Abstract
Background:
We previously reported that the cumulative risk of femoral fracture in patients treated with intensity-modulated radiation therapy (IMRT) for thigh and groin soft tissue sarcoma (STS) is low. In the current study, we sought to evaluate the effect of radiation dose constraints on the rate of femoral fracture in a more contemporary cohort.
Methods:
All patients treated with IMRT for soft tissue sarcoma of the thigh or groin from 2004–2016 were included (n = 145). Beginning in 2011, radiation dose was constrained to a mean dose < 37 Gy, volume of bone receiving ≥ 40 Gy (V40Gy) < 64%, and maximum dose < 59 Gy to limit the dose to the femur.
Results:
Sixty-one patients were treated before dose constraints were implemented, and 84 after. Median follow-up for patients treated before and after constraints were implemented was 6.1 and 5.7 years, respectively, and the two groups were demographically and clinically similar. On univariate analysis, the 5-year cumulative incidence of femoral fracture among patients treated with and without dose constraints was 1.8% (95% CI, 0.3–12.2%) versus 7.4% (95% CI, 3.1–17.6%), respectively (p = 0.11, non-significant). On multivariable analysis, only age ≥ 60 years was significantly associated with increased risk of fracture.
Conclusions:
The risk of femoral fracture after IMRT for soft tissue sarcoma of the thigh/groin is low, and with the implementation of radiation dose constraints, the risk is < 2%. Although longer follow-up is needed, our results support the utilization of extremity sarcoma IMRT-specific dose constraints for fracture prevention.
Keywords: Extremity sarcoma, femur, IMRT, fracture, dose constraints
Introduction
Standard treatment for soft tissue sarcoma of the extremity involves limb-sparing surgery with either preoperative or postoperative radiation. Although this combination results in excellent local control and survival, radiation-associated femoral fractures remain one of the more serious and disabling consequences of radiation for extremity soft tissue sarcoma, with an incidence of approximately 10%.1–4 The rate of radiation-induced femoral fractures is affected by clinical and treatment factors such as age, gender, tumor location, extent of periosteal stripping at the time of resection, receipt of chemotherapy, and radiation dose.1,5
Of these factors, radiation dose to the femur is easily controllable. Prior studies have shown an effect of dose on the risk of femoral fracture,5–7 suggesting that dose constraints may be beneficial. The adaptation and utilization of intensity-modulated radiation therapy (IMRT) for extremity soft tissue sarcoma allows delivery of therapeutic doses to the tumor while constraining radiation dose to the femur. However, there has never been a study specifically evaluating whether dose constraints to the femur that have been recently implemented are actually effective in minimizing the risk of bone fracture. In the current study, we thus sought to determine whether radiation dose constraints to the femur implemented in 2011 were effective in reducing the incidence of femoral fracture after radiation for extremity soft tissue sarcoma.
Methods
Patients
Following Institutional Review Board approval, we reviewed the prospectively maintained surgical database at Memorial Sloan Kettering Cancer Center (MSK) to identify patients with primary soft tissue sarcoma of the thigh or groin treated with limb-sparing surgery and radiation between January 2004 and December 2016. Patients who received prophylactic internal fixation were excluded.
Radiation therapy
From 2004–2010, radiation therapy was delivered without the utilization of dose constraints to the bone as previously described.1 From 2011–2016, bone constraints of mean dose (Dmean) < 37 Gy, V40Gy < 64% and maximum dose (Dmax) < 59 Gy were introduced and utilized with the goal of reducing the incidence of femoral fracture. Constraints were successfully met in all cases from 2011–2016. For all patients in our cohort, the clinical tumor volume (CTV) excluded nearby bone, and expansion from the CTV to planning target volume (PTV) was approximately 0.5–1 cm. Radiation dose was independent of time or constraint use, with a median dose of 50 Gy used in the preoperative setting and 63 Gy in the postoperative setting. All patients were treated with IMRT.
Statistics
The primary objective of this study was to determine the cumulative incidence of femoral fracture pre and post dose constraint implementation. Secondary objectives were to evaluate other risk factors associated with femoral fracture such as gender, age, tumor size, location of the tumor, presence of periosteal stripping during resection, radiation dose prescribed, and receipt of chemotherapy. All outcomes were measured from time of surgery to time of event or last follow-up. Cumulative incidence functions were used to calculate the incidence of femoral fracture. Multivariable analysis was performed using Cox proportional hazard regression models and included factors that were significant on univariate analysis. Baseline patient and tumor characteristics were compared between those treated before and after implementation of constraints using chi-square tests.
