Surgical Delay-Associated Mortality Risk Varies by Subtype in Loco-Regional Breast Cancer Patients in SEER-Medicare

Substantial evidence supports that delay of surgery after breast cancer diagnosis is associated with increased mortality risk, leading to the introduction of a new Commission on Cancer quality measure for receipt of surgery within 60 days of diagnosis for non-neoadjuvant patients. Breast cancer subtype is a critical prognostic factor and determines treatment options; however, it remains unknown whether surgical delay-associated breast cancer-specific mortality (BCSM) risk differs by subtype. This retrospective cohort study aimed to assess whether the impact of delayed surgery on survival varies by subtype (hormone [HR]+/HER2−, HR−/HER2−, and HER2+) in patients with loco-regional breast cancer who received surgery as their first treatment between 2010–2017 using the SEER-Medicare. Continuous time to surgery from diagnostic biopsy (TTS; days) in reference to TTS = 30 days. BCSM were evaluated as flexibly dependent on continuous time (days) to surgery from diagnosis (TTS) using Cox proportional hazards and Fine and Gray competing-risk regression models, respectively, by HR status. Inverse propensity score-weighting was used to adjust for demographic, clinical, and treatment variables impacting TTS. Adjusted BCSM risk grew with increasing TTS across all subtypes, however, the pattern and extent of the association varied. HR+/HER2− patients exhibited the most pronounced increase in BCSM risk associated with TTS, with approximately exponential growth after 42 days, with adjusted subdistribution hazard ratios (sHR) of 1.21 (95% CI: 1.06–1.37) at TTS = 60 days, 1.79 (95% CI: 1.40–2.29) at TTS = 90 days, and 2.83 (95% CI: 1.76–4.55) at TTS = 120 days. In contrast, both HER2 + and HR−/HER2− patients showed slower, approximately linear growth in sHR, although non-significant in HR−HER2−.


Introduction
A recent meta-analysis found that treatment delay is a critical factor contributing to mortality risk in multiple types of solid tumors, including breast cancer 1 .Extended time to surgery from diagnostic biopsy (TTS) in particular has negative survival implications in breast cancer [2][3][4][5][6] and studies have noted that both the frequency and length of delay are increasing [7][8][9] .Thus, in 2022 the Commission on Cancer (CoC) introduced a new quality measure for accredited facilities of receipt of surgery within 60 days of diagnosis for Stage I-III patients in the non-neoadjuvant setting 10 in order to address the negative survival impact of surgical delay.Breast cancer subtype is a critical prognostic factor that provides critical therapy-relevant information, since the underlying biologic differences re ect tumor behavior as well as treatment options 11 .Thus, questions remain as to whether all patients are predisposed to an equal level of mortality risk posed by extended TTS or whether risk differs with intrinsic properties such as subtype.
Only a few studies have explored outcomes in relation to TTS by subtype.In a retrospective study of 351,087 Stage I-III breast cancer patients, Mateo et al. reported a consistent 10% increase in risk of overall mortality each subsequent month after the rst 30 days post-diagnosis in a cohort of 351,087 Stage I-III patients that did not change based on subtype 12 .In contrast, in a cohort of 90,405 T1N0 breast cancer patients who received breast conserving surgery, Hills et al. found that risk of TTS-associated disease progression was con ned to only hormone receptor (HR) + disease, with 18% and 47% higher likelihood of tumor size progression for patients who waited between 61 and 90 days and over 90 days, respectively 13 .
While research thus far has shed critical light on the steady increase of surgical delay among breast cancer patients [7][8][9] and the risk it presents for mortality outcomes [1][2][3] , the commonly adopted approach of examining TTS as xed monthly or bi-monthly increments hinders vital understanding of how breast cancer-speci c mortality (BCSM) risk may exibly change with increasing TTS by subtype.Thus, the key objective of this study was to gain a comprehensive picture of whether TTS differentially impacts BCSM by subtype through exible modeling of daily estimates of risk in women with loco-regional breast cancer in the non-neoadjuvant setting using the Surveillance, Epidemiology, and End Results (SEER)-Medicare database.

Cohort
A retrospective cohort of women diagnosed by needle or incisional biopsy with loco-regional invasive, non-in ammatory breast cancer between 2010-2017 in the SEER-Medicare database who received surgery as their rst treatment was selected.The SEER-Medicare linked database combines Medicare Parts A and B claims with clinical and outcome data from SEER cancer registries 14 .All data were deidenti ed and met the criteria for exempt review by the University of Oklahoma Health Sciences Center Institutional Review Board (IRB7446).Patients who had HMO coverage or did not have continuous Part A and B coverage for at least one year prior through one year after diagnosis were excluded due to the inability to accurately ascertain diagnostic or treatment claims for the primary course of treatment.
Patients who received surgery within 7 days of diagnostic biopsy were excluded since the time required for pathologic molecular diagnosis commonly takes up to one week 15 .Additionally, patients that did not receive surgery until over ≥ 120 after diagnosis, had a time of death less than one year, SEER reported follow-up shorter than TTS, non-de nitive initial surgery (i.e.re-excisions), prior cancer diagnosis, nonlocoregional disease (i.e. in situ, regional direct extension, or distant metastatic spread), or missing information were excluded (Fig. 1).

Exposure
The primary exposure, time-to-surgery (TTS), was de ned as days from date of diagnostic biopsy to date of surgery.

Outcome
Breast cancer-speci c mortality (BCSM) in the presence of competing events (i.e.death from other causes) was assessed and survival times were calculated from the date of surgery to death or last contact (censored).
Race/ethnicity was categorized as non-Hispanic Black (Black), other (Asian, Hispanic, Paci c Islander, American Indian/Eskimo/Aleutian), or non-Hispanic White (White).The Charlson Comorbidity Index was calculated for each patient using the SEER-Medicare developed Comorbidity SAS Macro (2021 version) to search for relevant claims in the year prior to diagnosis, and classi ed as 0, 1, or ≥ 2. Education (% of residents without high school degree) and residential median income were based on census tract level information from the 2010 U.S. Census and the patient's census tract of residence at the time of diagnosis.Histology was categorized as ductal, lobular, or other by ICD-O-3 codes (Appendix 1).The HCPCS, ICD-9, and ICD-10 codes used to classify diagnosis, surgery, and adjuvant therapies in the Medicare claims are listed in Appendix 2. Surgery type was classi ed as breast conserving, mastectomy, or mastectomy with immediate reconstruction.

Statistical Methods
Time to death as a function of TTS was analyzed separately by subtype using Fine-Gray competing risk models for BCSM.All models were adjusted using inverse propensity score weights (IPW) to account for potential imbalances in covariates associated with TTS [16][17][18] .Covariate balancing propensity scores were computed using the R package "CBPS" with socio-demographic and clinical factors as predictors and logtransformed TTS as the response variable 18 .Pre-/post-weighting balance was assessed for each model using Love plots.Final survival models were adjusted by normalized IPW, with extreme weights beyond the 95th percentile winsorized, along with receipt of adjuvant radiation or systemic therapy, comorbidity score, and in HER2 + patients, hormone receptor status.B-splines were used to exibly model the subdistribution hazard of mortality as nonparametric functions of TTS.Subdistribution hazard ratio (sHR) estimates were calculated using TTS = 30 days as the reference point, since it is commonly used as the upper limit of the reference in categorical TTS studies 2,6,12 .Simultaneous 95% con dence intervals (CI) at each TTS point were computed using the Scheffe method.The association between TTS and sHR was considered signi cant when the simultaneous 95% CI did not include a subdistribution hazard ratio of 1.To provide estimates of the BCSM incidence at TTS of 30, 60, 90, and 120 days, the adjusted cumulative incidence function was derived from the Fine-Gray model conditioned on the subgroup of patients with most the most common characteristics.All statistical analyses were conducted using SAS (version 9.4; Cary, NC) and R software (version 4.0.4), and graphs were generated using JMP Pro 15.2.0 (SAS; Cary, NC).

