Sequence of therapy impact on older women with comorbidities and triple-negative or HER2-positive breast cancer

Prior studies have demonstrated age-related disparities in older patients with breast cancer, resulting in undertreatment with a detrimental impact on survival compared to younger counterparts¹. Older patients with comorbidities can pose a challenge for cancer care providers who must balance baseline life expectancy with the potentially debilitating quality of life consequences due to toxicity from standard treatment recommendations. This is particularly complicated for the approximately 25–30% of older breast cancer patients with HER2 positive or triple-negative disease who typically require surgical intervention and systemic therapy². Given that older patients are frequently underrepresented in clinical trials evaluating systemic therapy, there is a lack of robust evidence from prospective clinical trials to support clinical decisions for these patients, and clinicians must estimate a patient’s ability to tolerate treatment when offering recommendations.

In a population-based study evaluating the treatment of women 50 and older with stage I-III triple-negative breast cancer (TNBC), only 58% received surgery and chemotherapy and this rate decreased with increasing age, more so with regard to the use of adjuvant chemotherapy³. According to treatment guidelines, systemic therapy and surgical intervention should be offered on the basis of functional status and a comprehensive geriatric assessment rather than on chronological age⁴. Based on historical data from the NSABP B18 and B27 trials, there is no difference in survival between neoadjuvant or adjuvant chemotherapy approaches and these findings persist even in the utilization of modern systemic therapy⁵. In a more contemporary cohort of patients, several studies have demonstrated advantages to systemic therapy delivered in the neoadjuvant setting including downstaging of disease to allow for less invasive surgery of the breast and axilla without compromising oncologic outcome^6,7,8,9. Another reason to start with chemotherapy is to tailor treatment in the adjuvant setting based on the presence or absence of residual disease, however data to support this rationale for neoadjuvant treatment in older patients with comorbidities is limited⁹.

While the standard for patients with HER2 positive or triple-negative breast cancer is to receive both surgery and chemotherapy, complications from either treatment option could potentially delay or preclude next steps in care, resulting in patients who are either unable to undergo surgery after neoadjuvant chemotherapy or treatment with adjuvant systemic therapy after surgery. Additionally, prior studies demonstrate that delays to surgery or initiation of adjuvant chemotherapy adversely impact survival, even in older patients^10,11. We hypothesized that even in the older population of patients with comorbidities, patients who receive both surgery and chemotherapy may achieve a better outcome. We sought to evaluate rates of chemotherapy and surgery in older breast cancer patients with comorbidities, determine whether an adjuvant or neoadjuvant systemic therapy approach led to a higher likelihood of receiving both modalities of treatment, and assess for any differences in survival between these groups.

Patient characteristics

Of a total of 2,981,732 patients in the data set, 2911 patients met the inclusion criteria (mean age [SD], 78.5 [5.7] years). The majority of patients had cT1c/T2 disease (91.5%) and were cN0 (76.2%). The majority of patients underwent surgery (n = 2544) vs chemotherapy (n = 367) as the initial treatment. Of the patients treated with chemotherapy first, 77.9% (n = 286) went on to surgery. Of the patients who underwent surgery first, only 36.1% (n = 917) received adjuvant chemotherapy (Fig. 1). Overall distributions of patient and disease characteristics across treatment options

Consort diagram.

Full size image

Factors driving differences between initial treatment

The patients who were most likely to undergo surgery first were older (mean age [SD], 78.9 [5.7] vs 75.9 [4.4], p < 0.001) with an earlier stage of disease (cT1c 39.5% vs 16.4% & cN0 79.5% vs 53.4%). Rates of chemotherapy as the initial treatment increased over time and the degree of comorbidity was not significantly associated with the starting treatment (Table 1).

Patients who received chemotherapy as initial treatment

Among patients who started with chemotherapy, the majority of patients went on to surgery (n = 286, 77.9%). Patients who were clinically node negative were more likely to undergo surgery after chemotherapy compared to those who were clinically node positive (Table 2). Other factors, such as tumor size, year of diagnosis, status of hormonal receptors, Her2 status, and comorbidity were not significantly associated with receipt of surgery after chemotherapy.

