Abstract
Background:
Older women (>65 years) diagnosed with breast cancer may be at risk for chemotherapy dose reductions. We evaluated associations of age at diagnosis with two measures of chemotherapy dose reductions: first cycle dose proportion (FCDP) <90%, and average relative dose intensity (ARDI) <90%.
Methods:
From the Optimal Breast cancer Chemotherapy Dosing study, we included 10,166 women aged 18+ years treated with adjuvant chemotherapy for stage I-IIIA breast cancer at Kaiser Permanente Northern California (KPNC) and Washington (KPWA) between 2004-2019. We examined associations between age at diagnosis with FCDP<90% (reflecting clinician intent at chemotherapy initiation) and ARDI<90% (reflecting average dose across the chemotherapy course). We used generalized linear models of the Poisson family with a log-link function and robust standard errors to calculate prevalence ratios (PR) for FCDP<90% and ARDI<90% with 95% confidence intervals (CI) adjusted for patient and tumor characteristics, with and without adjusting for pre-existing comorbidities. All tests for statistical significance were 2-sided.
Results:
The proportion of women with FCDP<90% ranged from 2.9% among women aged 18-39 years to 18.6% among women aged 75+ years. Before adjusting for comorbidities, women aged 75+ years were more likely to have FCDP<90% (PR: 4.88; 95%CI: 3.58, 6.66) and ARDI<90% (PR: 1.91; 95%CI: 1.58, 2.32) versus women aged 40-49 years. Results were similar after adjusting for comorbidities as a composite comorbidity score or individual comorbidities.
Conclusion:
Older age at diagnosis was strongly associated with chemotherapy dose reductions in this population-based cohort, particularly at chemotherapy initiation but also across the course of treatment.
Introduction
Treatment advances in neoadjuvant and adjuvant chemotherapy have led to improved breast cancer survival rates, particularly for women with early-stage invasive disease.[1] Breast cancer chemotherapy regimen recommendations based on clinical trial evidence are outlined in the National Comprehensive Cancer Network (NCCN) guidelines, which also describe recommended doses for treatment generally based on Body Surface Area (BSA).[1-3] However, older women over the age of 65 years are often excluded from clinical chemotherapy trials because they do not meet stringent age or comorbidity enrollment criteria.[4] This has resulted in a critical evidence gap in determining appropriate chemotherapy regimens and dosing for older adults with cancer.
Chemotherapy has been shown to effectively prolong disease-free survival in older women with breast cancer in several [5-7] but not all [8] studies. Observational studies have shown older women are less likely to receive chemotherapy than younger women.[5, 6, 9, 10] Older women who receive chemotherapy are more likely to experience dose reductions, delays, and changes in regimens compared to younger women.[11-21] However, prior studies did not evaluate dosing across the full range of chemotherapy regimens available today nor did they distinguish between first cycle dose reductions and dose reductions across the chemotherapy course. This is important because lower chemotherapy dose may be associated with worse breast cancer survival, as demonstrated in one small study.[41]
In the population-based Optimal Breast cancer Chemotherapy Dosing (OBCD) study, we examined associations between age at diagnosis and chemotherapy dose reductions at treatment initiation and across the chemotherapy regimen. We evaluated associations between age and first cycle dose proportion <90% (FCDP, reflecting dosing intent at treatment initiation) and average relative dose intensity <90% (ARDI, reflecting average chemotherapy dose across the treatment course) across chemotherapy regimens and administration schedules used in real-world community settings. We explored how these associations changed after accounting for pre-existing comorbidities and performance status, and whether the associations between age and dose reductions differed by drug class.
Methods
Study population
The OBCD study included 34,109 women from Kaiser Permanente Northern California (KPNC) and Kaiser Permanente Washington (KPWA) diagnosed and treated for primary stage I-IIIA breast cancer between 2004-2019. Eligibility criteria have been described in detail elsewhere; stage was defined using the American Joint Committee on Cancer (AJCC) version in place at the time of diagnosis.[22, 23] Briefly, women were eligible if they were diagnosed with a primary breast cancer with no prior history or same-day diagnosis of any cancer (except non-melanoma skin cancer), enrolled at KPWA or KPNC at the time of diagnosis, had available medical records, and did not opt out of research studies. All study sites received IRB approval with a waiver of consent to collect and analyze patient data from KPNC and KPWA; the committees that provided approval were the KP Interregional IRB, KPNC IRB, Memorial Sloan Kettering IRB, and Rutgers University IRB.
We restricted this analysis to women receiving adjuvant chemotherapy after breast cancer surgery (N=11,839), which, by definition, excluded anyone who received neoadjuvant chemotherapy (N=1,122), no surgery (N=291), or no chemotherapy (N=20,857). We excluded women with neoadjuvant chemotherapy because the factors motivating treatment choice likely differed between people who received neoadjuvant and adjuvant chemotherapy, which would have necessitated stratification of our analysis. The number of older women receiving neoadjuvant chemotherapy (N=137 aged 65-74 years and N=30 aged 75+ years) was insufficient to analyze dose reductions by age. We included women who received monoclonal antibodies (e.g. trastuzumab, pertuzumab) because they are included in the NCCN systemic adjuvant treatment guidelines alongside chemotherapy for invasive HER2 positive breast cancer.[1] We excluded women who participated in randomized controlled trials (N=581), used a non-guideline drug combination or unidentifiable chemotherapy regimen (N=575), initiated chemotherapy after 2019 (N=159), were missing BSA (N=55), were missing dose or infusion date for any drug (N=127), underwent a regimen change from a guideline to non-guideline drug combination (N=96), or received oral cyclophosphamide (N=80). The final analytic sample included 10,166 women.
Exposures
Diagnostic and treatment data were obtained from the health plans’ virtual data warehouses (VDWs), which store electronic administrative data with standardized definitions.[24] Data were supplemented using electronic health records (EHR), including manual chart review, when needed.
We collected information on the following patient factors from the VDW at the time of breast cancer diagnosis: age, race and ethnicity, body mass index (BMI), and Charlson comorbidity index (using diagnosis codes in the 12 months prior to breast cancer diagnosis).[25] We collected block-level median household income using geocoded addresses from the time of each woman’s breast cancer diagnosis linked to 5-year income estimates from the American Community Survey data; at KPWA, we used a rolling 5-year average from 2010-2018, and at KPNC, we used 5-year estimates from 2010 only. We collected information on individual pre-existing comorbidities including renal disease, hepatic disease, diabetes, cardiovascular disease, neuropathy, anemia, and thrombocytopenia using a combination of electronic data (i.e., diagnosis codes and laboratory values, Table S1) up to the start date of chemotherapy.[26] We collected ECOG performance status, when available as a discrete field in the EHR from the 30 days prior to and including the first date of chemotherapy. Cancer characteristics were obtained from tumor registries and included year of diagnosis, AJCC stage, grade, tumor size, number of positive nodes, estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor-2 (HER2) status. Stage was defined using the AJCC staging version corresponding to diagnosis year using surgically derived pathology specimens.
Chemotherapy data
Data on chemotherapy received were obtained from EHRs from infusion and prescription medication databases and manual abstraction. We identified planned chemotherapy regimens using a methodology previously described.[27] Briefly, participants were grouped by drug combination received and then classified as receiving a drug combination within or outside of the NCCN guidelines for early-stage breast cancer at any point between 2004-2019. After excluding those with non-guideline-based drug combinations, women with guideline-based drug combinations were classified into intended administration schedules (by cycle number and interval) using algorithms that allowed for scheduling variation.[1] From these algorithms, we obtained information on the expected dosing and administration schedule. In cases of ambiguity, we used medical chart abstraction to identify the intended administration schedule.
