Practical Radiation Oncology (2015) xx, xxx–xxx
www.practicalradonc.org
Original Report
Postoperative radiation therapy for osseous metastasis: Outcomes and predictors of local failure Zachary D. Epstein-Peterson MD a , Adam Sullivan MS b , Monica Krishnan MD a, c , Julie T. Chen BA c , Marco Ferrone MD d , John Ready MD d , Elizabeth H. Baldini MD MPH a, c, e , Tracy Balboni MD MPH a, c, e,⁎ a
Harvard Medical School, Boston, Massachusetts Harvard School of Public Health, Boston, Massachusetts c Department of Radiation Oncology, Brigham and Women’s Hospital and Dana-Farber Cancer Institute, Boston, Massachusetts d Department of Orthopedic Surgery, Brigham and Women’s Hospital, Boston, Massachusetts e Department of Psychosocial Oncology and Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts b
Received 7 November 2014; revised 10 February 2015; accepted 13 February 2015
Abstract Purpose: To evaluate patterns and predictors of local failure in patients undergoing postoperative radiation therapy (RT) for osseous metastases. Methods and materials: Patients undergoing postoperative RT for bone metastases between June 2008 and January 2012 were retrospectively reviewed. Patterns of local failure were assessed, and Fine and Gray’s univariable and multivariable analyses (MVA) were used to evaluate factors associated with local progression, including dose intensity of RT (biological equivalent dose, BED, Gy10) and percent coverage of the surgical hardware by the RT fields. Additional predictors were similarly assessed, including patient (eg, age, performance status), disease (eg, tumor type, metastasis site), and treatment (eg, interval from surgery to RT) characteristics. Results: A total of 82 cases were followed for a median of 4.3 months (11.5 months among living patients) after treatment completion. Median BED was 39 Gy10 (range, 14-60), and RT fields covered an average of 71% (standard deviation, 26%) of the hardware. Fourteen cases (17%) experienced local progression. Although most (71%) failures occurred within the RT fields, 29% occurred marginally or out of field, but adjacent to surgical hardware. Increasing coverage of the surgical hardware by RT fields was associated with a reduced risk of local failure in MVA (hazard ratio [HR], 0.10; 95% confidence interval [CI], 0.012-0.82; P = .03), whereas a greater risk of failure was seen with increasing time between surgery and RT (HR, 1.03; 95% CI, 1.01-1.06; P = .01). Extremity rather than spinal site trended toward a greater risk of failure but did not reach significance (HR, 3.79; 95% CI, 0.96-14.89; P = .057). BED ≥ 39 Gy10 did not predict local failure (P = .51) in MVA.
Presented at the 2012 American Society for Radiation Oncology Conference, October 28-31, 2012, Boston, MA. Conflicts of interest: None. ⁎ Corresponding author. 450 Brookline Ave, Boston, MA, 02215. E-mail address:
[email protected] (T. Balboni). http://dx.doi.org/10.1016/j.prro.2015.02.006 1879-8500/© 2015 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
2
Z.D. Epstein-Peterson et al
Practical Radiation Oncology: Month 2015
Conclusions: Current strategies achieve good outcomes after postoperative RT for osseous metastases. Greater coverage of the surgical hardware with RT fields and avoiding delays between surgery and postoperative RT should be considered to reduce recurrence risk for patients with bone metastases requiring surgical stabilization. © 2015 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction Bone metastasis represents a frequent and severe complication of many cancers, occurring in up to 75% of patients with metastatic disease, 1 depending on site and stage of the primary disease. Bone metastases can be associated with pain, disability, and other clinical complications; carry a poor prognosis regardless of primary disease site 2-4 ; and create significant challenges in effective coordination of care. 5 More than half of patients with bone metastases will experience a significant event (eg, fracture, spinal cord compression) related to that disease site in the course of their illness. 6 Current treatments for osseous metastatic disease include systemic chemotherapy, radiation therapy (RT), surgical stabilization, bisphosphonates, monoclonal antibodies, hormonal therapy, and pharmacologic pain management. The goals of these modalities include halting disease progression, preventing or healing disability, and palliation of symptoms. The specific treatment plan chosen for a patient can depend on the location (eg, long bone vs vertebral column), distribution, and extent of metastasis; patient functional status and symptoms; clinical urgency; previous therapies used; fracture risk; and goals of care/estimated prognosis. Important subgroups of bone metastases necessitate operative management because of bony instability (ie, fracture, impending fracture) and/or the need for decompression of nearby structures, such as spinal cord. Several factors and algorithms are used to identify whether surgical stabilization is needed in these patients, including the Mirels scale, 7 both axial and cortical involvement of the bone metastasis, 8 and the Spinal Instability in Neoplasia score. 9 A single small series describing outcomes among patients who underwent surgical stabilization for bone metastases with and without postoperative RT showed improved patient functional status with RT. 10 Optimal postoperative RT treatment protocols, particularly RT dose, amount of hardware coverage by RT fields, and timing of treatments, remain undefined. The limited available evidence suggests that significant heterogeneity in these protocols exists. 10 For these patients, inadequate data are available to define the optimal timing of RT, extent of hardware coverage by RT, dose intensity, and fractionation schema. Given the frequency of this clinical scenario, data are required to guide treatment and to develop evidence-based therapeutic strategies for these patients. In this study, we evaluated a cohort of patients undergoing palliative postoperative RT,
specifically examining the relationship of RT dose intensity and hardware coverage to radiographic local disease progression. We also aimed to examine characteristics and frequency of local failure and acute treatment-related toxicities.
Methods and Materials Sample We analyzed all patients undergoing RT with palliative intent following surgical intervention for bone metastases from solid tumors, with or without placement of hardware, at the Dana-Farber Cancer Center/Brigham and Women’s Hospital, Boston, Massachusetts, between July 2008 and January 2012. Patients were excluded if they were younger than 18 years old, were incorrectly listed as having been treated with palliative intent, or had nonmetastatic disease. Institutional review board approval of this research study was obtained before data collection. The charts of all patients were assessed for multiple covariates, including location of treated bone metastases, presence of impending or pathological fracture before surgery, surgical type and location, days from surgery to first fraction of RT, and RT dose and fractionation. Records were also examined for acute toxicity, defined as any toxicity at the treated site within 12 weeks of completion of RT. Dose intensity of RT was assessed by calculating the biological equivalent dose (BED), assuming an α/β of 10 for tumor, based on the equation, BED = nd(1 + d/α/β), in which n is the number of fractions of RT and d is the dose per fraction. Based on RT treatment films, the extent of hardware coverage by RT was assessed, with the percent of hardware covered by RT quantified by dividing the length of hardware within the RT field by the total length of hardware. If all hardware was covered by the RT fields plus additional bone/soft tissue beyond the hardware, this was categorized as 100% coverage. Other variables assessed were sex, primary tumor type, age, and Eastern Cooperative Oncology Group performance status at time of consultation with the radiation oncologist, duration of time between metastatic cancer diagnosis and treatment, and duration of time between surgery and postoperative RT. The primary outcome was local progression of the bone metastasis at the treated site, assessed by radiographic imaging with the attending radiologist report of progression on these studies. Imaging
Practical Radiation Oncology: Month 2015
of local progressions was compared with corresponding RT treatment fields and local progressions were further defined as: (1) in-field (recurrence fully encompassed within RT fields); (2) marginal (tumor recurrence partially within and outside of RT fields); or (3) out of field (tumor entirely outside RT fields but adjacent to surgical site/hardware).
Statistical Analysis Descriptive statistics were used for sample, site, treatment, and local failure characteristics. Rates of local failures according to tumor type, median split hardware coverage, median split BED, and tumor location were examined with chi-square statistics. Similarly, the relationship of RT coverage to osseous site and the relationships of toxicities to osseous tumor site and disease type were assessed with chi-square statistics. In light of the high frequency of short survival times among these patients, Fine and Gray’s proportional hazards univariable and multivariable logistic regression models 11 were used to examine the relationship of BED (dichotomized by a median split at 39 Gy10) and percent coverage of the hardware by the RT fields to local progression. Dose intensity and hardware coverage were automatically included in the multivariable analysis as part of the a priori hypothesis that they would be related to local failures. Other potential predictors of local progression assessed were: sex, age, performance status, tumor type, osseous metastasis site (extremity vs spine sites), time from initial metastatic diagnosis to osseous metastasis surgery, and time from surgery to initiation of RT. These additional predictors were entered into the multivariable model where P b .2 on univariable analysis and retained when P b .1 on multivariable analysis. Two-sided P values b .05 were considered significant, and all analyses were performed using SAS, version 9.2.
