cord Dmax of 14 Gy with a 10% per millimeter falloff in radiation dose, there is a high probability of either underdosing at the margin of the spinal cord leading to epidural progression or surpassing the spinal cord constraint causing myelopathy.32 Secondarily, the resolution of epidural disease resulting from radioresistant tumors can take months resulting in ongoing spinal cord compression. Finally, in a single-center series of 62 patients receiving SRS alone without initial surgical decompression for high-grade ESCC, 9/62 patients (16%) had short term neurological progression, for which 5 patients (8%) went on to receive urgent surgical decompression, while surgical decompression prior to SRS generally prevents neurological deterioration.33 Separation Surgery Although SRS cannot be used in the setting of high-grade ESCC for radio-resistant tumors, it has become a very effective postoperative adjuvant. Using cEBRT in the postoperative setting resulted in recurrence rates as high as 70% at 1-yr despite gross total resection in many patients.28 With the integration of SRS, the surgical goals of spinal cord decompression and stabilization remain the same, but tumor control is predicated on the SRS response rather than the extent of tumor resection. With high enough single fraction or hypofractionated SRS (range 18-27Gy), ablative doses of radiation may be delivered to cancer histologies previously thought to be radioresistant (Table 1), obtaining 1- to 4-yr local progression free survival rates between 84% and 98%.24,25,29,30,34 These high rates of radiation-dependent local control afford the surgeon the ability to be less aggressive with surgery, decompressing the neural elements from tumor encroachment only as needed to create a target for SRS. Separation surgery, a technique whereby a posterolateral approach is utilized to obtain ventrolateral access to nerve roots, posterior longitudinal ligament, and, most importantly, ventral epidural disease has grown in popularity. No attempt is made to resect the tumor-infiltrated vertebral body or paraspinal disease: thus, anterior column reconstruction is rarely indicated. Posterolateral screw-rod fixation is performed in all patients (Figure 3). Circumferential decompression and reconstitution of the thecal sac is achieved such that high-dose single-fraction or hypofractionated radiosurgery can be given to the remaining disease. Separation surgery in a sense functions as a neoadjuvant treatment prior to more definitive radiation therapy in patients without a neurological deficit/symptomatic ESCC and continues to be the mainstay of treatment in the patient with rapidly progressive neurological deficit/symptomatic ESCC as rapid decompression of the spinal cord optimizes the restoration of neurological function.FIGURE 3: Diagram demonstrating circumferential spinal cord decompression from a posterior approach after pedicle screw fixation. © 2017 Ran Xu.In a recent single-center review of 186 patients undergoing separation surgery followed by SRS, 73% of patients had high-grade ESCC and 77% had radioresistant histology. In this cohort, 49% had failed prior radiotherapy. Of the 3 radiosurgical treatment strategies employed, patients receiving single fraction 24 Gy and high-dose hypofractionated 8 Gy × 3 fraction dose schedules had 1-yr recurrence rates of 9% and 4.1%, respectively.30 There was no impact of histology, prior failed radiation, or the degree of preoperative epidural extension on recurrence rates and no patient suffered a neurological complication. A major criticism of separation surgery is that failure to formally reconstruct the anterior column will result in hardware failure. In an analysis of 318 patients who underwent separation surgery for solid malignancies and stand-alone posterior instrumented stabilization, 9/318 patients (2.8%) experienced symptomatic hardware failure identified as rod or screw fracture, screw pull out, or significant vertebral body fracture.35 Risk factors for hardware failure were identified as junctional spine location and post-menopausal status. Further, in patients receiving preoperative cEBRT, the incidence of postoperative wound infection can be significant.36 There is no increased risk and, in fact, there is a trend towards less postoperative infection risk in patients who receive SRS prior to surgery as compared to patients who received cEBRT.36 Radiobiology of Single Fraction or Hypofractionated Radiation The improved local tumor control using high-dose single or hypofractionated SRS as either definitive or postoperative therapy is thought to be reflective of differences in radiobiology compared to conventional external beam radiation. Ionizing radiation can create a number of cellular aberrancies, amongst them double-stranded breaks within DNA.37,38 Double stranded breaks are recognized triggers for DNA repair processes, but also cell death via various cytotoxic cascades.37 In one pathway, SRS induces exocytosis of secretory lysosomes containing acid sphingomyelinase, an enzyme that catalyzes the conversion of extracellular membrane bound sphingomyelin to ceramide, a key component of a signal transduction platform that culminates in microvascular endothelial dysfunction and apoptosis, subsequent hypoperfusion of tumor tissue and ultimately tumor cell destruction.39 This route of radiation-induced microvascular dysfunction can be modulated with vascular endothelial growth factor receptor inhibitors to enhance the effects of high-dose single fraction radiotherapy on tumor growth delay and complete response rate in mouse models of sarcoma and melanoma.40 In a mouse model of whole body radiotherapy, cell death in the gastrointestinal lining of mice undergoing single fraction radiotherapy was assessed. Basic fibroblast growth factor (bFGF) was able to rescue endothelial cells from radiation-induced apoptosis up to doses of 15 Gy, thereby limiting gastrointestinal cell death, avoiding death in animal subjects. At doses of 17 to 18 Gy in a single fraction the authors’ findings suggest a switch to an endothelial cell-independent mechanism of gastrointestinal damage, potentially identifying a major difference in the radiobiology of high-dose per fraction radiation in comparison to lower dose alternatives.41 With the rise of biologics and immune modulators in cancer care has come a growing recognition of the complementary biology of radiotherapy and immune inhibitors T-lymphocyte-associated antigen 4 and death-1 single fraction or hypofractionated radiotherapy has been shown to tumor cell increase of antigen to cells and of tumor resulting from SRS the immune system and therapy tumor the immune system is and in this a systemic tumor the can be seen local high-dose radiation in with or PD-1 for high-dose radiotherapy be with the knowledge that effects to including the vertebral and spinal cord have been A number of centers have to dose for all organs at risk in to A is seen in Table One for tumor control is of the can the or a more favorable target for ablative SRS within normal tissue 3: A of for at Risk with and patients should take in the knowledge that spinal metastases are being increasingly care overall and cancer care in particular has led to increased survival amongst patients with There has been a trend towards minimizing the of surgical intervention in of less such as and continues in of dose reporting and can to effective for more patients as a will for a more data for research As of for target in SRS and postoperative SRS have been published by the International Spine Radiosurgery for the assessment of tumor response SRS, including have been as a way of reporting in clinical Separation surgery, the technique of a posterolateral surgical to circumferential tumor decompression of the neural elements and spinal fixation first described yr increases its in the treatment of spinal even more of the are being to spinal stabilization including and more percutaneous pedicle screw fixation with or without decompression are in routine As an to separation surgery in clinically appropriate patients both and high-dose rate interstitial with have been shown to be and effective In a cohort of spinal laser interstitial thermotherapy pain and improved quality of at 3 mo without of systemic At high-dose conformal radiation is being utilized for treatment of spine and with 2-yr local control rates of In the treatment of spinal metastases has over the past 2 decades to provide treatment to and continue to be and can potentially be to spinal tumor research in this field can have a impact on The authors have no or institutional in any of the or described in this