Check for updates
Diabetes mellitus and hyperglycemia are associated with inferior oncologic outcomes in adrenocortical carcinoma
Sean M. Wrenn 1,2,3 D . T. K. Pandian 2,3,4 D . Rajshri M. Gartland2,3 D . Zhi Ven Fong3 D . Matthew A. Nehs 2,5 D
Received: 15 June 2020 / Accepted: 14 December 2020 C Springer-Verlag GmbH Germany, part of Springer Nature 2021
Abstract
Purpose Prior literature suggests that cancer patients with hyperglycemia and type 2 diabetes mellitus (DM) exhibit worse oncologic and overall outcomes. Tumor metabolism and anabolism pathophysiology may explain this association, although this has not been adequately studied in adrenocortical carcinoma (ACC). We hypothesized that DM would be associated with worse oncological outcomes in ACC, and we utilized data from a national database and institutional sources for multimodal analysis.
Methods Both a multi-institutional database (the Collaborative Endocrine Surgery Quality Improvement Program or CESQIP) and a single-center longitudinal cohort (Dana Farber Cancer Institute or DFCI) were queried as unique retrospective cohorts to identify patients with ACC. Patient demographics, tumor characteristics, DM-specific variables, and oncologic outcome data were assessed. Results were analyzed via univariate analysis and multivariable linear regression analysis. Statistical significance was defined as p< 0.05.
Results Forty-eight CESQIP patients met inclusion criteria; 16 (33.0%) had DM. DM patients had a higher frequency of recurrence on longitudinal follow-up (12.5% v 0.0%, p =0.04). Persistent disease was observed in 68.8% of DM patients and 40.6% of non-DM patients (p=0.06). Patients in the DFCI cohort with lower average glucose values (<110 mg/dL) had a significant survival benefit (p <. 0001). A mean serum glucose > 110 mg/dL had increased risk (HR 36.3, 95% confidence interval 1.6, 831.3) for all-cause mortality.
Conclusions This multi-institutional, multimodal analysis suggests that patients with DM have worse oncologic and overall outcomes for ACC. While further study is warranted, consideration should be given among clinicians to optimize glycemic control as part of their ACC management.
Keywords Adrenal gland . Adrenocortical carcinoma . Endocrine surgery . Surgical metabolism . Quality · Outcomes · Diabetes mellitus · Hyperglycemia · Warburg effect
* ☒ Sean M. Wrenn Swrenn1@bwh.harvard.edu
1 Department of Surgery, Division of Surgical Oncology, Rush University Medical Center, 1725 W. Harrison St. Suite 810, Chicago, IL 60612, USA
2 Department of Surgery, Brigham and Women’s Hospital, Boston, MA, USA
3 Department of Surgery, Massachusetts General Hospital, Boston, MA, USA
4 Department of Surgery, Washington University in St. Louis, St. Louis, MO, USA
5 Surgical Oncology, Dana Farber Cancer Center, Boston, MA, USA
Abbreviations
| ACC | Adrenocortical carcinoma |
| CESQIP | Collaborative Endocrine |
| Surgery Quality Improvement Program | |
| DM | Diabetes mellitus |
| DFCI | Dana Farber Cancer Institute |
| AAES | American Association of Endocrine Surgeons |
| BMI | Body mass index |
Introduction
Although rare, adrenocortical carcinoma (ACC) is notable for poor outcomes and high mortality. While few prognostic var- iables are established, complete surgical (R0) resection and
low Ki-67 proliferation index (typically defined as < 10% [1, 2]) have been correlated with improved outcomes [3]. However, recurrence, progression, and metastasis remain common after initial presentation. It has previously been ob- served that ACCs that secrete cortisol have worse outcomes than those that do not exhibit hypercortisolemia [4]. Multiple proposed mechanisms for this association have been sug- gested, including immune system modulation and increased rates of venous thromboembolism. However, the association remains understudied and is assumed to be multifactorial.
Hypercortisolemia has a broad range of systemic effects, one of which is hyperglycemia and hyperinsulinemia with resultant proclivity to type 2 diabetes mellitus (DM). Hyperglycemia has been demonstrated to raise both the prev- alence and mortality for multiple types of cancer including breast, pancreatic, colorectal, bladder, liver, and endometrial [5]. Both in vivo and in vitro studies have demonstrated that hyperglycemia can promote tumorigenesis, increased tumor progression, and metastasis [6-8].