Patients who received surgery as the initial treatment

Among patients who underwent surgery first (n = 2544), 36% were treated with adjuvant chemotherapy. Those who received adjuvant chemotherapy were younger (mean age 75 vs 81 years), node positive (cN1-3: 27.0% vs 16.8%), Her2 positive (49.9% vs 42.9%), with moderate comorbidity (CCDM 2 69.9% vs 64.4%) compared to those treated with surgery alone (all p values < 0.01, Table 3). Tumor T stage, year of diagnosis, and status of hormonal receptors were not significantly associated with receipt of adjuvant chemotherapy.

Patients who received both modalities of treatment

A total of 1203 patients received complete treatment defined as chemotherapy followed by surgery or surgery followed by chemotherapy. In this group, compared to patients who underwent surgery first, those treated with chemotherapy first were more likely to have locally advanced stage disease (cT3: 15.7% vs 6% & cN1-3 42.0% vs 27.0%, p < 0.001) and were more likely to have been diagnosed in later years (2015–2017: 67.8% vs 49.7%) (all p values < 0.001, Table 4). Overall distributions of patient and disease characteristics across treatment options listed as row percentages are included in Supplementary Table 1.

Overall survival

There were 585 patients who had missing survival information. Among the 2326 patients with survival data, 47.3% (n = 1100) of patients died and 52.7% (n = 1226) of patients were alive at the last follow up. The median follow-up time was 49.3 months (95% CI = 47.1–50.7 months). The Kaplan Meier curves demonstrate survival outcomes by treatment type for all patients (Fig. 2a) and stratified by clinical T and N stage (Fig. 2b, c). Patients who received both modalities of treatment (neoadjuvant chemotherapy+surgery or surgery+adjuvant chemotherapy) had a significantly better survival compared to patients who received chemotherapy only or surgery only. With surgery alone as the reference group, these findings persisted in a multivariable Cox proportional hazard model adjusting for age, clinical T stage, clinical N stage, hormone receptors status, and comorbidity (chemotherapy followed with surgery: HR = 0.47, 95% CI = 0.36 to 0.61 and surgery followed with chemotherapy: HR = 0.55 95% CI of HR = 0.47 to 0.64). Patients who underwent chemotherapy alone demonstrated a significantly worse survival outcome compared to those who underwent surgery alone (HR = 1.46, 95% CI of HR = 1.05 to 2.04) (Table 5). Additional analysis comparing all patients who underwent surgery first (±adjuvant systemic therapy, n = 995) to patients treated with systemic therapy first (±surgery to follow, n = 274) demonstrated no difference in survival in both univariate (Fig. 2d) and multivariate (Table 6) analyses.

a All patients; b stratified by T stage; c stratified by nodal stage; d stratified by starting treatment.

Full size image

Women with HER2 positive or TNBC are often recommended for systemic therapy in addition to surgery. Approximately 10–15% of women with breast cancer over the age of 65 are diagnosed with HER2-positive disease². The International Society of Geriatric Oncology reviewed the literature on anti-HER2 treatments for patients over 70 and highlighted the lack of randomized control trial data in this population, as only 16% of patients in the key studies of adjuvant trastuzamab were over 60 years of age¹². Approximately 12–17% of older patients are diagnosed with TNBC, and similar to the HER2 population, representation of the older population with comorbidities in clinical trials for this population is scarce¹³. In light of the increasing population of older patients with breast cancer, understanding current treatment patterns and survival benefits of multimodal treatment especially among those with comorbidities is essential.