Results
Patients
From January 2004 to December 2016, 145 patients underwent limb-sparing surgery and IMRT for soft tissue sarcoma of the thigh or groin at MSK. Of these, 84 (57.9%) were treated between 2004 and 2010 prior to the implementation of dose constraints, and 61 (42.1%) were treated between 2011 and 2016, after constraint adoption. Median age for patients treated pre-constraints was 56.5 years (range, 19.2–88.5) versus 56.6 years (range, 26.4–80.5) post-constraints. Median follow-up was similar between the two groups: 6.1 years (range, 0.7–12.4) pre-constraints versus 5.7 years (range, 0.8–9.1) post-constraints. Overall, patients treated during the two periods were similar with regards to commonly associated risk factors for femoral fracture including gender, age, tumor location (anterior vs. non-anterior), tumor size, periosteal stripping, and receipt of chemotherapy (Table 1). Fewer patients treated post-constraint implementation received > 60 Gy compared with patients treated pre-constraints (41, 67.2% vs. 69, 82.1%; p = 0.04).
Table 1.
Demographic, clinical, tumor, and treatment characteristics of patients treated before (2004–2010) and after the implementation of dose constraints (2011–2016).
| Total cohort | 2004–2010 | 2011–2016 | p-value | ||
|---|---|---|---|---|---|
| n (%) | n (%) | n (%) | |||
| Gender | Female | 46 (31.7%) | 31 (36.9%) | 15 (24.6%) | 0.12 |
| Male | 99 (68.3%) | 53 (63.1%) | 46 (75.4%) | ||
| Age | < 60 years | 82 (56.6%) | 47 (56.0%) | 35 (57.4%) | 0.87 |
| ≥ 60 years | 63 (43.4%) | 37 (44.0%) | 26 (42.6%) | ||
| Compartment | Anterior | 62 (42.8%) | 38 (45.2%) | 24 (39.3%) | 0.48 |
| Posterior, medial, or groin | 83 (57.2%) | 46 (54.8%) | 37 (60.7%) | ||
| Tumor size | < 10 cm | 61 (42.1%) | 35 (41.7%) | 26 (42.6%) | 0.91 |
| ≥ 10 cm | 84 (57.9%) | 49 (58.3%) | 35 (57.4%) | ||
| Periosteal stripping | No | 113 (80.1%) | 63 (78.8%) | 50 (82.0%) | 0.64 |
| Yes | 28 (19.8%) | 17 (21.3%) | 11 (18.0%) | ||
| Radiation dose | ≤ 60 Gy | 35 (24.1%) | 15 (17.9%) | 20 (32.8%) | 0.04 |
| > 60 Gy | 110 (75.9%) | 69 (82.1%) | 41 (67.2%) | ||
| Chemotherapy | Yes | 51 (35.2%) | 30 (35.7%) | 21 (34.4%) | 0.87 |
| No | 94 (64.8%) | 54 (64.3%) | 40 (65.6%) |
Fractures
Among the entire cohort, 7 patients developed a femoral fracture. All 7 patients were 59 years or older, with anterior tumor location in the thigh in all patients except one, who had a medial thigh tumor. Six of 7 patients received postoperative radiation to a dose of 63 Gy in 35 fractions; the 7th patient received standard preoperative radiation to a dose of 50 Gy in 25 fractions. Five of 7 patients were female, and 3 patients had periosteal stripping at the time of surgery. All femoral fractures were within the radiation field, and 6 of 7 patients with fracture were treated before the implementation of dose constraints in 2011. Among the 6 patients treated pre dose constraints, the median Dmax to the femur was 65.9 Gy (range 65.5 – 76 Gy), the median Dmean to the femur was 42.5 Gy (range 31.2 – 60.2 Gy), and the median volume of the femur irradiated to 40 Gy or more was 63% (range 38% – 91.6%). Looking specifically at the location of the fracture in the femur, the median Dmax was 65.6 Gy (range, 62.9 – 74.1 Gy) and median Dmean was 52.2 Gy (range, 46.3 – 65.7 Gy). Median time to fracture was 2.3 years (range 0.6–7.4 years) in the patients treated pre-implementation of dose constraints versus 1.9 years in the patient treated post-implementation of dose constraints.
At 5 years, the cumulative incidence of femoral fracture among the entire cohort was 4.9% (95% CI 2.2–11.0%) (Figure 1). The cumulative incidence among patients treated without dose constraints from 2004–2010 was 7.4% (95% CI, 3.1–17.6%) versus 1.8% (95% CI, 0.3–12.2%) among patients treated after dose constraints were implemented from 2011–2016 (p = 0.11). Additionally, among the subset of patients who received > 60 Gy, there continued to be a trend toward decreased risk of femoral fracture with the implementation of dose constraints (5-year cumulative incidence of fracture of 7.0% without constraints versus 2.6% with constraints, p = 0.26). Other factors that were significantly associated with increased risk of fracture on univariate analysis included female gender, age 60 years or older, and tumor location within the anterior thigh (Table 2). Of note, prescribed dose, independent of dose constraints, had no effect on the risk of femoral fracture (5-year cumulative incidence of fracture of 3.3% for dose ≤ 60 Gy vs. 5.4% for > 60 Gy, p = 0.53). On multivariable analysis, only older age (≥ 60 years) remained significantly associated with increased risk of fracture (hazard ratio = 8.1, p = 0.05).
