ELSEVIER
Surgery
journal homepage: www.elsevier.com/locate/surg
SURGERY
NOVEMBER 2018
$ -
A
=
====
Identifying genomic signatures of recurrence in adrenocortical carcinoma after R0 resection
Check for updates
Benjamin C. Greenspun, MD*, Dawn Chirko, BS, Rajbir Toor, BA, Kyle Wierzbicki, BS, Teagan E. Marshall, MD, Abhinay Tumati, MD, Rasa Zarnegar, MD, FACS, Thomas J. Fahey III, MD, FACS, Brendan M. Finnerty, MD, FACS
Department of Surgery, Weill Cornell Medicine, New York, NY
ARTICLE INFO
Article history: Accepted 20 September 2024 Available online 20 October 2024
ABSTRACT
Background: Adrenocortical carcinoma (ACC) is a rare and aggressive malignancy with limited treatment options. Although there have been recent advancements revealing genomic drivers of these tumors, it remains unclear which genomic signatures are associated with recurrence, particularly following R0 resection.
Methods: Adrenocortical carcinoma patients treated with adrenalectomy in the Cancer Genome Atlas with recurrence data were identified using cBioPortal. Clinicopathologic variables, genomics, treatment patterns, and outcomes were retrospectively analyzed.
Results: Among 92 adrenocortical carcinoma patients, 84 had recurrence data, with 52% experiencing tumor recurrence. Age and sex were not significantly different between recurrent and nonrecurrent groups. Nonrecurrent patients had a significantly longer overall survival (54 months vs 35 months, P = . 0036). Adjuvant radiation was administered similarly in both groups (25.0% vs 16.2%, P = . 4164). There were no differences in capsular or venous invasion or median tumor size. Sixty-two patients had RO resection and 40.3% (n = 25/62) recurred. Multivariate logistic regression in this cohort, when con- trolling for vascular invasion, venous invasion, and capsular invasion, revealed that the WNT (odds ratio 4.43 [1.09-18.0], P = . 034), PI3K (odds ratio 7.80 [1.33-45.65], P = . 023), and cell cycle (odds ratio 6.81 [1.43-32.30], P = . 016) pathways were significantly associated with recurrence. Median time to recur- rence was 7.9 months; early recurrence (<7.9 months) was associated with MYC pathway alterations (40.9% vs 9.1%, P = . 0339).
Conclusion: This study identified genomic signatures in the PI3K, WNT, and cell cycle pathways associ- ated with adrenocortical carcinoma recurrence, including in those who underwent R0 resection. In- vestigations regarding the utility of these signatures as a prognostic tool to dictate adjuvant therapies or targeted treatment are warranted.
@ 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
Introduction
Adrenocortical carcinoma (ACC) remains a major challenge among solid organ malignancies, even in era of targeted therapies. Despite modern therapeutics, the only curative therapy for ACC is en bloc surgical resection with R0 margins while ensuring an intact capsule. Index surgery at a multidisciplinary center is critical to
preventing recurrence; however, even still, the recurrence rate remains high.1 This is particularly concerning given its aggressive disease biology. A recent large US series demonstrated a 5-year survival of only 38.6% among resected patients, with up to 35% of patients presenting with distant metastases at the time of diagnosis.1,2
To understand the tumorigenesis of these malignancies and to identify future targeted therapies, considerable work has been made to identify ACC driver mutations. A subset of ACCs are known to harbor germline variants associated with familial syndromes that are the presumptive initiators of carcinogenesis in these pa- tients. These include Li-Fraumeni (TP53), Lynch syndrome (mismatch repair genes), Beckwith-Weidman (IGF1), Familial
This work was previously presented as a poster at the 2024 American Associ- ation of Endocrine Surgeons (AAES) annual meeting.
* