Background
Squamous cell carcinoma of the anus (SCCA) is largely a disease of the elderly, with a median age at diagnosis of 62 years.1 The incidence of SCCA is increasing, particularly among patients aged ≥50 years, and a corresponding increase in mortality has been observed, with a nearly 5% increase in SCCA-based mortality in patients aged 60 to 69 years.2
The current standard of care (SOC) for treatment of localized SCCA is definitive chemoradiation (CRT) with concurrent multiagent chemotherapy,3 initially pioneered by Dr. Nigro and refined through multiple randomized trials.4–6 Despite the elderly representing a large proportion of patients with SCCA, they are disproportionately underrepresented in trials that helped define SOC treatment. Notably, 75% of patients in the ACT II trial were aged <65 years, inconsistent with the epidemiology of SCCA in the current population.6 Given the paucity of data regarding optimal management of the elderly, there is temptation to deliver non-SOC therapy, such as radiation therapy (RT) alone or CRT with single-agent chemotherapy, to patients with otherwise curable disease in order to reduce toxicity. The limited retrospective studies examining definitive treatment of SCCA in the elderly have shown mixed results, although data from other disease sites, such as non–small cell lung cancer, suggest that the elderly, if eligible, should receive SOC therapy.7–10 In common malignancies (eg, non–small cell lung cancer), few prospective studies have focused on curative treatment of the elderly.11,12 For SCCA, which represents 0.5% of new cancer cases in the United States, prospective trials to determine optimal treatment of the elderly would be challenging to conduct. However, because the elderly will represent an increasingly significant proportion of the population in the upcoming decades, coupled with the increasing incidence of SCCA in this age group, there is a pressing need to evaluate how the elderly are currently being treated and determine whether there is an optimal regimen for definitive treatment of this patient population.
In the present study, we set out to fully characterize and compare patterns of care in elderly (age ≥65 years) versus nonelderly (age <65 years) patients with stages I–III SCCA. We hypothesized that elderly patients would be less likely than younger patients to receive any treatment or curative-intent CRT. We further set out to compare overall survival (OS) in elderly patients treated with SOC CRT versus CRT with single-agent chemotherapy, which may represent an acceptable compromise in these patients.
Methods
Patient Selection
We performed a retrospective review using the National Cancer Database (NCDB), a combined effort of the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society. The NCDB is a nationwide hospital-based database with deidentified hospital registry data from >1,500 accredited facilities and represents >70% of newly diagnosed cancer cases in the United States.13 The CoC’s NCDB and the hospitals participating in the CoC NCDB are the source of the deidentified data, and they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions presented in this study.
Data on patients diagnosed with stages I–III SCCA from 2004 through 2015 were collected from the participant user file, with additional inclusion and exclusion criteria summarized in Figure 1. Although the definition of “elderly” varies, we defined it as age ≥65 years, similar to other studies.14–16 Patients treated with abdominoperineal resection, a non-SOC treatment of primary disease, represented a small proportion of this patient population and were excluded (n=981) (Figure 1). We also excluded patients with unknown disease stage, stage 0 or IV disease, nonsquamous histology, unknown RT and chemotherapy treatment details, and RT doses <8 or >59.4 Gy.
Study flow diagram of analytic cohorts.
Abbreviation: NCDB, National Cancer Database.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Study flow diagram of analytic cohorts.
Abbreviation: NCDB, National Cancer Database.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Study flow diagram of analytic cohorts.