To our knowledge, ours is the first study to evaluate receipt of chemotherapy and surgery in a population of older breast cancer patients with comorbidities. We found that of the 2911 patients who met the inclusion criteria, the majority underwent surgery (n = 2544) as the initial treatment. Of the patients treated with chemotherapy first, 77.9% proceeded to surgery and of the patients treated with surgery first, 36.1% received adjuvant chemotherapy. While there was no difference in survival between patients who received adjuvant vs neoadjuvant chemotherapy, patients who received both modalities of treatment had significantly improved survival compared to patients who received single treatment.

Per National Comprehensive Cancer Network guidelines for women > 70, recommendations for adjuvant chemotherapy should be tailored based on estimated benefit vs impact on quality of life secondary to existing comorbidities and ability to tolerate treatment¹⁴. Our findings of a majority of older patients undergoing surgery as the initial treatment aligns with prior studies^13,15. We found that patients who were most likely to undergo surgery first were older with earlier stage of disease (cT1c (39.5% vs 16.4%) & cN0 (79.5% vs 53.4%)). American Society of Clinical Oncology (ASCO) guidelines recommend against neoadjuvant therapy in patients with cT1a-bN0 disease, but for patients with cT1c disease, treatment recommendations are at the discretion of the multidisciplinary team¹⁶. Especially for older patients with early stage disease, an upfront surgery approach delineating true extent of disease may allow for potential de-escalation of systemic therapy.

Among patients who underwent surgery first, we found that those who were healthier and younger with positive nodes were more likely to receive adjuvant chemotherapy compared to clinically node-negative patients who were older with multiple comorbidities. This is similar to findings from a retrospective study by Giordano et al. evaluating outcomes of adjuvant chemotherapy in older women with breast cancer. Increased use of adjuvant chemotherapy was seen among estrogen receptor-negative and lymph node-positive patients with comorbidity scores of 0-1, and use significantly diminished with increasing age¹⁷. For the patients included in our study, the 10-year life expectancy rates range from 21% for patients 70–79 years old with a Charlson comorbidity score of 2 vs 0% for patients ≥79 with a comorbidity score of 3. Perceived lack of benefit of toxic therapy that could impact the quality of life for patients with already limited life expectancy may have resulted in physicians avoiding or patients refusing systemic therapy in this setting.

Similarly, we found that among patients who received chemotherapy first, younger patients with node-negative disease were more likely to have surgery compared to those who were older or had positive nodes. Patients with HER2 positive or TNBC, clinically node-positive disease are commonly treated with up front systemic therapy despite the lack of randomized trial data demonstrating a survival difference between neoadjuvant vs adjuvant chemotherapy. In the approximately 20% of patients who did not proceed to surgery after chemotherapy, it is possible that complications from treatment or deconditioning precluded surgical intervention. In a study by Williams et al. evaluating neoadjuvant chemotherapy in elderly patients, 78% of patients > 70 were unable to complete their prescribed neoadjuvant chemotherapy regimen. However, even with a de-escalated approach and the inability to complete neoadjuvant treatment, in the modern era, patients > 70 experience high rates of breast and axillary downstaging, similar to their younger counterparts¹⁸. Consistent with prior studies¹⁵, we found that of the patients who received both modalities of treatment, women with advanced stage of disease treated in recent years were more likely to receive neoadjuvant compared to adjuvant chemotherapy.

When stratified by stage, patients who were able to undergo both modalities of treatment had significantly better survival compared to patients who received a single treatment (chemotherapy only or surgery only). Given our inability to account for unmeasured variables that impact decisions to proceed with treatment, these findings could be attributed in part to selection bias. At the very least, our data suggest that physicians are offering treatment to patients whom they consider clinically healthy enough to tolerate surgery and neoadjuvant or adjuvant systemic therapy. In our cohort of older patients who might otherwise be considered frail secondary to comorbidities, these findings are significant especially in the setting of recent advances in systemic therapy demonstrating improved survival. In 2021, ASCO published guidelines on neoadjuvant chemotherapy in which patients with triple-negative or HER2-positive breast cancer were strongly recommended to proceed with neoadjuvant chemotherapy based on data from the KEYNOTE 522 and KATHERINE trials.^8,9,16. Since older, frail patients were not well represented in these trials, our findings suggest that this approach can be feasible in very carefully selected patients and efforts should be focused on geriatric directed support. In a randomized controlled trial published by Li et al. in JAMA Oncology, patients >65 with a cancer diagnosis were randomized to assessment by a geriatric-focused multidisciplinary team vs self-reported geriatric assessments which were interpreted by the oncologist prior to the initiation of chemotherapy. Both assessments prompted referrals to specialties such as physical therapy or nutrition to optimize patients and found that the multidisciplinary input and interventions significantly reduced grade 3 or higher chemotherapy toxic effects compared to self-reported measures¹⁹. Implementation of geriatric-focused care, especially in patients requiring multimodal treatment, will be increasingly important as the aging population continues to rise.