Postoperative radiation for bone metastasis Table 1
3
Sample and treatment characteristics (N = 82)
Variable
Value
Mean age at time of RT, y (SD) Male patients (%) ECOG performance status (%) a 0 1 2 3 4 Primary disease (%) Lung Prostate Breast Other Site (%) Lower extremity a Upper extremity a Spine Fracture (%) Pathological fracture Impending pathological fracture No pathological/impending fracture Type of surgical procedure (%) Open reduction and internal fixation Intramedullary rod Spine with hardware Spine without hardware Other Median time between surgery and RT, d (range) Median time between diagnosis of metastases and surgery, y (range) Median total dose, Gy (range) Median BED, Gy10 (range) Median number of fractions (range) Mean coverage of the surgical hardware by RT fields (SD) Median survival, mo
62 (12) 40 (49) 5 (6) 36 (44) 28 (34) 12 (15) 1 (1) 27 (33) 8 (10) 11 (13) 36 (44) 38 (46) 14 (17) 30 (37) 45 (54) 20 (24) 17 (21) 11 (13) 37 (45) 26 (32) 4 (5) 4 (5) 20 (1-116) 1.2 (0-16.0) 30 (8-50) 39 (14-60) 10 (1-25) 71 (26) 6.7
BED, biological equivalent dose; ECOG, Eastern Cooperative Oncology Group; RT, radiation therapy; SD, standard deviation. a Lower extremity includes 34 femur and 5 tibia cases; upper extremity includes 14 humerus cases.
Results A total of 82 bone metastases were treated with surgical intervention followed by palliative intent RT at the Dana-Farber Cancer Center/Brigham and Women’s Hospital, Boston, Massachusetts, from July 2008 to January 2012. Table 1 displays the patient characteristics. Median follow-up time was 4.6 months (range, 0-34) in the full sample and 11.5 months among surviving patients.
Patient and Treatment Characteristics Table 1 shows treatment site characteristics. A majority of patients (78%) received operative stabilization for pathological or impending fracture. The lower extremity was the most commonly treated site (46%), with most
being femur metastases (n = 34, 89%) and the remainder tibia metastases (n = 4, 11%). The most common surgical procedure performed was cephalomedullary fixation (45%). Approximately half (53%) of RT courses were completed with the dose fractionation schedule of 30 Gy in 10 fractions (BED = 39 Gy10). Twenty-one cases (27%) were treated with BED Gy10 values b 39 Gy10, with the most common regimens being 8 Gy × 1 (n = 5, 24%), 4 Gy × 5 (n = 10, 48%), and 4 Gy × 6 (n = 4, 19%). Eighteen (22%) were treated with regimens of BED values N 39 Gy10, with the most common regimens being 2.5 Gy × 14 (n = 5, 28%), 2.5 Gy × 15 (n = 3, 17%), and 3 Gy × 12 (n = 4, 22%). On average, 71% (standard deviation [SD], 26%; median, 71%) of the surgical hardware was covered within the RT fields.
4
Z.D. Epstein-Peterson et al Table 2
Practical Radiation Oncology: Month 2015
Characteristics of local failures
Variable
Local failure (%)
Full sample 14 (17) Anatomic site (% of failures) Lower extremity a 11 (79) Upper extremity 0 (0) Spine 3 (21) Location, relative to RT field (% of failures) In-field 10 (71) Marginal 1 (7) Out of field, adjacent to hardware 3 (21) RT, radiation therapy. a Nine of 11 lower extremity failures were in the femur, 2 of 11 were in the tibia.
Characteristics of Local Failures Table 2 characterizes the local failures in this cohort. Local failure occurred in 17% of cases at a median time of 3.8 months (range, 1-54) following RT. Most occurred in the lower extremity (79%) and most (71%) were found to be in-field; 1 (7%) was marginal and 3 (21%) were out-of-field but adjacent to surgical hardware. One patient experienced hardware failure because of local progression. Local failures were managed with repeat local therapy in 50% of cases: surgery alone in 2 cases, surgery and reirradiation in 1 case, and reirradiation alone in 4 cases.