Abbreviation: NCDB, National Cancer Database.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Treatment Definitions
We defined 4 treatment groups: SOC CRT with multiagent chemotherapy, non-SOC local therapy, palliative treatment, and no treatment. Patients were considered to have received SOC CRT if they (1) underwent concurrent CRT with multiagent chemotherapy delivered within 14 days of RT initiation and (2) received a radiation dose of 45 to 59.4 Gy. This 14-day window was based on the criteria for discontinuing protocol treatment in RTOG 0529. We used 45 Gy based on the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Anal Carcinoma.3 The non-SOC local therapy group consisted of CRT with single-agent chemotherapy, RT alone to a dose of 45 to 59.4 Gy, CRT with chemotherapy delivered >14 days from initiation of RT, and local tumor excision or excisional biopsy for patients with T1N0 disease. The palliative treatment group consisted of chemotherapy alone, RT or CRT to doses <45 Gy, and local tumor excision/excisional biopsy for patients with stage II–III disease, and non-SOC surgical procedures (supplemental eTable 1, available with this article at JNCCN.org). The no-treatment group consisted of patients who did not receive RT, chemotherapy, or any surgical procedure.
Study Variables
We dichotomized the following baseline covariates: elderly (age ≥65 vs <65 years), sex (male vs female), race (White vs non-White), median income (≥$46,000 vs <$46,000), insurance status (insured vs uninsured), and facility type (academic vs nonacademic). We also analyzed the effect of age by decade (<50 years, 50–59 years, 60–69 years, and ≥70 years) and by using 70 years of age as a cutoff for elderly (age ≥70 vs <70 years). Clinical stage group was categorized as T1N0, T2N0, and T3N0/stage III. The Charlson-Deyo comorbidity score was dichotomized as 0 (no comorbidities) or 1 (≥1 comorbidity). Distance to the nearest facility was analyzed as a continuous variable.
Statistical Methods
The primary objective of this study was to evaluate patterns of care in elderly versus nonelderly patients with localized SCCA. Differences in treatment groups between elderly and nonelderly patients were tested using the chi-square test. Logistic regression was used to identify predictors of (1) SOC CRT versus other treatments/no treatment and (2) multiagent versus single-agent chemotherapy in patients receiving CRT to doses of 45 to 59.4 Gy and chemotherapy initiated within 14 days of RT. Variables with P ≤ 10 in univariate analysis were included in the multivariate logistic regression model. We hypothesized that elderly patients would be less likely to receive SOC CRT and, when SOC CRT was delivered, that they would be less likely than nonelderly patients to receive multiagent chemotherapy.
We further set out to determine OS in elderly patients treated with SOC CRT with multiagent chemotherapy versus non-SOC CRT with single-agent chemotherapy in a propensity-matched cohort. All baseline covariates mentioned in the “Study Variables” section were included in the propensity score model. Patients receiving single-agent chemotherapy were matched with those receiving multiagent chemotherapy using a 1:1 nearest available neighbor match without replacement17 using a caliper size calculated as 20% of the standard deviation of the propensity score.18 Common support of the propensity score distributions was evaluated graphically, and balance was evaluated by computing the standardized difference of the covariates across the 2 groups.19 After propensity score matching, OS between treatment groups was estimated using the Kaplan-Meier method, and the effect of multiagent chemotherapy was evaluated with a Cox proportional hazards model with robust standard errors to account for clustering in matched pairs. All statistical analyses were performed using SAS 9.4 (SAS Institute Inc).
Results
Baseline Characteristics
We identified 9,156 elderly patients and 17,640 nonelderly patients who met the study criteria (Figure 1). Patient characteristics for the entire cohort are shown in Table 1. Patients excluded because of unknown stage (n=6,444) were of similar age to the study cohort (mean [SD] age, 60.3 [13.9] years vs 60.3 [12.4] years; P=.9163) and had a slightly higher proportion of patients with ≥1 comorbidity (22.0% vs 20.2%; P=.0014). Overall, elderly patients were less likely than nonelderly patients to have no comorbidities, be treated at academic centers, be uninsured, and have higher-stage disease. Elderly patients were more likely than nonelderly patients to be female and White.
