Thyrotropin Suppression for Papillary Thyroid Cancer: A Physician Survey Study

Abstract

Background: Current guidelines recommend against thyrotropin (TSH) suppression in low-risk differentiated thyroid cancer patients; however, physician practices remain underexplored. Our objective was to understand treating physicians' approach to TSH suppression in patients with papillary thyroid cancer.

Methods: Endocrinologists and surgeons identified by thyroid cancer patients from the Surveillance, Epidemiology, and End Results registries of Georgia and Los Angeles were surveyed in 2018–2019. Physicians were asked to report how likely they were to recommend TSH suppression (i.e., TSH <0.5 mIU/L) in three clinical scenarios: patients with intermediate-risk, low-risk, and very low-risk papillary thyroid cancer. Responses were measured on a 4-point Likert scale (extremely unlikely to extremely likely). Multivariable logistic regressions were performed to determine physician characteristics associated with recommending TSH suppression in each of the aforementioned scenarios.

Results: Response rate was 69% (448/654). Overall, 80.4% of physicians were likely/extremely likely to recommend TSH suppression for a patient with an intermediate-risk papillary thyroid cancer, 48.8% for a patient with low-risk papillary thyroid cancer, and 29.7% for a patient with very low-risk papillary thyroid cancer. Surgeons were less likely to recommend TSH suppression for an intermediate-risk papillary thyroid cancer patient (odds ratio [OR] = 0.36 [95% confidence interval, CI, 0.19–0.69]) compared with endocrinologists. Physicians with higher thyroid cancer patient volume were less likely to suppress TSH in low-risk and very low-risk papillary thyroid cancer patients (i.e., >40 patients per year, OR = 0.53 [CI 0.30–0.96]; OR = 0.49 [CI 0.24–0.99], respectively, compared with 0–20 patients per year). Physicians who estimated higher likelihood of recurrence were more likely to suppress TSH in a patient with very low-risk papillary thyroid cancer (OR = 2.34 [CI 1.91–4.59]).

Conclusions: Many patients with low-risk thyroid cancer continue to be treated with suppressive doses of thyroid hormone, emphasizing the need for more high-quality research to guide thyroid cancer management, as well as better understanding of barriers that hinder guideline adoption.

Introduction

Papillary thyroid cancer is the most common and least aggressive type of thyroid malignancy, accounting for ∼90% of differentiated thyroid cancers (DTC) (1). Owing to the widespread use of ultrasonography and fine-needle aspiration biopsy of incidentally detected thyroid nodules in recent years, the most commonly occurring papillary thyroid cancers in the United States are low-risk papillary thyroid cancers and very low-risk papillary thyroid microcarcinomas (2–6), with risk denoting the risk of cancer recurrence after initial treatment as outlined in the 2015 American Thyroid Association (ATA) guidelines (7). Specifically, very low-risk and low-risk papillary thyroid cancers carry a ≤5% risk of recurrence, while intermediate-risk papillary thyroid cancers carry a recurrence risk of ∼5% to 30% (7).

Traditionally, adequate surgery with or without radioactive iodine ablation has been followed by long-term thyrotropin (TSH) suppression therapy for patients with DTC, such as papillary thyroid cancer, based on the rationale that higher TSH may stimulate thyroid cancer cell proliferation and thyroglobulin production in these patients (7). More recently, the ATA guidelines on DTC proposed a graded tailored approach to TSH suppression based on initial risk of recurrence, subsequent response to therapy, and risk of harm from thyroid hormone-suppressive therapy, in an effort to decrease unnecessary adverse effects from thyroid hormone overtreatment (7,8).

Specifically, recommendations advise against initial TSH suppression therapy, that is, TSH <0.5 mIU/L, in patients with very low-risk and low-risk papillary thyroid cancer (7). These recommendations are based on prior studies failing to document improved clinical outcomes, such as overall and disease-free survival, except in patients with high-risk advanced disease (9–14). Despite consensus by experts that TSH suppression therapy in patients with low-risk papillary thyroid cancer does not lead to improved outcomes and may lead to patient harm (7,14–17), physician practices regarding TSH suppression in these patients remain both underexplored and difficult to study.