Our primary outcome of interest was FCDP <90%, reflecting clinician dosing intent at chemotherapy initiation. FCDP was calculated by dividing the observed chemotherapy dose in the first cycle for each drug by the expected dose in the first cycle (based on BSA from the Mostellar formula) using data solely from the first day of chemotherapy administration.[3, 28] The FCDP was calculated separately for each drug given in the first cycle of chemotherapy and then averaged to determine FCDP. We categorized FCDP as a binary variable reflecting receipt of <90% or ≥90% of first cycle dose.[2, 3]
We also evaluated ARDI across all cycles of chemotherapy, which reflects additional dose reductions and delays over the course of chemotherapy. We first calculated relative dose intensity (RDI)[29] of each drug by dividing the cumulative dose received by weeks of treatment and subsequently dividing this by the cumulative intended (or standard) dose by intended weeks of treatment as shown below:
RDI was calculated for each drug separately, averaged across drugs to obtain ARDI, and categorized as ARDI <90% or ARDI ≥90%. When participants discontinued drugs early, the ARDI only included data for completed chemotherapy cycles. Thus, the numerator and denominator were censored at the time of drug discontinuation in order not to obscure intentional dose reductions. The ARDI inherently accounted for treatment delays by incorporating treatment duration into calculations.[27]
Statistical Analyses
We described patient and tumor characteristics, overall and stratified by age at diagnosis (18-39, 40-49, 50-64, 65-74, and 75+ years). We used generalized linear models of the Poisson family with a log link-function and robust standard errors to calculate prevalence ratios (PR) and corresponding 95% confidence intervals (CI) for age at diagnosis (using 40-49 years as the reference group) associated with each of the two binary outcomes, FCDP and ARDI <90% vs ≥90%. Models were minimally-adjusted for BMI (<18.5, 18.5-<25, 25-<30, 30-<35, 35-<40, 40+ kg/m2), stage (I, II, IIIA), diagnosis year (January 2004-December 2007, January 2008-December 2011, January 2012-December 2014, January 2015-December 2019), and study site (KPNC, KPWA). We then examined fully-adjusted multivariable models including all variables in the minimally-adjusted models plus additional statistically significant patient or tumor characteristics from the minimally adjusted models (race and ethnicity [Hispanic, non-Hispanic Asian, non-Hispanic Black, non-Hispanic White, all others combined including Native Hawaiian, Pacific Islander, American Indian, Alaskan Native, and multi-racial)], income [quartiles], HER2 positivity [yes, no], grade [1, 2, 3], ER/PR positivity [ER+ and/or PR+, ER− and PR−], and type of surgery [mastectomy, breast conserving surgery]). All adjusted models used a complete-case analysis, excluding individuals with missing data.
To understand how comorbidities might account for the association between age and FCDP or ARDI, we examined the fully adjusted models with additional adjustment for comorbidities. We examined these as three separate models adjusting for 1) Charlson comorbidity score, 2) individual comorbidities, and 3) both Charlson and individual comorbidities. To further assess the robustness of our analyses, we evaluated whether age at diagnosis was associated with FCDP or ARDI limited to people with a Charlson comorbidity score of 0. In addition, we evaluated associations between age and FCDP or ARDI adjusted for ECOG score (categorized as 0 vs 1+), and associations between ECOG score and FCDP or ARDI adjusted for age. These analyses were limited to 5,045 women with ECOG performance data available.
In sensitivity analyses, we limited the population to women with at least 12 months of continuous enrollment in the KP health plan before diagnosis because the Charlson Comorbidity score is based on clinical encounters within the year prior to diagnosis. In addition, we conducted exploratory analyses stratified by the four most common anticancer drug classes in our study (anthracyclines, taxanes, monoclonal antibodies, alkylating agents) to assess whether associations between age and dosing differed by drugs received. For the drug-class-specific analyses, FCDP was calculated using the first cycle of the drug in that drug class even if not given in the first chemotherapy cycle. For example, if someone received the AC-T regimen (Adriamycin + Cyclophosphamide followed by Paclitaxel), their first cycle of Paclitaxel was incorporated into the FCDP for taxanes even though they would have received their first dose of paclitaxel later in their course of chemotherapy. Finally, we conducted sensitivity analyses using a dose threshold of 85% for FCDP and ARDI to reflect a commonly used cutoff in the literature. Throughout, we presented p-trends for exposures modelled as grouped linear and global p-values for categorical variables. All tests for statistical significance were 2-sided with values of p<0.05 considered statistically significant. All analyses were conducted using R Statistical Software (v.4.3.1; R Core Team 2023) and Stata.v.17 (StataCorp, College Station, TX).
Results
Among 10,166 women who received adjuvant chemotherapy, descriptive results showed nearly half (N=4,787, 47.1%) were aged 50-64 years, 1,810 (17.8%) were 65-74 years, and 286 (2.8%) were 75+ years (Table 1). Older women had a higher comorbidity burden than younger women with 53.5% of women 75+ years having a Charlson score ≥1 compared with 9.8% of women 18-39 years. The most common pre-diagnostic comorbidity was diabetes (10.8% overall), followed by renal impairment (8.9%), and both increased with age. Among 5,045 women with ECOG data, 15.6% had a score of 1 or higher, increasing to 26.6% among women aged 75+ years.
Table 1:
Select demographic and clinical characteristics of study participants, according to age at diagnosis
| Age at diagnosis (years) | ||||||
|---|---|---|---|---|---|---|
|
|
||||||
| Characteristic | Overall N = 10,166 (100%) |
18-39 N = 796 (7.8%) |
40-49 N = 2,487 (24.5%) |
50-64 N = 4,787 (47.1%) |
65-74 N = 1,810 (17.8%) |
75+ N = 286 (2.8%) |
| Race/ethnicity | ||||||
| Hispanic | 1,537 (15.2%) | 199 (25.1%) | 483 (19.5%) | 627 (13.1%) | 204 (11.3%) | 24 (8.4%) |
| Non-Hispanic American Indian or Alaskan Native, more than one race, or other | 53 (0.5%) | 8 (1.0%) | 8 (0.3%) | 28 (0.6%) | 8 (0.4%) | <5 (0.3%) |
| Non-Hispanic Asian | 1,824 (18.0%) | 203 (25.6%) | 541 (21.8%) | 835 (17.5%) | 214 (11.8%) | 31 (10.8%) |
| Non-Hispanic Black | 812 (8.0%) | 50 (6.3%) | 183 (7.4%) | 418 (8.7%) | 144 (8.0%) | 17 (5.9%) |
| Non-Hispanic Native Hawaiian or Other Pacific Islander | 87 (0.9%) | 11 (1.4%) | 25 (1.0%) | 43 (0.9%) | 7 (0.4%) | <5 (0.3%) |
| Non-Hispanic White | 5,829 (57.5%) | 322 (40.6%) | 1,238 (50.0%) | 2,827 (59.2%) | 1,230 (68.1%) | 212 (74.1%) |
| Unknown | 24 | 3 | 9 | 9 | 3 | 0 |
| Median neighborhood household income | ||||||
| Quartile 1: <66,392 | 2,460 (24.4%) | 185 (23.3%) | 590 (23.9%) | 1,164 (24.5%) | 457 (25.4%) | 64 (22.5%) |
| Quartile 2: $66,392-<$89,107 | 2,484 (24.6%) | 199 (25.1%) | 598 (24.2%) | 1,190 (25.0%) | 432 (24.0%) | 65 (22.9%) |
| Quartile 3: $89,107-<$117,278 | 2,596 (25.7%) | 209 (26.3%) | 639 (25.9%) | 1,237 (26.0%) | 442 (24.5%) | 69 (24.3%) |
| Quartile 4: ≥$117,278 | 2,561 (25.4%) | 201 (25.3%) | 644 (26.1%) | 1,160 (24.4%) | 470 (26.1%) | 86 (30.3%) |
| Unknown | 65 | 2 | 16 | 36 | 9 | 2 |
| BMI (kg/m2) | ||||||
| <18.5 | 113 (1.1%) | 18 (2.3%) | 22 (0.9%) | 48 (1.0%) | 19 (1.0%) | 6 (2.1%) |
| 18.5-<25 | 3,200 (31.5%) | 354 (44.5%) | 900 (36.2%) | 1,362 (28.5%) | 498 (27.5%) | 86 (30.1%) |
| 25-<30 | 3,170 (31.2%) | 235 (29.5%) | 770 (31.0%) | 1,500 (31.3%) | 565 (31.2%) | 100 (35.0%) |
| 30-<35 | 1,978 (19.5%) | 116 (14.6%) | 422 (17.0%) | 990 (20.7%) | 392 (21.7%) | 58 (20.3%) |
| 35-<40 | 1,025 (10.1%) | 51 (6.4%) | 223 (9.0%) | 533 (11.1%) | 189 (10.4%) | 29 (10.1%) |
| 40+ | 680 (6.7%) | 22 (2.8%) | 150 (6.0%) | 354 (7.4%) | 147 (8.1%) | 7 (2.4%) |
| Year of diagnosis | ||||||
| Jan 2004 - Dec 2007 | 1,404 (13.8%) | 109 (13.7%) | 391 (15.7%) | 681 (14.2%) | 193 (10.7%) | 30 (10.5%) |
| Jan 2008 - Dec 2011 | 2,897 (28.5%) | 182 (22.9%) | 743 (29.9%) | 1,439 (30.1%) | 467 (25.8%) | 66 (23.1%) |
| Jan 2012 - Dec 2014 | 2,302 (22.6%) | 172 (21.6%) | 542 (21.8%) | 1,089 (22.7%) | 443 (24.5%) | 56 (19.6%) |
| Jan 2015 - Dec 2019 | 3,563 (35.0%) | 333 (41.8%) | 811 (32.6%) | 1,578 (33.0%) | 707 (39.1%) | 134 (46.9%) |
| Surgery type | ||||||
| Mastectomy | 4,744 (46.7%) | 529 (66.5%) | 1,271 (51.1%) | 2,070 (43.2%) | 745 (41.2%) | 129 (45.1%) |
| Breast conserving surgery | 5,419 (53.3%) | 267 (33.5%) | 1,214 (48.9%) | 2,717 (56.8%) | 1,064 (58.8%) | 157 (54.9%) |
| Unknown | 3 | 0 | 2 | 0 | 1 | 0 |
| Stage | ||||||
| Stage I | 3,616 (35.6%) | 261 (32.8%) | 911 (36.6%) | 1,770 (37.0%) | 596 (32.9%) | 78 (27.3%) |
| Stage II | 5,400 (53.1%) | 443 (55.7%) | 1,320 (53.1%) | 2,502 (52.3%) | 981 (54.2%) | 154 (53.8%) |
| Stage IIIA | 1,150 (11.3%) | 92 (11.6%) | 256 (10.3%) | 515 (10.8%) | 233 (12.9%) | 54 (18.9%) |
| Grade | ||||||
| 1 | 1,056 (10.6%) | 47 (6.0%) | 295 (12.1%) | 502 (10.7%) | 192 (10.8%) | 20 (7.1%) |
| 2 | 4,278 (42.9%) | 287 (36.5%) | 1,032 (42.4%) | 2,092 (44.6%) | 756 (42.6%) | 111 (39.5%) |
| 3 | 4,635 (46.5%) | 453 (57.6%) | 1,107 (45.5%) | 2,100 (44.7%) | 825 (46.5%) | 150 (53.4%) |
| Unknown | 197 | 9 | 53 | 93 | 37 | 5 |
| Tumor size | ||||||
| 0.1-0.5cm | 165 (1.6%) | 17 (2.2%) | 41 (1.7%) | 80 (1.7%) | 22 (1.2%) | 5 (1.7%) |
| >0.5-1cm | 894 (8.8%) | 51 (6.5%) | 218 (8.8%) | 451 (9.4%) | 160 (8.9%) | 14 (4.9%) |
| >1-2cm | 3,992 (39.4%) | 259 (32.8%) | 968 (39.1%) | 1,969 (41.3%) | 703 (39.0%) | 93 (32.5%) |
| >2-5cm | 4,561 (45.0%) | 415 (52.5%) | 1,112 (44.9%) | 2,051 (43.0%) | 829 (46.0%) | 154 (53.8%) |
| >5cm | 516 (5.1%) | 48 (6.1%) | 137 (5.5%) | 222 (4.7%) | 89 (4.9%) | 20 (7.0%) |
| Unknown | 38 | 6 | 11 | 14 | 7 | 0 |
| Number of nodes | ||||||
| All nodes negative | 5,262 (52.3%) | 440 (55.9%) | 1,302 (53.0%) | 2,477 (52.2%) | 916 (51.1%) | 127 (45.0%) |
| 1-3 nodes | 3,743 (37.2%) | 261 (33.2%) | 914 (37.2%) | 1,792 (37.8%) | 667 (37.2%) | 109 (38.7%) |
| 4-9 nodes | 1,001 (10.0%) | 81 (10.3%) | 227 (9.2%) | 458 (9.7%) | 197 (11.0%) | 38 (13.5%) |
| 10+ nodes | 54 (0.5%) | 5 (0.6%) | 14 (0.6%) | 16 (0.3%) | 11 (0.6%) | 8 (2.8%) |
| Unknown | 106 | 9 | 30 | 44 | 19 | 4 |
| Hormone-receptor positivity | ||||||
| ER− and PR− | 2,670 (26.3%) | 193 (24.2%) | 550 (22.2%) | 1,293 (27.1%) | 535 (29.6%) | 99 (34.6%) |
| ER+ and/or PR+ | 7,482 (73.7%) | 603 (75.8%) | 1,931 (77.8%) | 3,487 (72.9%) | 1,274 (70.4%) | 187 (65.4%) |
| Unknown | 14 | 0 | 6 | 7 | 1 | 0 |
| HER2 positivity | ||||||
| HER2− | 7,703 (76.1%) | 591 (74.6%) | 1,928 (77.9%) | 3,644 (76.5%) | 1,351 (74.9%) | 189 (66.3%) |
| HER2+ | 2,413 (23.9%) | 201 (25.4%) | 546 (22.1%) | 1,118 (23.5%) | 452 (25.1%) | 96 (33.7%) |
| Unknown | 50 | 4 | 13 | 25 | 7 | 1 |
| Charlson comorbidity index | ||||||
| 0 | 7,661 (75.4%) | 718 (90.2%) | 2,148 (86.4%) | 3,606 (75.3%) | 1,056 (58.3%) | 133 (46.5%) |
| 1 | 1,659 (16.3%) | 69 (8.7%) | 276 (11.1%) | 839 (17.5%) | 417 (23.0%) | 58 (20.3%) |
| 2 | 480 (4.7%) | 8 (1.0%) | 41 (1.6%) | 214 (4.5%) | 173 (9.6%) | 44 (15.4%) |
| 3+ | 366 (3.6%) | <5 (0.1%) | 22 (0.9%) | 128 (2.7%) | 164 (9.1%) | 51 (17.8%) |
| ECOG Scorea | ||||||
| 0 | 4,258 (84.4%) | 392 (89.9%) | 1,018 (87.7%) | 1,951 (84.9%) | 762 (78.8%) | 135 (73.4%) |
