Key Points
Question
Is the receipt of neoadjuvant chemoradiation therapy (NACRT) followed by surgery for locally advanced rectal cancer (LARC) associated with improvements in overall survival among older patients (80 years and older)?
Findings
In this cohort study including 3868 patients with LARC who underwent surgical resection, 2273 (58.8%) received NACRT followed by surgical resection. NACRT was independently associated with a 25% decreased risk of death compared with alternative treatment sequences.
Meaning
This study highlights the importance of therapy sequencing on outcomes in the treatment of LARC in older adults, concluding that NACRT was associated with improved R0 resection rates and a significant survival benefit when compared with treatment with surgery with or without adjuvant therapy.
This cohort study assesses the trends in management of older patients diagnosed with locally advanced rectal cancer who had a surgical resection.
Abstract
Importance
The number of older patients (80 years and older) diagnosed with locally advanced rectal cancer (LARC) is expected to increase. Although current guidelines recommend neoadjuvant chemoradiation therapy (NACRT) followed by resection, little is known about management and outcomes in this older population.
Objective
To assess the trends in management of older patients diagnosed with LARC who had a surgical resection.
Design, Setting, and Participants
Patients 80 years and older who had a surgical resection for LARC were identified in the 2004-2016 National Cancer Database. Patients were grouped based on therapy sequence: (1) surgery followed by adjuvant therapy (AT), ie, chemotherapy or radiation; (2) surgery alone; or (3) NACRT followed by surgical resection. Data were analyzed in May 2021.
Exposures
NACRT followed by surgery, and surgery with or without AT.
Main Outcomes and Measures
Overall survival (OS) was assessed using Kaplan-Meier analyses with inverse probability of treatment weighting (IPTW) and Cox proportional hazards regression were performed to examine the association of NACRT with the risk of death.
Results
Of 3868 patients with LARC who underwent surgical resection, 2042 (52.8%) were male, and the mean (SD) age was 83.4 (3.0) years. A total of 2273 (58.8%) received NACRT followed by surgical resection. Factors independently associated with NACRT were more recent diagnosis, age 80 to 85 years (vs 86 years and older), fewer comorbidities, larger tumors, and node-positive disease. The Kaplan-Meier analyses with IPTW showed that 3-year and 5-year OS for NACRT (3-year: 68.9%; 95% CI, 67.0-70.8; 5-year: 51.1%; 95% CI, 49.0-53.4) vs surgery with AT (3-year: 64.4%; 95% CI, 59.0-70.2; 5-year: 43.0%; 95% CI, 37.4-49.5) vs surgery alone (3-year: 55.8%; 95% CI, 52.0-60.0; 5-year: 34.7%; 95% CI, 30.8-39.0) was significantly different (P < .001). After adjusting for confounders, patients who received NACRT were more likely to undergo an R0 resection (adjusted odds ratio, 2.16; 95% CI, 1.62-2.88), which independently improved OS (P < .001). Moreover, receipt of NACRT was independently associated with a 25% decreased risk of death (adjusted hazard ratio, 0.75; 95% CI, 0.69-0.82) compared with alternative treatment sequences.
Conclusions and Relevance
Approximately 40% of older patients with LARC did not receive the current standard of care. In this cohort, NACRT was associated with a higher likelihood of an R0 resection and improved OS. Clinicians should advocate for receipt of NACRT in older patients with LARC.
Introduction
Colorectal cancer is the third most common cancer diagnosed in both men and women in the US. Because of the aging population, the number of patients 80 years and older diagnosed with locally advanced rectal cancer (LARC) is expected to increase.1,2 In the US, it is estimated there will be 44 580 new cases of rectal cancer diagnosed this year, with approximately half of these cases occurring in patients older than 65 years.3 Furthermore, it is projected that there will be more than 53 200 deaths attributable to colorectal cancer this upcoming year, with more than 36 180 occurring in older patients.4
The management of rectal cancer is complex and requires a multidisciplinary team approach. For most patients with LARC, treatment will consist of a combination of chemoradiation, systemic chemotherapy, and a transabdominal resection, including a total mesorectal excision (TME). The optimal order in which these treatments are sequenced remains a matter of ongoing debate.5 Designing a balanced approach of local, systemic, and surgical therapy for the treatment of older patients with multiple comorbidities becomes even more complex.