We hypothesized that DM would be associated with worse oncological outcomes in ACC, by providing additional tumor substrate in the form of free plasma glucose. We also hypoth- esized that hyperglycemia within the general population, in- cluding those with no previous diagnosis of DM, would also be associated with worse outcomes via the same proposed mechanism. To answer this question, we utilized two unique cohorts of patients independently: a single-center longitudinal cohort from Dana Farber Cancer Institute (DFCI) and a multi- institutional database cohort from the Collaborative Endocrine Surgery Quality Improvement Program (CESQIP).
Material and methods
CESQIP database analysis
CESQIP is a Center for Medicare and Medicaid Services- approved Qualified Clinical Data Registry, founded in 2012 by members of the American Association of Endocrine Surgeons [9, 10]. It focuses on over 300 endocrine surgery- specific variables with a goal towards quality improvement. Deidentified data provided securely via Arbormetrix was pro- vided after a written proposal outlining the research aims was provided to the CESQIP committee of the American Association of Endocrine Surgeons (AAES), along with a signed data use agreement. Appropriate institutional IRB ap- proval was obtained.
The CESQIP database contains a distinct adrenal surgery module from which we obtained deidentified data for analysis. The database was analyzed from its inception in January 2014 to April 2019 for patients with pathology-proven ACC utiliz- ing diagnosis codes. Clinical data from each CESQIP partic- ipating site is obtained via specialized research data
abstractors who typically obtain the data directly from the electronic medical record; however, criteria for defining each variable are left to the discretion of the abstractors. The co- morbidity of DM is recorded as a dichotomous variable (yes/ no) within CESQIP. A variety of clinical demographic and outcome variables were assessed, which are noted in Table 1.
Dana Farber Cancer Institute cohort analysis
In order to study longitudinal efforts for survival analysis and to study the effect of glycemic control on ACC outcomes, a cohort of patients who received care at DFCI and/or Brigham and Women’s Hospital was additionally reviewed. Appropriate Institutional Review Board approval was obtain- ed prior to the study. Included patients had a diagnosis of ACC and were seen by one of the adrenal carcinoma special- ists. This cohort of patients was diagnosed during a period spanning 2000-2018 and followed longitudinally.
We analyzed this cohort within our institution to study longitudinal and granular details not achievable within the CESQIP database. Patient charts were retrospectively reviewed from the point of their initial evaluation by either a surgeon, endocrinologist, or medical endocrinologist that composed the multidisciplinary adrenal cancer team. Patient oncological outcomes were measured and followed longitudi- nally from diagnosis until the present time, unless they be- came deceased or were lost to follow-up. Patients were assessed for the presence of DM, and their medications were reviewed for the presence of exogenous corticosteroid, mitotane, and insulin. Laboratory markers associated with DM, specifically hemoglobin A1c, fasting glucose, and ran- dom glucose values, were recorded for each patient (when available). All the patients in the cohort had pathologically confirmed ACC.
To facilitate quantification and comparison, some of the Ki-67 and mitoses values had to be augmented and standard- ized. All Ki-67 values that were reported as a range (e.g., 40- 50%) were assigned a value at the median of the upper and lower aspects of the range (e.g., 45%). Mitoses that were re- ported as less than 50 high powered fields (the standard for ACC) were extrapolated to 50 high powered fields via multi- plication (e.g., 5 mitoses per 10 high powered fields was ex- trapolated to 25 mitoses per 50 high powered fields).
A univariate statistical analysis was performed to deter- mine whether differences in dependent variables between the defined groups (glucose <110, glucose 110-140, glucose > 140) were statistically significant (as defined as p ≤.05). The categorical variables were analyzed using the chi-squared test. The continuous variables that were presumed to be normal in distribution were analyzed using parametric testing (unpaired or Student’s t test), whereas continuous variables with non- normal or unknown to be normal distribution were analyzed using non-parametric testing (Kruskal-Wallis H test).
| Variable | Entire population | Diabetes mellitus | No Hx of DM | p value |
|---|---|---|---|---|
| n | 48 | <20 | 32 | |
| Patient demographics | ||||
| Female gender (%) | 62.5% | 81.3% | 53.1% | 0.05 |
| BMI >40 (%) | 12.5% | 25.0% | 6.3% | 0.06 |
| Mean age (SD) | 52.2 (15.6) | 54.0 (12.7) | 51.0 (15.9) | 0.48 |
| Hypertension (%) | 60.4% | 81.3% | 50.0% | 0.03 |
| Anticoagulation | 18.8% | 18.8% | 18.8% | > .99 |
| Race (%) | ||||
| African American | 2.1% | 6.3% | 0.0% | 0.09 |
| White | 81.3% | 68.8% | 87.5% | |
| Asian | 0.0% | 0.0% | 0.0% | |
| Hispanic | 12.5% | 25.0% | 6.3% | |
| Not recorded | 4.2% | 0.0% | 6.3% | |
| Disease characteristics | ||||
| Documented hypercortisolism (%) | 39.6% | 56.3% | 31.3% | 0.06 |
| Operative approach (%) | ||||
| Laparoscopic | 18.8% | 31.3% | 12.5% | 0.24 |
| Laparoscopic converted to open | 2.1% | 0.0% | 3.1% | |
| Open | 79.2% | 68.8% | 84.4% | |
| Tumor size in cm (SD) | 9.0 (4.6) | 8.4 (4.0) | 9.3 (4.9) | 0.5 |
| M1 disease (%) | 18.8% | 12.5% | 21.9% | 0.43 |
| Stage IV disease (%) | 31.3% | 37.5% | 28.1% | 0.5 |
| N1 disease (%) | 14.6% | 18.8% | 12.5% | 0.56 |
| Perioperative outcomes (%) | ||||
| SSI | 4.2% | 12.5% | 0.0% | 0.04 |
| Non-specific complication | 10.4% | 18.8% | 6.3% | 0.18 |
| Readmission | 4.2% | 6.3% | 3.1% | 0.6 |
| Unplanned return to OR | 2.1% | 0.0% | 3.1% | 0.47 |
| Renal failure | 2.1% | 0.0% | 3.1% | 0.47 |
| Cardiac arrest | 4.2% | 0.0% | 6.3% | 0.3 |
| VTE | 4.2% | 6.3% | 3.1% | 0.61 |
| Mortality | 2.1% | 0.0% | 3.1% | 0.47 |
| Oncologic outcomes (%) | ||||
| Positive tumor margin/R1 resection (%) | 14.6% | 25.0% | 9.4% | 0.14 |
| Recurrent disease | 4.2% | 12.5% | 0.0% | 0.04 |
| Persistent disease | 50.0% | 68.8% | 40.6% | 0.06 |
Patients were compared based on the independent variable of diabetes mellitus and based on a dichotomous variable measuring this within CESQIP. Univariate statistical analysis was performed comparing outcomes between the diabetes mellitus group and the non-diabetes mellitus group using chi-squared or unpaired 2 tailed t test. CESQIP Collaborative Endocrine Surgery Quality Improvement Program, DM diabetes mellitus, Hx history, SD standard deviation, BMI body mass index, OR operating room, VTE venous thromboembolism
Survival analysis
The 61 patients in the DFCI cohort with at least one re- corded plasma glucose value recorded in their health re- cord after their diagnosis of ACC were stratified into three tiers to represent their glycemic control: normoglycemia (glucose <110 mg/dL, n =38), mild hyperglycemia
(glucose 110-140 mg/dL, n = 15), and moderate-severe hyperglycemia (glucose > 140 mg/dL, n = 8). Survival curves were formulated by determining time in days for each subject to death or censure (lost to follow-up). Statistical analysis was performed with log-rank (Mantel-Cox) test on GraphPad Prism (GraphPad Prism LLC, Version 8.4.2) software.
Multivariable analysis
In order to control for pre-existing differences between groups that might confound the result, we performed a multivariable Cox regression comparing the euglycemia (glucose < 110 mg/ dL) and hyperglycemia (glucose ≥ 110 mg/dL) groups. In order to have adequate power for multivariable analysis, we selected 6 additional predictor variables to control for that were (1) felt to most likely to confound the association be- tween hyperglycemia and mortality and/or (2) demonstrated significant differences between groups on univariate analysis (age, presence of hypertension, presence of coronary artery disease, mean serum cortisol levels, presence of chronic ste- roid use, Ki-67 proliferation index of tumor). Variables found to be colinear such as hyperlipidemia and coronary artery disease were addressed by eliminating one correlated variable (presence of hypertension) from analysis. Two additional sen- sitivity analyses were performed, one substituting tumor laterality for serum cortisol levels, and one running fewer variables (5). Significance was defined by p ≤ 0.05, and 95% confidence intervals were obtained for hazard ratio for the outcome of all-cause mortality. Statistical analysis was performed using Stata/SE 14 (StataCorp LLC, College Station, TX).