Patient Characteristics
Table 1 provides a summary of treatments delivered in the elderly and nonelderly patients, and Figure 2 shows a more detailed description by stage and age. The proportion of patients receiving SOC CRT increased with increasing stage in each age group. Notably, among patients aged ≥70 years with higher than T1N0 disease, the proportion receiving SOC CRT was near 50% (49.6% for T2N0; 52.6% for T3N0/stage III). In contrast, these rates were >70% for patients aged 50–59 years (70.2% for T2N0; 70.6% for T3N0/stage III). The lowest SOC CRT rates were seen in patients with T1N0 disease, in whom >25% of patients aged ≥70 and <50 years underwent excision as their primary treatment, and the overall proportion treated with non-SOC local therapies was numerically higher than that of those treated with SOC CRT in these age groups (44.1% vs 43.3% for <50 years; 45.6% vs 41.3% for ≥70 years) (Figure 2A). Although the number of patients who did not receive any treatment was low, Table 1 shows that the elderly were significantly more likely than nonelderly patients to not receive any treatment (4.3% vs 2.4%; P<.0001). This effect is illustrated in Figure 2C, which shows that 6.3% of patients aged ≥70 years with T3N0/stage III disease received no treatment compared with 3.2% for 60–69 years, 2.5% for 50–59 years, and 3.2% for <50 years.
Distribution of treatments delivered by age group in patients with (A) T1N0 disease, (B) T2N0 disease, and (C) T3N0/stage III disease.
Abbreviations: CRT, chemoradiation; LT, local therapy; SA, single-agent chemotherapy; SOC, standard of care.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Distribution of treatments delivered by age group in patients with (A) T1N0 disease, (B) T2N0 disease, and (C) T3N0/stage III disease.
Abbreviations: CRT, chemoradiation; LT, local therapy; SA, single-agent chemotherapy; SOC, standard of care.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Distribution of treatments delivered by age group in patients with (A) T1N0 disease, (B) T2N0 disease, and (C) T3N0/stage III disease.
Abbreviations: CRT, chemoradiation; LT, local therapy; SA, single-agent chemotherapy; SOC, standard of care.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
SOC CRT Versus Other/No Treatment
A significantly lower proportion of elderly versus nonelderly patients received SOC CRT (54.5% vs 65.0%; P<.0001). In multivariate analysis, elderly patients were 38% less likely than nonelderly patients (odds ratio [OR], 0.62; 95% CI, 0.58–0.65; P<.0001) to receive SOC CRT (Table 2). We found similar trends when categorizing age by decade or defining elderly as ≥70 years of age. Duration of SOC CRT was similar between elderly and nonelderly patients (56.6 vs 56.9 days; P=.86). In addition, patients with more comorbidities, with T1N0 disease (vs T2N0), and treated at academic centers were less likely to receive SOC CRT, whereas women, patients of White race, and those with T3N0/stage III (vs T2N0) disease were more likely to receive SOC CRT. When categorizing age by decade, we found that patients aged <50 years were less likely than patients aged 50–59 years to receive SOC CRT.
Logistic Regression Analysis of Factors Associated With SOC CRT Versus Other/No Treatment
Multiagent Versus Single-Agent Chemotherapy
Although 4,994 elderly patients and 11,467 nonelderly patients received SOC CRT, an additional 1,014 elderly and 1,541 nonelderly patients received CRT with single-agent chemotherapy. A significantly higher proportion of the elderly were treated with single-agent chemotherapy (16.9% vs 11.8%; P<.0001), resulting in their having a >1.5-fold increase in the likelihood of receiving single-agent chemotherapy (OR, 1.52; 95% CI, 1.39–1.66) in multivariate analysis (Table 3). These trends were similar when categorizing age by decade or defining elderly as ≥70 years of age. Female patients, White patients, and those with T3N0/stage III disease were more likely to receive multiagent chemotherapy in multivariate analysis.
