SPINE Volume 40, Number 8, pp 544-549 ©2015, Wolters Kluwer Health, Inc. All rights reserved.
CLINICAL CASE SERIES
Postoperative Proton Therapy for Chordomas and Chondrosarcomas of the Spine Adjuvant Versus Salvage Radiation Therapy Emma B. Holliday, MD,* Hari S. Mitra, BS,† Jeremy S. Somerson, MD,‡ Laurence D. Rhines, MD,§ Anita Mahajan, MD,* Paul D. Brown, MD,* and David R. Grosshans, MD, PhD*
Study Design. Retrospective comparative cohort series. Objective. The aim of this study was to evaluate patients treated with proton therapy for chordoma and chondrosarcoma of the spine in the postoperative setting and to report local control, relapse-free, and overall survival outcomes. Summary of Background Data. Margin-negative resection of spinal chordomas and chondrosarcomas can be challenging, so adjuvant radiotherapy is often recommended. However, delivery of adequate radiotherapy is complicated by the relative radioresistance of these tumors, necessitating high doses, as well as the proximity of the spinal cord and exiting nerve roots increasing the risk for toxicity. Proton radiotherapy has favorable physical properties for avoiding nearby nontarget structures and is increasingly used for such lesions. Methods. Nineteen patients who underwent postoperative proton therapy at a single institution from 2006 to 2012 were identified including 13 with chordoma and 6 with chondrosarcoma. Surgical approach varied by tumor location in the cervical (n = 3), thoracic (n = 1), lumbar (n = 2), or sacral (n = 13) spine. Eight patients were categorized as receiving “early adjuvant” and 11 patients as receiving “salvage” treatment, as determined by initiation of radiation therapy after primary surgery or local recurrence, respectively. The median radiation dose delivered was 70 Gy relative biologic effectiveness (range: 56–78 Gy relative biologic effectiveness).
From *Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX; †School of Medicine, The University of Texas Health Science Center San Antonio, San Antonio, TX; ‡Department of Orthopaedic Surgery, The University of Texas Health Science Center San Antonio, San Antonio, TX; and §Department of Neurosurgery, The University of Texas MD Anderson Cancer Center, Houston, TX. Acknowledgment date: August 21, 2014. First revision date: December 1, 2014. Second revision date: January 12, 2015. Acceptance date: January 16, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No relevant financial activities outside the submitted work. Address correspondence and reprint requests to David R. Grosshans, MD, PhD, Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center, 1840 Old Spanish Trail, Unit 1150, Houston, TX 77054; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000804
544
Results. For the entire cohort, 2-year local control, relapsefree survival, and overall survival were 58%, 51.9%, and 93.3%, respectively. The early adjuvant group had significantly higher 2-year local control (80% vs. 45.5%; P = 0.024). Conclusion. Patients referred early for primary adjuvant radiation therapy after surgery had higher rates of disease control than those referred for salvage treatment of recurrent disease. Recurrence rates in our cohort were higher overall than other published series, indicating that even higher radiation doses may be helpful for further improving local control in the presence of gross or recurrent disease. Key words: chordoma, chondrosarcoma, spine, adjuvant radiotherapy, salvage, proton. Level of Evidence: 3 Spine 2015;40:544–549
C
hordomas and chondrosarcomas are rare, slow-growing neoplasms that are difficult to control because of their invasive nature and proximity to critical structures precluding widely negative surgical margins.1 The methods taken to treat these tumors have evolved over time. Intralesional resections were once considered to be the standard of care of spinal oncology surgery, but they have fallen out of favor for management of chordomas because of their inability to ensure complete removal of the tumor with negative margins. En bloc resection is now the preferred approach as it allows preservation of the tumor margin, ensuring a complete oncologic resection.2 Although ideal, en bloc resections carry great risks to many important structures and thus are not always feasible. Adjuvant treatment with radiation therapy (RT) may improve local control (LC) and potentially overall survival (OS)3,4; however, the proximity of the operative bed to critical structures such as the spinal cord and exiting nerve roots has the potential to add considerable toxicity. In addition to location, the relative resistance of chordomas and chondrosarcomas to radiation necessitates higher doses of RT in order to achieve LC.5,6 The dose of radiation able to be delivered to the tumor is limited by the tolerance of the spinal cord, which is typically limited to a maximum of 50 Gray (Gy), a dose level with a minimal risk of paralysis.7 Chordomas and
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 544
14/03/15 1:46 PM
CLINICAL CASE SERIES chondrosarcomas have been shown to require doses in excess of this spinal cord tolerance. Treatment with photon RT to doses of less than 60 Gy results in eventual recurrence rates of 50% to 100% and 5-year progression-free survival rates of less than 25%.8,9 Although 60 Gy is considered inadequate for durable LC, that dose nevertheless risks significant toxicity.10 Particle therapy, including proton beam radiation (PBR), has unique physical properties that may spare normal tissues from unnecessary exposure. Protons have a characteristic depth dose profile notable for a sharp increase in dose deposited at the end of the particle range, a phenomenon known as the Bragg peak. Because protons stop in tissue at the Bragg peak, essentially no exit dose is delivered beyond the target. This is especially helpful when treating tumors adjacent to critical structures. The aim of this study was to evaluate patients treated with PBR for chordomas and chondrosarcomas of the spine in the postoperative setting and to report LC, relapse-free survival (RFS), and OS outcomes. Prognostic factors for LC were also evaluated, specifically, the timing of PBR after surgery.
MATERIALS AND METHODS Patients with chordomas or chondrosarcomas of the cervical, thoracic, lumbar, and sacral spine treated with postoperative PBR at a single, tertiary referral institution between January 1, 2006, and April 1, 2012, were identified from a prospectively collected, institutional review board–approved database. Nineteen patients were identified with tumors in the cervical (n = 3), thoracic (n = 2), lumbar (n = 1), and sacral (n = 13) spine. Prior to treatment, all had their pathology reviewed at our institution: 13 with chordoma and 6 with chondrosarcoma. Patient demographics such as age at first treatment, sex, and race were collected from the medical record. The median radiation dose delivered was 70 Gy relative biologic effectiveness (GyRBE) (range: 56–78 GyRBE) delivered in 2 GyRBE per daily fraction. In 5 patients, RT was delivered using a combined passive scatter proton and photon-based intensity modulated radiation therapy. One patient received 44 GyRBE using a passive scatter proton technique, followed by a stereotactic photon boost of 12 Gy in 3 fractions. Fourteen patients received all of their radiation treatment with passive scatter protons. Dosimetric data were reviewed for 17 of the 19 patients because the 2 patients treated with combined proton-photon plans did not have composite plans available. The target was defined as gross tumor volume in the case of patients with gross residual disease and the postoperative bed or clinical target volume in the case of patients without gross residual disease. Median and range for mean dose to the target and minimum point dose to the target were 73.6 GyRBE (61.7–80.1 GyRBE) and 61.9 GyRBE (11.3–78.7 GyRBE), respectively. Median and range maximum point dose to the spinal cord for the 7 treatment plans treating tumors at the level of the cord were 52.3 GyRBE (43.1–59.1 GyRBE). Disease target was designated as gross disease or microscopic disease based on information from operative notes, pathology reports, and postoperative magnetic resonance images. There were no patients who had a margin-negative Spine
Postoperative Protons for Spine Chordoma • Holliday et al
resection. Finally, the presence or absence of implanted surgical instrumentation was recorded from operative notes. Patients were categorized as having received “early adjuvant” radiation if they received PBR after primary surgery and had obtained a postoperative magnetic resonance image showing a gross total resection or had known gross residual disease that was radiographically stable after surgery. Patients were categorized as having received “salvage” radiation if they received radiation after recurrence was documented by new or progressive disease visible on magnetic resonance image. By this definition, 8 patients (42.1%) received PBR in the early adjuvant setting and 11 (57.9%) received PBR in the salvage setting after documented recurrence or progression. Longer-term clinical outcomes at last follow-up including pain level on the 10-point comparative pain scale, narcotic requirement, the presence or absence of incontinence, Karnofsky Performance Status, and whether or not extremity weakness caused functional impairment were also collected when available. LC, RFS, and OS were measured from the start date of radiation until the event of interest. Chi-square test was used for between group comparisons and Mann-Whitney U test was used to compare medians. Data were analyzed using JMP, Version 7 (SAS Institute Inc., Cary, NC).