Figure 1.
Cumulative rate of fracture over time among patients with soft tissue sarcoma of the thigh treated by limb-sparing surgery and IMRT with (n = 61) or without (n = 84) radiation dose constraints.
Table 2.
Risk of femoral fracture based on patient, tumor, and treatment variables.
| Crude risk | Cumulative risk over 5 years | 95% CI | P value | |
|---|---|---|---|---|
| Gender | ||||
| Female | 5/46 | 11.4% | 4.4–29.8% | 0.02 |
| Male | 2/99 | 2.3% | 0.6–9.0% | |
| Age | ||||
| Age < 60 years | 1/82 | 1.2% | 0.2–8.7 | 0.02 |
| Age ≥ 60 years | 6/63 | 10.3% | 4.4–24.2% | |
| Compartment | ||||
| Anterior | 6/62 | 9.7% | 4.1–22.8% | 0.02 |
| Other | 1/83 | 1.4% | 0.2–10.1% | |
| Tumor size | ||||
| < 10 cm | 1/61 | 0% | n/a | 0.07 |
| ≥ 10 cm | 6/84 | 9.2% | 4.2–20.2% | |
| Periosteal stripping | ||||
| Present | 3/28 | 12.6% | 4.2–37.5% | 0.06 |
| Absent | 4/117 | 3.3% | 1.1–10.1% | |
| Dose | ||||
| ≤ 60 Gy | 1/35 | 3.3% | 0.5–22.9% | 0.53 |
| > 60 Gy | 6/110 | 5.4% | 2.2–12.8% | |
| Chemotherapy | ||||
| No chemotherapy | 5/94 | 5.3% | 2.0–13.9% | 0.82 |
| Chemotherapy | 2/51 | 4.0% | 1–15.6% | |
| IMRT constraints utilized | ||||
| No (2004–2010) | 6/84 | 7.4% | 3.1–17.6% | 0.11 |
| Yes (2011–2016) | 1/61 | 1.8% | 0.3–12.2% |
Discussion
We found the overall incidence of femoral fracture after IMRT for soft tissue sarcoma of the thigh or groin to be < 2% following the implementation of radiation dose constraints. Compared with those treated at our institution prior to the implementation of dose constraints, and compared with rates in other large series including patients treated without constraints, this low rate of fracture suggests the utility of dose constraints specific to extremity sarcoma IMRT for fracture prevention.
This study builds on our previous finding that IMRT lowers the risk of femoral fracture from 25.6% as calculated via the Princess Margaret Hospital (PMH) nomogram to only 6.7%.1 As IMRT allows for lower circumferential dosing to the femur, it follows that such adaptation of IMRT should reduce the risk of femoral fracture. However, the efficacy of constraining the dose to the femur had not been directly tested.
The dose constraints utilized in our cohort (Dmean < 37 Gy, V40Gy < 64%, and Dmax < 59 Gy) were based on previous retrospective studies evaluating the radiation dose and volume parameters that affect the risk of fracture. Specifically, Dickie et al. showed that the patients who developed fracture had a Dmean and Dmax to the femur of 44 Gy and 65 Gy, versus 38 Gy and 59 Gy in those who did not,6 and Holt et al. reported a fracture rate of 9% with > 60 Gy versus 1% for those treated to 50 Gy.7 Pak et al. found that all fractures occurred after a mean dose exposure of more than 40 Gy to the femur;8 and Bishop et al. found the rate of femur fracture to be as high as 37% when the 50-Gy isodose line encompassed the entire bone circumference in the presence of other risk factors such as bone exposure, periosteal stripping, and perioperative chemotherapy.5 In our cohort of patients who were treated before dose constraint implementation and developed fracture, the mean radiation dose to the femur was 42.5 Gy, the Dmax was 65.9 Gy, and V40Gy was 63% (median values). Overall, although the median V40Gy < 64% was achieved in this group, the Dmax and Dmean would not have met the dose constraints implemented in 2011. Dmax, Dmean, and V40Gy constraints all prioritize different dose parameters and are likely all important for contributing to the risk for fracture. Therefore, ideally all should be achieved.
