ELSEVIER
Cancer Treatment and Research Communications
journal homepage: www.elsevier.com/locate/ctarc
cancer TREATMENT COMMUNICATIONS
Smoking is associated with adrenal adenomas and adrenocortical carcinomas: a nationwide multicenter analysis
Check for updates
Ahmed Yousafª, Jessica Pattersonª, Gerald Hobbsb, Stephen M. Davisc,d, Muhammad Yousafe, Maria Hafez“, Heidar Albandarf, Thomas Hoganf, Joanna Kolodneyf,*
ª Department of Dermatology, West Virginia University, Morgantown, WV, USA
b Department of Statistics, West Virginia University, Morgantown, WV, USA
” Department of Health Policy, Management & Leadership, West Virginia University, Morgantown, WV, USA
d Department of Emergency Medicine, West Virginia University, Morgantown, WV, USA
e West Virginia University School of Medicine, Morgantown, WV, USA
Department of Medicine, Section of Hematology / Oncology, West Virginia University, Morgantown, WV, USA
ARTICLE INFO
Keywords:
Adrenal Adrenal adenoma Adrenocortical Carcinoma Tobacco Use SEER TriNetX National
ABSTRACT
Microabstract: The effect of smoking on adrenal cancer is poorly understood. A clear association of adrenal adenoma and adrenocortical carcinoma with smoking among the United States population is observed. This association points to the possibility of environmental carcinogenic and/or lifestyle factors contributing to adrenal cancer formation. Our results support the association of tobacco use with adrenal adenomas and adrenal cortical carcinoma.
Background: Smoking has been suggested as a risk factor for adrenal cortical carcinoma (ACC), but this hy- pothesis has only been inferred from a single study using all types of adrenal cancers including pheochromo- cytoma, neuroblastoma, as well as ACC. Given the high rate of tobacco use in West Virginia, we hypothesized that smoking might contribute to increased prevalence of ACC.
Materials and Methods: De-identified institutional review board-exempted records were analyzed in the Surveillance, Epidemiology, and End Results (SEER) Program from 2001-2016 and in patients from the United States nationwide, multicenter TriNetX database of 41,063,707 patients from 2008-2018. In addition, the state- level ratio of smoking to ACC prevalence was computed in all 50 states using data from SEER and the Center for Disease Control. West Virginia Health System data from 2008-2018 was extracted to confirm population-level findings. Melanoma was used as a cancer control in both databases.
Results: 6,946 ACC cases were identified. West Virginia had the highest smoking rate and the second highest rate of ACC. A significant association was found between smoking and ACC (Pearson correlation coefficient r = 0.4887, p =. 0004). From 2008 to 2018 using TriNetX, 846 ACC and 36,434 AA were extracted. Both adrenal neoplasm cohorts had increased prevalence of tobacco use compared with melanoma controls, where 23.5% were smokers compared to 36.4% and 33.9% in the ACC and AA groups, respectively (p<0.0001 each).
Conclusion: To our knowledge, this is the first United States population-based study supporting smoking as a risk factor for adrenal carcinogenesis and ACC.
Introduction
Adrenocortical carcinoma (ACC) is a rare and often fatal malignant neoplasm with an estimated incidence of 0.7-2.0 cases per million
persons per year [1-3]. A bimodal distribution exists with a primary peak in the fourth decade and a secondary peak in the first decade of life [4, 5]. ACCs share a female preponderance with a 1.5:1 female-to- male ratio [5, 6]. Currently, most ACCs are thought to originate
Abbreviations: ACC, adrenal cortical carcinoma/adrenocortical carcinoma; SEER, the Surveillance, Epidemiology, and End Results; NPCR, National Program of Cancer Registries; ICD-10-CM, International Classification of Diseases, 10th Revision, Clinical Modification; ICDO-3, International Classification of Diseases for Oncology, third revision; BRFSS, Behavioral Risk Factor Surveillance System; AOR, adjusted odd’s ratio
* Corresponding author at: WVU School of Medicine, Department of Medicine, Section of Hematology / Oncology, PO Box 9162, 64 Medical Center Drive, Morgantown 26506, WV, USA
E-mail address: Joanna.kolodney@hsc.wvu.edu (J. Kolodney).
https://doi.org/10.1016/j.ctarc.2020.100206
sporadically although they can be found in rare genetic syndromes such as Beckwith-Wiedemann syndrome, Li-Fraumeni syndrome, Carney syndrome, and MEN 1 [7, 8].
