TURKISH JOURNAL OF ONCOLOGY

Summary

OBJECTIVE
To evaluate the dosimetric characteristics and clinical outcomes of metastasis-directed therapy (MDT) using stereotactic body radiotherapy (SBRT) for contiguous vertebral metastases in patients with oligometastatic disease.

METHODS
Clinical and dosimetric data from 58 patients (64 lesions) treated between December 2012 and December 2020 were retrospectively reviewed. All patients had ≤5 metastases at diagnosis or during follow-up and were treated with single-fraction SBRT. Primary endpoints were local control (LC) and progression-free survival (PFS), and the secondary endpoint was overall survival (OS).

RESULTS
Mean clinical tumor volume was 27.55 cc, mean spinal cord volume was 4.41 cc, and the mean maximum spinal cord dose was 11.68 Gy. Median age was 54 years and median follow-up was 12 months; 83% of patients had died at last follow-up. Two-year OS and PFS rates were 65% and 45%, while 1- and 2-year LC rates were 81% and 70%. Local recurrence occurred in 17% of patients at a median of 6 months. Recurrence within 12 months was associated with worse survival (p=0.013). No grade ≥3 toxicity was observed.

CONCLUSION
Single-fraction SBRT for contiguous vertebral metastases is feasible and provides acceptable local control with low toxicity; however, distant progression and recurrence warrant further studies of combined systemic therapy.

Summary

Introduction

Spinal metastasis is a common and emergent problem in patients with advanced cancer with an incidence up to 40%.[1] Pain, mechanical instability, and neurological deficit are the most common symptoms that often require intervention. Local treatment of spinal metastasis includes surgery, radiotherapy (RT) and invasive procedures such as regional injections of anesthetics or steroids. Surgery has a limited role in the palliative treatment of metastatic spinal disease except for acute neurological circumstances or vertebral compression fracture (VCF). Traditionally, patients with spinal metastasis have been treated with conventional three dimensional (3-D) external beam radiotherapy (EBRT) with the aim of palliation. Conventional EBRT of 30 Gy in 10 fractions or single dose 8 Gy can be effective in acute palliation of symptoms such as pain or neurological symptoms of spinal metastasis.[2]

Stereotactic body radiotherapy (SBRT) is a non-invasive, image-guided EBRT method used to very precisely deliver higher doses of RT to a small extracranial metastasis using single fraction or small number of fractions with high precision and rapid dose falloff gradients defined by American Society for Radiation Oncology (ASTRO).[3] Accurate positioning and immobilization of patient is quite essential while delivering increased load of radiation to a small target in SBRT. There are various dose ranges in the literature for single fraction SBRT with local control (LC) rates up to 93-100%.[4-6]

Most of the SBRT studies investigated single vertebral level while multilevel treatment data are scarce.[7-9] The data of multilevel spine SBRT is coming from small subgroups of large series. Ryu et al.[7] investigated single fraction spinal SBRT for localized (1-3 separate, 2 contiguous) spine metastasis in 46 patients and only 8 (18%) of the cohort had 2 involvement. Similarly, Beeler et al.[9] analyzed consecutive cases of 101 single level and 84 multilevel (2-7 contiguous) spinal SBRT. Eighty-two of 84 patients (98%) were treated with 3-5 fractions and 4.8-9 Gy per fraction and only 25% of single level patients treated with 1 fraction 18-24 Gy. The authors reported that multilevel SBRT provides high rates of LC similar to single level treatment. Finally, Sellin et al.[8] reported 37 consecutive patients undergoing spinal SBRT for contiguous sites. The local progression, time between the first metastasis and primary diagnosis, performance status, presence of metastatic disease and progressive metastatic disease were found to be prognostic factors affecting survival.

In the absence of contiguous spinal SBRT in a homogenous group of patients treated per protocol (single fraction 16-18 Gy), we sought to perform both dosimetric analysis and clinical outcomes of such patients. We also aimed to analyze the factors affecting local control and survival.

Introduction

Methods

Ethics committee approval was waived due to the retrospective nature of the study and absence of direct patient intervention.

Patient Characteristics
The clinical data of 58 patients with oligometastatic primary and contiguous vertebral metastasis who were treated with SBRT for metastatic sites between December 2012 and December 2020 were retrospectively analyzed. Patients either had oligometastasis at diagnosis, which is called de novo oligometastatic disease (35 patients, 60%) or had oligoprogression (23 patients, 40%) after completion of primary treatment. Patients were included if they had biopsy proven primary carcinoma, had <5 metastasis at diagnosis or progression, and had received treatment with SBRT for oligometastatic site. Patients with ≥5 metastasis or who underwent re-irradiation for the oligometastatic site were excluded. Patients who had surgery for the oligometastatic site and received postoperative SBRT were excluded. Also, different dose and fraction schemes rather than single fraction 16 or 18 Gy were excluded. All cases were presented and discussed at a multidisciplinary tumor board.

The Neurologic, Oncologic, Mechanical, and Systemic (NOMS) decision framework was retrospectively applied to characterize disease status at baseline. Neurologic status was evaluated using the Bilsky epidural spinal cord compression (ESCC) scale and baseline neurological findings. Mechanical stability was assessed using imaging-based risk features. Oncologic factors were considered based on tumor-related characteristics, and systemic status was defined using ECOG performance status and the presence of visceral metastases.

Treatment Planning
Treatment planning and delivery were described elsewhere.[10] Briefly, systemic treatment was not allowed during or 5-7 days before and after SBRT. Patients were positioned supine with arms above head. BodyFIX bluebag vacuum wrap (Elekta, Stockholm, Sweden) was used for immobilization. There was no gross tumor volume (GTV) delineated. All available imaging modalities such as computerized tomography (CT), magnetic resonance imaging (MRI) or positron emission tomography CT (PET-CT) with 18-fluorodeoxyglucose or Ga-68 prostate specific membrane antigen were used to define the extent of gross metastatic tumor, where appropriate. Clinical tumor volume (CTV) encompassed the gross metastatic tumor involvement plus associated entire vertebral body, pedicle transverse process, lamina, spinous process as defined in the International Spine Radiosurgery Consortium consensus.[11] No planning tumor volume (PTV) margin was allowed. The dose constraints were detailed elsewhere.[10] Briefly, D1cc<18 Gy and V10Gy<0.35 cc for spinal cord for each vertebral level. Treatment was delivered in a single day. The dose was prescribed to 90% isodose line CTV coverage was aimed >95% of prescribed dose. The prescribed SBRT doses were 1×16 Gy or 1×18 Gy according to the anatomic localization, extent of gross metastatic tumor and invading structure and dose constraints of neighboring structures. Fractionated delivery of SBRT was not allowed. All treatment plans were performed for delivery with an Axesse linear accelerator (Elekta, Stockholm, Sweden). Volumetric modulated arc therapy (VMAT) plans were used for double or triple 360° arcs, where appropriate (Fig. 1). Response assessment with PET-CT was performed according to the PERCIST v1.0 criteria.[12] RECIST v1.1 criteria was used for patients evaluated through MRI.[13] Toxicity outcomes were recorded during treatment and at each follow-up using the Common Terminology Criteria for Adverse Events version 4.03.

Fig 1: A representative 50% and 95% dose distribution of a contiguous vertebral stereotactic body radiotherapy in axial (a, d), coronal (b, e), and sagittal planes (c, f), respectively.

Follow-up
Patients were followed up at 3-month intervals after the completion of SBRT. Response assessment was performed with MRI if the baseline tool before SBRT was MRI. Similarly, PET-CT was used for response assessment according to the initial management with PET-CT. Patients were followed up at clinically appropriate intervals for both local recurrence and metastatic spread, based on the characteristics of their primary tumours. All patients continued the required systemic therapies after SBRT.

Statistical Analysis
Statistical analyses were performed using the Statistical Package for the Social Sciences software, version 20.0 (SPSS, Chicago, IL, USA). Kaplan-Meier analyses were used to calculate LC, overall survival (OS), and progression-free survival (PFS). OS and PFS were defined as the intervals between the first day of SBRT and the date of death or last follow-up, and the date of any local, regional, or distant disease progression, respectively. Categorical and quantitative variables were described as frequency distributions and by means, medians, and ranges, respectively. The chi-square test was used for subgroup comparisons regarding demographic characteristics and toxicity incidence. To evaluate the relationship between different variables and survival, a Cox proportional hazards model was applied. All p values <0.05 were considered statistically significant.

Methods

Results

Patient Characteristics
Patient and tumor characteristics of the entire cohort are summarized in Table 1. A total of 64 lesions in 58 patients were analyzed. The median age at diagnosis was 54 years (range, 28-85 years). Half of the patients (29 patients, 50%) had breast cancer, 15 patients (25%) had prostate cancer, 5 patients (8%) had lung cancer, and the remaining 9 patients (17%) had various primary tumors (rectal, cervical, gastric, etc.). Fifty-two patients (90%) had a single target volume encompassing contiguous vertebrae, while six patients (10%) had two separate metastatic sites. Thirty-three patients (57%) had thoracic metastases, and twenty-five patients (43%) had lumbar metastases. ESCC was low-grade in 69% of lesions and high-grade in 31%. Mechanical instability was present in 67.2% of lesions. Most patients had good performance status (ECOG 0-1: 96.6%), and visceral metastases were present in 56.9% of patients. Thirty-five patients (60%) had de novo oligometastatic disease, whereas twenty-three patients (40%) had oligoprogression. The SBRT dose was 1×16 Gy in the majority of patients (40 patients, 69%), while eighteen patients (31%) were treated with 1×18 Gy.

Table 1: Patient, tumor, and NOMS-related characteristics

When NOMS-related parameters were evaluated in relation to progression, no significant association was observed between ESCC group and progression (p=0.28). Although progression was numerically higher in the high ESCC group (83.3% vs. 70%), this difference did not reach statistical significance. Similarly, mechanical instability was not associated with progression (p=0.96), with nearly identical progression rates observed between low- and high-risk groups (73.7% vs. 74.4%). ECOG performance status was not significantly associated with progression (p=0.43), although interpretation is limited by the small number of patients with ECOG 2-3. Likewise, the presence of visceral metastases was not associated with progression (p=0.78), with similar progression rates observed in both groups.

All patients were treated with curative intent for their primary tumors and had received surgery, radiotherapy (RT), systemic chemotherapy, or a combination of these modalities. Following SBRT for the oligometastatic lesions, systemic chemotherapy, hormonotherapy, and immunotherapy were continued according to the treating clinicians' discretion. Among these patients, thirty-four (59%) received systemic chemotherapy and twenty-two (38%) received hormonotherapy. Two patients (3%) with low-volume metastatic prostate cancer who achieved a complete response (CR) did not receive any adjuvant treatment after SBRT.

Dosimetry
The dosimetric analyses of the CTV and spinal cord are summarized in Table 2 and Table 3, respectively. All statistical data are presented as mean±standard deviation (SD). Comparisons between different dose levels (1×16 Gy vs. 1×18 Gy) were performed using percentage-based analyses. Briefly, the mean tumor volume of the metastatic sites treated with SBRT was 27.6±11.6 cc. For the entire cohort, the minimum dose (Dmin), maximum dose (Dmax), and mean dose (Dmean) were 14.18±1.02, 20.99±1.95, and 18.95±3.26, respectively. The mean doses to 2% (D2), 50% (D50), and 98% (D98) of the CTV were 20.37±1.70, 19.24±1.46, and 16.61±0.93, respectively. The Vtarget and Vbody were 27.19±11.40 cc and 29.02±19.50 cc, respectively. The spinal cord volume was 4.41±1.65 cc. The maximum dose to the spinal cord (Dmax), the dose to 0.35 cc of the spinal cord (D0.35), the dose to 10% of the spinal cord (D10%), and the volume of the spinal cord receiving 10 Gy (V10) were 11.68±0.85, 8.91±0.93, 8.53±0.87, and 0.14±0.11, respectively.