To better understand treating physicians' approach toward TSH suppression in patients with papillary thyroid cancer following initial treatment, we surveyed a diverse group of physicians caring for a population-based cohort of thyroid cancer patients using real-life clinical scenarios. We specifically focused on delineating physician recommendations for TSH suppression in papillary thyroid cancer patients with intermediate-risk, low-risk, and very low-risk disease and determining physician characteristics associated with recommending TSH suppression in these scenarios.

Methods

Study population

Patients with DTC affiliated with the Surveillance, Epidemiology, and End Results (SEER) registries of Georgia and Los Angeles County were asked to report the endocrinologists and surgeons involved in their thyroid cancer care (18–22). Of 699 physicians identified, 45 were ineligible due to no longer being in practice or retired, being deceased before the initial mailing, unable to be located, or not meeting screening criteria after initial contact. Surveys were mailed to 654 response-eligible physicians between October 2018 and December 2019. Surveys were mailed to all physicians identified by more than one patient (N = 482). Additionally, the survey was mailed to 172 physicians randomly sampled out of the cohort of physicians identified by only one patient.

As previously reported in prior work with physician surveys, we used the modified Dillman method of survey administration to enhance response rates (23–26). A $50 cash incentive was included in the survey mailing. Follow-up for nonresponders included additional mailings and contact by phone, fax, or e-mail to facilitate survey completion. Double data entry was used to ensure a <1% error rate. A total of 448 physicians completed the survey resulting in a 69% response rate (448/654) (27).

The study was approved by the University of Michigan (HUM00113715), the University of Southern California (HS-16-00646), the Committee for the Protection of Human Subjects (California State Institutional Review Board), the Georgia Department of Public Health, and the Emory University Institutional Review Boards (IRB00093983). The study also received approval from the California Cancer Registry. A waiver of written informed consent was obtained for all subjects in this study.

Survey questionnaire design and content

The content of the survey questionnaire, which included clinical scenarios, was designed by a multidisciplinary team of endocrinologists, surgeons, survey methodologists, and health services researchers, based on a systematic review of the literature, a conceptual framework, and expertise in providing thyroid cancer patient care. In addition to multiple reviews by design and content experts, the survey was pilot tested in a select cohort of nine endocrinologists and surgeons in the state of Michigan to assess reliability and content validity.

Measures

Physicians were asked to report how likely they were to recommend TSH suppression, defined as a TSH <0.5 mIU/L, in three different clinical scenarios: (a) a 55-year-old female patient with a 1.5 cm classic papillary thyroid cancer with microscopic extrathyroidal extension, no vascular invasion, but with 1 of 10 central neck lymph nodes positive for papillary thyroid cancer (intermediate risk); (b) a 65-year-old female patient with a 1.5 cm classic papillary thyroid cancer without extrathyroidal extension, vascular invasion, or known lymph node metastases (low risk); and (c) a 40-year-old female patient with a 0.9 cm classic papillary thyroid cancer without extrathyroidal extension, vascular invasion, or known lymph node metastases (very low risk). Classification of scenarios in corresponding risk groups was based on the 2015 ATA guidelines for risk of recurrence for adult patients with DTC (7). Responses were based on a 4-point Likert scale as follows: extremely unlikely, unlikely, likely, and extremely likely. These were then categorized as extremely unlikely/unlikely and likely/extremely likely before analyses.

If physicians were likely to recommend TSH suppression, they were also asked to report how many years they would continue TSH suppression for in each scenario. Answer options included 1 year, 2–5 years, 6–10 years, >10 years, or not applicable. Answers were categorized as ≤5 and >5 years.