| 1+ | 787 (15.6%) | 44 (10.1%) | 143 (12.3%) | 346 (15.1%) | 205 (21.2%) | 49 (26.6%) |
| Diabetes | ||||||
| No | 9,064 (89.2%) | 783 (98.4%) | 2,359 (94.9%) | 4,213 (88.0%) | 1,470 (81.2%) | 239 (83.6%) |
| Yes | 1,102 (10.8%) | 13 (1.6%) | 128 (5.1%) | 574 (12.0%) | 340 (18.8%) | 47 (16.4%) |
| CVD Comorbidity | ||||||
| No | 9,780 (96.2%) | 788 (99.0%) | 2,446 (98.4%) | 4,604 (96.2%) | 1,689 (93.3%) | 253 (88.5%) |
| Yes | 386 (3.8%) | 8 (1.0%) | 41 (1.6%) | 183 (3.8%) | 121 (6.7%) | 33 (11.5%) |
| Renal Comorbidity | ||||||
| No | 9,265 (91.1%) | 792 (99.5%) | 2,459 (98.9%) | 4,488 (93.8%) | 1,398 (77.2%) | 128 (44.8%) |
| Yes | 901 (8.9%) | 4 (0.5%) | 28 (1.1%) | 299 (6.2%) | 412 (22.8%) | 158 (55.2%) |
| Hepatic Comorbidity | ||||||
| No | 9,960 (98.0%) | 779 (97.9%) | 2,435 (97.9%) | 4,687 (97.9%) | 1,782 (98.5%) | 277 (96.9%) |
| Yes | 206 (2.0%) | 17 (2.1%) | 52 (2.1%) | 100 (2.1%) | 28 (1.5%) | 9 (3.1%) |
| Anemia Comorbidity | ||||||
| No | 9,569 (94.1%) | 732 (92.0%) | 2,270 (91.3%) | 4,573 (95.5%) | 1,726 (95.4%) | 268 (93.7%) |
| Yes | 597 (5.9%) | 64 (8.0%) | 217 (8.7%) | 214 (4.5%) | 84 (4.6%) | 18 (6.3%) |
| Neuropathy Comorbidity | ||||||
| No | 9,389 (92.4%) | 764 (96.0%) | 2,366 (95.1%) | 4,410 (92.1%) | 1,602 (88.5%) | 247 (86.4%) |
| Yes | 777 (7.6%) | 32 (4.0%) | 121 (4.9%) | 377 (7.9%) | 208 (11.5%) | 39 (13.6%) |
| Thrombocytopenia Comorbidity | ||||||
| No | 10,102 (99.4%) | 788 (99.0%) | 2,474 (99.5%) | 4,757 (99.4%) | 1,800 (99.4%) | 283 (99.0%) |
| Yes | 64 (0.6%) | 8 (1.0%) | 13 (0.5%) | 30 (0.6%) | 10 (0.6%) | <5 (1.0%) |
Abbreviations: BMI (body mass index), CVD (cardiovascular disease), ECOG (Eastern Cooperative Oncology Group), ER (estrogen receptor), HER2 (human epidermal growth factor receptor 2), PR (progesterone receptor)
ECOG score among 5045 women where ECOG score was available in the 30 days prior to and including date of first chemotherapy.
The proportion of women with a FCDP<90% ranged from 2.9% among women 18-39 years to 18.6% among women 75+ years (Table 2). In all models, we observed that increasing age was statistically significantly associated with increasing prevalence of FCDP<90% (p-trend<0.001). In multivariable models (minimally adjusted models plus adjustment for race and ethnicity, income, HER2 positivity, grade, ER/PR positivity, and surgery type but not comorbidities), women 50-64 years had a 34% greater prevalence of FCDP<90% (PR=1.34, 95% CI=1.10, 1.64) compared to women 40-49 years. In addition, women 65-74 years had a 74% greater prevalence of FCDP<90% (PR=1.74, 95% CI=1.38, 2.20), and women 75+ years had a 388% greater prevalence of FCDP<90% (PR=4.88, 95% CI=3.58, 6.66) compared to women 40-49 years. Women 18-39 years had a similar prevalence of FCDP<90% compared to women 40-49 years (PR=0.80, 95%CI= 0.52, 1.21). Additionally adjusting models for Charlson comorbidity score, individual comorbidities, or both Charlson score and individual comorbidities did not substantially change PRs. Limiting the multivariable model to women with a Charlson score of 0 slightly increased the PR for women 75+ years to 6.04 (95% CI=4.05, 8.99).
Table 2.
Age at diagnosis associated with first cycle dose proportion (FCDP) <90% of chemotherapy regimen
| FCDP ≥90% (n, row %) |
FCDP <90% (n, row %) |
Minimally adjusted PRa (95% CI) |
Multivariable adjusted PRb (95% CI) |
Multivariable additionally adjusted for Charlson PR (95% CI) |
Multivariable additionally adjusted for individual comorbidities PR (95% CI) |
Multivariable additionally adjusted for Charlson and individual comorbidities PR (95% CI) |
Multivariable adjusted limited to people with Charlson score of 0 PR (95% CI) |
|
|---|---|---|---|---|---|---|---|---|
| TOTALc | 9,080 (93.5) | 626 (6.5) | N = 9,706 | N = 9,706 | N = 9,706 | N = 9,706 | N = 9,706 | N = 7,290 |
| AGE AT DIAGNOSIS (years) | ||||||||
| 18-39 | 745 (97.1) | 22 (2.9) | 0.84 (0.54, 1.30) | 0.80 (0.52, 1.21) | 0.81 (0.53, 1.23) | 0.83 (0.54, 1.27) | 0.84 (0.55, 1.28) | 0.92 (0.59, 1.44) |
| 40-49 | 2,249 (95.4) | 109 (4.6) | REF | REF | REF | REF | REF | REF |
| 50-64 | 4,268 (93.3) | 305 (6.7) | 1.33 (1.09, 1.62) | 1.34 (1.10, 1.64) | 1.31 (1.07, 1.61) | 1.32 (1.08, 1.61) | 1.31 (1.07, 1.60) | 1.36 (1.06, 1.73) |
| 65-74 | 1,595 (92.0) | 139 (8.0) | 1.74 (1.38, 2.20) | 1.74 (1.38, 2.20) | 1.62 (1.28, 2.07) | 1.65 (1.29, 2.10) | 1.61 (1.26, 2.05) | 1.88 (1.40, 2.52) |
| 75+ | 223 (81.4) | 51 (18.6) | 5.15 (3.79, 7.00) | 4.88 (3.58, 6.66) | 4.49 (3.24, 6.23) | 4.04 (2.89, 5.67) | 3.93 (2.77, 5.57) | 6.04 (4.05, 8.99) |
| p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | |||
Abbreviations: BMI (body mass index), CI (confidence interval), ER/PR (estrogen/progesterone receptor), FCDP (first cycle dose proportion), HER2 (human epidermal growth factor receptor 2), PR (prevalence ratio), REF (reference group)
Minimally-adjusted models adjusted for BMI (<18.5, 18.5-<25, 25-<30, 30-<35, 35-<40, 40+), stage at diagnosis (Stage I, Stage II, Stage IIIA), year of diagnosis (Jan 2004-Dec 2007, Jan 2008-Dec 2011, Jan 2012-Dec 2014, Jan 2015-Dec 2019), and study site (KPNC, KPWA)
Multivariable-adjusted models additionally adjusted for race and ethnicity (Hispanic; Non-Hispanic Asian; Non-Hispanic Black; Non-Hispanic White; all others combined), income (Q1: <$66,392; Q2: $66,392-<$89,107; $89,107-<$117,278; ≥$117,278) , HER-2 positivity (HER2−, HER2+), grade (1, 2, 3), ER/PR positivity (ER− and PR−, ER+ and/or PR+), surgery type (Mastectomy, Breast-conserving surgery)
All adjusted models used a complete-case analysis, excluding individuals with missing data.
