Abstract
Background:
Surgical guidelines for soft tissue sarcoma (STS) emphasize pretreatment evaluation, and reports of the perils of unplanned excision exist. Given the paucity of population-based data on this topic, our objective was to analyze clinical outcomes and costs of planned versus unplanned STS excisions in the Medicare population.
Methods:
We analyzed 3,913 surgical STS patients ≥66 years old from 1992-2011 using SEER-Medicare. Planned excisions were classified based on preoperative MRI and/or biopsy, whereas unplanned excisions were classified by excision as the first procedure. Inverse probability of treatment weighting with propensity scores was used to adjust for clinicopathologic differences. Re-excisions, complications, and Medicare payments were compared with multivariate models. Overall (OS) and disease specific survival (DSS) were analyzed using cox proportional hazards and competing risk models.
Results:
Prior to the first excision, 24.3% had a MRI & biopsy, 27.3% had a MRI, 11.4% had a biopsy, and 36.9% were unplanned. Re-excision rates were highest for unplanned excisions: 46.3% compared to 18.1%, 36.4%, and 29.7% for other groups (p<0.0001). There was no difference in DSS or OS between groups (p>0.05). Planned excisions were associated with increased Medicare costs (p<0.05), with the first resection contributing to the majority of costs. Subgroup analyses by histologic grade and tumor size revealed similar results.
Conclusion:
Survival was comparable with greater healthcare costs in elderly patients undergoing planned STS excision. Although unplanned excisions remain a quality of care issue with high re-excision rates, these data have important implications for the surgical management of STS in the elderly.
Keywords: MRI, Biopsy, Sarcoma, Survival, Cost
Introduction
Given the rarity and heterogeneous behavior of soft tissue sarcomas (STS), current clinical practice guidelines consistently emphasize magnetic resonance imaging (MRI), biopsy, and evaluation by a STS multidisciplinary team prior to surgical resection.1-4 However, benign soft tissue masses markedly outnumber STS. As a result, STS are commonly mistaken for benign tumors, frequently leading to excision without a preoperative diagnosis or pursuit of tumor-free margins.5-7 Such practice is referred to as an “unplanned excision” and, although common with rates as high as 39-50%, is widely regarded unfavorably by sarcoma experts.6, 8, 9 Rates of residual disease among unplanned excisions have been reported as high as 65%, and consensus recommendations following unplanned excision endorse re-resection of the tumor bed to reduce risk of local recurrence and optimize oncological outcomes.5, 9-13
Despite concerns that unplanned excisions lead to worse morbidity and oncologic outcome, research on this topic has yielded inconsistent results. Although multiple studies show worse oncologic outcome in the form of increased local recurrence after unplanned excision, these findings were limited to subset analyses focusing on higher risk groups such as those with high-grade tumors.12, 14, 15 In fact, most studies have shown comparable outcomes among STS patients undergoing planned versus unplanned excisions with prognosis dependent on final margin status.5-9, 11 A notable study by Lewis et al. demonstrated that patients who initially underwent an unplanned excision followed by subsequent re-resection at a sarcoma referral center had superior disease-specific survival (DSS) and metastases-free survival compared to patients who underwent a planned excision, although these results were attributed to the impact of margin status and selection bias as smaller, more superficial tumors tend to be treated with unplanned excision at non-sarcoma centers.5
The oncologic impact of unplanned excision remains a subject of debate and unplanned excision is an important quality of care issue given the frequent recommendation for a salvage procedure which exposes patients to additional surgical morbidity and healthcare costs. Therefore, the purpose of this study was to further analyze this controversy using a U.S. population-based dataset, the Surveillance, Epidemiology, and End Results (SEER) Medicare linked dataset, to compare overall survival (OS), DSS, and healthcare costs among elderly STS patients who undergo planned versus unplanned STS resections. Given the heterogeneity of these patients and influence of selection bias in determining outcomes, we utilized inverse probability of treatment (i.e. preoperative MRI and/or biopsy) weighting (IPTW) with propensity scores to control for the non-random allocation of pre-operative evaluation strategy. We sought to examine the impact of unplanned excisions on healthcare cancer-related costs since the nature of this relationship is unknown. This is particularly important as healthcare cancer-related costs continue to rise, with significant impact on patients, families, and the economy.16-18 We hypothesized that survival would be comparable for STS patients regardless of preoperative evaluation strategy, but that associated healthcare costs would be higher following unplanned excision given the recommendation for re-excision and potential for greater operative morbidity.
Methods
We performed a retrospective analysis of elderly STS patients using the SEER-Medicare linked datafiles. SEER is estimated to include 28% of the United States population19 and Medicare covers approximately 93% of those 65 years of age and older.20 SEER-Medicare has been shown to be a reliable method to identify cancer-related surgical procedures, radiotherapy and chemotherapy.21-24 As such, it has been widely used to assess surgical and cancer-related outcomes, including patterns of care and healthcare costs, for a wide range of cancer-types including STS.16, 18,25-27
We identified 8,931 patients ≥66 years of age diagnosed with truncal or extremity STS from 1992-2011 from histology, site, and behavior codes (see Supplement for coding details, Figure 1). We evaluated corresponding patient Medicare files (i.e. MEDPAR, Carrier, and Outpatient claims files), for services provided from 1991-2013, ensuring adequate follow-up. Exclusion criteria included stage IV disease, missing/unknown staging information, and absence of continuous Medicare A/B coverage and/or HMO coverage within 12 months of STS diagnosis (to ensure complete medical record documentation). We further excluded patients without documentation of a STS resection within 6 months of diagnosis as defined by corresponding Current Procedural Terminology (CPT) and International Classification of Diseases 9th edition (ICD-9) codes (see Supplement for coding details). Using a 6-month time interval from cancer diagnosis to surgery has been estimated to identify more than 90% of cancer-related surgical resections22 and was selected to reduce likelihood of including operations performed for treatment of local reoccurrence, palliation, and/or unrelated reasons.
Figure 1.
Selection of the cohort of elderly patients with soft tissue sarcomas who underwent surgical resection from SEER-Medicare. *≤10 patients excluded due to a prior excision within 6 months of the 1st identified excision based on the inclusion time-frame consistent; reported as ≤10 in accordance with the Center for Medicare Services reporting guidelines.