Table 2: Dosimetric analysis of the clinical target volume according to prescribed dose

Table 3: Dosimetric analysis of the spinal cord according to prescribed dose

When CTV dosimetry was compared according to SBRT dose, no significant differences were observed between the 1×16 Gy and 1×18 Gy groups. Similarly, spinal cord dosimetry did not differ between the two dose groups. An additional comparison was performed based on metastatic site location. A significant difference was observed in both CTV volume and spinal cord volume between thoracic and lumbar metastases. The mean CTV volume of thoracic metastases (22.04±9.31 cc) was significantly lower than that of lumbar metastases (37.02±8.65 cc) (p<0.001). Similarly, the mean spinal cord volume in thoracic metastases (4.08±1.13 cc) was significantly lower than in lumbar metastases (6.13±1.59 cc) (p<0.001). In addition, Vtarget (1.28±0.52 vs. 2.18±0.50; p<0.001) and Vbody (1.30±0.40 vs. 2.87±1.40; p<0.002) were significantly lower in thoracic metastases compared with lumbar metastases.

In addition, conformity index (CI) and homogeneity index (HI) were available for a subset of 34 lesions and were analyzed accordingly. Median HI and CI values were similar between thoracic and lumbar vertebral groups, with no statistically significant differences (p=0.32 and p=0.92, respectively). Likewise, median CTV Dmax and spinal cord Dmax values were comparable between the two groups (p=0.14 and p=0.89, respectively) (Table 4).

Table 4: Comparison of dosimetric parameters between thoracic and lumbar vertebral groups

Patient Outcomes and Survival
The median follow-up of the entire cohort was 12 months (range, 1-56 months). The vast majority of patients (48 patients, 83%) had died by the time of the last visit. Response assessment was performed using either PET/CT or MRI at a median of 3 months (range, 1-7 months) after completion of SBRT. For the entire cohort, complete response (CR), partial response (PR), and progressive disease (PD) rates were 31%, 59%, and 10%, respectively. Forty-two patients (72%) experienced disease recurrence at a median of 7 months (range, 1-21 months) during follow-up. Local recurrence was observed in ten patients (17%) at a median of 6 months (range, 4-17 months) after completion of SBRT.

The 1- and 2-year OS rates were 84% and 65%, respectively, and the 1- and 2-year PFS rates were 73% and 45%, respectively. The 1- and 2-year LC rates were 81% and 70%, respectively. We investigated the prognostic significance of lesion location (thoracic vs. lumbar), lesion number (1 vs. 2), primary histology (breast vs. others), SBRT timing (de novo vs. progression), SBRT dose (1×16 Gy vs. 1×18 Gy), SBRT response (CR vs. PR/PD), and time to recurrence (≤12 months vs. >12 months) with respect to survival and local recurrence. We found that patients with primary breast histology had better OS than those with other primary tumors (24.3 vs. 13.9 months, p=0.024). There was no significant advantage in PFS or LC for patients with breast cancer compared with other primary histologies. OS (18.7 vs. 38.8 months, p=0.013) was worse in patients who experienced recurrence within 12 months compared to counterparts. No significant difference in LC was observed according to recurrence pattern. No additional prognostic factors were identified in the univariate analyses of OS, PFS, and LC (Table 5). Multivariate analysis was not performed because of the limited significance observed in the univariate analysis and the small sample size. None of the patients experienced grade 3 or higher acute or late toxicity.

Table 5: Univariate analysis related with OS, PFS and LC

Results

Discussion

In the present study, we analyzed the dosimetric aspects and clinical outcomes of contiguous vertebral SBRT in the treatment of oligometastatic patients and demonstrated its effectiveness and safety in this setting. The 1- and 2-year LC rates were 81% and 70%, respectively. None of the patients experienced grade 3 or higher acute or late toxicity. Forty-two patients (72%) developed disease progression at a median of 7 months, and among these, ten patients (17%) experienced local recurrence at a median of 6 months after completion of SBRT. Patients with early recurrence (≤12 months) after completion of SBRT had worse survival compared with their counterparts.