Despite ongoing debate regarding the optimal sequencing of systemic therapy, the benefit of neoadjuvant chemoradiation therapy (NACRT) and a standardized, high-quality surgery has been established and widely accepted as the treatment for LARC for approximately 20 years. The initial benefit was reported from the Swedish Rectal Cancer Trial in 1997, where patients who received preoperative radiation therapy experienced not only an improvement in local recurrence rates but also a survival benefit.6 This finding was further validated in a meta-analysis published shortly thereafter.7 In 2001, the Dutch trial found that the addition of a TME to preoperative radiation further improved local recurrence rates.8 And in 2004, the German Rectal Cancer Trial introduced the concept of NACRT and found that an NACRT approach followed by TME decreased local recurrence rates by more than half, from 13% to 6%.9
At the time of the present study, the National Comprehensive Cancer Network (NCCN) guidelines for an LARC would recommend a course of NACRT followed by TME, with subsequent adjuvant systemic chemotherapy once fully recovered from surgery.10 Despite the guidelines, a National Cancer Database (NCDB) study from 201511 looked at the management of LARC in patients older than 80 years and found that more than half of these patients received a surgery-only approach and only 34% of patients received NACRT prior to surgical resection. Only 4% of patients in the study received the complete standard of care, consisting of NACRT followed by TME and subsequent adjuvant systemic therapy. This finding is in line with other studies that have shown that older patients diagnosed with colorectal cancer are less likely to be referred to medical oncology and to receive standard of care chemotherapy and more likely to undergo dose reductions and early termination of therapy.12,13,14,15
Unfortunately, many of the trials that have evaluated the role of NACRT in LARC have either excluded patients older than 80 years or those patients are vastly underrepresented in the study population.16,17 Yet the treatment strategies adopted in those studies are then extrapolated to the older adult population. The objective of this study is therefore to evaluate the trends in management of patients 80 years and older with LARC who underwent a surgical resection. Secondary outcomes are to identify factors associated with the receipt of NACRT, receipt of adjuvant therapy (AT), and achieving a negative surgical margin (R0) as well as the association with overall survival (OS).
Methods
The NCDB is the largest cancer registry in the US and is maintained by the American College of Surgeons and the American Cancer Society. It captures 70% of cancers in the US. Reporting facilities are required to have at least 90% patient follow-up over 5 years.18 All patient-level and hospital-level data in the NCDB are deidentified; therefore, analysis of this data set was exempt from institutional review board approval.
Patients included for analysis were those 80 years or older who had histologically confirmed rectal adenocarcinoma as defined by International Classification of Diseases for Oncology (ICD-O-3) histology codes 8140-8147, 8160-8162, 8180-8221, 8250-8506, 8520-8550, 8560, 8570-8573, and 8940-8941.19 These lesions were confined to the rectum (ICD-O-3 topography code 20.9) and patients had either a partial proctectomy, total proctectomy, pelvic exenteration, or proctectomy, not otherwise specified (surgical codes 30, 40, 50, 60, 70 80, and 90) between 2004 and 2016. Only patients with locally advanced rectal adenocarcinoma were included, which was defined as a clinical tumor stage of 3 to 4 (cT3 to cT4) and clinically node negative (cN0) without any evidence of distant metastases (cM0) or as cT1 to cT4, cN1 to cN2, and cM0.
Patients were grouped by treatment type into 1 of 3 groups: (1) those who received surgery followed by AT (chemotherapy or radiation), (2) those who received surgery alone, and (3) those who received NACRT followed by surgical resection (eFigure 1 in the Supplement).
Baseline characteristics were compared using χ2 tests and analysis of variance analyses for categorical and continuous variables, respectively. Trends over time in utilization of NACRT vs surgery with and without AT were plotted and compared using χ2 tests for trend. Multivariable logistic regression analyses were performed to examine factors that contributed to patients receiving NACRT as well as the factors associated with R0 margin.
OS was calculated from diagnosis to the time of death or last known follow-up and compared using the Kaplan-Meier method with log-rank test. To examine the association between the treatment groups and OS, univariate Cox proportional hazards regression analysis was performed, followed by multivariable risk-adjusted Cox proportional hazards regression and Kaplan-Meier analyses with inverse probability of treatment weighting using propensity score method.20,21 The stabilized inverse probability weights were derived from the generalized boosting method average treatment effect among the treated predicted probabilities of NACRT on diagnosis year, age, facility type, facility location, residence area (rural or urban), Charlson Comorbidity Index score, T stage, N stage, margin, carcinoembryonic antigen, perineural invasion (PNI), and circumferential resection margin (CRM) using R package twang. The assessments for the balance of the covariates in the unweighted and weighted samples were between surgery with AT vs NACRT and between surgery alone vs NACRT (eFigure 2 in the Supplement). To account for immortal time bias, 3-month conditional landmark analyses were performed, where only patients who survived at least 3 months after diagnosis were included as a conditional landmark in Cox proportional hazards regression.22 Our choice of a 3-month landmark related to the rationale that on average, patients who received NACRT took 3 months longer to undergo surgery compared with patients who received surgery with or without AT. All tests were 2-sided, and statistical significance was set at P < .05. All statistical analyses were performed using R version 3.6.3 (The R Foundation).
Results
Patient Characteristics
Between 2004 and 2016, 3868 patients 80 years and older underwent surgery for locally advanced adenocarcinoma of the rectum. Of 3868 patients with LARC who underwent surgical resection, 2042 (52.8%) were male, and the mean (SD) age was 83.4 (3.0) years. A total of 404 patients (10.4%) were treated by surgery followed by AT, 1191 (30.8%) received surgery alone, and 2273 (58.8%) received NACRT followed by surgical resection. The use of NACRT in older patients remained relatively constant until 2011, after which time there was increased utilization of NACRT; only 1247 patients (55.8%) were treated with NACRT from 2004 to 2011, whereas 1001 (62.8%) received NACRT from 2012 to 2016 (Figure 1).
Figure 1. Number and Percentage in the Utilization of Surgery Alone, Surgery With Adjuvant Therapy (AT), and Neoadjuvant Chemoradiation Therapy (NACRT) Among Patients With Locally Advanced Rectal Cancer Over Time.