We defined preoperative planning as MRI and/or biopsy. We selected to not include computed tomography (CT) scans in this definition as MRI is the current standard of care for preoperative surgical planning based on expert consensus guidelines due to the greater delineation of muscle and facial groups, compartments, bone, neurovascular structures and tumor with MRI compared to CT scan.2, 28 The final cohort consisted on 3,913 patients (Figure 1). Patients were assigned to one of the following four groups based on preoperative evaluation: MRI and biopsy, MRI alone, biopsy alone, and neither (i.e. unplanned). Since all patient data were de-identified, this study protocol was reviewed by the University of California, Davis Institutional Review Board and deemed exempt.
Patient demographics and clinicopathologic characteristics obtained from SEER included age, gender, race, histology, tumor grade, size, depth, AJCC and SEER historical staging. Chemotherapy, radiotherapy, MRI, and biopsy information was obtained from Medicare data by ICD-9 and CPT codes and defined as chemotherapy within 6 months, radiotherapy within 4 months, and MRI and/or biopsy of the associated cancer site 1 year prior to the primary resection. We excluded MRIs/biopsies performed >1 year prior to the index procedure to reduce likelihood of including unrelated imaging/biopsies. The Charlson comorbidity index (CCI) was created using Medicare ICD-9 diagnosis codes.29, 30 Reconstructive surgery included tissue transfers, graft and flap creation based on ICD-9 codes and CPT codes within 1 year of the primary resection.
The primary outcomes were total surgical cancer-related healthcare costs, DSS, and OS. Costs were estimated from Medicare payments,31, 32 and were price standardized, adjusting for geographic variation, according to methods described by the Dartmouth Atlas of Healthcare. 32, 33 Due to the unavailability of Center for Medicare and Medicaid Services files required for payment standardization prior to 2003, cost analyses were only performed on those receiving all STS-related care from 2003-2013 (N=2,403). Total surgical healthcare costs comprised of: (1) preoperative MRI, (2) preoperative biopsy, (3) index surgery, associated hospitalization, and services provided within 30-days of discharge, (4) re-excision surgeries, associated hospitalizations and services provided within 30-days of discharge, and (5) reconstructive surgery payments including associated hospitalizations. Inpatient hospitalization costs were estimated using diagnosis-related group (DRG) and outlier payments. Skilled nursing and rehabilitation facility payments were prorated so that payments were limited to services provided within 30-days of discharge. Costs were adjusted for inflation to 2013 dollars using the Medicare Prospective Payment System adjuster and Medicare Economic Index for Medicare parts A and B claims, respectively.34 Survival was defined as months alive after STS diagnosis, which was abstracted from SEER. The median follow-up for survival for the cohort was 42 months (range 0-251, interquartile range [IQR] 19-86).
Secondary outcomes were re-excisions and postoperative complications. Re-excisions were defined as reoperation for resection of tumor within 1 year of the first resection. Postoperative complications were identified based on ICD-9 codes within 30-days of the index operation and included pneumonia, respiratory failure, cardiac arrest, myocardial infarction, pulmonary embolism, deep vein thrombosis, hematoma, acute kidney failure, shock, and surgical site/wound infections.35, 36
Statistical analysis
Patient demographic and clinicopathologic characteristics are presented as means and standard deviations (SD), medians and IQR, or percentages as appropriate. Analysis of variance analyses with Tukey pairwise comparisons and χ2 were performed to determine differences in patient and tumor characteristics between groups. Tumor size was log-transformed to create normality. Multiple imputation was performed to address missing data for tumor size (n=485, 12.4% missing).
We performed IPTW with propensity scores in our analyses of primary and secondary endpoints to control for selection bias.37 Propensity scores were created estimating the probability of selection into the four groups (MRI and biopsy, MRI only, biopsy only, and unplanned) with a multinomial logistic regression model with age, gender, race, CCI, tumor histology, size, grade, site, depth, and year of diagnosis as covariates. Covariates were selected based on clinical significance and/or statistically significant differences between groups (p≤0.10). We were unable to include National Cancer Institute (NCI) cancer center designation as a covariate for the propensity score as hospital-level data was not available from Medicare for the years 1991-1995, 1997, and 1999. Patients were weighted by the inverse probability of selection into the preoperative planning groups. Covariates were evaluated using standardized differences and determined to be balanced across groups after IPTW adjustment.
To address confounding from treatment effects of chemotherapy, radiotherapy, and reconstruction, multivariate regression analysis with IPTW was performed in addition to univariate unweighted analysis of our primary outcomes. Logistic regression was performed to compare odds of re-excision and complications. Generalized linear models with gamma distribution and log link were performed to compare actual and standardized payments between groups. Fine and Gray competing risks analyses and cumulative incidence function curves were performed to compare DSS among groups.38 Kaplan-Meier curves were created for OS. The log-rank test and Cox proportional hazards regression analyses were performed to compare OS between groups. Subgroup analyses was performed for high-grade tumors and large tumors (≥8cm). Statistical analysis was performed using SAS 9.4 (SAS Institute, Cary, NC) and Stata 15 (StataCorp, College Station, TX) software. All statistical tests were two-sided and p-values<0.05 were considered significant.
Results
Table 1 depicts unadjusted patient demographics and clinicopathological characteristics. Of the 3,913 patients, 24.3% had a preoperative MRI and biopsy, 27.3% had a MRI only, 11.4% had a biopsy only, and 36.9% were unplanned. There were significant differences between groups with respect to age, gender, CCI, tumor grade, size, histology, depth, site, and stage, history of preoperative chemotherapy, pre- and postoperative radiotherapy, and rates of reexcision, limb-sparing, and reconstructive surgery (p<0.05 all). As shown in Figure 2, there was a temporal trend with an increasing rate of preoperative MRI and biopsy and a decreasing rate of unplanned excisions over time (p=0.007).
Table 1.