Our analysis has several limitations. Given that the NCDB is a hospital-based registry including facilities accredited by the Commission on Cancer, only 30% of hospitals in the United States are included, which may not represent the general population of breast cancer patients. Especially in the treatment of older, frail patients, there are unmeasured variables that cannot be accounted for such as patient preference, which may have impacted treatment decisions and thus leading to selection bias. Additionally, physicians may recommend against treatment when taking into account the functional or the overall health status of a patient, which cannot be fully captured by the Charlson Comorbidity Score, as found in prior studies²⁰. Alternatively, patients in a better baseline state of health may have been offered both chemotherapy and surgery given a perceived overall benefit when weighed against risk of complications from the treatment itself. Patient-reported data are not included in the NCDB and may have impacted treatment decisions and outcomes, particularly in this group of vulnerable patients. Therefore, given the retrospective nature of the study and especially among older patients, it is difficult to delineate whether treatment or baseline health status contributed to the observed improved overall survival. Although 65 is also used to define older age, the risk of frailty increases by at least 4.1% for each year of age after 70⁴. In order to create a homogenous cohort and mitigate selection bias, we limited the cohort to those patients > 70. While the NCDB documents receipt of systemic therapy and timing with respect to surgery, completion of treatment or the number of cycles of treatment received are not recorded. We cannot accurately account for patients who may have had to stop treatments earlier than anticipated secondary to toxicities. The exact systemic treatment regimen is not recorded and we cannot account for those patients who may have had a de-escalated approach to therapy or dose reductions in anticipation of chemotherapy-related complications. While we attempted to limit analysis to women with reported comorbidity scores of 2 or 3, these confounders may have skewed the population to a healthier cohort that could tolerate treatment with improved survival outcomes based on clinical judgement. The NCDB does not include breast cancer specific survival data, however, in this population, overall survival may serve as a surrogate for effectiveness of treatment in the context of competing comorbidities. All patients with missing survival information were diagnosed in 2017 for whom the NCDB did not provide vital status at 5 years and we could not include those patients in the Kaplan-Meier estimates. Geriatric parameters such as caregiver support and cognitive decline are not included as variables in the database and may account for reasons why systemic therapy was either declined by patients or not offered by providers.

In older triple-negative or HER2-positive breast cancer patients with comorbidities, we found that receipt of chemotherapy and surgery was associated with improved survival. Patients treated with systemic therapy first were almost twice as likely to proceed with surgery compared to patients treated with surgery first who then went on to receive adjuvant chemotherapy. While we aimed to identify which strategy optimized for receipt of both chemotherapy and surgery, it should be noted that as a single modality, patients treated with surgery alone fare better than those treated with chemotherapy alone. There is a small likelihood that patients treated with systemic therapy first may not undergo surgery, and this consideration in conjunction with patient preferences on quality of life should all be included as part of shared decision making.