Baseline characteristics for the 3 treatment groups are summarized in Table 1. Most patients underwent a partial proctectomy or a total proctectomy/abdominoperineal resection. Most patients were non-Hispanic White and male. Additionally, most patients had no comorbidities and received their care at a community facility in a metropolitan area, located within 50 miles of their residence. There was a higher percentage of patients older than 85 years in the surgery alone group. The treatment groups were similar in respect to cT stage; however, there was a greater percentage of patients with cN-positive disease in the NACRT group (983 [43.2%]) vs surgery with or without AT group (553 [34.7%]). Pathologic outcomes also differed between groups. The R0 resection rate was higher in the NACRT group (2101 [92.4%]) compared with the surgery with AT group (322 [79.7%]) and the surgery alone group (1037 [87.1%]). Furthermore, there were fewer patients with an involved CRM and PNI in the surgical specimen in the NACRT group (82 [3.6%] and 150 [6.6%], respectively) compared with the surgery with AT group (45 [11.1%] and 60 [14.9%]).
Table 1. Baseline Characteristics.
| Characteristic | No. (%) | P value | ||
|---|---|---|---|---|
| NACRT (n = 2273) | Surgery with AT (n = 404) | Surgery alone (n = 1191) | ||
| Procedure | ||||
| Partial procectomy | 1402 (61.7) | 274 (67.8) | 789 (66.2) | .03 |
| Total proctectomy/APR | 710 (31.2) | 106 (26.2) | 313 (26.3) | |
| Total proctocolectomy | 47 (2.1) | 4 (1.0) | 35 (2.9) | |
| Pelvic exenteration | 48 (2.1) | 8 (2.0) | 26 (2.2) | |
| Surgery, NOS | 64 (2.8) | 12 (3.0) | 28 (2.4) | |
| Year of diagnosis | ||||
| 2004-2010 | 1040 (45.8) | 216 (53.5) | 578 (48.5) | .01 |
| 2011-2016 | 1233 (54.2) | 188 (46.5) | 613 (51.5) | |
| Age, mean (SD), y | 83.1 (2.7) | 83.1 (2.8) | 84.9 (3.3) | |
| Age, range, y | ||||
| 80-85 | 1834 (80.7) | 326 (80.7) | 693 (58.2) | <.001 |
| 86-90 | 439 (19.3) | 78 (19.3) | 498 (41.8) | |
| Sex | ||||
| Male | 1202 (52.9) | 222 (55) | 618 (51.9) | .56 |
| Female | 1071 (47.1) | 182 (45) | 573 (48.1) | |
| Race and ethnicity | ||||
| American Indian | 8 (0.4) | 0 | 1 (0.1) | .27 |
| Asian or Pacific Islander | 74 (3.3) | 11 (2.7) | 26 (2.2) | |
| Black | 105 (4.6) | 20 (5.0) | 53 (4.5) | |
| Hispanic | 72 (3.2) | 5 (1.2) | 37 (3.1) | |
| White | 1875 (82.5) | 348 (86.1) | 995 (83.5) | |
| Other race | 139 (6.1) | 20 (5.0) | 79 (6.6) | |
| Insurance status | ||||
| Private | 197 (8.7) | 35 (8.7) | 108 (9.1) | .49 |
| Uninsured | 7 (0.3) | 1 (0.2) | 8 (0.7) | |
| Public or government | 2052 (90.3) | 366 (90.6) | 1061 (89.1) | |
| Other or unknown | 17 (0.7) | 2 (0.5) | 14 (1.2) | |
| Average educationa | ||||
| ≥17.6% (Lowest quartile) | 328 (14.4) | 63 (15.6) | 189 (15.9) | .02 |
| 10.9%-17.5% (Low quartile) | 533 (23.4) | 104 (25.7) | 300 (25.2) | |
| 6.3%-10.8% (Moderate quartile) | 654 (28.8) | 111 (27.5) | 337 (28.3) | |
| <6.3% (High quartile) | 566 (24.9) | 107 (26.5) | 303 (25.4) | |
| Other or unknown | 192 (8.4) | 19 (4.7) | 62 (5.2) | |
| Facility typeb | ||||
| Academic | 679 (29.9) | 99 (24.5) | 372 (31.2) | .04 |
| Community | 1594 (70.1) | 305 (75.5) | 819 (68.8) | |
| Facility location | ||||
| Northeast | 503 (22.1) | 107 (26.5) | 298 (25.0) | .002 |
| South | 575 (25.3) | 114 (28.2) | 335 (28.1) | |
| Midwest | 812 (35.7) | 110 (27.2) | 349 (29.3) | |
| Pacific | 295 (13.0) | 51 (12.6) | 165 (13.9) | |
| Mountain | 88 (3.9) | 22 (5.4) | 44 (3.7) | |
| City type | ||||
| Metropolitan | 1807 (79.5) | 329 (81.4) | 980 (82.3) | .28 |
| Urban | 359 (15.8) | 57 (14.1) | 151 (12.7) | |