Logistic Regression Analysis of Receiving Single-Agent Versus Multiagent Chemotherapy in Patients Who Received CRT to 45–59.4 Gy
We next evaluated OS in the elderly treated with SOC CRT versus single-agent CRT. With propensity score matching, a multiagent chemotherapy match (using a caliper size of 0.0148 based on a propensity score SD of 0.0742) was identified for 1,011 of the possible 1,014 patients treated with single-agent CRT (supplemental eTable 2). The propensity score distributions showed nearly ideal common support (supplemental eFigure 1). Median follow-up was 38.2 months (interquartile range, 18.4–68.2 months), at which point 852 patients had died. The 3-year OS was significantly higher in patients who received SOC CRT (77.1% vs 67.5%; hazard ratio, 0.78; 95% CI, 0.68–0.89; P=.0002) (Figure 3A). We performed a similar propensity score–matched analysis defining elderly as ≥70 years of age (n=743 treated with single-agent CRT; n=2,989 treated with SOC CRT). A total of 728 of a possible 743 matches were identified (supplemental eTable 3). The 3-year OS with SOC CRT versus single-agent CRT was 69.8% versus 63.1% (hazard ratio, 0.85; 95% CI, 0.74–0.98; P=.0268) (Figure 3B).
Overall survival of propensity score–matched elderly patients, defined as age (A) ≥65 years or (B) ≥70 years, treated with definitive chemoradiation with single-agent versus multiagent chemotherapy. Curves represent actual survival as estimated by the Kaplan-Meier method.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Overall survival of propensity score–matched elderly patients, defined as age (A) ≥65 years or (B) ≥70 years, treated with definitive chemoradiation with single-agent versus multiagent chemotherapy. Curves represent actual survival as estimated by the Kaplan-Meier method.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Overall survival of propensity score–matched elderly patients, defined as age (A) ≥65 years or (B) ≥70 years, treated with definitive chemoradiation with single-agent versus multiagent chemotherapy. Curves represent actual survival as estimated by the Kaplan-Meier method.
Citation: Journal of the National Comprehensive Cancer Network 20, 6; 10.6004/jnccn.2020.7691
Discussion
We found substantial disparities in the management of stages I–III SCCA based on age (≥65 vs <65 years). First, compared with nonelderly patients, elderly patients were 38% less likely to receive SOC CRT. Next, when definitive CRT was delivered, the elderly had a >1.5-fold increased odds of receiving concurrent single-agent rather than multiagent chemotherapy, which was associated with worse survival.
Definitive treatment of SCCA with SOC CRT is associated with significant toxicity, and de-escalation of therapy or choosing an alternative local therapy may be a consideration when evaluating patients who may be particularly susceptible to certain adverse effects of treatment. Our results showed that, compared with the nonelderly, a lower proportion of the elderly receive treatment with SOC CRT, whereas a higher proportion are treated with non-SOC local therapy, including single-agent CRT or excision. The most striking observation was in patients aged ≥70 years, among whom the proportion receiving SOC CRT was close to 50%, including those with T3N0/stage III disease. Notably, the results of our propensity score–matched analysis showed improved OS with SOC CRT using multiagent chemotherapy when including patients aged ≥65 years and those aged ≥70 years, suggesting that these patients should receive SOC CRT, if feasible. In treating patients with early-stage disease, practitioners may have a preference to forgo CRT to avoid treatment toxicities. Interestingly, we found that in patients with T1N0 disease, the lowest rates of SOC CRT were observed in patients aged ≥70 and <50 years. The underuse of SOC CRT in younger patients warrants further study.
The optimal management of localized SCCA in the elderly is controversial. A recent study by Raphael et al20 explored rates of CRT treatment interruption and noncompletion. Patients aged >70 years were 40% less likely than those aged <50 years to complete CRT. In addition, patients who did not complete CRT had a higher risk of salvage abdominoperineal resection or death. Interestingly, we found that duration of SOC CRT (45–59.4 Gy) in our study was equivalent in the elderly and nonelderly. However, it is possible that some elderly patients did not complete the planned CRT course, which could partly explain the lower SOC CRT rate seen in our elderly population.