Covariates

Covariates in the analyses included physician characteristics obtained from the survey. These were physician specialty (endocrinology and surgery, with surgery including general surgery [n = 134] and otolaryngology [n = 130]), practice setting (academic medical center, large medical group or staff-model Health Maintenance Organization, private practice or community health clinic), years in practice (<20, ≥20), number of thyroid cancer patients seen in the past year (0–20, 21–40, >40), and SEER site (Georgia, Los Angeles County). As endocrine surgery and surgical oncology are general surgery subspecialties, physicians who reported being endocrine surgeons (n = 16) or surgical oncologists (n = 13) were categorized as general surgeons. Physicians who could not be categorized as endocrinologists or surgeons (n = 8) and those who reported “other” as their practice setting (n = 4) were excluded from the final analytic cohort.

Additionally for each scenario, that is, intermediate-risk, low-risk, and very low-risk papillary thyroid cancer, physicians were asked to state the number of patients out of 100 they thought will have thyroid cancer recurrence in the next 10 years (categories included 5 or fewer, 10, 20, and then multiples of 10). Responses were categorized as ≤5 or ≥10.

Statistical analyses

Descriptive statistics were generated for all categorical variables. Nonweighted frequencies and weighted percentages are reported. After frequencies were examined, univariate analyses were performed using Rao–Scott chi-square tests. Separate multivariable weighted regression analyses were then performed to determine physician characteristics associated with being likely/extremely likely to recommend TSH suppression for each clinical scenario, that is, intermediate-risk papillary thyroid cancer, low-risk papillary thyroid cancer, and very low-risk papillary thyroid cancer.

Missing data were <5% per survey item. All statistical analyses incorporated weights to adjust for nonresponse, which were computed using the variables of physician specialty (identified by patients), patient number (number of patients who identified the physicians), and SEER site. This method allows adjusting for potential nonresponse bias to improve the representativeness of the sample studied (i.e., physicians treating patients with thyroid cancer), as it accounts for disproportionate nonresponse rates across physician subgroups (18,28). Statistical analyses were conducted using R version 3.6.1. and Stata 15.1 (StataCorp LLC, College Station, TX). The Wald 95% confidence interval (CI) not including the null value and a p-value of <0.05 were used to determine statistical significance.

Results

Sample characteristics

Physician respondent characteristics are shown in Table 1. A total of 41.7% of respondents were endocrinologists and 58.3% were surgeons (general surgery [n = 134] and otolaryngology [n = 130]). The majority of physicians reported private practice or community health clinic as their practice setting (55.6%). In regard to thyroid cancer patient volume in the past year, 51.9% cared for 0–20 patients, 21.0% cared for 21–40 patients, and 27.1% cared for >40 patients. When asked about the number of patients out of 100 whom physicians think will have recurrence in the next 10 years, 63.3% estimated that 10 or more patients with intermediate-risk papillary thyroid cancer would have recurrence, 21.6% estimated that 10 or more patients with low-risk papillary thyroid cancer would have recurrence, and 10.7% estimated that 10 or more patients with very low-risk papillary thyroid cancer would have recurrence. The other respondents estimated that 5 or fewer patients would have recurrence.

Table 1.

Physician Characteristics (N = 448)

Physician characteristic	n (weighted %)
Specialty
Endocrinology	176 (41.7)
Surgery	264 (58.3)
Years in practice
<20	226 (51.8)
≥20	213 (48.2)
Practice setting
Private practice or community health clinic	244 (55.6)
Large medical group or staff-model HMO	114 (25.2)
Academic medical center	83 (19.2)
Thyroid cancer patient volume in past year
0–20	232 (51.9)
21–40	94 (21.0)
>40	117 (27.1)
SEER site
Georgia	205 (49.2)
Los Angeles County	243 (50.8)
Number of patients with intermediate-risk thyroid cancer out of 100 whom physicians think will have recurrence in the next 10 years
≤5	160 (36.7)
≥10	278 (63.3)
Number of patients with low-risk thyroid cancer out of 100 whom physicians think will have recurrence in the next 10 years
≤5	340 (78.4)
≥10	94 (21.6)
Number of patients with very low-risk thyroid cancer out of 100 whom physicians think will have recurrence in the next 10 years
≤5	392 (89.3)
≥10	47 (10.7)

HMO, Health Maintenance Organization; SEER, Surveillance, Epidemiology, and End Results.