Patient Demographics and Clinicopathologic Characteristics.
| Preoperative Planning Approach | |||||
|---|---|---|---|---|---|
| Characteristic | MRI & Biopsy N=952 |
MRI Only N=1,069 |
Biopsy Only N=447 |
Unplanned N=1,445 |
P-value |
| Age (years, mean±SD) | 76.6±6.8 | 77.0±6.9 | 78.7±7.2 | 77.6±7.4 | P<0.05* |
| Male Gender | 459 (48.2%) | 520 (48.6%) | 256 (57.2%) | 787 (54.5%) | P=0.0003 |
| Race | P=0.12 | ||||
| Caucasian | 856 (89.9%) | 939 (87.8%) | 414 (92.6%) | 1,290 (89.3%) | |
| Black | 62 (6.5%) | 67 (6.3%) | 15 (3.4%) | 82 (5.7%) | |
| Asian/Pacific Islander | 32 (3.4%) | 56 (5.2%) | 16 (3.6%) | 67 (4.6%) | |
| Hispanic Ethnicity | 42 (4.4%) | 39 (3.7%) | 16 (3.6%) | 75 (5.2%) | P=0.33 |
| Charlson Comorbidity Index | P=0.03 | ||||
| 0 | 437 (45.9%) | 520 (49.6%) | 204 (45.6%) | 735 (50.9%) | |
| 1 | 286 (30.0%) | 317 (29.7%) | 129 (28.9%) | 374 (25.9%) | |
| 2 | 133 (14.0%) | 124 (11.6%) | 51 (11.4%) | 164 (11.4%) | |
| ≥3 | 96 (10.1%) | 108 (10.1%) | 63 (14.1%) | 172(11.9%) | |
| Grade | |||||
| I | 126 (13.2%) | 212 (19.8%) | 59 (13.2%) | 177 (12.3%) | P<0001; |
| II | 125 (13.1%) | 143 (13.4%) | 74 (16.6%) | 196 (13.6%) | |
| III | 219 (23.0%) | 191 (17.9%) | 95 (21.3%) | 286 (19.8%) | |
| IV | 362 (38.0%) | 358 (33.5%) | 134 (30.0%) | 392 (27.1%) | |
| Unknown | 120 (12.6%) | 165 (15.4%) | 85 (19.0%) | 394 (27.3%) | |
| Tumor Size (cm, median, IQR) | 9.7 (5.9-15.0) | 8.0 (4.6-13.0) | 6.0 (4.0-11.0) | 5.5 (3.5-9.0) | P<0.05 |
| Depth | P<0.0001 | ||||
| Superficial | 142 (14.9%) | 167 (15.6%) | 92 (20.6%) | 258 (17.9%) | |
| Deep | 418 (43.9%) | 374 (35.0%) | 132 (29.5%) | 346 (23.9%) | |
| Unknown | 392 (41.2%) | 528 (49.4%) | 223 (49.9%) | 841 (58.2%) | |
| Histology | P<0.0001 | ||||
| Liposarcoma | 219 (23.0%) | 307 (28.7%) | 73 (16.3%) | 287 (19.9%) | |
| Fibrosarcoma | 84(8.8%) | 78 (7.3%) | 44 (9.1%) | 134 (8.9%) | |
| Leiomyosarcoma | 84 (8.8%) | 115 (10.8%) | 69 (15.4%) | 202 (14.0%) | |
| HGUPS | 297 (31.2%) | 329 (30.8%) | 129 (28.9%) | 457 (31.6%) | |
| Synovial | ≤10‡ | 15 (1.4%) | ≤10‡ | 17 (1.2%) | |
| Vascular | 21 (2.2%) | 11 (1.0%) | 23 (5.2%) | 44 (3.0%) | |
| MPNST | 17(1.8%) | 22 (2.1%) | ≤10‡ | 47 (3.3%) | |
| Sarcoma Nos | 224 (23.5%) | 192 (18.0%) | 99 (22.1%) | 257 (17.8%) | |
| Site | P<0.0001 | ||||
| Upper extremity | 182 (19.1%) | 209 (20.1%) | 90 (20.1%) | 347 (24.0%) | |
| Lower extremity | 641 (67.3%) | 705 (66.0%) | 164 (36.7%) | 523 (36.2%) | |
| Trunk | 125 (13.1%) | 147 (13.8%) | 186 (41.6.1%) | 560 (38.8%) | |
| SEER Stage | P=0.03 | ||||
| Local | 672 (70.6%) | 761 (71.2%) | 332 (74.3%) | 1090 (75.4%) | |
| Regional | 280 (29.4%) | 308 (28.8%) | 115 (25.7%) | 355 (24.6%) | |
| Chemotherapy | |||||
| Preoperative | 70 (7.4%) | 26 (2.4%) | 19 (4.3%) | 32 (2.2%) | P<0.0001 |
| Postoperative | 66 (6.9%) | 77 (7.2%) | 26 (5.8%) | 88 (6.1%) | P=0.60 |
| Radiotherapy | |||||
| Preoperative | 230 (24.2%) | 66 (6.2%) | 33 (7.4%) | 34 (2.4%) | P<0.0001 |
| Postoperative | 482 (50.6%) | 573 (53.6%) | 185 (41.4%) | 618 (42.8%) | P<0.0001 |
| First Surgery | |||||
| Limb-sparing | 893 (93.8%) | 1,045 (97.8%) | 431 (96.4%) | 1,419 (98.2%) | P<0.0001 |
| Amputation | 59 (6.2%) | 24 (2.3%) | 16 (3.6%) | 26 (1.8%) | |
| Reconstruction | 277 (29.1%) | 148 (13.8%) | 97 (21.7%) | 143 (9.9%) | P<0.0001 |
| 30-Day Complications | 233 (24.5%) | 187 (17.5%) | 84 (18.8%) | 204 (14.1%) | P<0.0001 |
| First Surgery Hospital** | |||||
| Teaching Facility | 508 (87.4%) | 472 (74.6%) | 170 (71.7%) | 405 (58.3%) | P<0.0001 |
| NCI Cancer Center† | 168 (31.4%) | 109 (19.6%) | 40 (19.6%) | 43 (7.4%) | P<0.0001 |
| Re-excision Surgery | P<0.0001 | ||||
| 1 | 155 (16.3%) | 312 (29.2%) | 112 (25.1%) | 494 (34.2%) | |
| 2 | 18 (1.9%) | 68 (6.4%) | 16 (3.6%) | 158 (10.9%) | |
| ≥3 | ≤10‡ | 12 (1.1%) | ≤10‡ | 32 (2.2%) | |
| Limb-sparing | 152 (84.4%) | 342 (87.7%) | 129 (94.2%) | 637 (93.4%) | P=0.0001 |
| Amputation | 28 (15.6%) | 48 (12.3%) | ≤10‡ | 45 (6.6%) | |
| Reconstruction | 61 (33.9%) | 109 (27.8%) | 50 (36.2%) | 206 (30.1%) | P=0.21 |
| 30-Day Complications | 57 (31.7%) | 98 (25.0%) | 30 (21.7%) | 139 (20.3%) | P=0.01 |
MRI, magnetic resonance imaging; SD standard deviation; IQR interquartile range; NOS not otherwise specified; HGUPS high-grade undifferentiated pleomorphic sarcoma, MPNST malignant peripheral nerve sheath tumor
All pairwise comparisons were significant, except for MRI only vs. unplanned excision.