Given that SCCA is rare, much of the available guidance in elderly patients is retrospective. In an early study, Allal et al7 reported outcomes of 58 patients with SCCA aged ≥75 years. Although most patients were treated with split-course RT using outdated techniques, the outcomes and toxicity compared favorably with rates reported for younger patients. Saarilahti et al10 reported on 62 patients treated with CRT for SCCA (37% aged ≥70 years). Treatment was more modern, including use of intensity-modulated radiation therapy (IMRT). The rates of locoregional control and disease-free survival (DFS) were similar between the 2 age groups, despite a higher rate of required chemotherapy course changes and acute toxicity in the elderly. Fallai et al21 reported outcomes of 62 patients aged ≥70 years treated with RT or CRT for SCCA. Patients receiving CRT had improved local control, locoregional control, and DFS compared with the RT-alone group, with higher rates of grade ≥3 toxicity.
Other studies have shown less favorable outcomes with definitive treatment in elderly patients with SCCA. Reporting on 115 patients with SCCA, 30% of whom were aged ≥70 years, Claren et al8 found that elderly patients were treated with CRT less often than their younger counterparts, with fewer elderly patients completing the prescribed treatment course. Among elderly patients treated with definitive CRT, inferior DFS and metastasis-free survival was observed compared with younger patients, likely secondary to a de-escalation of therapy in that patient group.
Because of concerns regarding increased toxicity and the potential for inferior outcomes, modifications to the SOC CRT regimen have been suggested in the elderly. Charnley et al22 reported on 16 elderly patients with SCCA believed not to be suitable for standard CRT because of their age, performance status, or other comorbidity. Patients were treated with a reduced RT dose and single-agent chemotherapy. Treatment was well tolerated, with no patients experiencing hematologic toxicity. At a median follow-up of 16 months, the local control rate was 73%, disease-specific survival was 86%, and OS was 69%. De-escalation of SOC CRT for SCCA is being investigated currently in the prospective DECREASE study (ClinicalTrials.gov identifier: NCT04166318), which is evaluating the use of lower-dose CRT for patients with early-stage SCCA. The results of such studies are vitally important for identification of effective treatment regimens that might be better tolerated by vulnerable patient groups such as the elderly.
One limitation to the aforementioned studies focusing on outcomes in the elderly is the routine use of 2D or 3D conformal RT as opposed to more modern treatment planning with dose-painting IMRT. Reported results of the RTOG 0529 trial showed a significant reduction in acute toxicities compared with conventional radiation delivery reported in RTOG 98-11.23 Treatment with IMRT may be particularly valuable in this vulnerable patient population.24–26
Along with the morbidity and mortality associated with SCCA in the elderly, the economic burden is also substantial. Deshmukh et al27 analyzed data from the SEER-Medicare linked database to assess the economic impact of SCCA in the elderly (age ≥66 years) in the United States. The cost of cancer-related care increased as age and stage increased, with a lifetime economic burden of $112 million. The investigators point out that although the economic impact of anal cancer is small compared with the overall cost of cancer care in the United States, the overall financial burden of SCCA is considerable, given that it represents a very small proportion (0.003%) of all cancer subsites. The economic and health impacts of SCCA on the elderly represents an area ripe for improvement. The routine use of comprehensive geriatric assessment in elderly patients diagnosed with SCCA may help in the development of an optimal plan of care and identify health issues that may otherwise be overlooked.28 Increasing efforts to prevent this malignancy or detect disease earlier also need to be considered, including improving rates of HPV vaccination and expanding screening of at-risk populations.2
Although this is a comprehensive analysis of patterns of care in 26,796 patients with SCCA, there are several limitations. First, it is possible that a subset of patients included in our analysis had T1–2N0 perianal cancer and were treated with local excision alone rather than with CRT, which is considered SOC. The NCDB provides detailed treatment data, including radiation dose, treatment delivery dates, and limited systemic therapy details, which are necessary in demarcating definitive from palliative treatment. However, although the NCDB offers detailed treatment data, information regarding performance status and medical comorbidities, including HIV status, are not included aside from the Charlson-Deyo comorbidity score. Therefore, other potential confounders that influence treatment decision-making and survival likely persist. One of the major limitations of the NCDB is that no other endpoints are captured, such as local control or cancer-specific survival. Additional limitations include the definition of “elderly” and “curative treatment.” We believed that the age of 65 years was a reasonable threshold, given the recent publication showing the increase in SCCA incidence among patients aged ≥60 years.2 We also found similar results when we used 70 years of age as a cutoff for the definition of “elderly.” Our definition of curative therapy for SCCA was based on the currently available NCCN Guidelines and the definition of per-protocol treatment used in RTOG 0529. Another important consideration is that the NCDB does not capture the specific chemotherapy agents used, so we were unable to compare patterns of cisplatin/5-FU versus mitomycin C/5-FU administration in this study.