| Rural | 46 (2.0) | 10 (2.5) | 29 (2.4) | |
| Unknown | 61 (2.7) | 8 (2.0) | 31 (2.6) | |
| Charlson Comorbidity Index score | ||||
| 0 | 1628 (71.6) | 279 (69.1) | 749 (62.9) | <.001 |
| 1 | 477 (21.0) | 85 (21) | 294 (24.7) | |
| ≥2 | 168 (7.4) | 40 (9.9) | 148 (12.4) | |
| Travel distance, milesc | ||||
| <50 | 1883 (82.8) | 361 (89.4) | 1031 (86.6) | .001 |
| 50-100 | 141 (6.2) | 14 (3.5) | 67 (5.6) | |
| >100 | 77 (3.4) | 10 (2.5) | 40 (3.4) | |
| Unknown | 172 (7.6) | 19 (4.7) | 53 (4.5) | |
| Clinical stage group | ||||
| Stage 2 | 1290 (56.8) | 214 (53.0) | 828 (69.5) | <.001 |
| Stage 3 | 983 (43.2) | 190 (47.0) | 363 (30.5) | |
| cT stage | ||||
| T1 | 10 (0.4) | 11 (2.7) | 13 (1.1) | <.001 |
| T2 | 90 (4.0) | 22 (5.4) | 50 (4.2) | |
| T3 | 2025 (89.1) | 326 (80.7) | 1028 (86.3) | |
| T4 | 148 (6.5) | 45 (11.1) | 100 (8.4) | |
| cN stage | ||||
| N0 | 1290 (56.8) | 214 (53) | 828 (69.5) | <.001 |
| N1 | 867 (38.1) | 129 (31.9) | 270 (22.7) | |
| N2 | 116 (5.1) | 61 (15.1) | 93 (7.8) | |
| Tumor grade | ||||
| Well or moderately differentiated | 1675 (73.7) | 301 (74.5) | 953 (80.0) | <.001 |
| Poor or undifferentiated | 303 (13.3) | 90 (22.3) | 210 (17.6) | |
| Unknown | 295 (13.0) | 13 (3.2) | 28 (2.4) | |
| Surgical margin | ||||
| Negative | 2101 (92.4) | 322 (79.7) | 1037 (87.1) | <.001 |
| Positive | 125 (5.5) | 72 (17.8) | 137 (11.5) | |
| Unknown | 47 (2.1) | 10 (2.5) | 17 (1.4) | |
| Regional nodes examined | ||||
| 0 | 188 (8.3) | 23 (5.7) | 41 (3.4) | <.001 |
| 1-14 | 1241 (54.6) | 150 (37.1) | 493 (41.4) | |
| ≥15 | 803 (35.3) | 229 (56.7) | 653 (54.8) | |
| Unknown | 41 (1.8) | 2 (0.5) | 4 (0.3) | |
| Path T | ||||
| T0/Tis | 183 (8.1) | 1 (0.2) | 1 (0.1) | <.001 |
| T1 | 98 (4.3) | 10 (2.5) | 19 (1.6) | |
| T2 | 509 (22.4) | 32 (7.9) | 160 (13.4) | |
| T3 | 930 (40.9) | 250 (61.9) | 720 (60.5) | |
| T4 | 74 (3.3) | 43 (10.6) | 94 (7.9) | |
| Tx | 479 (21.1) | 68 (16.8) | 197 (16.5) | |
| Path N | ||||
| N0 | 1296 (57.0) | 107 (26.5) | 596 (50.0) | <.001 |
| N1 | 389 (17.1) | 127 (31.4) | 241 (20.2) | |
| N2 | 121 (5.3) | 95 (23.5) | 141 (11.8) | |
| Nx | 467 (20.5) | 75 (18.6) | 213 (17.9) | |
| Surgical approach | ||||
| Open | 652 (28.7) | 111 (27.5) | 415 (34.8) | <.001 |
| Laparoscopic or robotic | 498 (21.9) | 70 (17.3) | 281 (23.6) | |
| MIS converted to open | 64 (2.8) | 9 (2.2) | 45 (3.8) | |
| Unknown | 1059 (46.6) | 214 (53.0) | 450 (37.8) | |
| CEA | ||||
| ≤3 | 24 (1.1) | 4 (1.0) | 11 (0.9) | .003 |
| >3 | 980 (43.1) | 147 (36.4) | 461 (38.7) | |
| Unknown | 1269 (55.8) | 253 (62.6) | 719 (60.4) | |
| CRM | ||||
| Not involved | 1979 (87.1) | 323 (80.0) | 1041 (87.4) | <.001 |
| Involved | 82 (3.6) | 45 (11.1) | 86 (7.2) | |
| Unknown | 212 (9.3) | 36 (8.9) | 64 (5.4) | |
| PNI | ||||
| No | 1114 (49) | 151 (37.4) | 598 (50.2) | <.001 |
| Yes | 150 (6.6) | 60 (14.9) | 127 (10.7) | |
| Unknown | 1009 (44.4) | 193 (47.8) | 466 (39.1) | |
| Treatment | ||||
| Radiation | ||||
| No | 56 (2.5) | 139 (34.4) | 1191 (100) | <.001 |
| Yes | 2200 (96.8) | 257 (63.6) | 0 | |
| Unknown | 17 (0.7) | 8 (2.0) | 0 | |
| Chemotherapy | ||||
| No | 130 (5.7) | 41 (10.1) | 767 (64.4) | <.001 |
| Yes | 2051 (90.2) | 338 (83.7) | 0 | |
| Contraindicated | 46 (2.0) | 11 (2.7) | 172 (14.4) | |
| Patient deceased prior to start | 0 | 0 | 12 (1.0) | |
| Refused | 38 (1.7) | 12 (3.0) | 222 (18.6) | |
| Unknown | 8 (0.4) | 2 (0.5) | 18 (1.5) | |
| Chemotherapy regimen | ||||
| Single agent | 1525 (67.1) | 212 (52.5) | 0 | <.001 |
| Multiagent | 307 (13.5) | 96 (23.8) | 0 | |
| Yes | 219 (9.6) | 30 (7.4) | 0 | |