Patient population

The NCDB, a joint project of the Commission on Cancer of the American College of Surgeons and the American Cancer Society, is a hospital-based registry that captures approximately 70% of newly diagnosed cancer cases in the United States and Puerto Rico and draws data from >1500 commission-accredited cancer programs. Originating in 1989, the NCDB now contains approximately 34 million records. Data registries contain patient characteristics, cancer stage, tumor histologic characteristics, type of first treatment administered, and outcomes. The American College of Surgeons and the Commission on Cancer have not verified and are not responsible for the analytic or statistical methods used, or for the conclusions drawn from these data by the investigators of this study. Because patient-level identifiers are not available to NCDB users, this study was exempted from institutional review board review and approval. The 2017 NCDB Participant User File was used for this study.

Patients aged >70 years diagnosed with clinical T1c-3/N0-3, HER2 positive or triple negative invasive breast cancer from 2010 through 2017 were selected from the NCDB. Patients were included if they had a Charlson comorbidity score of 2 or 3, and those with a comorbidity score of 0 or 1 were excluded. We included patients treated with chemotherapy or surgery at the primary site within 9 months of diagnosis to account for those patients being treated with curative intent and not for palliation. Patients who received chemotherapy prior to surgery were defined as neoadjuvant and patients treated with chemotherapy within 6 months after surgery were defined as adjuvant. Patients were excluded if they were diagnosed with metastatic (M1) disease. Patients who received neither chemotherapy nor surgery and patients who had missing or unknown surgery or chemotherapy data were excluded. Patients with hormone receptor-positive/HER2 negative disease and those who had missing hormone receptor/HER2 status were excluded. The start date of 2010 was the first year that data on HER2 status were captured in the NCDB.

Charlson/Deyo comorbidity score

The Charlson/Deyo value is a weighted score derived from the sum of the scores for comorbid conditions. Comorbid conditions with a score of 1 include myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatologic disease, peptic ulcer disease, mild liver disease, or diabetes. Comorbid conditions with a score of 2 include a single diagnosis of diabetes with complications, hemiplegia/paraplegia, or renal disease. Comorbid conditions with a score of 3 include moderate-to-severe liver disease or AIDS. Unique to the NCDB classification, a cumulative score ≥3 is classified as a Charlson/Deyo score of 3, which is not how the score was originally described. Although a patient’s cancer diagnosis is included in a true Charlson/Deyo comorbidity score, the cancer diagnosis is not included in the NCDB’s reporting of this score. In addition, patients with a score of 0 may have comorbidities not included among those listed.

Outcome measures

The aim of the analysis was to evaluate outcomes in older breast cancer patients with comorbidities in whom chemotherapy and surgery are indicated. Patients who met inclusion criteria were classified by the initial treatment, chemotherapy or surgery. Within the chemotherapy group, patients were further subdivided into those who went on to surgery vs no surgery; and within the surgery group, patients were subdivided into those who received adjuvant chemotherapy vs no further systemic therapy. Factors driving differences between treatment groups and survival differences stratified by stage of disease were evaluated as secondary outcome measures between the 4 groups—neoadjuvant chemotherapy+surgery, surgery+adjuvant chemotherapy, chemotherapy alone, and surgery alone.

Statistical analyses

Summary statistics were provided, including mean, standard deviation, median, and range for continuous variables. The frequency counts and percentages for categorical variables were summarized. A chi-square test was used to evaluate associations between categorical variables. The Wilcoxon rank sum test or t-test was used to evaluate differences in continuous variables between different patient groups. Kaplan-Meier curves were estimated by prognostic factors of interest. The Log-rank test was used to test the difference in survival distributions among the subgroups. Univariate Cox proportional hazard models were used to determine the effects of prognostic factors on survival distributions. A Multivariable Cox proportional hazards model was used to evaluate the association between overall survival and treatment after adjusting for age, clinical T stage, clinical N stage, hormonal receptors status (ER and PR), HER2 status, year of diagnosis, and comorbidity status.

All tests were two-sided. P-values less than 0.05 were considered statistically significant. All analyses were conducted using the SAS (version 9.4; SAS Institute, Cary, NC) and S-plus (version 8.04, TIBCO Software Inc., Palto Alto, CA) statistical software.