Hospital information only available for years 1996, 1998 and ≥2000.
NCI National Cancer Institute only available for years ≥2002.
indicates ≤10 patients consistent with Center for Medicare Services reporting guidelines.
Figure 2.
Percent of planned excisions (i.e. MRI and biopsy, MRI alone, and biopsy alone) and unplanned excisions over time. There was a temporal trend with an increasing percentage of patients who underwent preoperative MRI and biopsy and a decreasing percentage of patients who underwent unplanned excisions over the time period from 1992 to 2011 (p=0.007).
Unplanned excisions were associated with greater odds of re-excision in the IPTW multivariate (Table 2) and unadjusted models (p<0.001). Rates of re-excision were 47.3% among unplanned excisions compared to 18.9% for those with MRI and biopsy, 36.7% with MRI alone and 30.9% with biopsy alone. Omitting “unplanned excision” as a complication, the odds of complications after the initial excision and in total (combined first excisions and re-excisions) were greater for patients with MRI and biopsy and MRI alone compared to patients with unplanned excision in the IPTW multivariate (p<0.05, Table 2) and unadjusted models. Given the tendency for larger tumors to undergo preoperative evaluation, we performed a subgroup analysis of patients with tumors ≥8 cm (n=1,588). In IPTW multivariate models, we observed a higher odds of re-excision in the unplanned group (p<0.05) and similar trends with respect to complications as observed in our primary analysis (Table 3).
Table 2.
Inverse Probability of Treatment Weighted* Multivariate Models for Re-excision and Complications for Planned versus Unplanned Excisions.
| Re-Excisions | First Excision Complications |
Re-Excision Complications | Total Complications | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-Value | OR | 95% CI | P-Value | OR | 95% CI | P-Value | OR | 95% CI | P-Value | |||||
| Preoperative Planning | ||||||||||||||||
| Unplanned (ref) | ||||||||||||||||
| MRI & Biopsy | 0.26 | 0.20 | 0.34 | <0.001 | 1.59 | 1.18 | 2.13 | 0.002 | 1.60 | 0.97 | 2.65 | 0.07 | 1.33 | 1.01 | 1.76 | 0.04 |
| MRI only | 0.61 | 0.50 | 0.75 | <0.001 | 1.38 | 1.07 | 1.79 | 0.01 | 1.18 | 0.84 | 1.66 | 0.35 | 1.30 | 1.03 | 1.64 | 0.03 |
| Biopsy only | 0.53 | 0.40 | 0.70 | <0.001 | 1.52 | 1.08 | 2.14 | 0.02 | 1.24 | 0.73 | 2.10 | 0.43 | 1.31 | 0.96 | 1.80 | 0.09 |
| Chemotherapy | ||||||||||||||||
| Preoperative | 0.60 | 0.35 | 1.01 | 0.6 | 1.18 | 0.67 | 2.09 | 0.57 | 0.31 | 0.10 | 0.92 | 0.04 | 0.98 | 0.56 | 1.70 | 0.93 |
| Postoperative | 1.78 | 1.20 | 2.65 | 0.004 | 0.89 | 0.53 | 1.49 | 0.65 | 1.27 | 0.62 | 2.60 | 0.52 | 1.01 | 0.67 | 1.53 | 0.97 |
| Radiotherapy | ||||||||||||||||
| Preoperative | 0.80 | 0.49 | 1.30 | 0.37 | 1.38 | 0.94 | 2.04 | 0.10 | 1.24 | 0.52 | 2.97 | 0.63 | 1.43 | 0.99 | 2.06 | 0.06 |
| Postoperative | 1.26 | 1.03 | 1.53 | 0.02 | 0.69 | 0.54 | 0.88 | 0.003 | 0.72 | 0.49 | 1.05 | 0.09 | 0.68 | 0.55 | 0.85 | 0.001 |
| Reconstruction | ||||||||||||||||
| First Excision | 1.31 | 0.99 | 1.74 | 0.06 | 1.12 | 0.85 | 1.47 | 0.43 | ||||||||
| Re-excision | 1.73 | 1.17 | 2.56 | 0.01 | 2.16 | 1.62 | 2.87 | <0.001 | ||||||||
OR, Odds Ratio; CI, confidence interval; Ref, reference; MRI, magnetic resonance imaging
Inverse probability of treatment (i.e. preoperative planning) weighted adjustment with propensity scores (covariates: age, gender, race, Charlson comorbidity index, tumor histology, size, depth, site, grade, and diagnostic year).
Table 3.
Inverse Probability of Treatment Weighted* Models for Odds of Complications for Planned versus Unplanned Excisions of Tumors ≥ 8 cm.
| First Excision Complications |
Re-Excision Complications |
Total Complications | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-Value | OR | 95% CI | P-Value | OR | 95% CI | P-Value | ||||
| Preoperative Planning | ||||||||||||
| Unplanned (ref) | ||||||||||||
| MRI & Biopsy | 1.70 | 1.17 | 2.49 | 0.01 | 1.32 | 0.63 | 2.78 | 0.46 | 1.43 | 0.98 | 2.07 | 0.06 |
| MRI only | 1.53 | 1.06 | 2.22 | 0.02 | 1.55 | 0.86 | 2.78 | 0.15 | 1.52 | 1.07 | 2.17 | 0.02 |
| Biopsy only | 1.40 | 0.80 | 2.45 | 0.24 | 1.23 | 0.47 | 3.18 | 0.67 | 1.32 | 0.77 | 2.25 | 0.32 |
| Chemotherapy | ||||||||||||
| Preoperative | 0.69 | 0.35 | 1.35 | 0.27 | 0.36 | 0.10 | 1.27 | 0.11 | 0.65 | 0.34 | 1.24 | 0.19 |
| Postoperative | 0.98 | 0.55 | 1.74 | 0.93 | 1.55 | 0.66 | 3.65 | 0.31 | 1.03 | 0.61 | 1.72 | 0.92 |
| Radiotherapy | ||||||||||||
| Preoperative | 0.97 | 0.54 | 1.74 | 0.91 | 1.86 | 0.77 | 4.48 | 0.17 | 1.00 | 0.57 | 1.75 | 1.00 |
| Postoperative | 0.60 | 0.42 | 0.85 | 0.004 | 0.69 | 0.39 | 1.21 | 0.19 | 0.59 | 0.43 | 0.83 | 0.002 |
| Reconstruction | ||||||||||||
| First Excision | 2.13 | 1.31 | 3.47 | 0.002 | 1.83 | 1.13 | 2.96 | 0.01 | ||||
| Re-excision | 2.30 | 1.33 | 4.00 | 0.003 | 2.99 | 1.88 | 4.76 | <0.001 | ||||
OR, Odds Ratio; CI, confidence interval; Ref, reference; MRI, magnetic resonance imaging
Inverse probability of treatment (i.e. preoperative planning) weighted adjustment with propensity scores using covariates: age, gender, race, Charlson comorbidity index score, tumor histology, size, grade, site, depth, and diagnostic year.
Unadjusted and IPTW Kaplan Meier and cumulative incidence curves for OS and DSS are illustrated in Figure 3. Adjusted median OS was 57 months for MRI and biopsy, 66 months for MRI alone, 50 months for biopsy alone and 59 months for unplanned excisions. Although there were significant differences in DSS and OS between groups in the univariate analysis (p<0.05), these differences were not significant with IPTW adjustment (Figure 3). Notably, in a subgroup analysis of unplanned excisions (n=1,445), re-excision was not associated with improved DSS or OS with IPTW (DSS: HR 0.94, 95%CI 0.72-1.23, p=0.64; OS: HR 0.95, 95%CI 0.83-1.09, p=0.45). Given the tendency for patients with more aggressive biology to undergo pretreatment MRI and/or biopsy, we performed a subgroup analyses restricted to patients with high-grade tumors. In this analysis of high-grade tumors only (n=2,025), we observed no significant difference in OS by pre-operative evaluation strategy (p>0.05; Table 4). Subgroup analysis of tumors ≥8cm also showed no differences in DSS or OS between groups with IPTW (p>0.05 all).
Figure 3.
Survival for planned versus unplanned excisions with (A) unweighted and (B) inverse probability of treatment weighted (IPTW) Kaplan Meier curves for overall survival and (C) unweighted and (D) IPTW cumulative incidence curves for disease specific survival. *Model adjusted with IPTW (using propensity scores created from model with covariates: age, gender, race, Charlson comorbidity index score, tumor histology, size, grade, site, depth, and year of diagnosis) and for chemotherapy and radiotherapy. SHR, subhazard ratio.
Table 4.
Inverse Probability of Treatment Weighted* Multivariate Survival Analyses with Fine and Gray’s Competing Risks Model for Disease Specific Survival and Cox Proportional Hazards Model for Overall Survival for High-Grade Tumors†
| High-Grade Tumors Disease Specific Survival |
High-Grade Tumors Overall Survival |
|||||||
|---|---|---|---|---|---|---|---|---|
| Subhazard Ratio |
95%CI | P-value | Hazard Ratio |
95% CI | P-value | |||
| Preoperative Planning | ||||||||
| Unplanned (ref) | ||||||||
| MRI & Biopsy | 1.34 | 1.01 | 1.79 | 0.05 | 1.05 | 0.89 | 1.23 | 0.57 |
| MRI only | 1.27 | 0.98 | 1.64 | 0.07 | 0.97 | 0.83 | 1.13 | 0.71 |
| Biopsy only | 1.02 | 0.68 | 1.54 | 0.91 | 1.03 | 0.82 | 1.29 | 0.79 |
| Chemotherapy | ||||||||
| Preoperative | 0.95 | 0.41 | 2.24 | 0.91 | 1.03 | 0.76 | 1.41 | 0.83 |
| Postoperative | 1.55 | 1.07 | 2.25 | 0.02 | 1.00 | 0.79 | 1.26 | 0.98 |
| Radiation Therapy | ||||||||
| Preoperative | 0.85 | 0.61 | 1.20 | 0.36 | 0.75 | 0.60 | 0.95 | 0.02 |
| Postoperative | 0.75 | 0.59 | 0.95 | 0.02 | 0.66 | 0.57 | 0.77 | <0.001 |
Ref, reference; MRI, magnetic resonance imaging; CI, confidence interval.
Inverse probability of treatment (i.e. preoperative planning) weighted adjustment with propensity scores using covariates: age, gender, race, Charlson comorbidity index score, tumor histology, size, grade, site, depth, and diagnostic year.
High-grade tumors defined as grade 3 and 4 (N=2,025).
When analyzing costs, actual and standardized total Medicare payments were greater for planned excisions compared to unplanned excisions in the unadjusted and multivariate IPTW analyses (p<0.05, Table A1, Supplement). When evaluating the components of total Medicare payments (Table 5), the first STS excision including hospitalization and services provided within 30-day after discharge were observed to comprise the majority of the total costs for all groups. Therefore, we included facility type and NCI center designation of the first STS resection in a second multivariate IPTW model and found no difference in actual or standardized Medicare payments based on type of preoperative evaluation (p>0.05, Table A1, Model 2). Reconstructive surgery was associated with greater costs in all models (p<0.001 all). Additionally, an inpatient facility location for the first excision, compared to an outpatient/ambulatory surgery location, was associated with greater costs for both actual and standardized Medicare payments (Table A1, p<0.001 all). In the subgroup analysis of tumors ≥8 cm, we found similar results with higher actual and standardized costs among planned excisions in multivariable IPTW models (p<0.05).
Table 5.