Physician recommendations regarding TSH suppression therapy

As demonstrated in Figure 1, 80.4% of physicians were likely/extremely likely to recommend TSH suppression for a patient with intermediate-risk papillary thyroid cancer, 48.8% for a patient with low-risk papillary thyroid cancer, and 29.7% for a patient with very low-risk papillary thyroid cancer. Of the physicians who were likely/extremely likely to recommend TSH suppression therapy, 39.7% of physicians reported that they would continue TSH suppression for >5 years in a patient with intermediate-risk papillary thyroid cancer, while similarly 41.5% and 45.8% reported that they would suppress TSH for >5 years in a patient with low-risk papillary thyroid cancer and very low-risk papillary thyroid cancer, respectively.

FIG. 1.

Percentage of physicians likely to extremely likely to recommend TSH suppression in intermediate-risk, low-risk, and very low-risk papillary thyroid cancer patients. TSH, thyrotropin.

Intermediate-risk papillary thyroid cancer

Table 2 shows the results from the multivariable regression analysis to determine physician characteristics associated with being likely/extremely likely to recommend TSH suppression in a patient with intermediate-risk papillary thyroid cancer. Surgeons were less likely to recommend TSH suppression for a patient with intermediate-risk papillary thyroid cancer (odds ratio [OR] = 0.36 [CI 0.19–0.69]) compared with endocrinologists.

Table 2.

Physician Characteristics Associated with Recommending Thyrotropin Suppression in a Patient with Intermediate-Risk Papillary Thyroid Cancer

	n (%) of physicians likely to extremely likely to suppress TSH	OR [CI]	p
Specialty
Endocrinology	155 (88.0)	Ref.
Surgery	182 (74.7)	0.36 [0.19–0.69]	0.002
Years in practice
<20	176 (80.9)	Ref.
≥20	161 (80.6)	1.29 [0.77–2.17]	0.327
Practice setting
Academic medical center	69 (84.6)	Ref.
Large medical group or staff-model HMO	87 (82.2)	0.78 [0.34–1.77]	0.553
Private practice or community health clinic	181 (77.7)	0.59 [0.28–1.25]	0.165
Thyroid cancer patient volume in past year
0–20	168 (76.8)	Ref.
21–40	75 (82.9)	1.24 [0.61–2.56]	0.552
>40	96 (84.5)	0.89 [0.44–1.80]	0.739
Number of patients out of 100 whom physicians think will have recurrence in the next 10 years
<5	124 (79.3)	Ref.
≥10	214 (80.8)	1.15 [0.68–1.96]	0.52
SEER site
Georgia	160 (80.7)	Ref.
Los Angeles County	183 (80.2)	0.85 [0.51–1.43]	0.546

CI, 95% confidence interval; OR, odds ratio; TSH, thyrotropin.

Low-risk papillary thyroid cancer

Table 3 demonstrates that compared with low-volume physicians (i.e., those who reported caring for 20 or fewer thyroid cancer patients in the past year), high-volume physicians seeing >40 thyroid cancer patients in the past year were less likely to recommend TSH suppression for a patient with low-risk papillary thyroid cancer (OR = 0.53 [CI 0.30–0.96]).

Table 3.