Adrenal adenomas are benign adrenal cortex neoplasms that re- present 50-80% of all adrenal neoplasms [9]. They can be unilateral or bilateral, most commonly found on CT scans, frequently termed “in- cidentalomas” in reference to their incidental discovery. 6% are func- tional producing an excess of adrenal hormones [10]. When there is no patient history of malignancy, incidentalomas are classified as ade- nomas [11].
Currently, three reports have raised suspicion for tobacco use as a preventable risk factor associated with adrenal carcinogenesis, but population-level data are still lacking [12-14]. At our own institution at West Virginia University with a smoking-heavy population, we ob- served eight cases of ACC between 2017 and 2019. For West Virginia with a population of 1.8 million, we expected to see at most 4 cases of ACC. This led us to clarify the association of ACC and adrenal adenoma with smoking using two United States population-level databases and the West Virginia Health System database. We hypothesized that an increased number of ACC cases in West Virginia might be associated with state-level smoking prevalence.
Materials and Methods
Data sources
Due to the rarity of ACC, population-level data were deemed most suitable for a statistical comparison between ACCs, adrenal adenomas, and a control population. Therefore, the Surveillance, Epidemiology, and End Results (SEER) database and the TriNetX database were se- lected for review. To validate these population-level databases, our electronic medical record system was used to look at individual-level data and control for confounders of age, gender, race, and region.
The SEER database, compiled by the National Cancer Institute, contains cancer incidence and survival information specific to the United States from population-based cancer registries covering ap- proximately 34 percent of the United States population, including pa- tient demographics, primary tumor site, tumor morphology, stage at diagnosis, first course of treatment, and follow-up for survival. These data are collected on every cancer case reported from 19 United States geographic areas [15].
TriNetX is a global federated health research network providing access to statistics on electronic medical records including diagnoses, procedures, medications, laboratory values, and genomic information from approximately 41,063,707 patients from 2008-2018 across 31 large United States academic institutions [16]. TriNetX converts data from local providers by standardizing data structure and terminology to its own model. Importantly, TriNetX allows ICD-10-CM queries for comorbid diagnoses including ACC, adrenal adenoma, melanoma, and tobacco use. As a federated network, TriNetX received an IRB waiver since only aggregated counts, statistical summaries of de-identified information, but no protected health information was received and no study-specific activities were performed in retrospective analyses.
Cancer cases
National estimates of ACC and adrenal adenoma cases were queried using the CDC’s National Program of Cancer Registries (NPCR) and the SEER Incidence - United States Cancer Statistics Public Use Database, Nov. 2018 submission. Melanoma cases were used as a comparison to a cancer not associated with smoking [17]. This dataset included cancer incidence data from central cancer registries reported to NPCR in 46 states, the District of Columbia, and to SEER in 4 states. With special request, this database was publicly accessible and included demo- graphic characteristics including, age, sex, and race, and tumor char- acteristics, including year of diagnosis, site, histology, stage, and
behavior. Cases were extracted using AYA Site Recode International Classification of Diseases for Oncology, third revision (ICD-O-3) 8.7.1 for ACC and 7.1 for melanoma. No code for adrenal adenoma was identified, so in order to include them, TriNetX data were analyzed. Cases were extracted using International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) codes. ACC (c74.0), adrenal adenoma (d35.0), and melanoma (C43.0, C43.10, C43.20, C43.39, C43.4, C43.59, C43.60, C43.70, C43.9, and C43.52) controls were extracted.
To validate the population-level data provided by SEER and TriNetX, the West Virginia Health System electronic medical records system was utilized. TriNetX data did not provide the tools to control for confounders including age, gender, race, and region. These con- founders were controlled for using West Virginia Health System’s electronic medical records, apart from race due to poor data entry. Cases were extracted using the same methodology to identify TriNetX cases with the same date range and ICD-10-CM codes.