Mean Actual and Standardized Medicare Payments for Planned versus Unplanned Excisions.
| MRI & Biopsy N=632 |
MRI Only N=666 |
Biopsy Only N=276 |
Unplanned N=829 |
|||||
|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| Total Actual Payments ($) | 28,421 | 29,014 | 23,061 | 24,494 | 24,451 | 26,874 | 18,151 | 21,470 |
| Adjusted Total Actual Payments($)* | 25,119 | 31,653 | 23,881 | 26,260 | 28,656 | 50,951 | 20,537 | 25,002 |
| MRI($) | 714 | 415 | 735 | 415 | ||||
| Biopsy($) | 1,486 | 4,065 | 1,641 | 6,553 | ||||
| First Resection†($) | 23,112 | 24,180 | 15,576 | 18,361 | 18,024 | 22,055 | 10,926 | 14,348 |
| Re-excision†($) | 4,787 | 15,172 | 8,232 | 17,290 | 6,658 | 15,798 | 8,689 | 17,790 |
| Total Standardized Payments($) | 27,068 | 27,489 | 21,901 | 22,259 | 23,993 | 25,853 | 17,918 | 20,821 |
| Adjusted Total Standardized Payments($)* | 23,562 | 30,251 | 22,563 | 23,995 | 27,413 | 51,901 | 20,059 | 23,313 |
| MRI($) | 715 | 406 | 746 | 401 | ||||
| Biopsy($) | 1,405 | 3,733 | 1,692 | 6,536 | ||||
| First Resection†($) | 21,889 | 22,945 | 14,861 | 16,455 | 17,713 | 21,128 | 10,936 | 14,549 |
| Re-excision†($) | 4,629 | 14,690 | 7,688 | 16,066 | 6,516 | 15,251 | 8,474 | 17,029 |
MRI, Magnetic resonance imaging; SD, standard deviation
Adjusted costs estimated from generalized linear models with gamma distribution and log-link with inverse probability of treatment weighting (covariates: age, gender, race, Charlson comorbidity index, tumor histology, size, grade, site, depth, and diagnostic year) and adjusting for chemotherapy, radiotherapy, and reconstruction.
Includes costs of surgeries, associated hospitalization, and services within 30-day after discharge.
Discussion
In this SEER-Medicare analysis of elderly STS patients, contrary to expectations, we observed that preoperative evaluation strategy, whether MRI, biopsy, MRI with biopsy, or initial “unplanned” excision, was not associated with significant differences in oncologic outcome, but was associated with lower healthcare costs when MRI and/or biopsy were not performed. Although our a priori hypothesis was that preoperative MRI and biopsy would improve DSS and OS by permitting optimal surgical planning leading to a reduction in positive margins, fewer operations, and better compliance with multimodality care, we did not observe such findings. In addition, although re-excision rates were significantly higher for unplanned excisions (but only 47% suggesting a selective approach to tumor bed excision), we importantly did not find that this translated into higher overall complication rates or healthcare costs. In fact, total surgical cancer-related costs were significantly higher when these preoperative evaluation strategies were employed.
Our findings therefore raise important questions regarding the oncologic and healthcare implications of unplanned excisions and suggest that efforts to develop a selective approach to preoperative STS evaluation and re-excision after unplanned excision are indicated among the elderly. Ultimately, we were unable to show meaningful oncological benefits associated with preoperative MRI/biopsy or re-excision following unplanned excisions, despite extensive analyses including subset analyses focusing on higher risk groups such as those with high-grade tumors and those with tumors greater than 8 cm. Although several retrospective studies have observed similar findings regarding the oncologic outcomes following unplanned STS excision, our study is the first to provide a population-based analysis of a large generalizable cohort (although restricted to elderly patients) and the first to examine associated costs.5, 6,9, 11 Though Lewis et al. observed that STS patients who underwent re-resection after unplanned excision had superior survival compared to patients who underwent planned excision,5 a key limitation of that study was the difference in final surgical margins between the groups. Tumor-positive margins were higher (26%) after planned excision compared to those who underwent re-resection after unplanned excision (9%), reinforcing the premise that tumor-free surgical margins are a key predictor of STS outcome. Yet, the impact of more aggressive local treatment on ultimate STS outcomes is controversial, and similar to recent key publications regarding the role of re-excision in breast carcinoma,39 a more selective approach to re-excision in STS may be justified, even in cases of unplanned excision which historically have been considered “non-oncologic.” In fact, a key implication of our data is that similar to recent trends in the surgical management of breast cancer where less aggressive pursuit of margins has replaced prior recommendations for aggressive re-excisions and widely clear margins,40, 41 an examination of comparable multi-modality treatment paradigms in soft tissue sarcoma may be indicated.
In addition to an apparent lack of significant oncologic benefit in our analysis, preoperative MRI and biopsy also failed to show significant benefits with respect to operative morbidity. Although we expected fewer complications after planned excisions given the role of MRI for preoperative surgical planning, in fact, we observed the opposite.28 Planned excisions had higher rates of postoperative complications compared to unplanned excisions for the initial procedure and overall, even when analyzing a subset of patients with larger tumors known to have an elevated risk of surgical complications. Although the underlying cause of these findings is unknown, it may suggest a tendency for planned operations to be overly aggressive in an attempt to obtain wide negative margins, especially in the era of combined modality therapy where neo-adjuvant and adjuvant radiotherapy have allowed for greater function-preserving and less radical operations.
Importantly, in addition to a lack of significant improvement in oncologic and morbidity endpoints, we observed greater total surgical and perioperative costs among patients undergoing preoperative MRI and/or biopsy. Our results suggest that performing the initial STS operation in an outpatient setting is significantly less expensive than performing the surgery in an inpatient hospital setting. This likely contributed to the increased costs among planned excisions, as there was no significant difference in actual or standardized total costs when facility type was added to the model. Considering the negligible differences in oncologic outcome and perioperative complications between STS patients having planned versus unplanned excision as the initial surgical procedure, the greater costs associated with planned excision (at least in the setting of the US healthcare system) raises further key questions regarding the optimal preoperative evaluation and surgical management of extremity and truncal STS in cases of both planned and unplanned excision. Further research is also needed to examine the impact of local recurrence on healthcare costs associated with planned versus unplanned excisions in STS.