Conclusions
Our results showed that elderly patients with localized SCCA are less likely than nonelderly patients to receive SOC CRT. When SOC CRT was delivered, the elderly were >1.5 times more likely to receive single-agent versus multiagent chemotherapy. However, we found that SOC CRT with multiagent chemotherapy was associated with improved survival compared with CRT with single-agent chemotherapy in elderly patients. Therefore, our results suggest that CRT with multiagent chemotherapy, which is the SOC treatment regimen for SCCA, should be offered to eligible elderly patients. Future studies should prioritize inclusion of the elderly to help define an optimal regimen for this expanding patient population.
References
- 1.↑
NCI Surveillance, Epidemiology, and End Results Program. Cancer stat facts: anal cancer. Accessed July 31, 2020. Available at: http://seer.cancer.gov/statfacts/html/anus.html
- 2.↑
Deshmukh AA, Suk R, Shiels MS, et al. Recent trends in squamous cell carcinoma of the anus incidence and mortality in the United States, 2001–2015. J Natl Cancer Inst 2020;112:829–838.
- 4.↑
Nigro ND, Vaitkevicius VK, Considine B Jr. Combined therapy for cancer of the anal canal: a preliminary report. Dis Colon Rectum 1974;17:354–356.
- 5.↑
Gunderson LL, Winter KA, Ajani JA, et al. Long-term update of US GI intergroup RTOG 98-11 phase III trial for anal carcinoma: survival, relapse, and colostomy failure with concurrent chemoradiation involving fluorouracil/mitomycin versus fluorouracil/cisplatin. J Clin Oncol 2012;30:4344–4351.
- 6.↑
James RD, Glynne-Jones R, Meadows HM, et al. Mitomycin or cisplatin chemoradiation with or without maintenance chemotherapy for treatment of squamous-cell carcinoma of the anus (ACT II): a randomised, phase 3, open-label, 2 × 2 factorial trial. Lancet Oncol 2013;14:516–524.
- 7.↑
Allal AS, Obradovic M, Laurencet F, et al. Treatment of anal carcinoma in the elderly: feasibility and outcome of radical radiotherapy with or without concomitant chemotherapy. Cancer 1999;85:26–31.
- 8.↑
Claren A, Doyen J, Falk AT, et al. Results of age-dependent anal canal cancer treatment: a single centre retrospective study. Dig Liver Dis 2014;46:460–464.
- 9.↑
Miller ED, Fisher JL, Haglund KE, et al. The addition of chemotherapy to radiation therapy improves survival in elderly patients with stage III non-small cell lung cancer. J Thorac Oncol 2018;13:426–435.
- 10.↑
Saarilahti K, Arponen P, Vaalavirta L, et al. Chemoradiotherapy of anal cancer is feasible in elderly patients: treatment results of mitomycin-5-FU combined with radiotherapy at Helsinki University Central Hospital 1992–2003. Acta Oncol 2006;45:736–742.
- 11.↑
Atagi S, Kawahara M, Tamura T, et al. Standard thoracic radiotherapy with or without concurrent daily low-dose carboplatin in elderly patients with locally advanced non-small cell lung cancer: a phase III trial of the Japan Clinical Oncology Group (JCOG9812). Jpn J Clin Oncol 2005;35:195–201.