Results
CESQIP
Of a total 1510 adrenalectomies within the CESQIP database during the study period, 48 carried a diagnosis of ACC and met criteria for inclusion. Included patients (n =48) underwent adrenalectomy at < 20 unique sites by 22 surgeons. The DM group had higher levels of female (81.3% v 53.1%, p =0.05) and hypertensive (81.3% v 50.0%, p=0.03) sub- jects. DM patients also had a higher frequency of postopera- tive superficial skin infection (12.5% v 0.0%, p = 0.04); how- ever, other perioperative outcomes were similar. Tumor size was similar between groups, as was the incidence of locoregional and distant spread of disease. Oncologic out- comes were notable for a significantly (p=0.04) lower risk of recurrent disease in the non-DM group. Perioperative mor- tality was similar between groups; however, long-term overall and disease-specific survival were not followed in the data- base. Comprehensive univariate comparison in demographics, perioperative outcomes, and oncologic (longitudinal) out- comes can be viewed in both Table 1 and Fig. 1.
DFCI cohort
A total of 67 patients were analyzed as patients with con- firmed pathological diagnosis of adrenocortical carcinoma
whom had been seen within the multidisciplinary adrenal group between 2012 and 2019. Of these, 61 patients had at least one recorded serum glucose measured after their initial diagnosis and were included for analysis (population medi- an = 23 recorded plasma glucose values per subject, popula- tion mean = 34.3). Their initial diagnosis of ACC spanned from 2018 to as far back as 2000. The complete univariate analysis of their outcome variables can be seen in Table 2. Multivariable analysis comparing the hyperglycemia group (≥ 110 mg/dL) with euglycemia group (plasma glucose < 110 mg/dL) demonstrated a hazard ratio (HR) of 36.3 (95% CI 1.6, 831.3) for all-cause mortality in the hyperglycemia group (Z-score 2.25, p= 0.025). Additional sensitivity analy- ses (one incorporating tumor laterality as a variable, and one with fewer variables) demonstrated consistent results. Multivariable results can be seen in more depth in Supplemental Table 1, which includes both the index results (1a) with two additional sensitivity analyses (1b, 1c).
Survival analysis of DFCI cohort
Log-rank test demonstrated a significant (p =. 0002) survival advantage for the subject group with the most optimal glyce- mic control (glucose < 110 mg/dL) as seen in Fig. 2.
Discussion
While numerous studies have demonstrated associations be- tween DM and poor oncological outcome in other forms of malignancy [5-8, 11], here we build on this important associ- ation in ACC. Given this common association in other malig- nancies, we sought to characterize the outcomes in patients with ACC who had known diagnosis of DM using the CESQIP database and our institutional (DFCI) cohort. We further tested this hypothesis using plasma glucose levels as a proxy for glycemic control since hemoglobin A1C levels were not routinely obtained in these patients. We believe that DM, hyperglycemia, and insulin resistance may represent im- portant physiological links between altered tumor metabo- lism, hypercortisolism, and tumor aggressivity in ACC.
The association between hyperglycemia and tumor anabo- lism was established in the early days of metabolic oncology research. In 1956, Otto Warburg famously noted the enhanced ability of cancer cells to utilize glucose via aerobic glycolysis [12]. This key observation has been fortified over time with molecular research demonstrating that key glycolytic genes and proteins responsible for glucose metabolism are upregu- lated in multiple malignancies [13-16]. For example, Fenske et al. demonstrated that increased expression of hypoxia- responsive glucose transporter isoform 1 (GLUT1) was asso- ciated with a worse prognosis in ACC [17]. The implication of this is that many cancers, including ACC, depend on abundant
100
80.
p=0.06
Percent %
60
40
p=0.04
p=0.14
*
20-
0
Positive tumor margin
Recurrent disease
Persistent disease
DM
No history of DM
glucose in their environment, and specific mutations can help them rapidly procure bioavailable glucose which in turn ac- celerates disease progression. Additionally, chronic hyperinsulinemia has been linked to an increase risk in cancer wherein increased bioavailability of insulin-like growth factor leads to cellular changes that favor tumor formation (mitogenesis, anti-apoptosis) [11]. The increased risk of cancer-related mortality from hyperinsulinemia has been ob- served in both obese and non-obese populations [18]. Together, these studies suggest that insulin and glucose me- tabolism are critical in the development and promulgation of multiple malignant tumors.