Predictors of Local Failures Results of Fine and Gray univariable and multivariable models assessing predictors of failure are shown in Table 3. In addition to extent of hardware coverage and dose intensity, predictors assessed included: age, Eastern Cooperative Oncology Group performance status, sex, tumor type, treatment site, time from metastatic diagnosis to surgery, and duration of time between surgery and RT delivery. Increasing coverage of the hardware by RT fields was associated with a decreased risk of local failure (HR, 0.10; 95% CI, 0.012-0.82; P = .03). Furthermore, increasing time (days) between surgery and RT initiation was also associated with a greater risk of local recurrence (OR, 1.03; 95% CI, 1.01-1.06; P = .01). Extremity sites trended toward a greater risk of recurrence as compared to spine sites in MVA (OR, 3.79; 95% CI, 0.96-14.89; P = .057). Radiation dose intensity was not associated with local recurrence (P = .51).
Local Toxicities Within the first 3 months of surgery and RT, 16 local toxicities occurred in 14 patients (17%), as shown in Table 4. The most common local toxicity was wound infection (n = 7, 44% of local toxicities) followed by
hardware failure (n = 5, 31% of local toxicities). There were no statistically significant differences in local toxicities by site (lower extremity, 21%; spine, 17%; upper extremity, 7%; P = .50). Additionally, tumor type was not associated with rates of local toxicities (P = .28).
Discussion In this study examining outcomes among a cohort of patients who underwent surgery followed by postoperative RT for osseous metastases, 17% were found to have local recurrences, with half requiring local surgical and/or radiotherapeutic interventions. In the assessment of predictors of local recurrence, this study demonstrated that greater coverage of the surgical hardware within the RT treatment fields is associated with a reduced risk of local recurrence, whereas dose intensity of RT was not related to recurrence. Finally, greater delay between surgery and RT initiation was associated with a greater risk of recurrence at the treatment site. Townsend et al 10 examined the role of postoperative RT for pathological nonspinal bone fractures and determined that the addition of RT to surgery contributed positively to regaining functional status and reducing subsequent orthopedic procedures when compared with surgery alone. The local failure rate for patients treated with surgery and RT in this report was 3%, which is lower than the 17% rate reported for our study. However, unlike our study, Townsend et al. did not specifically review radiographic imaging for all patients and thus may have underreported rates of local failures. A recent series by Alvi et al. showed a similar 12% incidence of local failure among 96 patients with bone metastases who underwent operative management followed by RT, although this cohort included lymphoma and multiple myeloma tumor types, whereas ours did not. 12 In the present study, the majority, 71%, of treatment failures were located within the RT fields, whereas the remaining 29% were marginal and/or out of field but adjacent to surgical hardware. Prior randomized trials have examined dose and fractionation schemes in management of bone metastases and their effect on endpoints such as pain improvement and need for retreatment with RT. 13 However, to our knowledge, no data specifically examine local failure of surgically managed osseous metastases according to extent of surgical hardware coverage by the RT fields. Townsend et al 10,14 only had data on hardware coverage for 25 of their 64 patients. Among those 25, most (84%) had full hardware coverage by the RT fields. Hence, given their limited data, they were unable to assess failures according to extent of coverage of the surgical hardware. In contrast, within our cohort, there was a wide variation in extent of coverage of the surgical hardware by RT fields. Notably, we found increasing coverage of the surgical
Practical Radiation Oncology: Month 2015
5
Postoperative radiation for bone metastasis
Table 3 Fine and Gray univariable and multivariable analyses of predictors of local failure after postoperative radiation therapy for osseous metastases (N = 82) Variable BED ≥ 39 Gy10 (vs b 39 Gy10) Percent hardware coverage (continuous) Extremity site (vs spine) Time from surgery to RT (days) Breast disease type (vs nonbreast) Age N 62 y (vs ≤ 62 y) Male sex ECOG performance status 0-1 (vs 2-4) Time from metastatic diagnosis to surgery (y)
Univariable analyses a
Multivariable analysis a
HR (95% CI)
P
HR (95% CI)
P
1.78 (0.39-8.17) 0.12 (0.01-1.09) 2.42 (0.74-7.98) 1.02 (1.00-1.03) 3.60 (1.13-11.47) 0.72 (0.26-1.97) 1.16 (0.42-3.23) 1.36 (0.49-3.77) 1.18 (1.05-1.33)
.46 .06 .15 .11 .03 .52 .78 .56 .007
1.64 0.10 3.79 1.03
.51 .03 .06 .01
(0.37-7.22) (0.01-0.82) (0.96-14.89) (1.01-1.06)
BED, biological equivalent dose; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio. a Univariable and multivariable competing-risks regression was based on Fine and Gray’s proportional hazards model. Multivariable model inclusive of primary predictors of interest: BED and hardware coverage. Potential predictors (primary disease site, performance status, age, sex, anatomic site, time from metastatic diagnosis to surgery, and time for surgery to RT) were entered into the model where P b .20 on univariable analysis and retained in the model where P b .10. Using these criteria, time from metastatic diagnosis to surgery, breast cancer type vs other cancer types, time from surgery to initiation of RT, and extremity site (eg, femur, tibia, humerus) versus spine site were entered into the model; breast and time from metastatic diagnosis to surgery not retained in the final model due to failure to reach multivariable model criteria of P b .10.