Physician Characteristics Associated with Recommending Thyrotropin Suppression in a Patient with Low-Risk Papillary Thyroid Cancer

	n (%) of physicians likely to extremely likely to suppress TSH	OR [CI]	p
Specialty
Endocrinology	88 (49.9)	Ref.
Surgery	119 (48.4)	0.66 [0.41–1.07]	0.093
Years in practice
<20	105 (47.5)	Ref.
≥20	102 (50.8)	1.19 [0.77–1.82]	0.432
Practice setting
Academic medical center	32 (38.6)	Ref.
Large medical group or staff-model HMO	51 (47.0)	1.16 [0.62–2.16]	0.643
Private practice or community health clinic	121 (52.1)	1.44 [0.82–2.53]	0.198
Thyroid cancer patient volume in past year
0–20	112 (51.3)	Ref.
21–40	50 (55.5)	1.11 [0.64–1.94]	0.708
>40	47 (40.2)	0.53 [0.30–0.96]	0.035
Number of patients out of 100 whom physicians think will have recurrence in the next 10 years
<5	161 (48.3)	Ref.
≥10	48 (52.8)	1.16 [0.71–1.88]	0.563
SEER site
Georgia	94 (47.5)	Ref.
Los Angeles County	116 (50.1)	1.11 [0.74–1.68]	0.603

Very low-risk papillary thyroid cancer

The results of the multivariable logistic regression analyzing the association of physician characteristics with recommending TSH suppression in a patient with very low-risk papillary thyroid cancer are shown in Table 4. High-volume physicians (>40 thyroid cancer patients in the past year) were less likely to recommend TSH suppression (OR = 0.49 [CI 0.24–0.99], compared with 0–20 patients in the past year), while those who estimated higher likelihood of recurrence in the next 10 years (i.e., ≥10%) were more likely to recommend TSH suppression in this scenario (OR = 2.34 [CI 1.91–4.59], compared with ≤5%) (Table 4).

Table 4.

Physician Characteristics Associated with Recommending Thyrotropin Suppression in a Patient with Very Low-Risk Papillary Thyroid Cancer (Microcarcinoma)

	n (%) of physicians likely to extremely likely to suppress TSH	OR [CI]	p
Specialty
Endocrinology	42 (23.5)	Ref.
Surgery	83 (34.2)	1.09 [0.62–1.93]	0.764
Years in practice
<20	53 (24.0)	Ref.
≥20	72 (35.9)	1.39 [0.88–2.18]	0.154
Practice setting
Academic medical center	14 (17.2)	Ref.
Large medical group or staff-model HMO	32 (29.6)	1.71 [0.79–3.73]	0.175
Private practice or community health clinic	79 (33.7)	1.91 [0.93–3.91]	0.077
Thyroid cancer patient volume in past year
0–20	78 (35.7)	Ref.
21–40	26 (29.1)	0.79 [0.42–1.48]	0.465
>40	22 (18.5)	0.49 [0.24–0.99]	0.048
Number of patients out of 100 whom physicians think will have recurrence in the next 10 years*
<5	107 (27.9)	Ref.
≥10	21 (48.0)	2.34 [1.91–4.59]	0.014
SEER site
Georgia	62 (30.9)	Ref.
Los Angeles County	66 (28.6)	0.96 [0.61–1.50]	0.854

^* Note: Statistically significant in univariate analysis, p = 0.006.

Discussion

In this large survey study of physicians caring for a population-based cohort of thyroid cancer patients, we found that in addition to the majority of physicians recommending TSH suppression in patients with intermediate-risk papillary thyroid cancer, almost half reported that they recommend TSH suppression in patients with low-risk papillary thyroid cancer and almost one-third recommend TSH suppression in patients with very low-risk papillary thyroid cancer. Physicians with higher thyroid cancer patient volume were less likely to recommend TSH suppression in low-risk and very low-risk papillary thyroid cancer patients, while physicians who estimated higher likelihood of recurrence were more likely to recommend TSH suppression in a patient with very low-risk papillary thyroid cancer. Additionally, just under half of the physicians who recommended suppressing TSH in low-risk and very low-risk papillary thyroid cancer patients reported that they would continue this practice for longer than 5 years. Our results demonstrate that marked variation exists in physician recommendations regarding optimal TSH suppression therapy for low-risk and very low-risk papillary thyroid cancer.