The International Spine Radiosurgery Consortium consensus guidelines define target volume delineation in spinal SBRT with great precision.[11] In addition, the Radiation Therapy Oncology Group (RTOG) 0631 phase II/III trial protocol for spinal SBRT is highly informative with respect to treatment procedures and dose constraints.[7] However, multilevel or contiguous vertebral SBRT remains more challenging compared with single-level SBRT. Data on the dosimetry of contiguous vertebral SBRT are very limited, and most available dosimetric data are derived from heterogeneous patient cohorts in which single-level and multilevel spinal SBRT are analyzed together.[7,14] Schipani et al.[14] investigated the dosimetric characteristics of single-fraction 1×18 Gy spinal SBRT in 124 patients with 165 lesions. Their dosimetric analysis showed that the target volume received a median dose of 19 Gy, while the median spinal cord Dmax, D0.35, and V10 were 13.8 Gy, 8.9 Gy, and 0.33 cc, respectively. The authors reported a local control rate of 92% at a median follow-up of 7 months and identified a spinal cord Dmax of 14 Gy and D0.35 of 10 Gy as safe dose constraints. Ryu et al.[7,15] defined their dosimetric criteria as 80-90% target volume coverage, spinal cord D10% <10%, and spinal cord D0.35<0.35 cc. The authors recommended that these criteria be applied to each treated spinal level. In the present study, we strictly followed these criteria, and our dosimetric analyses demonstrated that both target coverage and spinal cord dose constraints were consistent with the existing literature. In addition, we evaluated the impact of lesion location (thoracic vs. lumbar) and SBRT dose (1×16 Gy vs. 1×18 Gy) on dosimetry. We found that dose escalation had no significant impact on dosimetric parameters, and that a dose of 1×18 Gy could be safely delivered instead of 1×16 Gy without compromising target coverage or increasing toxicity. Thoracic CTV volume, as well as the associated Vtarget and Vbody values, were significantly lower than those observed for lumbar metastases in our contiguous spinal SBRT series. Given that lumbar vertebrae are anatomically larger than thoracic vertebrae, we believe that this difference has no clinical relevance, and that spinal SBRT can be safely delivered regardless of lesion location, provided that essential dosimetric criteria are met.

Our cohort consisted of both de novo oligometastatic and oligoprogressive patients. Local recurrence rate of 17% is higher than those reported in single-level SBRT series, in which tumor control rates above 90% have been described.[4-6] Recently, Zelefsky et al.[16] reported the results of a phase III randomized trial comparing single-fraction SBRT (1×24 Gy) with a hypofractionated regimen (3×9 Gy) in the treatment of oligometastatic patients. The authors demonstrated that local recurrence and the cumulative incidence of distant metastases were significantly higher in the hypofractionated SBRT arm compared with the single-fraction SBRT arm, and concluded that single-fraction SBRT represents a superior ablative treatment approach. However, the difference in biologically effective dose (BED) between the 1×24 Gy and 3×9 Gy regimens should be kept in mind when interpreting these results. In our series, we specifically investigated single-fraction SBRT for contiguous vertebral metastases in order to maintain a homogeneous study population and minimize potential bias. Patients treated with fractionated SBRT regimens for contiguous vertebrae were therefore excluded. Murai et al.[17] investigated intensity-modulated radiotherapy with a simultaneous integrated boost (SIB-IMRT) for single or multiple vertebral metastases. The prescribed doses were 5×8 Gy or 3×16 Gy in a cohort of thirty patients with forty lesions. The authors reported a 1-year local control rate of 84% and demonstrated that up to four consecutive vertebrae could be successfully treated with SIB-IMRT. Similarly, Sellin et al.[8] reported 37 consecutive patients who received SBRT for contiguous vertebral sites. In 22 of the 37 patients (60%), disease involvement was limited to a single vertebral body. Patients were treated with SBRT doses of 1×24 Gy, 3×9 Gy, or 5×6 Gy. The authors reported a local control rate of 51% and found that local disease progression was associated with worse survival. Local progression, metastasis occurring within 12 months after primary diagnosis, impaired performance status, and progression of systemic disease were identified as significant factors affecting survival. Ryu et al.[7] reported the results of RTOG 0631, a phase II/III study of spinal SBRT, in which the prescribed SBRT dose was 1×16 Gy in forty-four patients. The number of patients with contiguous vertebral involvement was limited to 8 of 36 patients (18%). Beeler et al.[9] investigated 101 single-level and 85 multilevel (2-7 contiguous vertebrae) SBRT cases. In 98% of the multilevel cases, the prescribed SBRT doses were 3×9 Gy or 5×8 Gy. The authors reported no significant differences in local control or survival between single-level and multilevel SBRT. In contrast, our study focused exclusively on single-fraction SBRT for purely contiguous vertebral metastases and demonstrated acceptable local control rates. Time to progression was the only factor significantly associated with survival in the univariate analysis. Patients who experienced disease recurrence within 12 months had worse OS and PFS compared with their counterparts. In addition, patients with a primary breast cancer had better OS than those with other primary tumors; however, we believe that this finding should be interpreted with caution until confirmed in larger cohorts. No other factors were found to be significantly associated with OS, PFS, or LC.