| No | 130 (5.7) | 41 (10.1) | 767 (64.4) | |
| Refused | 38 (1.7) | 12 (3) | 222 (18.6) | |
| Unknown, contraindicated, or deceased | 54 (2.4) | 13 (3.2) | 202 (17) | |
| Outcomes | ||||
| LOS, d | ||||
| ≤7 | 1144 (50.3) | 211 (52.2) | 565 (47.4) | <.001 |
| 8-14 | 625 (27.5) | 105 (26.0) | 358 (30.1) | |
| ≥15 | 215 (9.5) | 36 (8.9) | 183 (15.4) | |
| Unknown | 289 (12.7) | 52 (12.9) | 85 (7.1) | |
| 30-d Readmission | ||||
| No | 2043 (89.9) | 359 (88.9) | 1057 (88.7) | .02 |
| Yes | 171 (7.5) | 33 (8.2) | 118 (9.9) | |
| Unknown | 59 (2.6) | 12 (3.0) | 16 (1.3) | |
| 30-d Mortality | ||||
| No | 2173 (95.6) | 403 (99.8) | 1097 (92.1) | <.001 |
| Yes | 89 (3.9) | 0 | 89 (7.5) | |
| Missing | 11 (0.5) | 1 (0.2) | 5 (0.4) | |
| 90-d Mortality | ||||
| No | 2097 (92.3) | 390 (96.5) | 1019 (85.6) | <.001 |
| Yes | 156 (6.9) | 13 (3.2) | 158 (13.3) | |
| Missing | 20 (0.9) | 1 (0.2) | 14 (1.2) | |
Abbreviations: APR, abdominoperineal resection; AT, adjuvant therapy; CEA, carcinoembryonic antigen; cN, clinical node stage; CRM, circumferential resection margin; cT, clinical tumor stage; LOS, length of stay; NACRT, neoadjuvant chemoradiation therapy; NOS, not other specified; PNI, perineural invasion.
This measure of educational attainment for each patient’s area of residence is estimated by matching the zip code of the patient recorded at the time of diagnosis against files derived from US Census data. This item provides a measure of the number of adults in the patient’s zip code who did not graduate from high school and is categorized as equally proportioned quartiles among all US zip codes (as taken from the National Cancer Database Participant User Data File dictionary).
Category classification by the Commission on Cancer Accreditation program.
Distance in miles between the patient’s residence and the hospital that reported the case (as taken from the National Cancer Database Participant User Data File dictionary).
Predictors of Receipt of NACRT, Receipt of AT, and R0 Resection
The utilization of NACRT gradually increased over the study period. The greatest rise was seen from 2011 to 2012; 207 patients (49.8%) received NACRT in 2011, with an increase to 229 (58.3%) in 2012. After adjusting for confounding factors, a lower Charlson Comorbidity Index score, later year of diagnosis (2011 to 2016), younger age (80 to 85 years), cT2 to cT4 tumors, and cN1 tumors were independently associated with the receipt of NACRT (Table 2). Patients were less likely to receive NACRT if they had clinical N2 disease (adjusted odds ratio [aOR], 0.65; 95% CI, 0.49-0.85) or a poorly differentiated tumor (aOR, 0.70; 95% CI, 0.50-0.97). Patients who received upfront surgery were less likely to receive AT if they were older (aOR, 0.43; 95% CI, 0.37-0.50), whereas they were more likely to receive AT if they had tumor-positive lymph nodes (aOR, 3.47; 95% CI, 2.57-4.68) or a positive surgical margin (R1 or R2) (aOR 1.57; 95% CI, 1.10-2.25). Patients who had received NACRT had a significantly greater likelihood of having an R0 resection (aOR, 2.16; 95% CI, 1.62-2.88). Factors independently associated with a positive surgical margin were clinical T4 tumors (aOR, 0.19; 95% CI, 0.05-0.81), poorly differentiated tumor grade (aOR, 0.70; 95% CI, 0.50-0.97), an involved CRM (aOR, 0.02; 95% CI, 0.02-0.04), and presence of PNI (aOR, 0.42; 95% CI, 0.28-0.65). On unadjusted analysis, the median OS for patients who achieved an R0 resection was 4.39 years (95% CI, 4.19-4.58), whereas those who did not had a median OS of 2.04 years (95% CI, 1.70-2.40) (eFigure 3 in the Supplement).
Table 2. Multivariable Logistic Regression Analysis for R0 Resection, Receipt of Neoadjuvant Chemoradiation Therapy (NACRT), and Receipt of Adjuvant Therapy (AT).
| Measure | OR (95% CI) | P value |
|---|---|---|
| R0 resection | ||
| Treatment | ||
| Surgery with or without AT | 1 [Reference] | NA |
| NACRT | 2.16 (1.62-2.88) | <.001 |
| City type | ||
| Metropolitan | 1 [Reference] | NA |
| Urban | 1.79 (1.16-2.76) | .009 |
| Rural | 0.78 (0.35-1.75) | .55 |
| Clinical T stage | ||
| T1 | 1 [Reference] | NA |
| T2 | 1.25 (0.25-6.23) | .78 |
| T3 | 0.67 (0.17-2.73) | .58 |
| T4 | 0.19 (0.05-0.81) | .03 |
| Tumor grade | ||
| Well or moderately differentiated | 1 [Reference] | NA |
| Poor or undifferentiated | 0.70 (0.50-0.97) | .03 |
| Receipt of NACRT | ||
| Year of diagnosis | ||
| 2004-2010 | 1 [Reference] | NA |
| 2011-2016 | 1.16 (1.01-1.33) | .03 |
| Age, y | ||
| 80-85 | 1 [Reference] | NA |
| >85 | 0.43 (0.37-0.50) | <.001 |
| Charlson Comorbidity Index score | ||
| 0 | 1 [Reference] | NA |
| ≥2 | 0.52 (0.41-0.66) | <.001 |
| Clinical T stage | ||
| T1 | 1 [Reference] | NA |
| T2 | 3.77 (1.60-8.89) | .003 |
| T3 | 6.14 (2.75-13.71) | <.001 |
| T4 | 4.56 (1.97-10.53) | <.001 |
| Clinical N stage | ||
| N0 | 1 [Reference] | NA |
| N1 | 1.93 (1.64-2.26) | <.001 |
| N2 | 0.65 (0.49-0.85) | .002 |
| Receipt of AT | ||
| Age, y | ||
| 80-85 | 1 [Reference] | NA |
| >85 | 0.30 (0.23-0.41) | <.001 |
| Margin | ||
| R0 | 1 [Reference] | NA |
| R1/2 | 1.57 (1.10-2.25) | .01 |
| Positive lymph nodes | ||
| 0 | 1 [Reference] | NA |
| ≥1 | 3.47 (2.57-4.68) | <.001 |
Abbreviation: OR, odds ratio.
Survival Analysis
When comparing patients who received both surgery and additional therapies, there was a significant survival advantage to NACRT over AT, with a median survival of 5.11 years (95% CI, 4.90-5.41) for those receiving NACRT compared with 4.19 years (95% CI, 3.78-4.91) for those who received surgery with AT (Figure 2). In multivariable risk-adjusted Cox proportional hazards regression analyses (Figure 3), receipt of NACRT was independently associated with a 25% decreased risk of death (adjusted hazard ratio, 0.75; 95% CI, 0.69-0.82). Age older than 85 years, a higher Charlson Comorbidity Index score, clinical N2 lesion, poorly differentiated tumor grade, and involved CRM and PNI were associated with an increased risk of death, whereas an R0 margin and female sex were associated with a decreased risk of death (Figure 3).
Figure 2. Inverse Probability of Treatment–Weighted Kaplan-Meier Curves for the Comparisons of Overall Survival Between Neoadjuvant Chemoradiation Therapy (NACRT), Surgery With Adjuvant Therapy (AT), and Surgery Alone.
Figure 3. Multivariable Cox Proportional Hazards Regression Analyses.
APR indicates abdominoperineal resection; AT, adjuvant therapy; CEA, carcinoembryonic antigen; cN, clinical node stage; CRM, circumferential resection margin; cT, clinical tumor stage; HR, hazard ratio; NACRT, neoadjuvant chemoradiation therapy; PNI, perineural invasion.
Discussion
Our study is one of the largest retrospective studies to focus exclusively on older patients undergoing surgical resection for LARC and compare their OS based on the sequencing of therapy. Of 3868 patients 80 years and older who received surgical resection for LARC, 58.7% received NACRT prior to surgical resection, with the remainder of patients receiving surgery with AT or surgery alone. NACRT conferred an independent survival advantage compared with an upfront surgery approach with or without the receipt of AT. This independent survival advantage remained even when patients receiving NACRT were compared only with those patients who successfully received AT after surgery.
The importance of our findings are 2-fold. First, we were able to demonstrate that approximately one-third of patients 80 years and older diagnosed with an LARC received a nonpalliative surgery–alone approach, where this number is close to 9% among a younger cohort.23 Although this number has improved over time, it also highlights that there is still marked room for improvement. Second, and more importantly, we present the novel finding that the sequence of therapy in this patient population appears to have implications for OS. This is in contrast to a prior retrospective study that looked at a very similar patient population11; however, with the addition of our updated cohort of patients (2012 to 2016), we have found that the receipt of NACRT is associated with improved OS.
According to the NCCN guidelines, patients diagnosed with an LARC should receive NACRT prior to TME.10 The benefits of this treatment strategy has resulted in improved local control and even improved OS.6,7,8,9 However, we found that among an older patient population, 41% of our cohort received a nonpalliative surgery-first approach, with more than 30% of patients receiving a surgery-alone approach. There appears to be greater deviation from the standard of care by physicians when caring for older patients; this is not an isolated finding to LARC. It is our feeling that while the underutilization of NACRT is likely multifactorial, the biggest driver in this patient population is because of the age of the patient and the reluctance of physicians to treat patients with NACRT. And while that trend appears to be decreasing as we look at the utilization of NACRT over time, there are still many patients older than 80 years who do not receive the standard of care.11 Future research should continue to track utilization of neoadjuvant therapy for LARC, and simultaneously, education efforts for clinicians should focus on the importance of following NCCN guidelines for all patients, regardless of age, until geriatric-specific trials are conducted.