Prior studies using a large thyroid cancer consortium, which includes patients from multiple institutions that are prospectively followed in the United States, have shown that aggressive TSH suppression therapy did not confer benefit in reducing risk of recurrence or in improving overall survival in patients with low-risk DTC (9,10,29). More recently, a single-center, retrospective, propensity score-matched cohort study from South Korea evaluated the effects of TSH suppression therapy in patients with low-risk DTC who underwent lobectomy and similarly found that there was no significant difference in disease-free survival between patients who underwent TSH suppression therapy and those who did not (N = 446; median follow-up 8.6 years) (12). In addition to the lack of benefit in patients with low-risk DTC, aggressive TSH suppression has been shown to lead to patient harm, especially in older patients, including increased risk for atrial fibrillation, cardiovascular and all-cause mortality, and osteoporosis in postmenopausal women (14–17,30–32).

Taking into consideration the high risk and low benefit of TSH suppression in patients with low-risk and very low-risk DTC, physicians' tendency for more aggressive treatment than guidelines recommend is likely unwarranted and may lead to adverse patient outcomes. Even more concerning than initiation of TSH suppression in patients with low-risk and very low-risk disease, is the continuation of long-term TSH suppression for many years as this increases likelihood of patient harm (15,33,34).

Furthermore, the intermediate-risk scenario presented in our survey lies on the lower end of the recurrence risk continuum. Thus, the finding of a high proportion of physicians recommending TSH suppression for this intermediate-risk patient is additionally worrisome as these patients may shift to the lower risk category in future guidelines. Contributory to this physician behavior could be the fact that due to low-quality evidence, some guideline recommendations surrounding initiation and length of TSH suppression rely on expert consensus (7). More high-quality studies on the role of TSH suppression, including studies focusing on optimal length of time to suppress TSH in specific scenarios, are needed to optimally update future guidelines.

It has previously been shown that physician adherence to guidelines is more likely when supported by high-quality evidence (35). Guideline adherence may also be influenced by several other barriers, including lack of familiarity with guidelines, lack of self-efficacy, disagreement with guidelines, inertia of previous practice, or difficulty in reconciling patient preferences with guideline recommendations (36). Although few studies have focused on TSH suppression, illustrative of the difficulty physicians face in adopting guidelines aiming to reduce use of low-value thyroid cancer care is a study by Lohia et al. (37) that found only minimal increase in guideline-recommended de-escalation of surgical and radioactive iodine treatment for low-risk papillary thyroid cancer several years after guideline recommendations shifted to less intensive treatment recommendations (38).

Our finding that physicians caring for a higher volume of thyroid cancer patients were less likely to recommend TSH suppression in patients with low-risk and very low-risk thyroid cancer highlights the importance of physician experience and expertise in regard to reducing low-value care. To the best of our knowledge, this is the first study to demonstrate an inverse relationship between patient volume and likelihood of recommending TSH suppression therapy in patients with low-risk and very low-risk papillary thyroid cancer. Even though higher surgeon volume has been shown to be an important predictor of improved patient outcomes after surgical management of thyroid cancer, such as fewer postoperative complications, studies on the association between patient volume and intensity of thyroid cancer medical management are lacking (39–44).

Similar to studies of other cancer types, our findings suggest that physicians who see more thyroid cancer patients may be more aware of the excellent prognosis of many thyroid cancer patients and of the potential harms of overtreatment, as well as more familiar with guidelines and thus more likely to adhere to guideline-concordant care (45–48). These findings suggest a new venue for reducing overtreatment with TSH suppression in low-risk thyroid cancer patients. In the future, it may be beneficial to target specific intervention efforts toward low-volume physicians. These future efforts will improve the quality of patient care and reduce costs.