- 12.↑
Atagi S, Kawahara M, Yokoyama A, et al. Thoracic radiotherapy with or without daily low-dose carboplatin in elderly patients with non-small-cell lung cancer: a randomised, controlled, phase 3 trial by the Japan Clinical Oncology Group (JCOG0301). Lancet Oncol 2012;13:671–678.
- 13.↑
Bilimoria KY, Stewart AK, Winchester DP, et al. The National Cancer Data Base: a powerful initiative to improve cancer care in the United States. Ann Surg Oncol 2008;15:683–690.
- 14.↑
Dutta SW, Trifiletti DM, Grover S, et al. Management of elderly patients with early-stage medically inoperable endometrial cancer: systematic review and National Cancer Database analysis. Brachytherapy 2017;16:526–533.
- 15.↑
Gibreel WO, Day CN, Hoskin TL, et al. Mastectomy and immediate breast reconstruction for cancer in the elderly: a National Cancer Data Base study. J Am Coll Surg 2017;224:895–905.
- 16.↑
Huang Y, Chen W, Haque W, et al. The impact of comorbidity on overall survival in elderly nasopharyngeal carcinoma patients: a National Cancer Data Base analysis. Cancer Med 2018;7:1093–1101.
- 17.↑
Coca-Perraillon M. Local and global optimal propensity score matching. Presented at the SAS Global Forum 2007 Conference; April 16–19, 2007; Orlando, Florida.
- 18.↑
Rosenbaum PR, Rubin DB. Constructing a control-group using multivariate matched sampling methods that incorporate the propensity score. Am Stat 1985;39:33–38.
- 19.↑
Lanehart RE, Rodriguez de Gil P, Kim ES, et al. Propensity score analysis and assessment of propensity score approaches using SAS procedures. Presented at the SAS Global Forum 2012 Conference; April 22–25, 2012; Orlando, Florida.
- 20.↑
Raphael MJ, Ko G, Booth CM, et al. Factors associated with chemoradiation therapy interruption and noncompletion among patients with squamous cell anal carcinoma. JAMA Oncol 2020;6:881–887.
- 21.↑
Fallai C, Cerrotta A, Valvo F, et al. Anal carcinoma of the elderly treated with radiotherapy alone or with concomitant radio-chemotherapy. Crit Rev Oncol Hematol 2007;61:261–268.
- 22.↑
Charnley N, Choudhury A, Chesser P, et al. Effective treatment of anal cancer in the elderly with low-dose chemoradiotherapy. Br J Cancer 2005;92:1221–1225.
- 23.↑
Kachnic LA, Winter K, Myerson RJ, et al. RTOG 0529: a phase 2 evaluation of dose-painted intensity modulated radiation therapy in combination with 5-fluorouracil and mitomycin-C for the reduction of acute morbidity in carcinoma of the anal canal. Int J Radiat Oncol Biol Phys 2013;86:27–33.
- 24.↑
Bazan JG, Hara W, Hsu A, et al. Intensity-modulated radiation therapy versus conventional radiation therapy for squamous cell carcinoma of the anal canal. Cancer 2011;117:3342–3351.
- 25.↑
Bryant AK, Huynh-Le MP, Simpson DR, et al. Intensity modulated radiation therapy versus conventional radiation for anal cancer in the Veterans Affairs system. Int J Radiat Oncol Biol Phys 2018;102:109–115.
- 26.↑
Dasgupta T, Rothenstein D, Chou JF, et al. Intensity-modulated radiotherapy vs. conventional radiotherapy in the treatment of anal squamous cell carcinoma: a propensity score analysis. Radiother Oncol 2013;107:189–194.
- 27.↑
Deshmukh AA, Zhao H, Das P, et al. Clinical and economic evaluation of treatment strategies for T1N0 anal canal cancer. Am J Clin Oncol 2018;41:626–631.
- 28.↑
Extermann M, Hurria A. Comprehensive geriatric assessment for older patients with cancer. J Clin Oncol 2007;25:1824–1831.