hardware to be associated with a reduced risk of local failure (HR, 0.10). This suggests that tumor seeding may occur along the tract of the hardware and that coverage within the RT fields should be strongly considered. Furthermore, it is notable that extremity sites, compared with spine sites, trended toward being at a greater risk of local failure in multivariable analysis (Table 3). The risk of recurrence within the surgical hardware may vary by differing bony locations (eg, from surgical practice within the extremities when nailing a rod through the metastatic lesion). Further studies are required with greater numbers to determine the relationship of treatment site and surgery type to local recurrence in the postoperative management of osseous metastases. Additionally, to our knowledge, no previous studies have examined the relationship between BED and local failure in the context of postoperative RT for osseous metastases. This study demonstrated no relationship between dose intensity and local failure, although just over half of patients received the regimen of 30 Gy in 10 fractions that, together with our modest sample size, limits our ability to detect a relationship between dose intensity and local failure. Randomized trials in the setting of painful, uncomplicated bone metastases (without prior surgery) have shown no difference in pain relief following 8 Gy in 1 fraction compared with longer, more dose-intense fractionation schemes (eg, 30 Gy in 10 fractions, 20 Gy in 5 fractions). 13 However, these data do consistently demonstrate higher rates of retreatment in the less doseintense, single-fraction arm, 5,13 which may be due to greater local tumor progression. Notably, it is unknown if higher retreatment rates are truly the result of local failures in the setting of suboptimal dose intensity rather than greater physician willingness to retreat with radiation therapy after single-fraction, low-dose-intensity RT. 13
Greater interval from surgical intervention to radiation therapy initiation was also found to be associated with a greater risk of recurrence. A longer interval from surgical intervention to postoperative RT, though to our knowledge not specifically studied in the setting of osseous metastases, has been found in other settings to confer a greater risk of local recurrence. For example, in breast cancer, increasing the interval from breast-conserving surgery to radiation therapy has been found to be associated with a greater risk of local recurrence. 15 Similar findings have been noted in the postoperative management of head and neck squamous cell carcinomas. 16 Hence, this finding supports the general postoperative RT management strategy of limiting the interval between surgery and RT to the timeframe required for adequate postsurgical healing. Local toxicities occurring in the first 3 months following surgery and postoperative RT occurred in 17% of cases, most commonly wound infection (9%) and hardware failure (6%). There were no significant differences in complications by treatment site or tumor type. Quinn et al. found a 10% complication rate over a median of 10.4 months in a cohort of patients who underwent operative management of femoral and acetabular lesions. 17 In another series, 18 patients with femoral metastases had a 20% rate of developing toxicities. Regarding hardware complications, Wedin et al. found a hardware failure rate of 11% in long bone disease, 19 with 24% of those deemed to result from local disease progression. Similarly, within our study, 20% (1 of 5) of hardware failures were associated with local progression. Our study has important limitations. First, the sample size is small and therefore may be underpowered to detect predictors of local failure. Additionally, many patients herein had short life expectancies in the setting of metastases; as a result, this study’s applicability to patients
6
Z.D. Epstein-Peterson et al
Practical Radiation Oncology: Month 2015
Table 4 Local toxicities within 12 weeks following surgery and postoperative radiation therapy for osseous metastases (N = 82) Toxicity
Frequency (%)
Patients experiencing any local toxicity Local toxicities a Wound infection Hardware failure Bleeding Wound breakdown Compartment syndrome
14 (17)
a
7 (9) 5 (6) 2 (2) 1 (1) 1 (1)
16 local toxicities experienced in 14 patients.
with long (eg, N 1 year) life expectancies may be limited. Future studies should assess outcomes in larger cohorts of patients receiving postoperative RT for osseous metastases, focusing on outcomes among particular treatments sites (eg, extremities) and patient populations, such as those with longer expected survival times. In conclusion, our study showed a low (17%) rate of radiographic local failure in patients undergoing palliative RT following surgical management of bone metastases, with half of these patients requiring local intervention because of failure. Greater coverage of the surgical hardware by RT fields was associated with a reduced risk of local failures, whereas increasing time from surgery to RT initiation was associated with a greater risk. Given the paucity of data regarding radiation therapy in the postoperative management of bone metastases, these data provide valuable information to guide radiation oncologists in their management of these patients.