Previous literature, such as a study by Bergquist et al,11 reported receipt of any therapy in addition to surgery, regardless of sequence, had an associated survival benefit compared with patients who had surgery alone. However, based on the results from our study, sequence of therapy has implications for patient outcomes. If a patient can tolerate a therapy in addition to their surgery, NACRT should be used, as opposed to AT. Additionally, recent publications have shown that frailty, as opposed to chronological age, is a better measure of a patient’s ability to tolerate therapy and should likely be used as a decision guide for therapy.24,25 Therefore, based on a prior NCDB study that did not find a significant difference in survival between definitive chemoradiation vs surgery alone for patients with LARC,23 for patients whom clinicians believe can only tolerate 1 modality of treatment, NACRT may be the best initial treatment as opposed to an upfront surgical resection.
Perhaps not surprising, we report that the receipt of NACRT was independently associated with an R0 resection. Among our entire cohort, the overall rate of a positive surgical margin was 8.6%. However, patients who had a surgery-first approach were almost 2.5-fold more likely to have a positive surgical margin than patients who received NACRT. Moreover, patients who had an R0 resection experienced a 29% decreased risk of death compared with an R1 or R2 resection. Additionally, we found that the median OS for patients who achieved an R0 resection was 4.39 years (95% CI, 4.19-4.58) compared with 2.04 years (95% CI, 1.70-2.40) for those with a positive surgical margin. This appears concordant with the report from the German Rectal Cancer Trial,9 which demonstrated NACRT followed by surgery resulted in a reduction in local recurrence compared with surgical resection followed by adjuvant chemotherapy with radiation. However, while they did not demonstrate an OS benefit associated with NACRT, our current study demonstrates a survival benefit for those patients who received NACRT.
Limitations
There are several limitations to our study, including the inherent limitations of the NCDB and any retrospective database. The NCDB does not have cancer-specific survival as an end point, making it impossible to perform competing-risk regression analysis for cause-specific death. Additionally, although the NCDB does specify if a patient is selected to receive NACRT, we are not able to identify patients who were selected to receive NACRT but, because of disease progression or a complication related to their treatment, were unable to proceed with surgical resection. In addition, since our study period ended in 2016, the regimens used are not completely reflective of the treatment strategies currently being applied today. Moreover, although we are able to determine if patients had access to a single agent or multiagent chemotherapy regimens, we lack any additional information about the specific systemic therapy that they received, duration of treatment, or any medication adjustments. Additionally, despite the risk-adjusted Cox proportional hazards regression and Cox regression with propensity score weighting to account for selection bias showing similar results and conclusion, there could have been other unmeasured confounders, and any database has the potential to underrepresent a patient’s comorbid status.
Conclusions
This study highlights the importance of therapy sequencing on outcomes in the treatment of LARC in older adults. NACRT was associated with improved R0 resection rates and a significant survival benefit compared with treatment with surgery with or without the addition of AT. Based on trial data, NACRT is generally well tolerated by patients with little added morbidity. If a patient is deemed a suitable candidate for surgical resection of LARC, it is our feeling that same patient will most likely tolerate NACRT with minimal complication. Although the percentage of older patients offered NACRT proportionally increased throughout our study period with a concomitant decrease in the percentage of patients offered a surgery-alone approach, there is still substantial room for improvement. Clinicians should offer the standard of care to patients and not deviate from guidelines solely based on a patient’s chronological age. Future clinical trials should aim to include older patients diagnosed with LARC to further validate these findings.
eFigure 1. Patient Selection Flowchart
eFigure 2. Absolute Standardized Differences for Unweighted and Weighted Covariates Between Surgery+AT vs NACRT and Surgery Alone vs NACRT
eFigure 3. KM Comparison of OS Between R0 and R1/2 Surgical Margin
References
- 1.Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021. CA Cancer J Clin. 2021;71(1):7-33. doi: 10.3322/caac.21654 [DOI] [PubMed] [Google Scholar]
- 2.Surveillance, Epidemiology, and End Results Program . Cancer Stat Facts: Colorectal Cancer. Accessed August 26, 2022. https://seer.cancer.gov/statfacts/html/colorect.html
- 3.American Cancer Society . Cancer Statistics Center. Accessed February 21, 2022. http://cancerstatisticscenter.cancer.org
- 4.Siegel RL, Miller KD, Goding Sauer A, et al. Colorectal cancer statistics, 2020. CA Cancer J Clin. 2020;70(3):145-164. doi: 10.3322/caac.21601 [DOI] [PubMed] [Google Scholar]