RESULTS Figure 1 depicts a representative sagittal (A) and axial (B) image from a treatment plan, delivered using passive scatter proton therapy for a sacral chordoma.For the group studied, acute radiation-related toxicities were mild and consisted of 13 patients with grade 1 radiation dermatitis (defined by the Radiation Therapy Oncology Group’s Acute Morbidity Scoring Criteria as mild, blanchable, erythema, or dry desquamation), 3 patients with grade 2 radiation dermatitis (defined as moist desquamation limited to the skin folds), and 4 patients treated for cervical spine tumors who developed grade 1 to 2 esophagitis (defined as mild to moderate dysphagia requiring non-narcotic analgesics and soft diet or narcotic analgesics and puree/liquid diet, respectively). No patient developed grade 3 to 5 toxicity or required a treatment break. Median follow-up from the time of PBR initiation to the time of last follow-up or death was 34.5 months. Overall, 7 patients experienced freedom from local failure at the time of data analysis in June 2014. Ten had progression of local disease, 1 patient had progression of local disease and distant failure, 1 patient experienced distant failure only, and 4 patients died of medical conditions not related to their cancer. Therefore, the overall 2-year LC, RFS, and OS for this cohort are 58%, 51.9%, and 93.3%, respectively. Eight patients received early adjuvant PBR and 11 patients received salvage PBR. Overall, the median time from surgery to initiation of radiation was 9.5 months (range: 1.3–110.2 mo). Table 1 outlines patient characteristics, surgery, and radiation details for both groups. Patients in the early adjuvant group had a shorter median interval from surgery to initiation of radiation (2.5 mo [range: 1.3–9.5 mo] vs. 32.6 mo [range: 3.7–110.2 mo]; P = 0.0003). In the early adjuvant group, there were significantly more nonsacral tumors (P = 0.0005). www.spinejournal.com
545
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 545
14/03/15 1:46 PM
CLINICAL CASE SERIES
Postoperative Protons for Spine Chordoma • Holliday et al
TABLE 1. Characteristics of Patients Receiving
Early Adjuvant Versus Salvage Proton Radiation Early Adjuvant (N = 8)
Salvage (N = 11)
P
46 (10–71)
57 (41–82)
0.12
Male, n (%)
4 (50%)
5 (54.5%)
Female, n (%)
4 (50%)
6 (45.5%)
Chordoma, n (%)
4 (50%)
9 (81.8%)
Chondrosarcoma, n (%)
4 (50%)
2 (18.2%)
7 (87.5%)
9 (72.7%)
0 (0%)
1 (9.1%)
1 (12.5%)
1 (9.1%)
Nonsacral
6 (75%)
0 (0%)
Sacral, n (%)
2 (25%)
11 (100%)
2.5 (1.3–9.5)
32.6 (3.7–110.2)
Gross disease, n (%)
3 (37.5%)
11 (100%)
Microscopic disease, n (%)
5 (62.5%)
0 (0%)
Absent
5 (62.5%)
11 (100%)
Present
3 (37.5%)
0 (0%)
Protons only, n (%)
5 (62.5%)
9 (81.8%)
Protons and photons, n (%)
3 (37.5%)
2 (18.2%)
70 (56–70.2)
70 (60–78)
No
6 (75%)
1 (9%)
Yes
2 (25%)
10 (91%)
Age at first treatment Median years (range) Sex 0.84
Histology 0.14
Self-reported race Caucasian non-Hispanic, n (%) Caucasian Hispanic, n (%) Asian, n (%)
0.67
Disease site Figure 1. Representative treatment plan using proton beam radiation in the postoperative setting. Sagittal (A) and axial (B) sections are depicted, which highlight sparing of normal tissues including bowel located distal to the beams.
Surgery to XRT interval Median months (range)
All 11 patients in the salvage group had gross disease at the time of radiation initiation whereas only 37.5% of patients in the early adjuvant group had gross disease (P = 0.0003). More patients in the early adjuvant group had metal instrumentation placed than patients in the salvage group (37.5% vs. 0%; P = 0.05), but there was no significant difference in the radiation technique used or the median dose delivered. Twenty-five percent of patients had a documented relapse of disease in the early adjuvant group compared with 91% in the salvage group at the time of data analysis (P = 0.006). Two-year LC was 88% in the early adjuvant group compared with 45% in the salvage group (P = 0.061). Five-year LC was 88% in the early adjuvant group compared with 9% in the salvage group (P < 0.001). Two-year RFS was 75% in the early adjuvant group compared with 45% in the salvage group (P = 0.198). Two-year OS was 88% in the early adjuvant group compared with 100% in the salvage group (P = 0.228). On univariate analysis, sacral disease site (P = 0.004), longer interval between surgery and initiation of radiation (P = 0.025), and gross disease present at the initiation of radiation (P = 0.024) were all found to be associated with higher rates of recurrence. There was a trend toward increased recurrence 546
0.0005
0.0007
Disease target 0.0003
Surgical Instrumentation 0.05
XRT technique 0.603
XRT dose Median (range) in Gy (RBE)
0.25
Ultimate relapse after XRT 0.006
The values in bold face indicate P ≤ 0.05. XRT indicates radiation therapy.
