Missing Components of Receptor Status Among Women With Invasive Breast Cancer

Key Points Question Do missing components of receptor status vary in the US by patient or disease characteristics? Findings In this cross-sectional study of 321 913 patients with invasive breast cancer, the rate of missing components of receptor status was higher in older women, Black women, and women from rural areas, as well as those with unstaged cancers, cases reported from facilities other than a hospital, and unknown insurance status. Meaning The results of this cross-sectional study suggest that there is a need to target some populations to ensure all individuals with invasive breast cancer are getting the testing needed to get the most accurate prognosis and treatment plan.


Introduction
An estimated 297 790 women will receive a diagnosis of breast cancer in the US in 2023, and nearly 43 170 women will die of the disease. 1 Assessment of possible expression of estrogen receptor (ER), progesterone receptor (PR), and erb-B2 receptor tyrosine kinase 2 (ERBB2 [formerly HER2]) is standard in identifying the most suitable breast cancer treatment and determining prognosis.Breast cancer is grouped into 3 subtypes: hormone receptor positive and ERBB2 negative (approximately 65%), ERBB2 amplified (approximately 5%-20%), and triple negative (approximately 5%). 2 The characterization of subtypes has changed how patients with breast cancer are treated. 3Because of this, testing for ER, PR, and ERBB2 expression is the recommended standard of care for all of those with a diagnosis of invasive breast cancer. 4porting of ER and PR data to the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) Program began in 1992.The proportion of cases with unknown ER and PR dropped considerably in 2003 when this information became required by Commission on Cancerapproved hospitals. 3Collection of ERBB2 data started in 2010, with the proportion of unknown ERBB2 decreasing steadily until 2016 (Figure).From 2012 through 2016, the proportion of missing ERBB2 was approximately 63% of higher than the proportion missing ER and PR (Figure).Missing ER, PR, and/or ERBB2 status is associated with biases in estimating SEER incidence rates and survival probabilities for specific breast cancer subtypes.Howlader et al 3 reported that ERBB2 is not missing entirely at random varying by age, stage, race and ethnicity, county-level socioeconomic status, and registry. 3This missingness has been followed by the development of multiple methods of imputation to estimate the missing receptor status before calculating incidence rates and survival estimates. 4,5om a public health perspective, it is essential that all individuals with a diagnosis of invasive breast cancer are tested for ER, PR, and ERBB2 receptor status because it is critical in making the best decisions regarding the most appropriate systemic and locoregional therapy choices and because it provides prognostic information. 6][7][8][9] However, to our knowledge, this study is the first to describe the population of women with invasive breast cancer reported to SEER with no record of being tested for ER, PR, or ERBB2 (missing components of receptor status [MCRS]).

Study Population
The SEER Program's collection of 18 registries (based on the November 2018 submission that was released on April 14, 2019) was used for our analyses.SEER's population-based cancer registries have PR ER

2016
The percentage of female patients with invasive breast cancer who were missing individual receptor components decreased with year of diagnosis.From 2012 through 2016, the proportion of missing ERBB2 status was approximately 63% higher than the proportion of missing estrogen receptor (ER) and progesterone receptor (PR).
a very high estimated completeness of reporting. 10Registries used in these analyses cover approximately 27.8% of the US population. 11The variables required to identify or distinguish whether hormone receptor tests (ER, PR, or ERBB2) were not completed (as opposed to being done but with unknown results) were only available for cases diagnosed through 2017.However, data through 2016 were used so that the type of insurance and urban/rural county of residence could be included in this Common patient and disease characteristics, as well as urban/rural county of residence, were examined for their associations with MCRS.Age at diagnosis was categorized as younger than 50 years, 50 to 64 years, 65 to 79 years, and 80 years or older.Race was identified and reported in the SEER data as the following 5 categories: American Indian or Alaska Native, Asian or Pacific Islander, Black, White, and unknown.Stage at diagnosis was classified as local, regional, distant, and unknown stage/unstaged based on the SEER summary stage.Insurance status at diagnosis categories were insured, insured/no specifics, any Medicaid, uninsured, and insurance status unknown.Countybased rural-urban continuum codes were used to classify each case as urban or rural. 14These codes broadly described counties as metropolitan or nonmetropolitan.6][17][18] The type of reporting source included 6 categories: hospital inpatient/ outpatient or clinic, laboratory only (hospital or private), nursing/convalescent home/hospice, other hospital outpatient unit or surgery center (2006+), physician office/private medical practitioner, and radiation treatment or medical oncology center (2006+).

Demographic and Clinical Characteristics Overall and by MCRS Status
The More than 90% (302 638) of the cases were reported to SEER from a hospital (inpatient or outpatient) or clinic (Table 1).A significant difference was found between those with MCRS and those with known receptor status for each demographic and clinical characteristic.Notable differences included the following groups with a higher proportion MCRS: 80 years of older, Black or unknown race, distant or unknown stage at diagnosis, reporting sources other than hospital or clinic, and unknown insurance status (Table 1).