The studies discussed above guided the creation and implementation of the current dose constraints utilized at our institution and nationally of Dmean < 37 Gy, V40Gy <64%, and Dmax < 59 Gy. This is the first study, though, to our knowledge, that has actually tested the efficacy of these constraints in preventing fracture. With the implementation of our dose constraints, it is encouraging that the risk of fracture is reduced to a level <2% in an otherwise at-risk population including many patients with other traditional risk factors associated with fracture. This suggests that even with risk factors such as older age, female gender, periosteal stripping, and receipt of chemotherapy, constraining radiation dose to the femur controls the rate of femoral fracture. Importantly, prescribed dose had no effect on risk of femoral fracture in our cohort, and among the subset of patients prescribed > 60 Gy, the benefit of dose constraints remained apparent. This suggests that the dose delivered to the femur (and not the prescribed dose to the tumor or tumor bed) is the contributing factor to the risk of femoral fracture.
Longer follow-up beyond 6 years is needed to definitively determine the long-term rate of femoral fracture using dose constraints. However, it is encouraging that the rate of femoral fracture with the implementation of dose constraints is < 2% in a more contemporary, but still at-risk cohort of patients. In conclusion, our results support the utilization of extremity sarcoma IMRT-specific dose constraints for fracture prevention. Given this low rate of fracture with IMRT and the ability to achieve dose constraints, an important next step is to re-consider who still remains at high enough risk for fracture to warrant prophylactic internal fixation.
Synopsis:
Femoral fractures represent one of the more serious consequences of radiation for extremity soft tissue sarcomas. Here, we present IMRT-specific radiation dose constraints that result in a fracture risk of < 2%, supporting their utilization for fracture prevention.
Acknowledgments
This research was supported by the NIH/NCI grant P50 CA217694 (SPORE in Soft Tissue Sarcoma) and the NIH/NCI Cancer Center Support Grant P30 CA008748.
Disclosures: This work was supported by the National Cancer Institute via the SPORE in Soft Tissue Sarcoma (P50 CA140146 to SS) and the Cancer Center Support Grant P30 CA008748. None of the authors have financial relationships to disclose. None of the authors have a personal or institutional financial interest in the materials or devices described in this submission.
Footnotes
Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.
References
- 1.Folkert MR, Casey DL, Berry SL, et al. Femoral Fracture in Primary Soft-Tissue Sarcoma of the Thigh and Groin Treated with Intensity-Modulated Radiation Therapy: Observed versus Expected Risk. Ann Surg Oncol. 2019;26(5):1326–1331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Blaes AH, Lindgren B, Mulrooney DA, Willson L, Cho LC. Pathologic femur fractures after limb-sparing treatment of soft-tissue sarcomas. J Cancer Surviv. 2010;4(4):399–404. [DOI] [PubMed] [Google Scholar]
- 3.Helmstedter CS, Goebel M, Zlotecki R, Scarborough MT. Pathologic fractures after surgery and radiation for soft tissue tumors. Clin Orthop Relat Res. 2001(389):165–172. [DOI] [PubMed]
- 4.Alektiar KM, Zelefsky MJ, Brennan MF. Morbidity of adjuvant brachytherapy in soft tissue sarcoma of the extremity and superficial trunk. Int J Radiat Oncol Biol Phys. 2000;47(5):1273–1279. [DOI] [PubMed] [Google Scholar]
- 5.Bishop AJ, Zagars GK, Allen PK, et al. Treatment-related fractures after combined modality therapy for soft tissue sarcomas of the proximal lower extremity: Can the risk be mitigated? Pract Radiat Oncol. 2016;6(3):194–200. [DOI] [PubMed] [Google Scholar]
- 6.Dickie CI, Parent AL, Griffin AM, et al. Bone fractures following external beam radiotherapy and limb-preservation surgery for lower extremity soft tissue sarcoma: relationship to irradiated bone length, volume, tumor location and dose. Int J Radiat Oncol Biol Phys. 2009;75(4):1119–1124. [DOI] [PubMed] [Google Scholar]
- 7.Holt GE, Griffin AM, Pintilie M, et al. Fractures following radiotherapy and limb-salvage surgery for lower extremity soft-tissue sarcomas. A comparison of high-dose and low-dose radiotherapy. J Bone Joint Surg Am. 2005;87(2):315–319. [DOI] [PubMed] [Google Scholar]
- 8.Pak D, Vineberg KA, Griffith KA, et al. Dose--effect relationships for femoral fractures after multimodality limb-sparing therapy of soft-tissue sarcomas of the proximal lower extremity. Int J Radiat Oncol Biol Phys. 2012;83(4):1257–1263. [DOI] [PubMed] [Google Scholar]