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 546
14/03/15 1:46 PM
CLINICAL CASE SERIES
Postoperative Protons for Spine Chordoma • Holliday et al
TABLE 2. Prognostic Factors for Recurrence-
TABLE 3. Clinical and Functional Outcomes
Free Survival
No Recurrence Recurrence (N = 12) (N = 7)
for Patients Receiving Early Adjuvant Versus Salvage Proton Radiation Early Adjuvant (n = 5)
Salvage (n = 10)
P
5 (0–6)
3 (0–8)
0.95
Yes
3 (60%)
7 (70%)
No
2 (40%)
3 (30%)
Yes
1 (20%)
4 (40%)
No
4 (80%)
6 (60%)
80 (60–100)
80 (70–100)
Yes
1 (20%)
6 (60%)
No
4 (80%)
4 (40%)
P
Histology Chordoma, n = 13
10 (76.9%)
3 (23.1%)
Chondrosarcoma, n=6
2 (33.3%)
4 (66.7%)
Pain level 0.07
Narcotic requirement, n (%)
Disease site Nonsacral, n = 6
1 (16.7%)
5 (83.3%)
Sacral, n = 13
11 (84.6%)
2 (15.4%)
24.75 (1.8–110.2)
2.5 (1.3–32.2)
0.025
Gross disease, n = 14
11 (78.6%)
3 (21.4%)
0.025
Microscopic disease, n=5
1 (20%)
4 (80%)
0.004
Surgical instrumentation 1 (33.3%)
2 (66.7%)
Absent, n = 16
11 (68.8%)
5 (31.2%)
Median (range)
0.62
Weakness affecting function 0.14
0.24
XRT indicates radiation therapy.
rates in chordoma patients when compared with chondrosarcoma patients (76.9% vs. 23.1%; P = 0.07). Table 2 outlines these prognostic factors. Clinical outcomes were available for 15 of the 19 patients included, 5 patients receiving early adjuvant and 10 patients receiving salvage radiation. There were no significant differences in any of the clinical outcomes collected (Table 3).
DISCUSSION This retrospective comparative series reports results of patients treated with postoperative PBR for chordomas and chondrosarcomas of the spine. Patients who received early adjuvant PBR were compared with those who were treated in the salvage setting. Patients treated in the early adjuvant setting had better 2-year LC. Both increased time between surgery and radiation as well as the presence of gross residual disease at the time of radiation were adverse prognostic factors. First studied in the base of skull,11–16 few series have evaluated PBR for chordomas and chondrosarcomas of the spine.17–22 Yasuda et al17 from the Institute Curie in France reported 75% LC for patients with chordomas of the skull base and spine who received PBR or combined photonproton RT to a mean dose of 68.9 Gy (RBE) after surgical resection. Rutz et al20 from the Paul Scherrer Institute in Switzerland reported 5 local failures out of 26 patients with extracranial chordomas who received postoperative Spine
0.44
Karnofsky performance status
Disease target
Present, n = 3
0.70
Incontinence, n (%)
Surgery to XRT interval Median (range) in months
Median (range)
spot-scanning PBR to a median dose of 72 Gy (RBE). Gross total resection was achieved in 69% of patients. The group from Massachusetts General Hospital has published most extensively on the subject.18,19,21,22 Hug et al18 reported 53% and 100% 5-year LC rates for chordoma and chondrosarcomas of the axial skeleton, respectively, with combined photonproton RT to a mean dose of 73.9 Gy (RBE). A more recent publication from Massachusetts General Hospital outlined their experience with definitive PBR in medically inoperable or otherwise unresected chordomas of the spine and reported 80% 5-year local progression–free survival.19 DeLaney et al reported results from a phase II study of 50 patients with spine sarcomas (29 had chordoma and 14 chondrosarcoma) treated with preoperative and/or postoperative PBR. Five-year LC was 78% for the entire cohort. Of note, 50% of patients in this cohort received a margin-negative resection. On the Massachusetts General Hospital phase II study, PBR doses were higher than those in the previously mentioned studies, with 70.2 Gy (RBE) prescribed for microscopic disease and 77.4 Gy (RBE) for gross disease.21 The same group recently published long-term results including 74% 8-year LC.22 Our overall LC results were lower than those in the majority of previously published series, with a 2-year LC rate of 58% for all patients. This may be partially attributable to the relatively low median dose of 70 Gy (RBE) that was delivered. Also potentially contributory is the fact that 58% (N = 11) of patients included in our series were treated for recurrence after primary resection. Indeed, patients receiving PBR in the salvage setting had a significantly lower 2-year LC of 45.5%. Other groups reported similarly poor LC for patients treated www.spinejournal.com
547
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 547
14/03/15 1:46 PM
CLINICAL CASE SERIES in the salvage setting. Delaney et al22 reported a 53% 5-year actuarial rate of local recurrence in previously recurrent tumors treated on protocol compared with 6% in those with primary tumors (P = 0.002). Patients in our series receiving early adjuvant PBR had a 2-year LC rate of 80%, which is more comparable with other published data. Finally, a margin-negative resection has been shown by several groups to be a positive prognostic factor,18,20–22 and none of the patients in our series received R0 resections compared with 50% to 75% in the other series cited. In addition to treatment in the salvage setting and increased interval from surgery to radiation, adverse prognostic factors identified in this study included the presence of gross residual disease18,20,22 and sacral tumor location,20 which have also been reported in other studies. In our series, patients with sacral tumors more often had gross disease present at the time of radiation initiation (n = 12, 92%) than patients with nonsacral tumors (n = 2, 33%; P = 0.005). Although some fear that giving higher doses of radiation could cause damage to the exiting sacral nerve roots, there was no difference in median dose delivered between patients with sacral and nonsacral tumors (70 [range: 60–78] and 70 [range: 54–70], respectively). Rutz et al20 reported the presence of implants placed at surgery to be associated with increased local failure. Several physical factors may explain this, including computed tomography artifacts that make accurate target delineation more difficult during PBR treatment planning as well as dose in homogeneities and “cold spots” caused by interactions with the proton beam and the implant itself. We did not find worse LC in patients with surgical implants; however, this may be due to low numbers in our series. Finally, chordomas and chondrosarcomas are distinct histological and clinical entities but were included together in this case series to be consistent with the existing, albeit limited, published data published from other proton therapy centers reporting on their patients with tumors of the spine and/or skull base.13–15,18,21 It should be emphasized that chondrosarcomas have a better prognosis than chordomas in terms of response to treatment and survival. Studies including patients with skull base sarcomas reported 5-year recurrence-free survival rates of 90% versus 65% for chordomas and chondrosarcomas, respectively; P = 0.09.1 Although numbers were too small and follow-up too short to detect any differences in survival in our series, patients with chondrosarcomas had a similar trend toward decreased local recurrence when compared with patients with chordomas (n = 2, 33% and n = 10, 77%, respectively [P = 0.07]). This was despite the trend toward a lower median dose given to patients with chondrosarcomas when compared with chordomas (65 (range: 54–70) and 70 (range: 66–78); P = 0.08). As with all retrospective case series, several inherent potential biases exist in the interpretation of these data. The potential for selection bias is high in this small, retrospectively analyzed cohort. Our status as a large, international cancer center attracts patients with multiply recurrent disease or those who were not offered treatment elsewhere, which may reflect inherently worse biology and contribute to our lower control rates compared with other published series. Recurrent tumor may 548
Postoperative Protons for Spine Chordoma • Holliday et al
also have inherently more resistant biology than primary disease, which introduces another potentially confounding factor. Moreover, many patients underwent surgical resections outside of our institution and as such the surgical approach was not standardized. The comparison between patients treated in the adjuvant setting versus in the salvage setting also has the potential for inherent bias, such that patients who recur after resection may also have more aggressive biology. Finally, the presence of gross disease at the time of radiation is a potential confounding factor in evaluating the importance of the timing of PBR after surgery because other studies have shown extent of disease and margin status to be adverse prognostic factors.22–24 The best way to minimize selection bias in future studies and to obtain higher quality evidence to assist in counseling patients with chordoma and chondrosarcoma would be to perform a randomized controlled trial in which patients are randomly assigned to receive adjuvant RT in the early postoperative period or in the salvage setting. However, the rarity of this tumor precludes such a study from being realistically performed. Therefore, we propose that patients with chordoma or chondrosarcoma of the spine be enrolled on a prospective, non–randomized study protocol from which consecutive patients could be analyzed. Ideally, this protocol would be opened at multiple centers to provide potentially more generalizable results. Chordomas and chondrosarcomas of the spine are challenging to manage because of their propensity for local recurrence. Postoperative radiation can decrease local recurrence, but a longer interval between surgery and radiation as well as gross disease present at the time of radiation are adverse prognostic factors. Therefore, to increase the effectiveness of this adjuvant therapy, we suggest that patients be referred for consideration of radiation in the early postoperative period prior to disease recurrence or progression. In addition, further consideration should be given to strategies that might further improve LC, including escalation of radiation dose and utilization of more conformal radiation techniques such as intensity-modulated proton therapy. In addition, longer follow-up is needed to monitor and report the late effects and functional outcomes of combined modality therapy.