Tobacco use prevalence
State-level estimates of adrenocortical carcinoma cases were cal- culated using the NPCR and SEER Incidence - U.S. Cancer Statistics Public Use Database, Nov 2018 submission (2001-2016). State-level adult smoking prevalence was determined using data provided by the 2017 Behavioral Risk Factor Surveillance System (BRFSS) Survey, a large annual telephone survey of U.S. households sponsored by the CDC designed to measure at the state level. Adult smoking prevalence was defined according to the BRFSS Survey as any individual of any race over the age of 18 responding positively to “do you now smoke cigar- ettes every day, some days, or not at all?” The state level modeled es- timates were controlled to the direct regional estimates so the ag- gregated state level estimates equal the corresponding regional estimates. The state-level adult population was calculated using data provided by the United States Census Bureau and the 2013- 2017 American Community Survey publicly accessible through the State Cancer Profiles database. State-level melanoma cases versus state- level smoking prevalence was used as a comparison to a cancer not associated with smoking [17].
Because SEER did not collect data on tobacco use, TriNetX data were used to determine comorbid tobacco use with melanomas, adrenal adenomas and adrenocortical carcinomas. Tobacco use was defined using ICD-10-CM codes Z72.0, Z71.6, F17, and Z87.891. West Virginia Health System data were used to control for confounders of age, gender, and region using the same ICD-10-CM codes used to search the TriNetX data.
Statistical analysis
Using the SEER program, population estimates for ACC and mela- noma were a modification of annual county population estimates by age, sex, bridged race, and ethnicity [18]. Modifications incorporated bridged, single-race estimates that were derived from multiple-race categories in the Census and accounted for known issues in certain counties. The modified county-level population estimates were summed to the state and national levels. State-level levels of ACC and melanoma cases were used in conjunction with state-level adult smoking pre- valence provided by the 2017 BRFSS survey to create bivariate scatter plots. Pearson correlation coefficients were calculated to determine the strength of correlation between cancer cases and smoking prevalence by state.
Using TriNetX and West Virginia Health System data, a chi-square analysis was conducted to determine significant differences amongst all independent variables when comparing adrenal adenoma with mela- noma and ACC with melanoma. Significance was set to an alpha level of 0.05.
| Characteristic | Melanoma | Adrenocortical Carcinoma No. (%) n=6,946 |
|---|---|---|
| No. (%) n=1,077,346 | ||
| Age | ||
| < 50 | 256,126 (23.8) | 2,270 (32.7) |
| 50 - 69 | 449,052 (41.7) | 3,061 (44.1) |
| ≥ 70 | 372,168 (34.5) | 1,615 (23.2) |
| Unknown | 0 | 0 |
| Gender | ||
| Male | 620,076 (57.6) | 2,956 (42.6) |
| Female | 457,270 (42.4) | 3,990 (57.4) |
| Unknown | 0 | 0 |
| Race | ||
| White | 1,027,274 (95.4) | 6,020 (86.7) |
| Black | 6,315 (0.6) | 628 (9.0) |
| Asian | 3,789 0.4) | 200 (2.9) |
| American Indian/Alaska | 2,491 (0.2) | 42 (0.6) |
| Native | ||
| Unknown | 37,477 (3.4) | 56 (0.8) |
| Region | ||
| Northeast | 206,934 (19.2) | 1,406 (20.2) |
| Midwest | 233,671 (21.7) | 1,592 (22.9) |
| South | 379,069 (35.2) | 2,486 (35.8) |
| West | 257,672 (23.9) | 1,462 (21.1) |
| Unknown | 0 | 0 |
Data are from population-based registries that participate in the Center for Disease Control’s National Program of Cancer Registries and/or the National Cancer Institute’s Surveillance, Epidemiology, and End Results Program and meet high-quality data criteria. These registries cover approximately 98% of the United States population. No SEER code for adrenal adenoma or related con- ditions was identified.