The proportion of women with ARDI<90% ranged from 12.8% in women 18-39 years to 32.1% in women 75+ years (Table 3). Patterns in the associations between age and ARDI<90% were similar to those for FCDP<90% (multivariable PR for women 75+ years vs 40-49 years=1.91, 95% CI=1.58, 2.32; multivariable PR for women 65-74 years vs 40-49 years=1.33, 95% CI=1.17, 1.51). Overall, the PRs for ARDI<90% were lower than the PRs for FCDP<90% for each age group as displayed in Table 2.
Table 3.
Age at diagnosis associated with average relative dose intensity (ARDI) <90% of chemotherapy regimen
| ARDI ≥90% (n, row %) |
ARDI <90% (n, row %) |
Minimally adjusted PRa (95% CI) |
Multivariable adjusted PRb (95% CI) |
Additionally adjusted for Charlson PR (95% CI) |
Additionally adjusted for individual comorbidities PR (95% CI) |
Additionally adjusted for Charlson and individual comorbidities PR (95% CI) |
Multivariable adjusted limited to people with Charlson score of 0 PR (95% CI) |
|
|---|---|---|---|---|---|---|---|---|
| TOTALc | 7,848 (80.9%) | 1,858 (19.1%) | N = 9,706 | N = 9,706 | N = 9,706 | N = 9,706 | N = 9,706 | N = 7,290 |
| AGE AT DIAGNOSIS (years) | ||||||||
| 18-39 | 669 (87.2) | 98 (12.8) | 0.87 (0.71, 1.06) | 0.82 (0.67, 1.01) | 0.83 (0.68, 1.02) | 0.84 (0.68, 1.03) | 0.84 (0.68, 1.03) | 0.90 (0.72, 1.12) |
| 40-49 | 1,973 (83.7) | 385 (16.3) | REF | REF | REF | REF | REF | REF |
| 50-64 | 3,677 (80.4) | 896 (19.6) | 1.16 (1.04, 1.28) | 1.15 (1.04, 1.28) | 1.13 (1.01, 1.26) | 1.13 (1.02, 1.26) | 1.13 (1.01, 1.25) | 1.14 (1.01, 1.29) |
| 65-74 | 1,343 (77.5) | 391 (22.6) | 1.36 (1.20, 1.54) | 1.33 (1.17, 1.51) | 1.25 (1.10, 1.42) | 1.27 (1.11, 1.44) | 1.24 (1.09, 1.41) | 1.36 (1.17, 1.59) |
| 75+ | 186 (67.9) | 88 (32.1) | 2.10 (1.73, 2.54) | 1.91 (1.58, 2.32) | 1.73 (1.41, 2.11) | 1.74 (1.41, 2.15) | 1.67 (1.35, 2.07) | 1.78 (1.32, 2.38) |
| p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | |||
Abbreviations: ARDI (average relative dose intensity), BMI (body mass index), CI (confidence interval), ER/PR (estrogen/progesterone receptor), HER2 (human epidermal growth factor receptor 2), PR (prevalence ratio), REF (reference group)
Minimally-adjusted models adjusted for BMI (<18.5, 18.5-<25, 25-<30, 30-<35, 35-<40, 40+), stage at diagnosis (Stage I, Stage II, Stage IIIA), year of diagnosis (Jan 2004-Dec 2007, Jan 2008-Dec 2011, Jan 2012-Dec 2014, Jan 2015-Dec 2019), and study site (KPNC, KPWA)
Multivariable-adjusted models additionally adjusted for race and ethnicity (Hispanic; Non-Hispanic Asian; Non-Hispanic Black; Non-Hispanic White; all others combined), income (Q1: <$66,392; Q2: $66,392-<$89,107; $89,107-<$117,278; ≥$117,278) , HER-2 positivity (HER2−, HER2+), grade (1, 2, 3), ER/PR positivity (ER− and PR−, ER+ and/or PR+), surgery type (Mastectomy, Breast-conserving surgery)
All adjusted models used a complete-case analysis, excluding individuals with missing data.
When limiting the multivariable models to women with ECOG scores (Table 4), women 75+ years were more likely to have FCDP<90% (PR= 6.14, 95% CI=3.61, 10.46) and ARDI<90% (PR=1.69, 95% CI=1.28, 2.23) compared to women 40-49 years. There was no association between ECOG score and prevalence of FCDP<90% or ARDI<90% in multivariable models.
Table 4.
ECOG score associated with first cycle dose proportion (FCDP) <90% and average relative dose intensity (ARDI) <90% of chemotherapy regimen
| FCDP ≥90% (n,%) |
FCDP <90% (n,%) |
Minimally adjusted PRa (95% CI) |
Multivariable adjusted PRb (95% CI) |
ARDI ≥90% (n,%) |
ARDI <90% (n,%) |
Minimally adjusted PRa (95% CI) |
Multivariable adjusted PRb (95% CI) |
|
|---|---|---|---|---|---|---|---|---|
| TOTALc | 4,714 (96.0) | 198 (4.0) | N = 4,912 | N = 4,912 | 4,092 (83.3) | 820 (16.7) | N = 4,912 | N = 4,912 |
| Age at diagnosis adjusted for ECOG score | ||||||||
| 18-39 | 419 (98.6) | 6 (1.4) | 0.64 (0.27, 1.52) | 0.63 (0.26, 1.49) | 379 (89.2) | 46 (10.8) | 0.80 (0.59, 1.09) | 0.76 (0.56, 1.03) |
| 40-49 | 1,100 (97.4) | 29 (2.6) | REF | REF | 964 (85.4) | 165 (14.6) | REF | REF |
| 50-64 | 2,160 (96.3) | 83 (3.7) | 1.32 (0.88, 1.99) | 1.32 (0.88, 1.99) | 1,851 (82.5) | 392 (17.5) | 1.15 (0.97, 1.36) | 1.12 (0.95, 1.32) |
| 65-74 | 881 (94.0) | 56 (6.0) | 2.19 (1.42, 3.38) | 2.13 (1.37, 3.31) | 769 (82.1) | 168 (17.9) | 1.17 (0.96, 1.42) | 1.13 (0.93, 1.37) |
| 75+ | 154 (86.5) | 24 (13.5) | 6.25 (3.72, 10.49) | 6.14 (3.61, 10.46) | 129 (72.5) | 49 (27.5) | 1.89 (1.44, 2.50) | 1.69 (1.28, 2.23) |
| p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | p-trend < 0.001 | |||||
| ECOG score adjusted for age at diagnosis | ||||||||
| 0 | 3,986 (96.1) | 160 (3.9) | REF | REF | 3,486 (84.1) | 660 (15.9) | REF | REF |
| 1+ | 728 (95.0) | 38 (5.0) | 0.85 (0.60, 1.22) | 0.82 (0.57, 1.16) | 606 (79.1) | 160 (20.9) | 1.15 (0.99, 1.35) | 1.07 (0.92, 1.25) |
Abbreviations: ARDI (average relative dose intensity), BMI (body mass index), CI (confidence interval), ECOG (Eastern Cooperative Oncology Group), ER/PR (estrogen/progesterone receptor), FCDP (first cycle dose proportion), HER2 (human epidermal growth factor receptor 2), PR (prevalence ratio), REF (reference group)
Minimally-adjusted models adjusted for BMI (<18.5, 18.5-<25, 25-<30, 30-<35, 35-<40, 40+), stage at diagnosis (Stage I, Stage II, Stage IIIA), year of diagnosis (Jan 2004-Dec 2007, Jan 2008-Dec 2011, Jan 2012-Dec 2014, Jan 2015-Dec 2019), study site (KPNC, KPWA), and ECOG score (0; 1+)
Multivariable-adjusted models additionally adjusted for race and ethnicity (Hispanic; Non-Hispanic Asian; Non-Hispanic Black; Non-Hispanic White; all others combined), income (Q1: <$66,392; Q2: $66,392-<$89,107; $89,107-<$117,278; ≥$117,278) , HER-2 positivity (HER2−, HER2+), grade (1, 2, 3), ER/PR positivity (ER− and PR−, ER+ and/or PR+), surgery type (Mastectomy, Breast-conserving surgery)
All adjusted models used a complete-case analysis, excluding individuals with missing data.