➢ Key Points In the absence of adequate therapy, chordomas and chondrosarcomas are prone to local recurrence; however, their anatomic location makes local control challenging. In this series, patients who received early adjuvant proton therapy had significantly better freedom from local failure than patients who received salvage radiotherapy after a local recurrence. Patients with sacral tumors, a longer surgery to radiation interval, and gross disease present at the time of radiation were associated with higher rates of recurrence.
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 548
14/03/15 1:46 PM
CLINICAL CASE SERIES References
1. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 patients. Neurosurgery 1995;36:887–97. 2. Chi JH, Sciubba DM, Rhines LD, et al. Surgery for primary vertebral tumors: en bloc versus intralesional resection. Neurosurg Clin N Am 2008;19:111–7. 3. Boriani S, Bandiera S, Biagini R, et al. Chordoma of the mobile spine: fifty years of experience. Spine (Phila Pa 1976) 2006;31: 493–503. 4. Fuller DB, Bloom JG. Radiotherapy for chordoma. Int J Radiat Oncol Biol Phys 1988;15:331–9. 5. Zagars GK, Ballo MT. Significance of dose in postoperative radiotherapy for soft tissue sarcoma. Int J Radiat Oncol Biol Phys 1995;31:1093–112. 6. DeLaney TF, Kepka L, Goldberg SI, et al. Radiation therapy for control of soft tissue sarcomas resected with positive margins. Int J Radiat Oncol Biol Phys 2007;67:1460–9. 7. Schultheiss TE, Kun LE, Ang KK, et al. Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 1995;31: 1093–112. 8. Catton C, O’Sullivan B, Bell R, et al. Chordoma: long-term follow-up after radical photon irradiation. Radiother Oncol 1996;41: 67–72. 9. Zorlu F, Gürkaynak M, Yildiz F, et al. Conventional external radiotherapy in the management of clivus chordomas with overt residual disease. Neurol Sci 2000;21:203–7. 10. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol 1999;175(suppl 2):57–63. 11. Noël G, Feuvret L, Calugaru V, et al. Chordomas of the base of the skull and upper cervical spine. One hundred patients irradiated by a 3D conformal technique combining photon and proton beams. Acta Oncol 2005;44:700–8. 12. McDonald MW, Linton OR, Moore MG, et al. Initial results of proton therapy for clival chordoma. Int J Radiat Oncol Biol Phys 2013;87:S277–8. 13. Ares C, Hug EB, Lomax AJ, et al. Effectiveness and safety of spot scanning proton radiation therapy for chordomas and chondrosarcomas of the skull base: first long-term report. Int J Radiat Oncol Biol Phys 2009;75:1111–8.
Spine
Postoperative Protons for Spine Chordoma • Holliday et al
14. Rutz HP, Weber DC, Goitein G, et al. Postoperative Spot-Scanning proton radiation therapy for chordoma and chondrosarcoma in children and adolescents: initial experience at Paul Scherrer Institute. Int J Radiat Oncol Biol Phys 2008;71:220–5. 15. Rombi B, Ares C, Hug EB, et al. Spot-scanning proton radiation therapy for pediatric chordoma and chondrosarcoma: clinical outcome of 26 patients treated at Paul Scherrer Institute. Int J Radiat Oncol Biol Phys 2013;86:578–84. 16. McDonald MW, Linton OR, Shah MV. Proton therapy for reirradiation of progressive or recurrent chordoma. Int J Radiat Oncol Biol Phys 2013;87:1107–14. 17. Yasuda M, Bresson D, Chibbaro S, et al. Chordomas of the skull base and cervical spine: clinical outcomes associated with a multimodal surgical resection combined with proton-beam radiation in 40 patients. Neurosurg Rev 2012;35:171–82; discussion 182–3. 18. Hug EB, Fitzek MM, Liebsch NJ, et al. Locally challenging osteoand chondrogenic tumors of the axial skeleton: results of combined proton and photon radiation therapy using three-dimensional treatment planning. Int J Radiat Oncol Biol Phys 1995;31:467–76. 19. Chen YL, Liebsch N, Kobayashi W, et al. Definitive high-dose proton/photon radiotherapy for unresected mobile spine and sacral chordomas. Spine (Philadelphia, PA 1976) 2013;38:E930–6. 20. Rutz HP, Weber DC, Sugahara S, et al. Extracranial chordoma: outcome in patients treated with function-preserving surgery followed by spot-scanning proton beam irradiation. Int J Radiat Oncol Biol Phys 2007;67:512–20. 21. DeLaney TF, Liebsch NJ, Pedlow FX, et al. Phase II study of high dose photon/proton radiotherapy in the management of spine sarcomas. Int J Radiat Oncol Biol Phys 2009;74:732–9. 22. DeLaney TF, Liebsch NJ, Pedlow FX, et al. Long-term results of phase II study of high dose proton/photon radiotherapy in the management of spine chordomas, chondrosarcomas and other sarcomas. J Surg Oncol 2014;110:115–22. 23. Lee J, Bhatia NN, Hoang BH, et al. Analysis of prognostic factors for patients with chordoma with use of the California Cancer Registry. J Bone Joint Surg Am 2012;94:356–63. 24. Bergh P, Kindblom LG, Gunterberg B, et al. Prognostic factors in chordoma of the sacrum and mobile spine: a study of 39 patients. Cancer 2000;88:2122–34.