Ethics approval
Approval of the study was obtained from the West Virginia University institutional review board (protocol number 1812393484). As a public register-linked study, the approval covered exemption from informed consent.
Results
Population characteristics using SEER
SEER national estimates of ACC and adrenal adenoma along with melanoma controls were queried (Table 1). No SEER code for adrenal adenoma or related conditions was identified. Overall, there were 6,946 ACC and 1,077,346 melanomas captured in the SEER program between the years 2001 - 2016. 67.2% of ACC and 76.2% of melanomas were identified in subjects aged greater than 50 years old. ACC was more prevalent in females (57.4%) compared to melanoma (42.4%). Both ACC and melanoma populations were mostly white (86.7% and 95.4%, respectively) and distributed across all regions of the United States, with most cases in the South (35.2% and 35.8%, respectively).
State-level estimates of ACC and smoking prevalence
State-level smoking prevalence, ACC, and melanomas per state were compiled and visualized for all 50 states in the SEER database. (Fig. 1). Comparing state-level smoking prevalence to state-level ACC cases, a statistically significant association was found (r=0.4887, p =. 0004). Melanoma was weakly negatively associated with smoking prevalence (r =- 0.1843, p =. 2099). The state of West Virginia had the highest smoking rate and the second highest rate of ACC.
Association of tobacco use with ACC and adrenal adenomas
Using TriNetX data, 98,747 melanomas, 36,434 adrenal adenomas, and 846 ACC were identified between 2008 and 2018 (Table 2). Sig- nificant differences were observed across age, gender, race, and region in adrenal adenoma and ACC cases compared against melanoma (p<0.001 each). Comorbid tobacco use was significantly associated with adrenal adenomas (p<0.0001, OR=1.86, 95% CI [1.82-1.91]) and ACC (p<0.001, OR=1.67, 95% CI [1.45-1.93]) when compared against comorbid tobacco use in melanoma.
Validating the association of tobacco use with ACC and adrenal adenomas from a single-site center
Using the West Virginia Health System electronic medical records system, 6,347 melanoma cases, 2,161 adrenal adenoma cases, and 16 ACC cases were identified between 2008 and 2018 (Table 3). All cases were extracted from the state of West Virginia, thus controlling for region. Significant differences were observed across age, gender, and tobacco use in adrenal adenoma cases compared against melanoma (p<0.001 each), but only age and tobacco use in ACC cases (p =. 0156, p =. 0011 respectively). When focusing on female cases controlling for age, comorbid tobacco use was significantly associated with adrenal adenoma (p<0.0001, OR=2.89, 95% CI [2.51-3.34]) and ACC cases (p = 0.0065, OR=6.52, 95% CI [1.69-25.14]) compared against mel- anoma. When focusing on male cases controlling for age, comorbid smoking was significantly associated with adrenal adenoma cases compared against melanoma (p<0.0001, OR=2.56, 95% CI [2.19- 3.00]), but no significant difference was found for ACC cases compared against melanoma (p = 0.2330, OR =16.85, 95% CI [0.16-1746.89]).
Discussion
The results of this multicenter population-based study suggest smoking is associated with an increased odds of adrenal adenomas and adrenocortical carcinomas. State-level estimates of ACC were found to be significantly associated with state-level smoking prevalence, con- firming our hypothesis. Moreover in a separate national TriNetX data- base, smoking was found to confer significantly higher odds of devel- oping adrenal adenoma and adrenocortical carcinoma compared to melanoma controls, with an 86% greater risk in adrenal adenoma to- bacco users and 68% increased risk of ACC in smokers. West Virginia Health System data assessed confounders in these population-level databases and found consistent results for a statistically significant higher odds of adrenal adenoma and ACC with smoking among age- controlled females; however, among age-controlled males, the West Virginia Health System data found smoking was associated with a sta- tistically significant higher odds of adrenal adenoma, but not with ACC. Reasons for this gender discrepancy remains unknown and further study is required to understand and confirm this relationship.