Discussion
Our study is the rst to provide dynamic insight into subtype-speci c differential patterns of BCSM risk associated with TTS.Through a robust statistical approach, capturing exible daily estimates of BCSM risk, rather than broadly grouped, discrete TTS intervals 2,6,12,13 , we found that patients with HR+/HER2breast cancer experienced rapid exponential trajectory of TTS-associated mortality risk, as opposed to the slower linear growth seen for patients with HER2 + and HR-/HER2-breast cancer.The adjusted TTSassociated BCSM cumulative incidence in HR+/HER2-patients was re ected by increasingly larger gaps, with approximately 5% higher 8-year mortality in patients with TTS = 120 days compared to 30 days (3.0% vs. 8.4%).This is important especially since patients with TTS beyond 60 days are continually exposed to growing risk across the follow-up period.Our data showed growing BCSM risk in all subtypes with increasing TTS, albeit non-signi cant in HR-/HER2-patients, emphasizing the bene t of timely surgery after biopsy diagnosis, consistent with the CoC's recent quality measure for surgery within 60 days of diagnosis for non-neoadjuvant Stage I-III patients 10 .
The observed difference by subtype was unexpected based on the prevailing view of favorable prognosis for HR+/HER2-disease and less favorable prognosis for HER2 + and HR-/HER2-breast cancer 19 .Death from breast cancer is primarily the result of metastatic outgrowth of disseminated cells in distant organs, thus provoking the question as to whether metastatic dissemination occurs differently by subtype over the duration of TTS.Breast cancer dissemination is proposed to occur by two key mechanisms: 1) linear progression in which cancer cells and the tumor microenvironment gradually acquire a phenotype conducive to metastasis or 2) parallel progression through early dissemination of inherently metastatic cancer cells in response to an angiogenic switch 20,21 .In this light, TTS may provide extra time for progressive phenotypic or microenvironment changes to occur.Mathematical simulation indicates that the natural breast tumor growth rate is relatively slow, taking approximately 1.7 years for a 1 cm tumor to double in size 22 .Thus, tumor size upstaging of T1N0M0 patients 13 as well as exponential BCSM risk in HR+/HER2-after a brief 42-day period following diagnosis may suggest the possibility of accelerated disease progression beyond the rate of natural linear progression after diagnosis 13 .On the other hand, TTS-associated mortality could re ect an intrinsic metastatic subpopulation within the highly heterogeneous HR+/HER2-subtype, since both Oncotype recurrence risk score and percentage of estrogen receptor/progesterone receptor positivity contribute to substantial prognostic diversity even in early stage disease [23][24][25] .Such differences should be further investigated to determine their role in TTSassociated outcomes.Along similar lines, adjuvant chemotherapy may potentially impact TTS-associated BCSM risk by attenuating the likelihood that cancer cells disseminated during the diagnosisto-surgery timeframe survive.Indeed, only 21% of HR+/HER2-patients in our cohort received systemic chemotherapy due to expected insu cient response 26 , as compared to 68% and 56%, in HER2 + and HR-/HER2-breast cancer, respectively (Table 1).Since subtype as well as Oncotype score are critical determinants of treatment course, we are unable to disentangle the potential effects of chemotherapy in the current study.Finally, a biologic basis for the observed increase in BCSM with longer surgical delays remains unknown and further studies are warranted, particularly to shed light on the high burden of TTSassociated mortality risk in HR+/HER2-patients.
The primary strength of this study lies in robust statistical modelling to identify the dynamic nature of mortality risk associated with TTS.Integrating the novel approach for propensity score calculation for non-parametric continuous variables developed by Imai and colleagues [16][17][18] to adjust for sociodemographic and clinical characteristics, Fine-Gray competing risk survival analysis delineated non-linear TTS-associated risk of BCSM in HR+/HER2-patients, which may not be visible when xed effect sizes across TTS monthly or bi-monthly increments are compared.Limitations of this study include the potential for confounders outside the scope of, or with incomplete reporting in, the databases (e.g.clinical staging, Ki67 status).Additionally, the cohort is composed of an elderly population with Medicare coverage and the sample size of patients with HER2 + or HR-/HER2-subtype, which typically make up a greater proportion of younger patients 27 , is limited, resulting large con dence bounds.Thus, further studies of differences in TTS-associated BCSM risk by subtype in younger women as well as across more diverse insurance types is recommended.Lastly, a plausible biologic mechanism to adequately explain such rapid mortality risk increases within a relatively short period of surgical delay following diagnosis is currently unavailable.Further study to elucidate the underlying reasons for TTS-associated mortality risk, including delay of adjuvant therapies, biologic changes 28 , or natural disease progression, is warranted.

Conclusion
This study identi ed that the association between surgical delay and BCSM risk varies by tumor subtype, with a rapid exponential increase in risk in HR+/HER2-patients and lesser linear increases in patients with HER2 + or HR-/HER2-breast cancer.Prevention of surgical delays holds the potential to improve survival outcomes for patients with locoregional breast cancer across multiple tumor subtypes.

Abbreviations
Figures

Table 1 .
Distribution of TTS by cohort demographic and clinical characteristics within subtypes

Table 2
Adjusted risk of breast cancer-speci c mortality at weekly points of TTS by subtype