www.spinejournal.com
549
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 549
14/03/15 1:46 PM
Copyright of Spine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
CLINICAL CASE SERIES
Postoperative Proton Therapy for Chordomas and Chondrosarcomas of the Spine Adjuvant Versus Salvage Radiation Therapy Emma B. Holliday, MD,* Hari S. Mitra, BS,† Jeremy S. Somerson, MD,‡ Laurence D. Rhines, MD,§ Anita Mahajan, MD,* Paul D. Brown, MD,* and David R. Grosshans, MD, PhD*
Study Design. Retrospective comparative cohort series. Objective. The aim of this study was to evaluate patients treated with proton therapy for chordoma and chondrosarcoma of the spine in the postoperative setting and to report local control, relapse-free, and overall survival outcomes. Summary of Background Data. Margin-negative resection of spinal chordomas and chondrosarcomas can be challenging, so adjuvant radiotherapy is often recommended. However, delivery of adequate radiotherapy is complicated by the relative radioresistance of these tumors, necessitating high doses, as well as the proximity of the spinal cord and exiting nerve roots increasing the risk for toxicity. Proton radiotherapy has favorable physical properties for avoiding nearby nontarget structures and is increasingly used for such lesions. Methods. Nineteen patients who underwent postoperative proton therapy at a single institution from 2006 to 2012 were identified including 13 with chordoma and 6 with chondrosarcoma. Surgical approach varied by tumor location in the cervical (n = 3), thoracic (n = 1), lumbar (n = 2), or sacral (n = 13) spine. Eight patients were categorized as receiving “early adjuvant” and 11 patients as receiving “salvage” treatment, as determined by initiation of radiation therapy after primary surgery or local recurrence, respectively. The median radiation dose delivered was 70 Gy relative biologic effectiveness (range: 56–78 Gy relative biologic effectiveness).
From *Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX; †School of Medicine, The University of Texas Health Science Center San Antonio, San Antonio, TX; ‡Department of Orthopaedic Surgery, The University of Texas Health Science Center San Antonio, San Antonio, TX; and §Department of Neurosurgery, The University of Texas MD Anderson Cancer Center, Houston, TX. Acknowledgment date: August 21, 2014. First revision date: December 1, 2014. Second revision date: January 12, 2015. Acceptance date: January 16, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No relevant financial activities outside the submitted work. Address correspondence and reprint requests to David R. Grosshans, MD, PhD, Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center, 1840 Old Spanish Trail, Unit 1150, Houston, TX 77054; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000804
544
Results. For the entire cohort, 2-year local control, relapsefree survival, and overall survival were 58%, 51.9%, and 93.3%, respectively. The early adjuvant group had significantly higher 2-year local control (80% vs. 45.5%; P = 0.024). Conclusion. Patients referred early for primary adjuvant radiation therapy after surgery had higher rates of disease control than those referred for salvage treatment of recurrent disease. Recurrence rates in our cohort were higher overall than other published series, indicating that even higher radiation doses may be helpful for further improving local control in the presence of gross or recurrent disease. Key words: chordoma, chondrosarcoma, spine, adjuvant radiotherapy, salvage, proton. Level of Evidence: 3 Spine 2015;40:544–549
C
hordomas and chondrosarcomas are rare, slow-growing neoplasms that are difficult to control because of their invasive nature and proximity to critical structures precluding widely negative surgical margins.1 The methods taken to treat these tumors have evolved over time. Intralesional resections were once considered to be the standard of care of spinal oncology surgery, but they have fallen out of favor for management of chordomas because of their inability to ensure complete removal of the tumor with negative margins. En bloc resection is now the preferred approach as it allows preservation of the tumor margin, ensuring a complete oncologic resection.2 Although ideal, en bloc resections carry great risks to many important structures and thus are not always feasible. Adjuvant treatment with radiation therapy (RT) may improve local control (LC) and potentially overall survival (OS)3,4; however, the proximity of the operative bed to critical structures such as the spinal cord and exiting nerve roots has the potential to add considerable toxicity. In addition to location, the relative resistance of chordomas and chondrosarcomas to radiation necessitates higher doses of RT in order to achieve LC.5,6 The dose of radiation able to be delivered to the tumor is limited by the tolerance of the spinal cord, which is typically limited to a maximum of 50 Gray (Gy), a dose level with a minimal risk of paralysis.7 Chordomas and
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 544
14/03/15 1:46 PM
CLINICAL CASE SERIES chondrosarcomas have been shown to require doses in excess of this spinal cord tolerance. Treatment with photon RT to doses of less than 60 Gy results in eventual recurrence rates of 50% to 100% and 5-year progression-free survival rates of less than 25%.8,9 Although 60 Gy is considered inadequate for durable LC, that dose nevertheless risks significant toxicity.10 Particle therapy, including proton beam radiation (PBR), has unique physical properties that may spare normal tissues from unnecessary exposure. Protons have a characteristic depth dose profile notable for a sharp increase in dose deposited at the end of the particle range, a phenomenon known as the Bragg peak. Because protons stop in tissue at the Bragg peak, essentially no exit dose is delivered beyond the target. This is especially helpful when treating tumors adjacent to critical structures. The aim of this study was to evaluate patients treated with PBR for chordomas and chondrosarcomas of the spine in the postoperative setting and to report LC, relapse-free survival (RFS), and OS outcomes. Prognostic factors for LC were also evaluated, specifically, the timing of PBR after surgery.
MATERIALS AND METHODS Patients with chordomas or chondrosarcomas of the cervical, thoracic, lumbar, and sacral spine treated with postoperative PBR at a single, tertiary referral institution between January 1, 2006, and April 1, 2012, were identified from a prospectively collected, institutional review board–approved database. Nineteen patients were identified with tumors in the cervical (n = 3), thoracic (n = 2), lumbar (n = 1), and sacral (n = 13) spine. Prior to treatment, all had their pathology reviewed at our institution: 13 with chordoma and 6 with chondrosarcoma. Patient demographics such as age at first treatment, sex, and race were collected from the medical record. The median radiation dose delivered was 70 Gy relative biologic effectiveness (GyRBE) (range: 56–78 GyRBE) delivered in 2 GyRBE per daily fraction. In 5 patients, RT was delivered using a combined passive scatter proton and photon-based intensity modulated radiation therapy. One patient received 44 GyRBE using a passive scatter proton technique, followed by a stereotactic photon boost of 12 Gy in 3 fractions. Fourteen patients received all of their radiation treatment with passive scatter protons. Dosimetric data were reviewed for 17 of the 19 patients because the 2 patients treated with combined proton-photon plans did not have composite plans available. The target was defined as gross tumor volume in the case of patients with gross residual disease and the postoperative bed or clinical target volume in the case of patients without gross residual disease. Median and range for mean dose to the target and minimum point dose to the target were 73.6 GyRBE (61.7–80.1 GyRBE) and 61.9 GyRBE (11.3–78.7 GyRBE), respectively. Median and range maximum point dose to the spinal cord for the 7 treatment plans treating tumors at the level of the cord were 52.3 GyRBE (43.1–59.1 GyRBE). Disease target was designated as gross disease or microscopic disease based on information from operative notes, pathology reports, and postoperative magnetic resonance images. There were no patients who had a margin-negative Spine
Postoperative Protons for Spine Chordoma • Holliday et al
resection. Finally, the presence or absence of implanted surgical instrumentation was recorded from operative notes. Patients were categorized as having received “early adjuvant” radiation if they received PBR after primary surgery and had obtained a postoperative magnetic resonance image showing a gross total resection or had known gross residual disease that was radiographically stable after surgery. Patients were categorized as having received “salvage” radiation if they received radiation after recurrence was documented by new or progressive disease visible on magnetic resonance image. By this definition, 8 patients (42.1%) received PBR in the early adjuvant setting and 11 (57.9%) received PBR in the salvage setting after documented recurrence or progression. Longer-term clinical outcomes at last follow-up including pain level on the 10-point comparative pain scale, narcotic requirement, the presence or absence of incontinence, Karnofsky Performance Status, and whether or not extremity weakness caused functional impairment were also collected when available. LC, RFS, and OS were measured from the start date of radiation until the event of interest. Chi-square test was used for between group comparisons and Mann-Whitney U test was used to compare medians. Data were analyzed using JMP, Version 7 (SAS Institute Inc., Cary, NC).