In the 1990s, an exploratory study utilized the 1986 National Mortality Followback Survey questionnaire to construct a case-control study analyzing adrenal cancer risk factors [13]. Including 176 adrenal cancer subjects and 352 controls without smoking history, the study found increased adrenal cancer risk in men who smoked more than 25 cigarettes daily. However, this study used a heterogeneous population of adrenal tumors including neuroblastoma, ganglioblastoma, pheo- chromocytoma, and an ill-defined tumor group termed “other malig- nancies.” In contrast, our study used homogeneous populations and quantifies the association of smoking in each group.
More recently, a report found cigarette smoking as a significant risk factor for ACC, associated with a 50% increased risk [14]. This hospital- based, case-control study included 432 ACC cases and 1204 controls, and found a markedly increased risk of ACC among male cigarette smokers (AOR = 1.8, 95% confidence interval = 1.2-2.9). Our results confirm these findings using US population-level data and additionally
40
80
r=0.4887, p=0.004
r=0.1843, p=0.2099
70
35
60
ACC/million
melanoma/ten thousand
30
50
25
40
20
30
15
20
A
10
15
20
25
10
Current smokers, %
B
15
20
25
Current smokers, %
| Characteristic | Melanoma No. (%) N=98,747 | Adrenal Adenoma No. (%) N=36,434 | P-Value | Adrenocortical Carcinoma No. (%) N=846 | P-Value |
|---|---|---|---|---|---|
| Age | p <. 0001 | p <. 0001 | |||
| < 50 | 13,896 (14.1) | 5,515 (15.1) | 173 (20.4) | ||
| 50 - 69 | 36,791 (37.3) | 17,060 (46.8) | 357 (42.2) | ||
| ≥ 70 | 47,641 (48.2) | 13,857 (38.1) | 315 (37.2) | ||
| Unknown | 419 (0.4) | 2 (0.0) | 1 (0.1) | ||
| Gender | p <. 0001 | p <. 0005 | |||
| Male | 53,320 (54.0) | 14,214 (39.0) | 406 (48.0) | ||
| Female | 45,427 (46.0) | 22,220 (61.0) | 440 (52.0) | ||
| Unknown | 0 (0.0) | 0 (0.0) | 0 (0.0) | ||
| Race | p <. 0001 | p <. 0001 | |||
| White | 84,922 (86.0) | 28,418 (78.0) | 635 (75.1) | ||
| Black | 980 (1.0) | 4,736 (13.0) | 118 (13.9) | ||
| Asian | 112 (0.1) | 401 (1.1) | 17 (2.0) | ||
| American Indian/Alaska Native | 59 (0.1) | 50 (0.1) | 8 (0.9) | ||
| Unknown | 12,674 (12.8) | 2,866 (7.8) | 68 (8.1) | ||
| Region | p <. 0001 | p <. 0001 | |||
| Northeast | 9,191 (9.3) | 2,611 (7.2) | 357 (42.2) | ||
| Midwest | 27,725 (28.1) | 9,499 (26.1) | 141 (16.7) | ||
| South | 32,581 (33.0) | 16,289 (44.7) | 188 (22.2) | ||
| West | 26,886 (27.2) | 5,919 (16.2) | 136 (16.1) | ||
| Unknown | 2,364 (2.4) | 2,116 (5.8) | 24 (2.8) | ||
| Tobacco use | p <. 0001 | p <. 0001 | |||
| Yes | 23,205 (23.5) | 13,262 (36.4) | 287 (33.9) | ||
| No | 75,542 (76.5) | 23,172 (63.6) | 559 (66.1) |
*Melanoma is used as a comparison to generate P-Values for adrenal adenoma and adrenocortical carcinoma characteristics. Significant P-Values in bold.
observed an association of smoking with adrenal adenomas.
In vivo data
In an animal study, rats exposed to 2.5 years of tobacco use at rates congruent to human consumption had significantly higher odds of de- veloping adrenal adenomas and ACC as compared to control rats [19]. Among the 80 rats exposed to chronic cigarette smoke, 1 developed an adrenal adenoma and 3 developed ACC compared to 0 of 93 control rats. Reactive oxygen species generated from tobacco use, genetic variations in carcinogen metabolism, cell-cycle regulation, and DNA repair may influence tobacco-related carcinogenesis [14, 20, 21]. Among tobacco-related carcinogenic agents, N-nitroso compounds have
been shown to induce adrenal dysplasia [22].