Sensitivity analyses limited to 8,945 women with 12 months of continuous health plan enrollment were consistent with the main analyses for FCDP (Table S2) and ARDI (Table S3). When we stratified results across four main drug classes, older age was associated with FCDP<90% in every drug class except monoclonal antibodies (Table S4). Women 75+ years who received taxanes had the highest PR for FCDP<90% of all the drug classes compared to women 40-49 years (PR=6.50, 95% CI=4.45, 9.49). PR patterns by age and drug class were similar when evaluating ARDI<90% (Table S5). When using 85% as the threshold for FCDP (Table S6) and ARDI (Table S7), the number of women with a dose reduction decreased compared to the main analyses using the 90% threshold. The PR patterns by age were similar to the patterns observed in the main analyses, although the PRs for women 75+ years increased when using the 85% threshold.
Discussion
In this population-based cohort of women with early-stage breast cancer, we showed that older age was strongly and statistically significantly associated with an increased likelihood of chemotherapy dose reductions. Age was a strong driver of dose reductions particularly at chemotherapy initiation, independent of comorbidities and ECOG score. To our knowledge, this is the largest study to date to evaluate associations between age at diagnosis and chemotherapy dose reductions, disaggregating between FCDP<90% and ARDI<90% in women with breast cancer.
Our findings add to the existing literature by providing information on chemotherapy treatment and dosing patterns for older women accounting for comorbidities. One prior large study showed that people with any comorbidities were more likely to have ARDI<85% compared to people with no comorbid disease.[20] However, that prior study did not include comorbidities in the final multivariable model, so it is unknown whether comorbidity adjustment would have changed their estimates. Our analysis showed that women 75+ years had a 4-5-fold increased prevalence of FCDP<90% and almost 2-fold increased prevalence of ARDI<90% compared to women 40-49 years, even after adjusting for comorbidities and BMI. Our results are consistent with other studies that have shown older age (defined as 65+ or 70+ years) is associated with chemotherapy modifications, including lower RDI, typically evaluated as <85% of the recommended dose.[13, 16-18, 20, 21, 30] We used 90% as the dose threshold to increase our sample size and statistical power, particularly of the oldest age group; however, sensitivity analyses using 85% as the dose threshold showed similar results. Overall, our results suggest that other factors such as comorbidities, cancer prognosis, performance status, and body size may play less of a role in chemotherapy dose reductions despite guidance to tailor initial chemotherapy dose to these factors.[31-33]
Like prior studies that examined specific drug classes,[11-13, 18, 20, 34] we noted older age was associated with an increased prevalence of FCDP<90% and ARDI<90% among women receiving anthracyclines. These analyses were exploratory, and results should be interpreted with caution given the small number of women in some age and anticancer drug class combinations. However, we noted the associations between older age and FCDP<90% were greatest among women 75+ years receiving taxanes. It is possible that we identified more dose reductions in women 75+ years receiving taxanes than anthracyclines simply because more women were prescribed taxanes. Among women 75+ years, 204 received taxanes and only 49 received anthracyclines. It’s also possible that the dose reductions for taxanes differed depending on other drugs a woman received, for example, taxanes after anthracycline (e.g. AC-T regimen) versus taxanes with cyclophosphamide (e.g. TC regimen). We did not stratify drug class analyses by regimen due to small numbers, but this is worth exploring further in a larger sample.
There are many valid reasons for initial and/or subsequent chemotherapy dose reductions in older women. Providers may have concerns about comorbidities, the ability to complete treatment as intended, potential treatment toxicities, patient performance score, and quality of life.[32, 35, 36] Older adults with frailty, disability, and/or extensive comorbidities may be at higher risk of chemotherapy-related toxicities and hospitalizations compared to younger adults with cancer or older adults without these conditions.[33, 35, 37, 38] As a result, older adults with cancer may not be able to tolerate the full chemotherapy regimen and could experience poorer quality of life and physical functioning, higher relapse rates, and worse survival outcomes.[35, 39-44] However, some older adults without frailty, disability, or comorbid disease may be able to tolerate and benefit from fully-dosed chemotherapy. The Cancer and Aging Research Group Toxicity Tool (CARG-TT) and Chemotherapy Risk Assessment Scale for High-Age Patients (CRASH) are two validated risk assessments that may be used to identify older adults at the highest risk of severe toxicity from chemotherapy,[41, 45, 46] where dose reductions may be appropriate. Surprisingly, we did not see any association between ECOG score and FCDP or ARDI in our analyses. This analysis was limited by ECOG data availability. However, it’s possible that women with higher ECOG scores were less likely to receive chemotherapy at all because providers were truly concerned about their ability to tolerate chemotherapy or potential for accelerated functional decline because of chemotherapy.[47] This may explain why ECOG score was not associated with dose reductions in the remaining women who did receive chemotherapy.
Our study had several additional limitations. While we collected detailed chemotherapy data, it is possible that some of the intended regimens were misclassified and thus misrepresented in the ARDI. We used extensive chart abstraction to mitigate this potential issue in ambiguous cases, but some misclassification may still have occurred.[22] We censored both the numerator and denominator of dose calculations when a woman discontinued a drug early, which resulted in people having an ARDI of 100% if they received their full, intended dose up to the point of discontinuation. Therefore, early discontinuation was not counted as a dose reduction and was a limitation of our analysis. Although we extensively adjusted for confounders in multivariable models, we may have residual confounding in our estimates from factors that we could not measure comprehensively in our cohort (e.g. performance score, Oncotypedx). We did not collect comprehensive data on physician experience or patient healthcare utilization – two additional factors that may affect chemotherapy dosing approaches in an older population. In addition, we collected income data at the neighborhood level, which may or may not accurately represent individual-level incomes. Income data were collected at one time point at one health system, which would not reflect changes over time due to inflation or other economic issues. We did not match individual comorbidities to specific chemotherapy drugs (e.g. cardiovascular disease and anthracyclines) to see whether adjustment for specific comorbidities altered analyses for specific drug classes. We only included data from two health plans and results may not be generalizable across the United States. Finally, we lacked data on social needs, social support, and physical function, all of which may impact chemotherapy dosing decisions and will be important to understand in future studies.