Despite these provocative findings, it is important to acknowledge the limitations of this study. Since our sample consisted of elderly STS patients, a key caveat of our findings is that they may not generalize to younger STS patients. Elderly STS patients have been shown to differ with respect to clinicopathologic characteristics and prognosis compared to younger STS cohorts.25, 42, 43 However, research investigating healthcare outcomes and costs in the elderly has become increasingly relevant as the elderly population in the US continues to grow. Individuals ≥65 years of age currently account for 15% of the total US population and are expected to increase by 105% by 2060.44 Regardless, we acknowledge that our findings may not be generalizable to younger patients and future research should investigate with a multicenter, population-based (consisting of sarcoma referral centers and community hospitals) study whether findings are similar for younger STS patient cohorts. As we also only included patients who underwent surgical resection, our findings are not applicable to STS patients with unresectable disease. Additionally, although we were able to analyze extensive administrative data, we were limited by a lack of notable information regarding involvement of neurovascular structures and margin status, and these factors may have contributed to differences in outcomes among cohorts. Furthermore, we were unable to assess additional clinically important outcomes for STS patients, specifically local recurrence and quality of life, due to limitations of the SEER-Medicare database. Moreover, as we did not include CT imaging in our definition of preoperative planning, patients who underwent a preoperative CT scan and subsequent oncologic resection potentially obtained adequate surgical margins, and therefore received an appropriate oncologic surgery despite non-standard preoperative planning. This potentially allowed for better oncologic outcomes in the unplanned excision group in this study. Finally, as with all analyses of STS, the diverse and heterogeneous nature of the histologic types involved may confound the results. Yet, despite these acknowledged limitations, we maintain that our use of IPTW with propensity scores substantially controlled for selection bias among the groups as it was created using multiple clinically relevant and validated clinicopathologic characteristics. This approach was further supported by our subset analyses restricted to higher risk patients where a positive effect(s) of preoperative MRI and/or biopsy were more likely to be observed.
Conclusions
In this SEER-Medicare analysis of elderly STS patients, we found comparable oncological outcomes and greater costs associated with planned excisions compared to unplanned excisions. These findings occurred despite greater rates of re-excision in the unplanned excision cohort. Overall, our findings suggest that a selective approach to preoperative evaluation and re-excision of unplanned excisions may be appropriate in the elderly, as more aggressive planning and treatment strategies were not associated with improved oncologic outcomes. This decision should be made carefully based on the surgeon’s clinical experience and expertise as well as the patients’ clinical presentation including tumor burden, behavior, and proximity to neurovascular structures.
Supplementary Material
Acknowledgments
Funding: National Center for Advancing Translational Sciences, NIH(UL1TR001860); Agency for Health Care Research and Quality(T32HS022236).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of interest: None
References
- 1.Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2017. CA Cancer J Clin. 2017;67: 7–30. [DOI] [PubMed] [Google Scholar]
- 2.Mehren M, L. RR, Benjamin RS, et al. National Comprehensive Cancer Network, NCCN Clinical Practice Guidelines in Oncology Soft Tissue Sarcoma. Version 2.2017., 2017. [DOI] [PubMed] [Google Scholar]
- 3.Andritsch E, Beishon M, Bielack S, et al. ECCO Essential Requirements for Quality Cancer Care: Soft Tissue Sarcoma in Adults and Bone Sarcoma. A critical review. Crit Rev Oncol Hematol. 2017;110: 94–105. [DOI] [PubMed] [Google Scholar]
- 4.European Sarcoma Network Working Group. Soft tissue and visceral sarcomas: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2014;25 Suppl 3: iii102–112. [DOI] [PubMed] [Google Scholar]
- 5.Lewis JJ, Leung D, Espat J, Woodruff JM, Brennan MF. Effect of reresection in extremity soft tissue sarcoma. Ann Surg. 2000;231: 655–663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Smolle MA, Tunn PU, Goldenitsch E, et al. The Prognostic Impact of Unplanned Excisions in a Cohort of 728 Soft Tissue Sarcoma Patients: A Multicentre Study. Ann Surg Oncol. 2017;24: 1596–1605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zagars GK, Ballo MT, Pisters PW, Pollock RE, Patel SR, Benjamin RS. Surgical margins and reresection in the management of patients with soft tissue sarcoma using conservative surgery and radiation therapy. Cancer. 2003;97: 2544–2553. [DOI] [PubMed] [Google Scholar]
- 8.Giuliano AE, Eilber FR. The rationale for planned reoperation after unplanned total excision of soft-tissue sarcomas. J Clin Oncol. 1985;3: 1344–1348. [DOI] [PubMed] [Google Scholar]
- 9.Fiore M, Casali PG, Miceli R, et al. Prognostic effect of re-excision in adult soft tissue sarcoma of the extremity. Ann Surg Oncol. 2006;13: 110–117. [DOI] [PubMed] [Google Scholar]
- 10.Noria S, Davis A, Kandel R, et al. Residual disease following unplanned excision of soft-tissue sarcoma of an extremity. J Bone Joint Surg Am. 1996;78: 650–655. [DOI] [PubMed] [Google Scholar]
- 11.Bianchi G, Sambri A, Cammelli S, et al. Impact of residual disease after "unplanned excision" of primary localized adult soft tissue sarcoma of the extremities: evaluation of 452 cases at a single Institution. Musculoskelet Surg. 2017. [DOI] [PubMed] [Google Scholar]
- 12.Qureshi YA, Huddy JR, Miller JD, Strauss DC, Thomas JM, Hayes AJ. Unplanned excision of soft tissue sarcoma results in increased rates of local recurrence despite full further oncological treatment. Ann Surg Oncol. 2012;19: 871–877. [DOI] [PubMed] [Google Scholar]
- 13.Chandrasekar CR, Wafa H, Grimer RJ, Carter SR, Tillman RM, Abudu A. The effect of an unplanned excision of a soft-tissue sarcoma on prognosis. J Bone Joint Surg Br. 2008;90: 203–208. [DOI] [PubMed] [Google Scholar]