Strengths
Strengths include the first United States population-level analysis associating tobacco use with adrenal adenoma and ACCs. A large number of both adrenal adenomas and ACC were included from two US national databases despite the rare incidence of both diseases. As a population-level analysis, these data sources avoid the pitfall of large single-site studies including referral bias.
| Characteristic | Melanoma No. (%) N=6,347 | Adrenal Adenoma No. (%) N=2,161 | P-Value | Adrenocortical Carcinoma No. (%) N=16 | P-Value ** |
|---|---|---|---|---|---|
| Age | p <. 0001 | p =. 0156 | |||
| < 50 | 1,009 (15.9) | 292 (13.5) | 5 (31.2) | ||
| 50 - 69 | 2,528 (39.8) | 990 (45.8) | 9 (56.2) | ||
| ≥ 70 | 2,810 (44.3) | 879 (40.7) | 2 (12.5) | ||
| Unknown | 0 (0.0) | 0 (0.0) | 0 (0.0) | ||
| Gender | p <. 0001 | p =. 3303 | |||
| Male | 3,185 (50.2) | 864 (40.0) | 6 (37.5) | ||
| Female | 3,162 (49.8) | 1,297 (60.0) | 10 (62.5) | ||
| Unknown | 0 (0.0) | 0 (0.0) | 0 (0.0) | ||
| Tobacco use | p <. 0001 | p =. 0011 | |||
| Yes | 1,502 (23.7) | 943 (43.6) | 10 (62.5) | ||
| No | 4,845 (76.3) | 1,218 (56.4) | 6 (37.5) |
*Melanoma is used as a comparison to generate P-Values for adrenal adenoma and adrenocortical carcinoma characteristics. Significant P-Values in bold.
** Fisher’s Exact test was used to generate P-Values.
Limitations
Retrospective analyses performed in this study are not proof of a cause-effect relationship, instead providing evidence of a possible as- sociation between smoking and adrenal adenomas and ACC. Using SEER data, results from the aggregated state-level level cannot be in- ferred to individuals. Therefore, this limits conclusions to an association between state-level ACC cases and state-level adult smoking prevalence. Using TriNetX data, the database structure prohibits adjustment of all significant demographic variables in a multiple logistic regression model. This was dealt with by using the West Virginia Health System data of the same date range to adjust for significant variables including age, gender, and region. However, race was unable to be assessed, as well as oral contraceptive use and alcohol use. Additionally, only 16 ACC cases were extracted from the West Virginia Health System, lim- iting the power of statistical analyses. Data coding conversion from local sites to TriNetX and SEER standards may result in data loss; however, both databases have been extensively tested across healthcare organizations to minimize data loss. Double counting ACC and adrenal adenomas between TriNetX and SEER data is possible since local data collection sites may be duplicated. Neither SEER nor TriNetX data allow the quantification of tobacco use in a dose-dependent manner. Moreover, adrenal adenomas and ACC are mostly discovered during radiological investigation for unrelated diseases and it remains possible advanced diagnostic approaches may be capturing more adrenal ade- nomas and ACC. Given smokers may undergo more imaging than non- smokers, smokers are more likely to be diagnosed with adrenal ade- nomas and ACC incidentally. In a population-based study examining the prevalence of smoking in those with adrenal incidentalomas, Olsen et al. reported 370 of 1044 (35%) patients were smokers [23]. In a separate large cohort study, Di Dalmazi et al. reported 60 of 193 (31%) smokers among patients with nonsecreting adrenal incidentalomas and 42 of 92 (45.7%) smokers among patients with autonomous cortisol secretion [24]. The observed prevalences may be due to causality or selection bias similar to the present study. Future population-level studies are needed to further adjust for socio-demographics as well as other potential confounders and test the dose-response relationship of adrenal adenomas and ACCs to smoking.