This study also had several strengths. The large study population and detailed chemotherapy data have potential to fill an evidence gap on appropriate chemotherapy regimens and dosing for older women with breast cancer. We were able to examine dose reductions at chemotherapy initiation and across the treatment course separately in a much larger, more contemporary patient sample than previous studies. We had the ability to control for comorbidities in multiple ways. Additionally, we conducted this research in an integrated healthcare setting that has been shown to be demographically similar to its underlying population, mitigating the potential for confounding by insurance coverage.[48]
We showed that older age at diagnosis, particularly 75+ years, was associated with increased prevalence of FCDP<90% and ARDI<90% in a large cohort of women with breast cancer, even after adjustment for comorbid disease. In general, associations were stronger between age and first cycle dose reductions than dose reductions throughout the chemotherapy course. This has clinical implications in that age alone appears to be a strong predictor of initial chemotherapy dose reductions beyond body size and comorbidity. However, some older women are not dose-reduced at the start of chemotherapy or across the regimen. This highlights the importance of conducting risk assessments before chemotherapy initiation in older women to gain a comprehensive understanding of women’s health and function beyond age. To further understand clinical impact, future research should focus on understanding the effects of lower FCDP and ARDI on breast cancer outcomes in contemporary populations of older women. We plan to evaluate associations between chemotherapy dosing and toxicity, recurrence, and mortality in our own work. Such studies should help understand potential tradeoffs between lowering chemotherapy dose to reduce toxicities in older women while also trying to maximize disease-free survival.
Supplementary Material
Acknowledgements:
The sponsors of this study had no role in the design of the study; the collection, analysis, and interpretation of the data; the writing of the manuscript; and the decision to submit the manuscript for publication. This study was previously presented at the American Society of Clinical Oncology Quality Care Symposium September 27-28, 2024, in San Francisco, CA.
Funding:
This work was supported by grants from the National Cancer Institute of the National Institutes of Health (grant numbers R37CA222793, U24CA171524, U01CA195565, R01CA214057, P30CA008748, R50CA211115, and P01CA154292), as well as the Geoffrey Beene Cancer Research Center at Memorial Sloan Kettering Cancer Center.
Footnotes
Conflict of interest statement: Dr. Bandera served as member of Pfizer’s Advisory Board to enhance minority participation in clinical trials (7/2021-8/2023). Dr. Liu served as member of Pfizer’s think tank on real-world evidence sponsored by Pfizer on 11/28/23 and 9/6/24. Dr. Liu received research funding unrelated to this current work from Genentech, AstraZeneca, Exact Sciences, Biotheranostics, and Beigene. Dr. Wang is a former employee of Daiichi Sankyo, Inc (6/2022-4/2023). Dr. Blinder, who is a JNCI Associate Editor and co-author on this paper, was not involved in the editorial review or decision to publish the manuscript. Other authors have no COI to declare.
Data availability statement:
Data contain potentially identifiable information (e.g. dates of diagnoses) that cannot be shared openly without appropriate human subjects approval and data use agreements. Individuals may request access to the data by emailing the principal investigator (kantore@mskcc.org) and obtaining necessary approvals and data use agreements from the participating study sites.
References
- 1.National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Breast Cancer Version 4.2024. https://www.nccn.org/professionals/physician_gls/pdf/breast.pdf. [Google Scholar]
- 2.Griggs JJ, Bohlke K, Balaban EP, et al. Appropriate Systemic Therapy Dosing for Obese Adult Patients With Cancer: ASCO Guideline Update. J Clin Oncol 2021;39(18):2037–2048. [DOI] [PubMed] [Google Scholar]
- 3.Griggs JJ, Mangu PB, Temin S, Lyman GH. Appropriate Chemotherapy Dosing for Obese Adult Patients With Cancer: American Society of Clinical Oncology Clinical Practice Guideline. J Oncol Pract 2012;8(4):e59–e61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Szlezinger K, Pogoda K, Jagiełło-Gruszfeld A, et al. Eligibility criteria in clinical trials in breast cancer: a cohort study. BMC Med 2023;21(1):240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Elkin EB, Hurria A, Mitra N, et al. Adjuvant chemotherapy and survival in older women with hormone receptor-negative breast cancer: assessing outcome in a population-based, observational cohort. J Clin Oncol 2006;24(18):2757–64. [DOI] [PubMed] [Google Scholar]
- 6.Giordano SH, Duan Z, Kuo YF, et al. Use and outcomes of adjuvant chemotherapy in older women with breast cancer. J Clin Oncol 2006;24(18):2750–6. [DOI] [PubMed] [Google Scholar]
- 7.Tamirisa N, Lin H, Shen Y, et al. Association of Chemotherapy With Survival in Elderly Patients With Multiple Comorbidities and Estrogen Receptor-Positive, Node-Positive Breast Cancer. JAMA Oncol 2020;6(10):1548–1554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wildiers H, Reiser M. Relative dose intensity of chemotherapy and its impact on outcomes in patients with early breast cancer or aggressive lymphoma. Crit Rev Oncol Hematol 2011;77(3):221–40. [DOI] [PubMed] [Google Scholar]
- 9.DeMichele A, Putt M, Zhang Y, et al. Older age predicts a decline in adjuvant chemotherapy recommendations for patients with breast carcinoma: evidence from a tertiary care cohort of chemotherapy-eligible patients. Cancer 2003;97(9):2150–9. [DOI] [PubMed] [Google Scholar]
- 10.Shayne M, Culakova E, Poniewierski MS, et al. Dose intensity and hematologic toxicity in older cancer patients receiving systemic chemotherapy. Cancer 2007;110(7):1611–20. [DOI] [PubMed] [Google Scholar]
- 11.Delgado-Ramos GM, Nasir SS, Wang J, Schwartzberg LS. Real-world evaluation of effectiveness and tolerance of chemotherapy for early-stage breast cancer in older women. Breast Cancer Res Treat 2020;182(2):247–258. [DOI] [PubMed] [Google Scholar]
- 12.Hurria A, Brogan K, Panageas KS, et al. Patterns of toxicity in older patients with breast cancer receiving adjuvant chemotherapy. Breast Cancer Res Treat 2005;92(2):151–6. [DOI] [PubMed] [Google Scholar]
- 13.Ladwa R, Kalas T, Pathmanathan S, et al. Maintaining Dose Intensity of Adjuvant Chemotherapy in Older Patients With Breast Cancer. Clin Breast Cancer 2018;18(5):e1181–e1187. [DOI] [PubMed] [Google Scholar]
- 14.Lyman GH, Dale DC, Tomita D, et al. A retrospective evaluation of chemotherapy dose intensity and supportive care for early-stage breast cancer in a curative setting. Breast Cancer Res Treat 2013;139(3):863–72. [DOI] [PubMed] [Google Scholar]
- 15.Nugent BD, Ren D, Bender CM, Rosenzweig M. The Impact of Age and Adjuvant Chemotherapy Modifications on Survival Among Black Women With Breast Cancer. Clin Breast Cancer 2019;19(4):254–258. [DOI] [PubMed] [Google Scholar]
- 16.Oladipo O, Coyle V, McAleer JJ, McKenna S. Achieving optimal dose intensity with adjuvant chemotherapy in elderly breast cancer patients: a 10-year retrospective study in a UK institution. Breast J 2012;18(1):16–22. [DOI] [PubMed] [Google Scholar]
- 17.Raza S, Welch S, Younus J. Relative dose intensity delivered to patients with early breast cancer: Canadian experience. Curr Oncol 2009;16(6):8–12. [Google Scholar]