In addition to these findings, we further evaluated our results within the NOMS framework. Overall, NOMS-related parameters were not consistently associated with progression. Neither ESCC-defined neurologic status nor mechanical instability showed a significant relationship with progression. Similarly, systemic factors, including ECOG performance status and the presence of visceral metastases, were not associated with outcomes in this cohort. Taken together, these findings suggest that while the NOMS framework remains valuable for clinical decision-making, its individual components may not uniformly predict outcomes in small retrospective cohorts treated with single-fraction SBRT.

Furthermore, we evaluated whether anatomical differences between thoracic and lumbar vertebrae influenced plan quality. Despite differences in vertebral and spinal cord volumes, median CI, HI, and maximum dose parameters were comparable between the two groups, with no statistically significant differences. These findings suggest that variations in vertebral inclination and anatomy did not have a measurable impact on dosimetric quality in single-fraction SBRT.

As mentioned above, most of the available data on contiguous vertebral SBRT are derived from subgroups of larger spinal SBRT series.[7-9,17] Therefore, it remains unclear whether the toxicity rates reported in these series accurately reflect the true incidence of adverse events specifically related to contiguous vertebral SBRT. Nevertheless, no grade 4-5 toxicity was reported in either single-level or multilevel spinal SBRT series. Beeler et al.[9] reported grade 3-4 late toxicity in the form of vertebral compression fracture in 9.2% of patients. Ryu et al.[7] reported grade 3 neck pain in only one of forty-four patients (2%). Murai et al.[17] reported an uncertain grade 3 ileus in one of forty patients (3%) with a history of colorectal cancer. In our series, no grade 3 or higher acute or late toxicity was observed.

Our study is not without limitations, primarily due to its retrospective design and the associated risk of selection bias. First, the small sample size limits the ability to clearly delineate differences in outcomes according to treatment site, dose, or disease pathology. Second, patients received different systemic chemotherapy or hormonotherapy regimens based on their primary tumor and disease stage, which may have directly influenced survival outcomes. Third, the relatively short follow-up period represents another limitation, although the vast majority of patients (48 patients, 83%) were dead at the last follow-up. Finally, CI and HI analyses were available only for a subset of 34 lesions, representing a limitation of the present study. This incomplete data availability may have introduced selection bias and limited the statistical power to detect potential differences between groups. Therefore, the absence of statistically significant differences in CI, HI, and maximum dose parameters between thoracic and lumbar vertebrae should be interpreted with caution. Despite these inherent limitations, this study is of particular importance, as it specifically investigates spinal SBRT for purely contiguous vertebral metastases in oligometastatic patients. We report both dosimetric analyses and clinical outcomes in a homogeneous patient cohort treated according to a single SBRT protocol.

Discussion

Conclusion

The use of single-fraction SBRT to treat contiguous spinal vertebrae is a feasible treatment approach. This technique provides acceptable local control with minimal toxicity. However, the relatively high rates of distant progression and local recurrence highlight the need for further research into the combination of SBRT and systemic therapies. Determining optimal strategies for patient selection, dose-fractionation schemes, and appropriate radiotherapy techniques remains an area of ongoing investigation.

Conclusion