Conclusion
Our findings strengthen an association between adrenal adenoma and ACC risk with smoking on a United States national scale. A sig- nificant association was found between state-level smoking prevalence and ACC cases with West Virginia having the highest smoking rate and the second highest ACC rate. Further research is needed to confirm
these findings, test the dose-response relationship between tobacco use and adrenal carcinogenesis, and understand the mechanism of these associations.
Clinical practice points
· What is already known about this subject: Smoking has been suggested as a risk factor for adrenal cortical carcinoma.
· What are new findings: Our findings represent the first United States population-level analysis associating tobacco use with adrenal ade- noma and ACCs. The main take-home points are:
O State-level smoking prevalence was associated with significant increases in ACC.
O Tobacco use was associated with significant increases in adrenal adenoma and ACC in two population-level databases.
· Impact of our findings: Currently, risk factors for adrenal carcino- genesis is limited. Our findings support the association of tobacco use with adrenal adenomas and adrenal cortical carcinoma.
Data accessibility
SEER data is available from the public domain upon request at: https://www.cdc.gov/cancer/uscs/public-use/index.htm. TriNetX and West Virginia Health System data are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Source of funding
No funding to declare.
CRediT authorship contribution statement
Ahmed Yousaf: Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Writing - original draft. Jessica Patterson: Data curation, Formal analysis, Investigation, Methodology. Gerald Hobbs: Formal analysis, Methodology, Software, Visualization, Writing - review & editing. Stephen M. Davis: Investigation, Methodology, Software, Validation. Muhammad Yousaf: Formal analysis, Investigation, Writing - original draft. Maria Hafez: Conceptualization, Methodology, Resources. Heidar Albandar: Conceptualization, Methodology, Supervision. Thomas Hogan: Resources, Supervision, Writing - review & editing. Joanna Kolodney: Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper.
Acknowledgments
This research was supported by the National Cancer Institute and the National Program of Cancer Registries. The interpretation and re- porting of these data are the sole responsibility of the authors.
References
[1] E. Kebebew, E. Reiff, Q .- Y. Duh, O.H. Clark, A. McMillan, Extent of disease at presentation and outcome for adrenocortical carcinoma: have we made progress? World J. Surg. 30 (2006) 872-878.
[2] M. Fassnacht, M. Kroiss, B. Allolio, Update in adrenocortical carcinoma, J. Clin. Endocrinol. Metab. 98 (2013) 4551-4564.
[3] T.M. Kerkhofs, R.H. Verhoeven, J.M. Van der Zwan, J. Dieleman, M.N. Kerstens, T.P. Links, L.V. Van de Poll-Franse, H.R. Haak, Adrenocortical carcinoma: a po- pulation-based study on incidence and survival in the Netherlands since 1993, Eur. J. Cancer 49 (2013) 2579-2586.
[4] B.L. Wajchenberg, M.A. Albergaria Pereira, B.B. Medonca, A.C. Latronico, P.C. Carneiro, V.A. Ferreira Alves, M.C.N. Zerbini, B. Liberman, G.C. Gomes, M.A. Kirschner, Adrenocortical carcinoma: clinical and laboratory observations, Cancer 88 (2000) 711-736.
[5] J .- P. Luton, S. Cerdas, L. Billaud, G. Thomas, B. Guilhaume, X. Bertagna, M .- H. Laudat, A. Louvel, Y. Chapuis, P. Blondeau, Clinical features of adrenocortical carcinoma, prognostic factors, and the effect of mitotane therapy, New Engl. J. Med. 322 (1990) 1195-1201.
[6] E.M. Caoili, M. Korobkin, I.R. Francis, R.H. Cohan, J.F. Platt, N.R. Dunnick, K.I. Raghupathi, Adrenal masses: characterization with combined unenhanced and delayed enhanced CT, Radiology 222 (2002) 629-633.
[7] S. Roman, Adrenocortical carcinoma, Curr. Opin. Oncol. 18 (2006) 36-42.