- 18.Sedrak MS, Sun CL, Ji J, et al. Low-Intensity Adjuvant Chemotherapy for Breast Cancer in Older Women: Results From the Prospective Multicenter HOPE Trial. J Clin Oncol 2023;41(2):316–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Vaz-Luis I, Keating NL, Lin NU, et al. Duration and toxicity of adjuvant trastuzumab in older patients with early-stage breast cancer: a population-based study. J Clin Oncol 2014;32(9):927–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shayne M, Crawford J, Dale DC, et al. Predictors of reduced dose intensity in patients with early-stage breast cancer receiving adjuvant chemotherapy. Breast Cancer Res Treat 2006;100(3):255–62. [DOI] [PubMed] [Google Scholar]
- 21.Lyman GH, Dale DC, Crawford J. Incidence and predictors of low dose-intensity in adjuvant breast cancer chemotherapy: a nationwide study of community practices. J Clin Oncol 2003;21(24):4524–31. [DOI] [PubMed] [Google Scholar]
- 22.Bhimani J, O’Connell K, Ergas IJ, et al. Methodology for Using Real-World Data From Electronic Health Records to Assess Chemotherapy Administration in Women With Breast Cancer. JCO Clin Cancer Inform 2024;8:e2300209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bhimani J, O’Connell K, Persaud S, et al. The landscape of use of NCCN-guideline chemotherapy regimens in stage I-IIIA breast cancer in an integrated healthcare delivery system. Breast Cancer Res Treat 2024; 10.1007/s10549-024-07433-4 [doi] 10.1007/s10549-024-07433-4 [pii]. [DOI] [Google Scholar]
- 24.Ross TR, Ng D, Brown JS, et al. The HMO Research Network Virtual Data Warehouse: A Public Data Model to Support Collaboration. EGEMS (Washington, DC) 2014;2(1):1049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.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(5):373–83. [DOI] [PubMed] [Google Scholar]
- 26.Wang P, O’Connell K, Bhimani J, et al. Methodologic Approach to Defining Comorbidities in a Cohort of Patients With Cancer: An Example in the Optimal Breast Cancer Chemotherapy Dosing Study. JCO Clin Cancer Inform 2025;9:e2400231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kantor ED, O’Connell K, Ergas IJ, et al. Assessment of breast cancer chemotherapy dose reduction in an integrated healthcare delivery system. Breast Cancer Res Treat 2024;203(3):565–574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mosteller RD. Simplified calculation of body-surface area. N Engl J Med 1987;317(17):1098. [DOI] [PubMed] [Google Scholar]
- 29.Griggs JJ, Sorbero ME, Lyman GH. Undertreatment of obese women receiving breast cancer chemotherapy. Arch Intern Med 2005;165(11):1267–73. [DOI] [PubMed] [Google Scholar]
- 30.Bhimani J, Wang P, Gallagher GB, et al. Patient factors and modifications to intended chemotherapy for women with Stages I-IIIA breast cancer. Int J Cancer 2025;157(7):1342–1353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Trapani D. Adjuvant Chemotherapy in Older Women With Early Breast Cancer. J Clin Oncol 2023;41(9):1652–1658. [DOI] [PubMed] [Google Scholar]
- 32.Biganzoli L, Battisti NML, Wildiers H, et al. Updated recommendations regarding the management of older patients with breast cancer: a joint paper from the European Society of Breast Cancer Specialists (EUSOMA) and the International Society of Geriatric Oncology (SIOG). Lancet Oncol 2021;22(7):e327–e340. [DOI] [PubMed] [Google Scholar]
- 33.Mohile SG, Dale W, Somerfield MR, et al. Practical Assessment and Management of Vulnerabilities in Older Patients Receiving Chemotherapy: ASCO Guideline for Geriatric Oncology. J Clin Oncol 2018;36(22):2326–2347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Bretzel RL Jr., , Cameron R, Gustas M, et al. Dose intensity in early-stage breast cancer: a community practice experience. J Oncol Pract 2009;5(6):287–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Baltussen JC, de Glas NA, van Holstein Y, et al. Chemotherapy-Related Toxic Effects and Quality of Life and Physical Functioning in Older Patients. JAMA Netw Open 2023;6(10):e2339116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Older Adult Oncology. https://www.nccn.org/professionals/physician_gls/pdf/older_adult.pdf. [Google Scholar]
- 37.Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicenter study. J Clin Oncol 2011;29(25):3457–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Nishijima TF, Deal AM, Williams GR, et al. Chemotherapy Toxicity Risk Score for Treatment Decisions in Older Adults with Advanced Solid Tumors. Oncologist 2018;23(5):573–579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Bonadonna G, Valagussa P. Dose-response effect of adjuvant chemotherapy in breast cancer. N Engl J Med 1981;304(1):10–5. [DOI] [PubMed] [Google Scholar]
- 40.Chirivella I, Bermejo B, Insa A, et al. Optimal delivery of anthracycline-based chemotherapy in the adjuvant setting improves outcome of breast cancer patients. Breast Cancer Res Treat 2009;114(3):479–84. [DOI] [PubMed] [Google Scholar]
- 41.Flannery MA, Culakova E, Canin BE, et al. Understanding Treatment Tolerability in Older Adults With Cancer. J Clin Oncol 2021;39(19):2150–2163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hryniuk W, Levine MN. Analysis of dose intensity for adjuvant chemotherapy trials in stage II breast cancer. J Clin Oncol 1986;4(8):1162–70. [DOI] [PubMed] [Google Scholar]
- 43.Qi W, Wang X, Gan L, et al. The effect of reduced RDI of chemotherapy on the outcome of breast cancer patients. Sci Rep 2020;10(1):13241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Wood WC, Budman DR, Korzun AH, et al. Dose and dose intensity of adjuvant chemotherapy for stage II, node-positive breast carcinoma. N Engl J Med 1994;330(18):1253–9. [DOI] [PubMed] [Google Scholar]
- 45.Extermann M, Boler I, Reich RR, et al. Predicting the risk of chemotherapy toxicity in older patients: the Chemotherapy Risk Assessment Scale for High-Age Patients (CRASH) score. Cancer 2012;118(13):3377–86. [DOI] [PubMed] [Google Scholar]
- 46.Hurria A, Mohile S, Gajra A, et al. Validation of a Prediction Tool for Chemotherapy Toxicity in Older Adults With Cancer. J Clin Oncol 2016;34(20):2366–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Sedrak MS, Sun CL, Bae M, et al. Functional decline in older breast cancer survivors treated with and without chemotherapy and non-cancer controls: results from the Hurria Older PatiEnts (HOPE) prospective study. J Cancer Surviv 2024;18(4):1131–1143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Davis AC, Voelkel JL, Remmers CL, et al. Comparing Kaiser Permanente Members to the General Population: Implications for Generalizability of Research. Perm J 2023;27(2):87–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Karter AJ, Ferrara A, Liu JY, et al. Ethnic disparities in diabetic complications in an insured population. JAMA 2002;287(19):2519–27. [DOI] [PubMed] [Google Scholar]
- 50.Greenlee H, Iribarren C, Rana JS, et al. Risk of Cardiovascular Disease in Women With and Without Breast Cancer: The Pathways Heart Study. J Clin Oncol 2022;40(15):1647–1658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Kwan ML, Cheng RK, Iribarren C, et al. Risk of Cardiometabolic Risk Factors in Women With and Without a History of Breast Cancer: The Pathways Heart Study. J Clin Oncol 2022;40(15):1635–1646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16(1):31–41. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data contain potentially identifiable information (e.g. dates of diagnoses) that cannot be shared openly without appropriate human subjects approval and data use agreements. Individuals may request access to the data by emailing the principal investigator (kantore@mskcc.org) and obtaining necessary approvals and data use agreements from the participating study sites.