RESULTS Figure 1 depicts a representative sagittal (A) and axial (B) image from a treatment plan, delivered using passive scatter proton therapy for a sacral chordoma.For the group studied, acute radiation-related toxicities were mild and consisted of 13 patients with grade 1 radiation dermatitis (defined by the Radiation Therapy Oncology Group’s Acute Morbidity Scoring Criteria as mild, blanchable, erythema, or dry desquamation), 3 patients with grade 2 radiation dermatitis (defined as moist desquamation limited to the skin folds), and 4 patients treated for cervical spine tumors who developed grade 1 to 2 esophagitis (defined as mild to moderate dysphagia requiring non-narcotic analgesics and soft diet or narcotic analgesics and puree/liquid diet, respectively). No patient developed grade 3 to 5 toxicity or required a treatment break. Median follow-up from the time of PBR initiation to the time of last follow-up or death was 34.5 months. Overall, 7 patients experienced freedom from local failure at the time of data analysis in June 2014. Ten had progression of local disease, 1 patient had progression of local disease and distant failure, 1 patient experienced distant failure only, and 4 patients died of medical conditions not related to their cancer. Therefore, the overall 2-year LC, RFS, and OS for this cohort are 58%, 51.9%, and 93.3%, respectively. Eight patients received early adjuvant PBR and 11 patients received salvage PBR. Overall, the median time from surgery to initiation of radiation was 9.5 months (range: 1.3–110.2 mo). Table 1 outlines patient characteristics, surgery, and radiation details for both groups. Patients in the early adjuvant group had a shorter median interval from surgery to initiation of radiation (2.5 mo [range: 1.3–9.5 mo] vs. 32.6 mo [range: 3.7–110.2 mo]; P = 0.0003). In the early adjuvant group, there were significantly more nonsacral tumors (P = 0.0005). www.spinejournal.com
545
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 545
14/03/15 1:46 PM
CLINICAL CASE SERIES
Postoperative Protons for Spine Chordoma • Holliday et al
TABLE 1. Characteristics of Patients Receiving
Early Adjuvant Versus Salvage Proton Radiation Early Adjuvant (N = 8)
Salvage (N = 11)
P
46 (10–71)
57 (41–82)
0.12
Male, n (%)
4 (50%)
5 (54.5%)
Female, n (%)
4 (50%)
6 (45.5%)
Chordoma, n (%)
4 (50%)
9 (81.8%)
Chondrosarcoma, n (%)
4 (50%)
2 (18.2%)
7 (87.5%)
9 (72.7%)
0 (0%)
1 (9.1%)
1 (12.5%)
1 (9.1%)
Nonsacral
6 (75%)
0 (0%)
Sacral, n (%)
2 (25%)
11 (100%)
2.5 (1.3–9.5)
32.6 (3.7–110.2)
Gross disease, n (%)
3 (37.5%)
11 (100%)
Microscopic disease, n (%)
5 (62.5%)
0 (0%)
Absent
5 (62.5%)
11 (100%)
Present
3 (37.5%)
0 (0%)
Protons only, n (%)
5 (62.5%)
9 (81.8%)
Protons and photons, n (%)
3 (37.5%)
2 (18.2%)
70 (56–70.2)
70 (60–78)
No
6 (75%)
1 (9%)
Yes
2 (25%)
10 (91%)
Age at first treatment Median years (range) Sex 0.84
Histology 0.14
Self-reported race Caucasian non-Hispanic, n (%) Caucasian Hispanic, n (%) Asian, n (%)
0.67
Disease site Figure 1. Representative treatment plan using proton beam radiation in the postoperative setting. Sagittal (A) and axial (B) sections are depicted, which highlight sparing of normal tissues including bowel located distal to the beams.
Surgery to XRT interval Median months (range)
All 11 patients in the salvage group had gross disease at the time of radiation initiation whereas only 37.5% of patients in the early adjuvant group had gross disease (P = 0.0003). More patients in the early adjuvant group had metal instrumentation placed than patients in the salvage group (37.5% vs. 0%; P = 0.05), but there was no significant difference in the radiation technique used or the median dose delivered. Twenty-five percent of patients had a documented relapse of disease in the early adjuvant group compared with 91% in the salvage group at the time of data analysis (P = 0.006). Two-year LC was 88% in the early adjuvant group compared with 45% in the salvage group (P = 0.061). Five-year LC was 88% in the early adjuvant group compared with 9% in the salvage group (P < 0.001). Two-year RFS was 75% in the early adjuvant group compared with 45% in the salvage group (P = 0.198). Two-year OS was 88% in the early adjuvant group compared with 100% in the salvage group (P = 0.228). On univariate analysis, sacral disease site (P = 0.004), longer interval between surgery and initiation of radiation (P = 0.025), and gross disease present at the initiation of radiation (P = 0.024) were all found to be associated with higher rates of recurrence. There was a trend toward increased recurrence 546
0.0005
0.0007
Disease target 0.0003
Surgical Instrumentation 0.05
XRT technique 0.603
XRT dose Median (range) in Gy (RBE)
0.25
Ultimate relapse after XRT 0.006
The values in bold face indicate P ≤ 0.05. XRT indicates radiation therapy.