- 5.Bahadoer RR, Dijkstra EA, van Etten B, et al. ; RAPIDO Collaborative Investigators . Short-course radiotherapy followed by chemotherapy before total mesorectal excision (TME) versus preoperative chemoradiotherapy, TME, and optional adjuvant chemotherapy in locally advanced rectal cancer (RAPIDO): a randomised, open-label, phase 3 trial. Lancet Oncol. 2021;22(1):29-42. doi: 10.1016/S1470-2045(20)30555-6 [DOI] [PubMed] [Google Scholar]
- 6.Cedermark B, Dahlberg M, Glimelius B, Påhlman L, Rutqvist LE, Wilking N; Swedish Rectal Cancer Trial . Improved survival with preoperative radiotherapy in resectable rectal cancer. N Engl J Med. 1997;336(14):980-987. doi: 10.1056/NEJM199704033361402 [DOI] [PubMed] [Google Scholar]
- 7.Cammà C, Giunta M, Fiorica F, Pagliaro L, Craxì A, Cottone M. Preoperative radiotherapy for resectable rectal cancer: a meta-analysis. JAMA. 2000;284(8):1008-1015. doi: 10.1001/jama.284.8.1008 [DOI] [PubMed] [Google Scholar]
- 8.Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. ; Dutch Colorectal Cancer Group . Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med. 2001;345(9):638-646. doi: 10.1056/NEJMoa010580 [DOI] [PubMed] [Google Scholar]
- 9.Sauer R, Becker H, Hohenberger W, et al. ; German Rectal Cancer Study Group . Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med. 2004;351(17):1731-1740. doi: 10.1056/NEJMoa040694 [DOI] [PubMed] [Google Scholar]
- 10.National Comprehensive Cancer Network . Rectal cancer. Accessed February 21, 2022. https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdf
- 11.Bergquist JR, Thiels CA, Shubert CR, et al. Is chemotherapy or radiation therapy in addition to surgery beneficial for locally advanced rectal cancer in the elderly? a National Cancer Data Base (NCDB) study. World J Surg. 2016;40(2):447-455. doi: 10.1007/s00268-015-3319-7 [DOI] [PubMed] [Google Scholar]
- 12.Abraham A, Habermann EB, Rothenberger DA, et al. Adjuvant chemotherapy for stage III colon cancer in the oldest old: results beyond clinical guidelines. Cancer. 2013;119(2):395-403. doi: 10.1002/cncr.27755 [DOI] [PubMed] [Google Scholar]
- 13.Luo R, Giordano SH, Freeman JL, Zhang D, Goodwin JS. Referral to medical oncology: a crucial step in the treatment of older patients with stage III colon cancer. Oncologist. 2006;11(9):1025-1033. doi: 10.1634/theoncologist.11-9-1025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Abrams TA, Brightly R, Mao J, et al. Patterns of adjuvant chemotherapy use in a population-based cohort of patients with resected stage II or III colon cancer. J Clin Oncol. 2011;29(24):3255-3262. doi: 10.1200/JCO.2011.35.0058 [DOI] [PubMed] [Google Scholar]
- 15.Schrag D, Gelfand SE, Bach PB, Guillem J, Minsky BD, Begg CB. Who gets adjuvant treatment for stage II and III rectal cancer? insight from surveillance, epidemiology, and end results—Medicare. J Clin Oncol. 2001;19(17):3712-3718. doi: 10.1200/JCO.2001.19.17.3712 [DOI] [PubMed] [Google Scholar]
- 16.Hurria A, Dale W, Mooney M, et al. ; Cancer and Aging Research Group . Designing therapeutic clinical trials for older and frail adults with cancer: U13 conference recommendations. J Clin Oncol. 2014;32(24):2587-2594. doi: 10.1200/JCO.2013.55.0418 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hathout L, Maloney-Patel N, Malhotra U, et al. Management of locally advanced rectal cancer in the elderly: a critical review and algorithm. J Gastrointest Oncol. 2018;9(2):363-376. doi: 10.21037/jgo.2017.10.10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Boffa DJ, Rosen JE, Mallin K, et al. Using the national cancer database for outcomes research: a review. JAMA Oncol. 2017;3(12):1722-1728. doi: 10.1001/jamaoncol.2016.6905 [DOI] [PubMed] [Google Scholar]
- 19.World Health Organization . International Statistical Classification of Diseases, Tenth Revision (ICD-10). 2nd ed. World Health Organization; 2004. [Google Scholar]
- 20.Austin PC. Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis. Stat Med. 2016;35(30):5642-5655. doi: 10.1002/sim.7084 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34(28):3661-3679. doi: 10.1002/sim.6607 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Giobbie-Hurder A, Gelber RD, Regan MM. Challenges of guarantee-time bias. J Clin Oncol. 2013;31(23):2963-2969. doi: 10.1200/JCO.2013.49.5283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sineshaw HM, Jemal A, Thomas CR Jr, Mitin T. Changes in treatment patterns for patients with locally advanced rectal cancer in the United States over the past decade: an analysis from the National Cancer Data Base. Cancer. 2016;122(13):1996-2003. doi: 10.1002/cncr.29993 [DOI] [PubMed] [Google Scholar]
- 24.Hägg S, Jylhävä J. Should we invest in biological age predictors to treat colorectal cancer in older adults? Eur J Surg Oncol. 2020;46(3):316-320. doi: 10.1016/j.ejso.2019.11.003 [DOI] [PubMed] [Google Scholar]
- 25.Ethun CG, Bilen MA, Jani AB, Maithel SK, Ogan K, Master VA. Frailty and cancer: implications for oncology surgery, medical oncology, and radiation oncology. CA Cancer J Clin. 2017;67(5):362-377. doi: 10.3322/caac.21406 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure 1. Patient Selection Flowchart
eFigure 2. Absolute Standardized Differences for Unweighted and Weighted Covariates Between Surgery+AT vs NACRT and Surgery Alone vs NACRT
eFigure 3. KM Comparison of OS Between R0 and R1/2 Surgical Margin