References 1. Coleman RE. Metastatic bone disease: Clinical features, pathophysiology and treatment strategies. Cancer Treat Rev. 2001;27:165-176. 2. Tong D, Gillick L, Hendrickson FR. The palliation of symptomatic osseous metastases: Final results of the Study by the Radiation Therapy Oncology Group. Cancer. 1982;50:893-899. 3. Sarahrudi K, Hora K, Heinz T, Millington S, Vecsei V. Treatment results of pathological fractures of the long bones: A retrospective analysis of 88 patients. Int Orthop. 2006;30:519-524.
4. Bohm P, Huber J. The surgical treatment of bony metastases of the spine and limbs. J Bone Joint Surg (Br). 2002;84:521-529. 5. Lutz S, Berk L, Chang E, et al. Palliative radiotherapy for bone metastases: an ASTRO evidence-based guideline. Int J Radiat Oncol Biol Phys. 2011;79(4):965-976. 6. Oster G, Lamerato L, Glass AG, et al. Natural history of skeletal-related events in patients with breast, lung, or prostate cancer and metastases to bone: A 15-year study in two large US health systems. Support Care Cancer. 2013;21:3279-3286. 7. Damron TA, Morgan H, Prakash D, Grant W, Aronowitz J, Heiner J. Critical evaluation of Mirels' rating system for impending pathologic fractures. Clin Orthop Relat Res. 2003(415 Suppl):S201-S207. 8. Van der Linden YM, Dijkstra PD, Kroon HM, et al. Comparative analysis of risk factors for pathological fracture with femoral metastases. J Bone Joint Surg (Br). 2004;86:566-573. 9. Fourney D, Frangou EM, Ryken TC, et al. Spinal Instability Neoplastic Score: An analysis of reliability and validity from the Spine Oncology Study Group. J Clin Oncol. 2011;29:3072-3077. 10. Townsend PW, Rosenthal HG, Smalley SR, Cozad SC, Hassanein RE. Impact of postoperative radiation therapy and other perioperative factors on outcome after orthopedic stabilization of impending or pathologic fractures due to metastatic disease. J Clin Oncol. 1994;12:2345-2350. 11. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496-509. 12. Alvi HM, Damron TA. Prophylactic stabilization for bone metastases, myeloma, or lymphoma: Do we need to protect the entire bone? Clin Orthop Relat Res. 2013;471:706-714. 13. Chow E, Harris K, Fan G, Tsao M, Sze WM. Palliative radiotherapy trials for bone metastases: A systematic review. J Clin Oncol. 2007;25:1423-1436. 14. Townsend PW, Smalley SR, Cozad SC, Rosenthal HG, Hassanein RE. Role of postoperative radiation therapy after stabilization of fractures caused by metastatic disease. Int J Radiat Oncol Biol Phys. 1995;31:43-49. 15. Punglia RS, Saito AM, Neville BA, Earle CC, Weeks JC. Impact of interval from breast conserving surgery to radiotherapy on local recurrence in older women with breast cancer: Retrospective cohort analysis. Br Med J. 2010;340:c845. 16. Parsons JT, Mendenhall WM, Stringer SP, et al. An analysis of factors influencing the outcome of postoperative irradiation for squamous cell carcinoma of the oral cavity. Int J Radiat Oncol Biol Phys. 1997;39:137-148. 17. Quinn RH, Drenga J. Perioperative morbidity and mortality after reconstruction for metastatic tumors of the proximal femur and acetabulum. J Arthroplasty. 2006;21:227-232. 18. Nilsson J, Gustafson P. Surgery for metastatic lesions of the femur: Good outcome after 245 operations in 216 patients. Injury. 2008;39: 404-410. 19. Wedin R, Bauer HC, Wersall P. Failures after operation for skeletal metastatic lesions of long bones. Clin Orthop Relat Res. 1999: 128-139.