Adrenocortical carcinoma (ACC) is a rare and aggressive neoplasm with often poor oncological outcomes [19]. Adrenal steroid hormone hyperproduction is a common feature of ACCs, and the hormonal profile will tend to dictate the phe- notypic disease presentation and prognosis. Adrenocortical carcinoma presents with significant clinical and prognostic heterogeneity, which is at least in part attributable to differ- ences in tumor hormonal production. In particular, the excess production of glucocorticoid resulting in Cushing’s syndrome is present in approximately 22-40% of ACC [19, 20]. The subgroup of ACC patients with hypercortisolism have been demonstrated in multiple studies to have worse outcomes, including shorter progression-free survival [20] and higher perioperative complication rate [20]. The deleterious out- comes in ACC associated with hypercortisolism have also been demonstrated in tumors that underwent radical resection [21].
The systemic manifestations of glucocorticoid excess are diverse and include hypertension, facial plethora, easy bruis- ing, proximal myopathy, diabetes mellitus, increased risk of venous thromboembolism, and weight gain. In women, it may mimic symptoms of polycystic ovarian syndrome by causing acne, hirsutism, or oligomenorrhea [19]. Because of the broad manifestations, there have been numerous hypotheses as to
why cortisol-producing ACC have worse outcomes: increased frailty, differences in mutation status leading to upregulated enzymatic machinery [22], and immunosuppression. Also of note, Cushing’s syndrome independent of ACC is associated with higher mortality than the general population with in- creased risk of myocardial infarction, stroke, infection, and venous thromboembolism [23]. While the ultimate reason for worse outcomes in hypercortisolemia is likely multifacto- rial, we hypothesize based on our data that the hyperglycemia that frequently results from hypercortisolism could be inde- pendently associated with disease progression and decreased survival. However, further studies are required to delineate and separate the myriad of physiological effects seen in Cushing’s hypercortisolism with hyperglycemic effects.
In this study, we found that random glucose levels, when measured serially, could help stratify outcomes in patients with ACC. This may suggest that controlling cortisol excess and hyperglycemia may improve outcomes in patients with ACC, but that would need to be addressed in a prospective clinical trial to definitively establish a causative link. Although hyperglycemic groups appeared to be associated with hypercortisolism (as expected physiologically), we can- not demonstrate or imply causation within these two associa- tions. Future studies that prospectively collect data on patient metabolism, tumor oncological status, cortisol production, and survival may further elucidate whether hyperglycemia is the mechanism leading the worse outcomes in ACC patients with hypercortisolism.
Limitations
The findings of this study must be interpreted in the context of several important limitations. First, based on our best available data, we included the use of random plasma glucose measure- ments as a proxy for each subject’s glycemic control over the stated study period. This is suboptimal given the significant
| Glucose <110 | Glucose 110- 140 | Glucose >140 | p | |
|---|---|---|---|---|
| Total subjects (n) | 38 | 15 | 8 | |
| 1. Patient characteristics and demographics | ||||
| Mean age at diagnosis (years) | 47.4 | 52.3 | 61.75 | 0.04 |
| Gender (% female) | 65.8% | 46.7% | 50.0% | 0.50 |
| Mean BMI | 27.2 | 27.3 | 27.8 | 0.96 |
| Metabolic syndrome diagnoses: | ||||
| Diabetes mellitus | 0% | 13.3% | 62.5% | < 0.0001 |
| Hyperlipidemia | 26.3% | 40.0% | 62.5% | 0.13 |
| Hypertension | 52.6% | 46.7% | 75% | 0.41 |
| Obstructive sleep apnea | 5% | 6.7% | 0% | 0.77 |
| Coronary or peripheral vascular disease | 16% | 0% | 12.5% | 0.27 |
| Chronic steroid use (% yes) | 60.5% | 60% | 37.5% | 0.48 |
| Mean serum cortisol level | 18.2 | 24.2 | 38.6 | 0.03 |
| Chronic insulin use (% yes) | 0% | 6.7% | 37.5% | 0.001 |
| 2. Oncologic parameters | ||||
| Laterality (% left) | 50.0% | 73.3% | 25.0% | 0.08 |
| Mean tumor size (cm) | 11.8 | 13.6 | 10.3 | 0.23 |
| % Underwent planned surgical resection | 92.1% | 80.0% | 75.0% | 0.28 |
| Additional organs resected (% of total surgeries) | ||||
| Kidney | 29.0% | 60.0% | 60.0% | 0.07 |
| Spleen | 2.6% | 40.0% | 20.0% | 0.002 |
| Distal pancreas | 0% | 26.7% | 20.0% | 0.01 |
| Liver (partial) | 21.0% | 0% | 20.0% | 0.16 |
| Inferior vena cava reconstruction (% Yes) | 13.0% | 0% | 0% | 0.19 |