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 546
14/03/15 1:46 PM
CLINICAL CASE SERIES
Postoperative Protons for Spine Chordoma • Holliday et al
TABLE 2. Prognostic Factors for Recurrence-
TABLE 3. Clinical and Functional Outcomes
Free Survival
No Recurrence Recurrence (N = 12) (N = 7)
for Patients Receiving Early Adjuvant Versus Salvage Proton Radiation Early Adjuvant (n = 5)
Salvage (n = 10)
P
5 (0–6)
3 (0–8)
0.95
Yes
3 (60%)
7 (70%)
No
2 (40%)
3 (30%)
Yes
1 (20%)
4 (40%)
No
4 (80%)
6 (60%)
80 (60–100)
80 (70–100)
Yes
1 (20%)
6 (60%)
No
4 (80%)
4 (40%)
P
Histology Chordoma, n = 13
10 (76.9%)
3 (23.1%)
Chondrosarcoma, n=6
2 (33.3%)
4 (66.7%)
Pain level 0.07
Narcotic requirement, n (%)
Disease site Nonsacral, n = 6
1 (16.7%)
5 (83.3%)
Sacral, n = 13
11 (84.6%)
2 (15.4%)
24.75 (1.8–110.2)
2.5 (1.3–32.2)
0.025
Gross disease, n = 14
11 (78.6%)
3 (21.4%)
0.025
Microscopic disease, n=5
1 (20%)
4 (80%)
0.004
Surgical instrumentation 1 (33.3%)
2 (66.7%)
Absent, n = 16
11 (68.8%)
5 (31.2%)
Median (range)
0.62
Weakness affecting function 0.14
0.24
XRT indicates radiation therapy.
rates in chordoma patients when compared with chondrosarcoma patients (76.9% vs. 23.1%; P = 0.07). Table 2 outlines these prognostic factors. Clinical outcomes were available for 15 of the 19 patients included, 5 patients receiving early adjuvant and 10 patients receiving salvage radiation. There were no significant differences in any of the clinical outcomes collected (Table 3).
DISCUSSION This retrospective comparative series reports results of patients treated with postoperative PBR for chordomas and chondrosarcomas of the spine. Patients who received early adjuvant PBR were compared with those who were treated in the salvage setting. Patients treated in the early adjuvant setting had better 2-year LC. Both increased time between surgery and radiation as well as the presence of gross residual disease at the time of radiation were adverse prognostic factors. First studied in the base of skull,11–16 few series have evaluated PBR for chordomas and chondrosarcomas of the spine.17–22 Yasuda et al17 from the Institute Curie in France reported 75% LC for patients with chordomas of the skull base and spine who received PBR or combined photonproton RT to a mean dose of 68.9 Gy (RBE) after surgical resection. Rutz et al20 from the Paul Scherrer Institute in Switzerland reported 5 local failures out of 26 patients with extracranial chordomas who received postoperative Spine
0.44
Karnofsky performance status
Disease target
Present, n = 3
0.70
Incontinence, n (%)
Surgery to XRT interval Median (range) in months
Median (range)
spot-scanning PBR to a median dose of 72 Gy (RBE). Gross total resection was achieved in 69% of patients. The group from Massachusetts General Hospital has published most extensively on the subject.18,19,21,22 Hug et al18 reported 53% and 100% 5-year LC rates for chordoma and chondrosarcomas of the axial skeleton, respectively, with combined photonproton RT to a mean dose of 73.9 Gy (RBE). A more recent publication from Massachusetts General Hospital outlined their experience with definitive PBR in medically inoperable or otherwise unresected chordomas of the spine and reported 80% 5-year local progression–free survival.19 DeLaney et al reported results from a phase II study of 50 patients with spine sarcomas (29 had chordoma and 14 chondrosarcoma) treated with preoperative and/or postoperative PBR. Five-year LC was 78% for the entire cohort. Of note, 50% of patients in this cohort received a margin-negative resection. On the Massachusetts General Hospital phase II study, PBR doses were higher than those in the previously mentioned studies, with 70.2 Gy (RBE) prescribed for microscopic disease and 77.4 Gy (RBE) for gross disease.21 The same group recently published long-term results including 74% 8-year LC.22 Our overall LC results were lower than those in the majority of previously published series, with a 2-year LC rate of 58% for all patients. This may be partially attributable to the relatively low median dose of 70 Gy (RBE) that was delivered. Also potentially contributory is the fact that 58% (N = 11) of patients included in our series were treated for recurrence after primary resection. Indeed, patients receiving PBR in the salvage setting had a significantly lower 2-year LC of 45.5%. Other groups reported similarly poor LC for patients treated www.spinejournal.com
547
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 547
14/03/15 1:46 PM
CLINICAL CASE SERIES in the salvage setting. Delaney et al22 reported a 53% 5-year actuarial rate of local recurrence in previously recurrent tumors treated on protocol compared with 6% in those with primary tumors (P = 0.002). Patients in our series receiving early adjuvant PBR had a 2-year LC rate of 80%, which is more comparable with other published data. Finally, a margin-negative resection has been shown by several groups to be a positive prognostic factor,18,20–22 and none of the patients in our series received R0 resections compared with 50% to 75% in the other series cited. In addition to treatment in the salvage setting and increased interval from surgery to radiation, adverse prognostic factors identified in this study included the presence of gross residual disease18,20,22 and sacral tumor location,20 which have also been reported in other studies. In our series, patients with sacral tumors more often had gross disease present at the time of radiation initiation (n = 12, 92%) than patients with nonsacral tumors (n = 2, 33%; P = 0.005). Although some fear that giving higher doses of radiation could cause damage to the exiting sacral nerve roots, there was no difference in median dose delivered between patients with sacral and nonsacral tumors (70 [range: 60–78] and 70 [range: 54–70], respectively). Rutz et al20 reported the presence of implants placed at surgery to be associated with increased local failure. Several physical factors may explain this, including computed tomography artifacts that make accurate target delineation more difficult during PBR treatment planning as well as dose in homogeneities and “cold spots” caused by interactions with the proton beam and the implant itself. We did not find worse LC in patients with surgical implants; however, this may be due to low numbers in our series. Finally, chordomas and chondrosarcomas are distinct histological and clinical entities but were included together in this case series to be consistent with the existing, albeit limited, published data published from other proton therapy centers reporting on their patients with tumors of the spine and/or skull base.13–15,18,21 It should be emphasized that chondrosarcomas have a better prognosis than chordomas in terms of response to treatment and survival. Studies including patients with skull base sarcomas reported 5-year recurrence-free survival rates of 90% versus 65% for chordomas and chondrosarcomas, respectively; P = 0.09.1 Although numbers were too small and follow-up too short to detect any differences in survival in our series, patients with chondrosarcomas had a similar trend toward decreased local recurrence when compared with patients with chordomas (n = 2, 33% and n = 10, 77%, respectively [P = 0.07]). This was despite the trend toward a lower median dose given to patients with chondrosarcomas when compared with chordomas (65 (range: 54–70) and 70 (range: 66–78); P = 0.08). As with all retrospective case series, several inherent potential biases exist in the interpretation of these data. The potential for selection bias is high in this small, retrospectively analyzed cohort. Our status as a large, international cancer center attracts patients with multiply recurrent disease or those who were not offered treatment elsewhere, which may reflect inherently worse biology and contribute to our lower control rates compared with other published series. Recurrent tumor may 548
Postoperative Protons for Spine Chordoma • Holliday et al
also have inherently more resistant biology than primary disease, which introduces another potentially confounding factor. Moreover, many patients underwent surgical resections outside of our institution and as such the surgical approach was not standardized. The comparison between patients treated in the adjuvant setting versus in the salvage setting also has the potential for inherent bias, such that patients who recur after resection may also have more aggressive biology. Finally, the presence of gross disease at the time of radiation is a potential confounding factor in evaluating the importance of the timing of PBR after surgery because other studies have shown extent of disease and margin status to be adverse prognostic factors.22–24 The best way to minimize selection bias in future studies and to obtain higher quality evidence to assist in counseling patients with chordoma and chondrosarcoma would be to perform a randomized controlled trial in which patients are randomly assigned to receive adjuvant RT in the early postoperative period or in the salvage setting. However, the rarity of this tumor precludes such a study from being realistically performed. Therefore, we propose that patients with chordoma or chondrosarcoma of the spine be enrolled on a prospective, non–randomized study protocol from which consecutive patients could be analyzed. Ideally, this protocol would be opened at multiple centers to provide potentially more generalizable results. Chordomas and chondrosarcomas of the spine are challenging to manage because of their propensity for local recurrence. Postoperative radiation can decrease local recurrence, but a longer interval between surgery and radiation as well as gross disease present at the time of radiation are adverse prognostic factors. Therefore, to increase the effectiveness of this adjuvant therapy, we suggest that patients be referred for consideration of radiation in the early postoperative period prior to disease recurrence or progression. In addition, further consideration should be given to strategies that might further improve LC, including escalation of radiation dose and utilization of more conformal radiation techniques such as intensity-modulated proton therapy. In addition, longer follow-up is needed to monitor and report the late effects and functional outcomes of combined modality therapy.
➢ Key Points In the absence of adequate therapy, chordomas and chondrosarcomas are prone to local recurrence; however, their anatomic location makes local control challenging. In this series, patients who received early adjuvant proton therapy had significantly better freedom from local failure than patients who received salvage radiotherapy after a local recurrence. Patients with sacral tumors, a longer surgery to radiation interval, and gross disease present at the time of radiation were associated with higher rates of recurrence.
www.spinejournal.com
April 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 548
14/03/15 1:46 PM
CLINICAL CASE SERIES References
1. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 patients. Neurosurgery 1995;36:887–97. 2. Chi JH, Sciubba DM, Rhines LD, et al. Surgery for primary vertebral tumors: en bloc versus intralesional resection. Neurosurg Clin N Am 2008;19:111–7. 3. Boriani S, Bandiera S, Biagini R, et al. Chordoma of the mobile spine: fifty years of experience. Spine (Phila Pa 1976) 2006;31: 493–503. 4. Fuller DB, Bloom JG. Radiotherapy for chordoma. Int J Radiat Oncol Biol Phys 1988;15:331–9. 5. Zagars GK, Ballo MT. Significance of dose in postoperative radiotherapy for soft tissue sarcoma. Int J Radiat Oncol Biol Phys 1995;31:1093–112. 6. DeLaney TF, Kepka L, Goldberg SI, et al. Radiation therapy for control of soft tissue sarcomas resected with positive margins. Int J Radiat Oncol Biol Phys 2007;67:1460–9. 7. Schultheiss TE, Kun LE, Ang KK, et al. Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 1995;31: 1093–112. 8. Catton C, O’Sullivan B, Bell R, et al. Chordoma: long-term follow-up after radical photon irradiation. Radiother Oncol 1996;41: 67–72. 9. Zorlu F, Gürkaynak M, Yildiz F, et al. Conventional external radiotherapy in the management of clivus chordomas with overt residual disease. Neurol Sci 2000;21:203–7. 10. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlenther Onkol 1999;175(suppl 2):57–63. 11. Noël G, Feuvret L, Calugaru V, et al. Chordomas of the base of the skull and upper cervical spine. One hundred patients irradiated by a 3D conformal technique combining photon and proton beams. Acta Oncol 2005;44:700–8. 12. McDonald MW, Linton OR, Moore MG, et al. Initial results of proton therapy for clival chordoma. Int J Radiat Oncol Biol Phys 2013;87:S277–8. 13. Ares C, Hug EB, Lomax AJ, et al. Effectiveness and safety of spot scanning proton radiation therapy for chordomas and chondrosarcomas of the skull base: first long-term report. Int J Radiat Oncol Biol Phys 2009;75:1111–8.
Spine
Postoperative Protons for Spine Chordoma • Holliday et al
14. Rutz HP, Weber DC, Goitein G, et al. Postoperative Spot-Scanning proton radiation therapy for chordoma and chondrosarcoma in children and adolescents: initial experience at Paul Scherrer Institute. Int J Radiat Oncol Biol Phys 2008;71:220–5. 15. Rombi B, Ares C, Hug EB, et al. Spot-scanning proton radiation therapy for pediatric chordoma and chondrosarcoma: clinical outcome of 26 patients treated at Paul Scherrer Institute. Int J Radiat Oncol Biol Phys 2013;86:578–84. 16. McDonald MW, Linton OR, Shah MV. Proton therapy for reirradiation of progressive or recurrent chordoma. Int J Radiat Oncol Biol Phys 2013;87:1107–14. 17. Yasuda M, Bresson D, Chibbaro S, et al. Chordomas of the skull base and cervical spine: clinical outcomes associated with a multimodal surgical resection combined with proton-beam radiation in 40 patients. Neurosurg Rev 2012;35:171–82; discussion 182–3. 18. Hug EB, Fitzek MM, Liebsch NJ, et al. Locally challenging osteoand chondrogenic tumors of the axial skeleton: results of combined proton and photon radiation therapy using three-dimensional treatment planning. Int J Radiat Oncol Biol Phys 1995;31:467–76. 19. Chen YL, Liebsch N, Kobayashi W, et al. Definitive high-dose proton/photon radiotherapy for unresected mobile spine and sacral chordomas. Spine (Philadelphia, PA 1976) 2013;38:E930–6. 20. Rutz HP, Weber DC, Sugahara S, et al. Extracranial chordoma: outcome in patients treated with function-preserving surgery followed by spot-scanning proton beam irradiation. Int J Radiat Oncol Biol Phys 2007;67:512–20. 21. DeLaney TF, Liebsch NJ, Pedlow FX, et al. Phase II study of high dose photon/proton radiotherapy in the management of spine sarcomas. Int J Radiat Oncol Biol Phys 2009;74:732–9. 22. DeLaney TF, Liebsch NJ, Pedlow FX, et al. Long-term results of phase II study of high dose proton/photon radiotherapy in the management of spine chordomas, chondrosarcomas and other sarcomas. J Surg Oncol 2014;110:115–22. 23. Lee J, Bhatia NN, Hoang BH, et al. Analysis of prognostic factors for patients with chordoma with use of the California Cancer Registry. J Bone Joint Surg Am 2012;94:356–63. 24. Bergh P, Kindblom LG, Gunterberg B, et al. Prognostic factors in chordoma of the sacrum and mobile spine: a study of 39 patients. Cancer 2000;88:2122–34.
www.spinejournal.com
549
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE141126_LR 549
14/03/15 1:46 PM
Copyright of Spine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