Multivariable Logistic Regression
As shown in

Discussion
This cross-sectional study investigated women with invasive breast cancer who were missing components of receptor status (ER, PR, or ERBB2) by analyzing information about 321 913 cases from the population-based SEER 18 We identified several subgroups of women with higher percentages of MCRS, including older women; Black and American Indian/Alaska Native women and women with unknown race; those with a diagnosis of distant stage or unknown stage/unstaged at diagnosis; those reported from nursing/convalescent homes/hospices, physician offices, or laboratory only; those with unknown insurance status or who were uninsured; and those in rural counties.Differences in MCRS were also found between the categories of subgroups of women in a multivariable analysis.For example, the odds of MCRS increased with age and advancing stage.The odds of MCRS were higher for Black women compared with White women and rural compared with urban residence.These findings were similar to those reported by Howlader et al, 4,19 although their definition of unknown cases was broader than that used in this study in that they included unknown results and missing data.In addition, the odds of MCRS were typically higher for women who had missing or unknown characteristics.These findings were similar to findings regarding factors associated with unknown stage, unknown grade, and/or incomplete records in central cancer registries, including SEER Program registries. 20e importance of establishing the receptor status in patients with breast cancer has been shown in previous studies.2][23][24] Hwang et al, 25 Howlader et al, 19 and Leone et al 24 reported that the subtype was a significant factor in breast cancer-specific and overall survival that were also based on SEER data.
Although the importance of testing has been established, some patients with severe comorbidities, deconditioning, and very poor prognosis may not need this testing, as it could be irrelevant for a patient who is only a candidate for best supportive care and hospice.These patients are typically older, with late or unknown stage, and reported by nonhospital facilities.However, for most patients with breast cancer, knowing the receptor status in breast cancer is imperative in determinin the best course of treatment for most other patients.Therapeutic trends have increasingly depended on selecting the most appropriate patients for particular systemic treatments based on a cancer's molecular makeup.Such treatments include endocrine therapy, small molecule tyrosine kinase inhibitors, chemotherapy, target-specific antibodies, antibody drug conjugates, immunotherapeutic agents, and cellular therapies.More often, such therapies are moving the needle of survival and quality of life.However, patients lacking critical biomarker information cannot benefit from use of these treatments.Conversely, therapeutic approaches that are used inappropriately can be associated with unnecessary harm, low efficacy, and excessive financial burden on patients and the health care system.

Limitations
Although the results of the present study are potentially informative, some limitations should be considered comorbidities, both of which may be associated with MCRS.The accuracy of MCRS could not be assessed due to the use of registry data; thus, some misclassification bias likely occurred.In addition, although insurance at the time of diagnosis was examined, there was no available information concerning changes in insurance status during the duration of treatment or details on the cost of testing throughout various areas of the US, and the associated financial burden of testing to individual women could not be analyzed.These factors may be barriers to testing.Lastly, from a public health perspective, we were interested in examining which factors were associated with missing values of receptor status.In this context, they could have been missing either because they was not ordered or for unknown reasons, and combining these 2 may have limited our capacity to make conclusions about whether the test was ordered.

Conclusions
The results of this cross-sectional study suggest that classifying invasive breast cancer by molecular subtypes based on hormone receptor and ERBB2 overexpression has been associated with several advancements in diagnosis and effective treatment in women as well as substantial improvements in quality of care and survival outcomes.Despite this testing being recommended by all expert breast cancer guidelines, this study showed that MCRS is still occurring, especially in some socioeconomic populations.The results of this study may help clinicians, public health practitioners, and policymakers target affected populations to minimize or eliminate this critical health disparity and help save more lives.

Figure .
Figure.Percentage of Missing Individual Receptor Components in Patients With Invasive Breast Cancer by Year of Diagnosis9 8 7 analysis cohort included 321 913 female patients with invasive breast cancer who received a diagnosis between 2012 and 2016.Most of the women were 50 years or older (260 077 [81%]),

Table 1 .
Demographic and Clinical Characteristics by MCRS in 321 913 Female Patients With Invasive Breast Cancer a,b a Missing is defined as the test not done (SSF code 998) or unknown/no information (SSF code 999).bA significant difference was found between those missing designation and those not for each demographic and clinical characteristics.Percentage

Table 3 ,
the associations between MCRS and factors of interest were found after adjusting for one another in a multivariable logistic regression model.The adjusted odds of MCRS among those 80 years and older was 1.75 times that of younger than 49 years.Compared with White women, Black women had higher adjusted odds of MCRS (adjusted odds ratio [aOR], 1.09; 95% CI, 1.04-1.16).Although a high percentage (14.4%) of those with unknown race had MCRS, those with unknown race had lower adjusted odds of MCRS compared with White individuals (aOR, 0.79; 95% CI, 0.68-0.91)which was due to small cell counts when race was combined with other variables in the model.Compared with those with local stage at diagnosis, the adjusted odds of MCRS for those with regional stage was 0.62 (95% CI, 0.60-0.65),while the adjusted odds of those with distant

Table 3 .
Adjusted Odds of MCRS Among Female Patients With Invasive Breast Cancer a and unknown stage or unstaged were more than 3 and 19 times that of local stage at diagnosis, respectively.With hospital inpatient/outpatient or clinic as the reference group, cases reported by laboratory only, nursing/convalescent home/hospice, and a physician's office were more likely to have MCRS (aOR, 1.42; 95% CI, 1.28-1.60;aOR, 9.37; 95% CI, 6.03-14.53;and aOR, 2.32; 95% CI, 2.06-2.62;respectively).The adjusted odds of MCRS was lower for those with any Medicaid coverage compared with those insured, but was higher for categories of insured/no specifics and insurance status unknown.The adjusted odds of MCRS was higher in rural areas compared with urban areas Abbreviations: aOR, adjusted odds ratio; MCRS, missing components of receptor status; NA, not applicable; SEER, Surveillance, Epidemiology, and End Results; SSF, site-specific factor.aMissing is defined as the test not done (SSF code 998) or unknown/no information (SSF code 999).stage Missing Components of Receptor Status Among Women With Invasive Breast Cancer . Our efforts are limited by the nature of the SEER data, which are deidentified and registry based, and these data do not include information about individual socioeconomic status and JAMA Network Open | Oncology JAMA Network Open.2023;6(8):e2330791.doi:10.1001/jamanetworkopen.2023.30791(Reprinted) August 24, 2023 7/10 Downloaded From: https://jamanetwork.com/ on 11/01/2023