[8] A.T. Phan, Adrenal cortical carcinoma-review of current knowledge and treatment practices, Hematol./Oncol. Clin. North Am. 21 (2007) 489-507.
[9] G.W. Boland, M.A. Blake, P.F. Hahn, W.W. Mayo-Smith, Incidental adrenal lesions: principles, techniques, and algorithms for imaging characterization, Radiology 249
(2008) 756-775.
[10] T.L. Mazzuco, I. Bourdeau, A. Lacroix, Adrenal incidentalomas and subclinical Cushing’s syndrome: diagnosis and treatment, Curr. Opin. Endocrinol. Diabetes Obes. 16 (2009) 203-210.
[11] J.H. Song, F.S. Chaudhry, W.W. Mayo-Smith, The incidental adrenal mass on CT: prevalence of adrenal disease in 1,049 consecutive adrenal masses in patients with no known malignancy, Am. J. Roentgenol. 190 (2008) 1163-1168.
[12] W .- H. Chow, A.W. Hsing, J.K. Mclaughlin, J. Fraumeni, Smoking and adrenal cancer mortality among United States veterans, Cancer Epidemiol. Prevent. Biomark. 5 (1996) 79-80.
[13] A.W. Hsing, J.M. Nam, H.T. Co Chien, J.K. Mclaughlin, J.F. Fraumeni Jr, Risk factors for adrenal cancer: an exploratory study, Int. J. Cancer 65 (1996) 432-436.
[14] M.A. Habra, M.A. Sukkari, A. Hasan, Y. Albousen, M.A. Elsheshtawi, C. Jimenez, M. Campbell, J.A. Karam, P.H. Graham, R.I. Hatia, Epidemiological risk factors for adrenocortical carcinoma: a hospital-based case-control study, Int. J. Cancer (2019).
[15] Surveillance, epidemiology, and end results (SEER) fact sheet. Available at https:// seer.cancer.gov/about/factsheets/, 2018.
[16] TriNetX, https://live.trinetx.com/, 2020.
[17] M.C. Kessides, L. Wheless, J. Hoffman-Bolton, S. Clipp, R.M. Alani, A.J. Alberg, Cigarette smoking and malignant melanoma: a case-control study, J. Am. Acad. Dermatol. 64 (2011) 84-90.
[18] Surveillance, Epidemiology, and End Results (SEER) Program. Population estimates used in NCI’s SEER*stat software. Available at https://seer.cancer.gov/popdata/ methods.html.
[19] W.E. Dalbey, P. Nettesheim, R. Griesemer, J.E. Caton, M.R. Guerin, Chronic in- halation of cigarette smoke by F344 rats, J. Natl. Cancer Inst. 64 (1980) 383-390.
[20] X. Wu, H. Zhao, R. Suk, D.C. Christiani, Genetic susceptibility to tobacco-related cancer, Oncogene 23 (2004) 6500-6523.
[21] H .- W. Lee, H .- T. Wang, M .- w. Weng, C. Chin, W. Huang, H. Lepor, X .- R. Wu, W.N. Rom, L .- C. Chen, M .- s. Tang, Cigarette side-stream smoke lung and bladder carcinogenesis: Inducing mutagenic acrolein-DNA adducts, inhibiting DNA repair and enhancing anchorage-independent-growth cell transformation, Oncotarget 6 (2015) 33226.
[22] K.D. Brunnemann, J.C. Scott, D. Hoffmann, N-Nitrosomorpholine and other volatile N-nitrosamines in snuff tobacco, Carcinogenesis 3 (1982) 693-696.
[23] H. Olsen, A. Kjellbom, M. Löndahl, O. Lindgren, High prevalence of smoking in patients with adrenal incidentalomas: causality or case selection?, 183 (2020) 335.
[24] G. Di Dalmazi, F. Fanelli, G. Zavatta, S. Ricci Bitti, M. Mezzullo, A. Repaci, C. Pelusi, A. Gambineri, P. Altieri, C. Mosconi, The steroid profile of adrenal incidentalomas: subtyping subjects with high cardiovascular risk, J. Clin. Endocrinol. Metab. 104 (2019) 5519-5528.