Additionally, physicians who overestimated risk of recurrence in patients with very low-risk thyroid cancer were more likely to recommend TSH suppression for their patients. This is consistent with other studies showing that overestimation of cancer recurrence risk is associated with more extensive treatment than necessary (21,49). Considering that risk of recurrence for papillary thyroid microcarcinomas is <5% (7,50,51), our results suggest a potential role for physician education as a method to optimize the understanding of recurrence risk and its implications for treatment decision making.

Strengths of our study include the diverse cohort of physicians, comprising both surgeons and endocrinologists and physicians in private practice and academics, inclusion of both high- and low-volume physicians, and a robust survey response rate. Additionally, unlike surgery and radioactive iodine, large cancer registries do not commonly capture information on TSH suppression in thyroid cancer patients. Therefore, we used a novel study design incorporating standardized clinical vignettes, which allows for case mix control, as a method of studying this important topic.

However, some limitations merit mention. Not all physicians responded to our survey, and nonresponse bias is possible. However, we had a strong response rate, and results were weighted for nonresponse to mitigate this bias. Additionally, our clinical vignettes did not include information on patient comorbidities or postoperative thyroglobulin, which may influence decision-making in real-world settings. Finally, our study population includes physicians caring for thyroid cancer patients in Los Angeles County, California, and Georgia, and therefore generalizability to other populations may be limited. However, we purposefully selected these SEER sites for their large catchment areas and diverse patient populations (52).

In an era of “less is more” with the pendulum swinging toward less intensive treatment of low-risk DTC in recent years, our study highlights the need to better understand physician behavior surrounding TSH suppression in patients with low-risk and very low-risk papillary thyroid cancer and to identify drivers and barriers to guideline adoption. To design effective and successful interventions to reduce variation in care and overtreatment, future efforts should focus on physician education, physician behavior change with data feedback, and high-quality thyroid cancer research that will inform future guidelines. Carefully designed, feasible and acceptable multilevel interventions to support de-escalation of TSH suppression therapy in patients with low-risk DTC should be tailored to pre-identified determinants for guideline adherence.

Authors' Contributions

M.R.H. had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: M.P. and M.R.H. Acquisition, analysis, or interpretation of data: M.P, D.W.C., D.R.-G., M.B., A.S.H., K.C.W., and M.R.H. Drafting of the article: M.P. and M.R.H. Critical revision of the article for important intellectual content: M.P., D.W.C., D.R.-G., M.B., A.S.H., K.C.W., and M.R.H. Statistical analysis: D.R.-G. and M.B. Obtained funding: M.R.H. Administrative, technical, or material support: A.S.H., K.C.W., and M.R.H. Study supervision: M.P. and M.R.H.

Author Disclosure Statement

The authors have nothing to disclose.

Funding Information

This study is supported by R01 CA201198 from the National Cancer Institute (NCI) to Principal Investigator, M.R.H.. M.R.H. also receives funding from the Agency for Healthcare Research and Quality (AHRQ) Grant No. R01 HS024512. M.P. receives funding from K08 AG049684 National Institute on Aging (NIA). D.W.C. receives funding from T32 DK007245 National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). The collection of cancer incidence data used in this study was supported by the California Department of Public Health pursuant to California Health and Safety Code Section 103885, Centers for Disease Control and Prevention (CDC) National Program of Cancer Registries, under cooperative agreement 5NU58DP006344, and the NCIs SEER Program, under contract HHSN261201800015I awarded to the University of Southern California. The collection of cancer incidence data in Georgia was supported by contract HHSN261201800003I, Task Order HHSN26100001, from the NCI and cooperative agreement 5NU58DP003875-04 from the CDC. The ideas and opinions expressed herein are those of the authors, and endorsement by the State of California and State of Georgia Departments of Public Health, the NCI, and the CDC or their contractors and subcontractors is not intended nor should be inferred.