| Diagnosed hereditary syndrome | 2.6% | 6.7% | 0% | 0.86 |
| Other concurrent malignancy | 2.6% | 26.7% | 0% | 0.01 |
| Mean Ki-67 (%) | 17.1% | 40.6% | 19.0% | 0.03 |
| Average mitoses per 50 high power fields | 39.0 | 45.2 | 73.1 | 0.20 |
| Functional hormone secretion of tumor | ||||
| Cortisol | 39.5% | 40.0% | 50.0% | 0.86 |
| DHEA/androgen | 29% | 53.3% | 37.5% | 0.84 |
| Aldosterone | 7.9% | 0% | 0% | 0.38 |
| Catecholamine | 2.6% | 0% | 0% | 0.74 |
| Adjuvant mitotane use (% yes) | 63.2% | 80.0% | 50.0% | 0.31 |
| % Not tolerated | 13.2% | 0% | 25.0% | 0.17 |
| 3. Oncologic outcomes | ||||
| Survival to completion (% yes) | 54.1% | 33.3% | 25.0% | 0.19 |
| Mean days to death or censure from | 1913 | 1470.7 | 366.8 | |
| diagnosis | ||||
| Median days to death or censure | 1414.5 | 835 | 231.5 | < . 001 |
| Recurrent or persistent disease (% yes) | 68.4% | 85.7% | 75.0% | 0.45 |
| Progressed (or presented) stage IV disease | 63.2% | 73.3% | 62.5% | 0.77 |
Patients diagnosed with and evaluated for adrenocortical carcinoma during the period of 2012-2019 were strat- ified into three groups based on the average of their recorded serum glucose values (<110 mg/dL, 110-140 mg/ dL, > 140 mg/dL) and compared based on their demographics and oncologic outcomes. Univariate analysis is again performed with chi-squared, Kruskal-Wallis H test, or one-way ANOVA
variability in the quantity and timing of these measurements, and also significant heterogeneity in the circumstances
surrounding each draw (postprandial, postoperative, fasting, inpatient, outpatient, etc.). However, prior studies have
Probability of Survival
100
— < 110
- 110-140
Log-rank p =. 0002
- >140
50
<110
>140
110-140
0
0
1000
2000
3000
4000
5000
Days
demonstrated that random plasma glucose, in lieu of fasting plasma glucose or hemoglobin A1c, can be a reliable indicator of poor glycemic control [24, 25]. While isolated individual glucose values can prove unreliable, large quantities of such “random” values measured serially over time can illuminate an aggregate picture of a patient’s glycemic control.
Second, the CESQIP database was limited in longitudinal follow-up and lacks additional variables in the database re- garding diabetes and metabolism. It tracks diabetes as a di- chotomous variable without accounting for severity of hyper- glycemia and disease control. Moving forward it would be informative to see how outcomes vary based on the spectrum of this disease. Third, the DFCI cohort may be affected by survivorship bias for the patients that were diagnosed with ACC prior to 2012, as we did not follow the patients that were diagnosed prior to 2012 and were deceased prior to the begin- ning of the inclusion period. Lastly, due to limitations in the statistical power of the study that stem from a relatively low number of subjects, we were unable to statistically control for all potential confounding variables accounting for differences in mortality between different levels of hyperglycemia.
Conclusion
This study demonstrates a strong association between hyper- glycemia and deleterious outcomes in adrenocortical carcino- ma and thus supports a potential mechanism for the worse outcomes seen in cortisol-producing ACC tumors. It remains unknown as to whether hyperglycemia is the primary causa- tive mechanism by which hypercortisolism confers worse out- comes or if it is a downstream consequence. Future studies should prospectively collect more granular metabolic data on all ACC patients, including fasting glucose values, fasting insulin levels, and hemoglobin A1c. In the interim, we suggest clinicians be mindful in optimizing glycemic control and met- abolic therapy (including a nutritional assessment) as part of the oncological care of ACC patients.
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00423-020-02061-0.
Acknowledgements The authors would like to thank the AAES research committee of CESQIP, particularly Dr. Dave Schneider for making these data available. We would like to thank Lia Wrenn MD for her critical appraisal of the manuscript and for her support.
Author’s contributions Sean Wrenn: Study conception and design, drafting of manuscript, analysis and interpretation of data, and critical revision of manuscript. TK Pandian: Study conception and design and critical revision of manuscript. Rajshri Gartland: Analysis and interpreta- tion of data and critical revision of manuscript. Zhi Ven Fong: Analysis and interpretation of data and critical revision of manuscript. Matthew Nehs: Study conception and design, analysis and interpretation of data, drafting of manuscript, and critical revision of manuscript.