- 14.Gustafson P, Dreinhofer KE, Rydholm A. Soft tissue sarcoma should be treated at a tumor center. A comparison of quality of surgery in 375 patients. Acta Orthop Scand. 1994;65: 47–50. [DOI] [PubMed] [Google Scholar]
- 15.Potter BK, Adams SC, Pitcher JD Jr., Temple HT. Local recurrence of disease after unplanned excisions of high-grade soft tissue sarcomas. Clin Orthop Relat Res. 2008;466: 3093–3100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Brown ML, Riley GF, Schussler N, Etzioni R. Estimating health care costs related to cancer treatment from SEER-Medicare data. Med Care. 2002;40: IV-104–117. [DOI] [PubMed] [Google Scholar]
- 17.Yabroff KR, Lund J, Kepka D, Mariotto A. Economic burden of cancer in the United States: estimates, projections, and future research. Cancer Epidemiol Biomarkers Prev. 2011;20: 2006–2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Yabroff KR, Lamont EB, Mariotto A, et al. Cost of care for elderly cancer patients in the United States. J Natl Cancer Inst. 2008;100: 630–641. [DOI] [PubMed] [Google Scholar]
- 19.National Cancer Institute. Surveillance, Epidemiology, and End Results (SEER), a program of the National Cancer Institute (NCI), provides information on cancer statistics to help reduce the burden of cancer on the U.S. Population. 2012. Available from URL: https://seer.cancer.gov/about/factsheets/SEERbrochure.pdf. [accessed September 22, 2017] [Google Scholar]
- 20.Barnett JC, Berchick ER. Health Insurance Coverage in the United States: 2016 Current Population Reports: United States Census Bureau, 2017:1–35. [Google Scholar]
- 21.Warren JL, Harlan LC, Fahey A, et al. Utility of the SEER-Medicare data to identify chemotherapy use. Med Care. 2002;40: IV-55–61. [DOI] [PubMed] [Google Scholar]
- 22.Cooper GS, Virnig B, Klabunde CN, Schussler N, Freeman J, Warren JL. Use of SEER-Medicare data for measuring cancer surgery. Med Care. 2002;40: IV-43–48. [DOI] [PubMed] [Google Scholar]
- 23.Virnig BA, Warren JL, Cooper GS, Klabunde CN, Schussler N, Freeman J. Studying radiation therapy using SEER-Medicare-linked data. Med Care. 2002;40: IV-49–54. [DOI] [PubMed] [Google Scholar]
- 24.Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40: IV-3–18. [DOI] [PubMed] [Google Scholar]
- 25.Al-Refaie WB, Habermann EB, Dudeja V, et al. Extremity soft tissue sarcoma care in the elderly: insights into the generalizability of NCI Cancer Trials. Ann Surg Oncol. 2010;17: 1732–1738. [DOI] [PubMed] [Google Scholar]
- 26.Nathan H, Atoria CL, Bach PB, Elkin EB. Hospital volume, complications, and cost of cancer surgery in the elderly. J Clin Oncol. 2015;33: 107–114. [DOI] [PubMed] [Google Scholar]
- 27.Guadagnolo BA, Xu Y, Zagars GK, et al. A population-based study of the quality of care in the diagnosis of large (>/=5 cm) soft tissue sarcomas. Am J Clin Oncol. 2012;35: 455–461. [DOI] [PubMed] [Google Scholar]
- 28.De La Hoz Polo M, Dick E, Bhumbra R, Pollock R, Sandhu R, Saifuddin A. Surgical considerations when reporting MRI studies of soft tissue sarcoma of the limbs. Skeletal Radiol. 2017;46: 1667–1678. [DOI] [PubMed] [Google Scholar]
- 29.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40: 373–383. [DOI] [PubMed] [Google Scholar]
- 30.Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45: 613–619. [DOI] [PubMed] [Google Scholar]
- 31.Birkmeyer JD, Gust C, Dimick JB, Birkmeyer NJ, Skinner JS. Hospital quality and the cost of inpatient surgery in the United States. Ann Surg. 2012;255: 1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Gottlieb DJ, Zhou W, Song Y, Andrews KG, Skinner JS, Sutherland JM. Prices don't drive regional Medicare spending variations. Health Aff (Millwood). 2010;29: 537–543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Gottlieb D, Zhou W, Song Y, Andrews KG, Skinner J, Sutherland J. Technical Report: A Standardized Method for Adjusting Medicare Expenditures for Regional Differences in Prices: The Darthmouth Institute for Health Policy & Clinical Practice, 2010. [Google Scholar]
- 34.Mariotto AB, Yabroff KR, Shao Y, Feuer EJ, Brown ML. Projections of the cost of cancer care in the United States: 2010-2020. J Natl Cancer Inst. 2011; 103: 117–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ghaferi AA, Birkmeyer JD, Dimick JB. Complications, failure to rescue, and mortality with major inpatient surgery in medicare patients. Ann Surg. 2009;250: 1029–1034. [DOI] [PubMed] [Google Scholar]
- 36.Iezzoni LI, Daley J, Heeren T, et al. Identifying complications of care using administrative data. Med Care. 1994;32: 700–715. [DOI] [PubMed] [Google Scholar]
- 37.Garrido MM, Kelley AS, Paris J, et al. Methods for constructing and assessing propensity scores. Health Serv Res. 2014;49: 1701–1720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. Journal of the American Statistical Association. 1999;94: 496–509. [Google Scholar]
- 39.Morrow M, Abrahamse P, Hofer TP, et al. Trends in Reoperation After Initial Lumpectomy for Breast Cancer: Addressing Overtreatment in Surgical Management. JAMA Oncol. 2017;3: 1352–1357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Gradishar WJ, Anderson BO, Balassanian R, et al. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Breast Cancer. Version 2.2017. Available from URL: https://www.nccn.org/professionals/physiciangls/pdf/breast.pdf. [accessed September 21, 2017] [Google Scholar]
- 41.Moran MS, Schnitt SJ, Giuliano AE, et al. Society of Surgical Oncology-American Society for Radiation Oncology consensus guideline on margins for breast-conserving surgery with whole-breast irradiation in stages I and II invasive breast cancer. J Clin Oncol. 2014;32: 1507–1515. [DOI] [PubMed] [Google Scholar]
- 42.Hoven-Gondrie ML, Bastiaannet E, Ho VK, et al. Worse Survival in Elderly Patients with Extremity Soft-Tissue Sarcoma. Ann Surg Oncol. 2016;23: 2577–2585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Biau DJ, Ferguson PC, Turcotte RE, et al. Adverse effect of older age on the recurrence of soft tissue sarcoma of the extremities and trunk. J Clin Oncol. 2011;29: 4029–4035. [DOI] [PubMed] [Google Scholar]
- 44.United States Census Bureau. Older Americans Month: May 2017. Available from URL: https://www.census.gov/content/dam/Census/newsroom/facts-for-features/2017/cb17-ff08.pdf [accessed September 22, 2017]. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.