Data availability Data is available via CESQIP.org and by written request. These data were previously presented in limited fashion at the virtual CESQIP session hosted by the American Association of Endocrine Surgeons (AAES) on May 18, 2020.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Ethical approval This retrospective chart review study involving human participants was in accordance with the ethical standards of the institu- tional and national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
This article does not contain any studies with animals performed by any of the authors.
Informed consent Given the retrospective nature of this research, a waiver of informed consent was obtained by the Institutional Review Board.
Disclaimer “CESQIP and the hospitals participating in CESQIP are the source of the data used herein; they have not verified and are not respon- sible for the statistical validity of the data analysis or the conclusions derived by the authors. The conclusions, findings, and opinions expressed by the authors do not necessarily reflect the official position of the AAES or CESQIP. Use of CESQIP data does not imply endorsement by any of the groups named above.”
References
1. Beuschlein F, Weigel J, Saeger W, Kroiss M, Wild V, Daffara F, Libé R, Ardito A, al Ghuzlan A, Quinkler M, Oßwald A, Ronchi CL, de Krijger R, Feelders RA, Waldmann J, Willenberg HS, Deutschbein T, Stell A, Reincke M, Papotti M, Baudin E, Tissier F, Haak HR, Loli P, Terzolo M, Allolio B, Müller HH, Fassnacht M (2015) Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resection. J Clin Endocrinol Metab 100(3):841-849. https://doi.org/10.1210/jc.2014-3182
2. Papathomas TG, Pucci E, Giordano TJ, Lu H, Duregon E, Volante M, Papotti M, Lloyd RV, Tischler AS, van Nederveen FH, Nose V, Erickson L, Mete O, Asa SL, Turchini J, Gill AJ, Matias-Guiu X, Skordilis K, Stephenson TJ, Tissier F, Feelders RA, Smid M, Nigg A, Korpershoek E, van der Spek PJ, Dinjens WNM, Stubbs AP, de Krijger RR (2016) An international Ki67 reproducibility study in adrenal cortical carcinoma. Am J Surg Pathol 40(4):569-576. https://doi.org/10.1097/PAS.0000000000000574
3. Terzolo M, Boccuzzi A, Bovio S, Cappia S, de Giuli P, Alì A, Paccotti P, Porpiglia F, Fontana D, Angeli A (2001) Immunohistochemical assessment of Ki-67 in the differential diag- nosis of adrenocortical tumors. Urology. 57:176-182
4. Berruti A, Terzolo M, Sperone P, Pia A, Della Casa S, Gross DJ, Carnaghi C, Casali P, Porpiglia F, Mantero F, Reimondo G, Angeli A, Dogliotti L (2005) Etoposide, doxorubicin and cisplatin plus mitotane in the treatment of advanced adrenocortical carcinoma: a large prospective phase II trial. Endocr Relat Cancer 12:657-666
5. . Li W, Zhang X, Sang H, Zhou Y, Shang C, Wang Y, Zhu H (2019) Effects of hyperglycemia on the progression of tumor diseases. J Exp Clin Cancer Res 38:327
6. Joshi S, Lie M, Turner N (2015) Diabetes and its link with cancer: providing the fuel and spark to launch an aggressive growth regi- men. Biomed Res Int 2015:390863
7. Hou Y, Zhou M, Xie J et al (2017) High glucose levels promote the proliferation of breast cancer cells through GTPases. Breast Cancer (Dove Med Press) 9:429-436
8. Han L, Ma Q, Li J, Liu H, Li W, Ma G, Xu Q, Zhou S, Wu E (2011) High glucose promotes pancreatic cancer cell proliferation via the induction of EGF expression and transactivation of EGFR. PLoS One 6:e27074
9. Talutis SD, Drake FT, Sachs T, Rao SR, McAneny D (2019) Evacuation of postoperative hematomas after thyroid and parathy- roid surgery: an analysis of the CESQIP database. Surgery. 165(1): 250-256. https://doi.org/10.1016/j.surg.2018.04.087
10. Homepage to the Collaborative Endocrine Surgery Quality Improvement Program. https://cesqip.org Site accessed May 13, 2020
11. Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M (2008) Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet. 371(9612):569-578. https://doi.org/10.1016/S0140-6736(08) 60269-X
12. Warburg O (1956) On the origin of cancer cells. Science. 123(3191):309-314. https://doi.org/10.1126/science.123.3191.309
13. Langbein S, Zerilli M, Zur Hausen A et al (2006) Expression of transketolase TKTL1 predicts colon and urothelial cancer patient survival: Warburg effect reinterpreted. Br J Cancer 94(4):578-585. https://doi.org/10.1038/sj.bjc.6602962
14. Zhou S, Wang S, Wu Q, Fan J, Wang Q (2008) Expression of glucose transporter-1 and -3 in the head and neck carcinoma-the correlation of the expression with the biological behaviors. ORL J Otorhinolaryngol Relat Spec 70(3):189-194. https://doi.org/10. 1159/000124293
15. Smith TA (1999) Facilitative glucose transporter expression in hu- man cancer tissue. Br J Biomed Sci 56(4):285-292
16. Reske SN, Grillenberger KG, Glatting G, Port M, Hildebrandt M, Gansauge F, Beger HG (1997) Overexpression of glucose trans- porter 1 and increased FDG uptake in pancreatic carcinoma. J Nucl Med 38(9):1344-1348
17. Fenske W, Völker HU, Adam P, Hahner S, Johanssen S, Wortmann S, Schmidt M, Morcos M, Müller-Hermelink HK, Allolio B, Fassnacht M (2009) Glucose transporter GLUT1 expression is an stage-independent predictor of clinical outcome in adrenocortical carcinoma. Endocr Relat Cancer 16:919-928
18. Tsujimoto T, Kajio H, Sugiyama T (2017) Association between hyperinsulinemia and increased risk of cancer death in nonobese and obese people: a population-based observational study. Int J Cancer 141(1):102-111. https://doi.org/10.1002/ijc.30729
19. Puglisi S, Perotti P, Pia A, Reimondo G, Terzolo M (2018) Adrenocortical carcinoma with hypercortisolism. Endocrinol Metab Clin N Am 47(2):395-407. https://doi.org/10.1016/j.ecl. 2018.02.003
20. Margonis GA, Kim Y, Tran TB, Postlewait LM, Maithel SK, Wang TS, Glenn JA, Hatzaras I, Shenoy R, Phay JE, Keplinger K, Fields RC, Jin LX, Weber SM, Salem A, Sicklick JK, Gad S, Yopp AC, Mansour JC, Duh QY, Seiser N, Solorzano CC, Kiernan CM, Votanopoulos KI, Levine EA, Poultsides GA, Pawlik TM (2016) Outcomes after resection of cortisol-secreting adrenocortical carci- noma. Am J Surg 211(6):1106-1113. https://doi.org/10.1016/j. amjsurg.2015.09.020
21. Berruti A, Fassnacht M, Haak H, Else T, Baudin E, Sperone P, Kroiss M, Kerkhofs T, Williams AR, Ardito A, Leboulleux S, Volante M, Deutschbein T, Feelders R, Ronchi C, Grisanti S, Gelderblom H, Porpiglia F, Papotti M, Hammer GD, Allolio B, Terzolo M (2014) Prognostic role of overt hypercortisolism in completely operated patients with adrenocortical cancer. Eur Urol 65(4):832-838. https://doi.org/10.1016/j.eururo.2013.11.006
22. Ronchi CL, Sbiera S, Leich E, Tissier F, Steinhauer S, Deutschbein T, Fassnacht M, Allolio B (2012) Low SGK1 expression in human adrenocortical tumors is associated with ACTH-independent gluco- corticoid secretion and poor prognosis. J Clin Endocrinol Metab 97(12):E2251-E2260. https://doi.org/10.1210/jc.2012-2669
23. Dekkers OM, Horváth-Puhó E, Jørgensen JO et al (2013) Multisystem morbidity and mortality in Cushing’s syndrome: a cohort study. J Clin Endocrinol Metab 98(6):2277-2284. https:// doi.org/10.1210/jc.2012-3582
24. el-Kebbi IM, Ziemer DC, Gallina DL, Phillips LS (1998) Diabetes in urban African-Americans. VI. Utility of fasting or random glu- cose in identifying poor glycemic control. Diabetes Care 21(4): 501-505. https://doi.org/10.2337/diacare.21.4.501
25. Gill GV, Hardy KJ, Patrick AW, Masterson A (1994) Random blood glucose estimation in type 2 diabetes: does it reflect overall glycaemic control? Diabet Med 11(7):705-708. https://doi.org/10. 1111/j.1464-5491.1994.tb00337.x
Publisher’s note Springer Nature remains neutral with regard to jurisdic- tional claims in published maps and institutional affiliations.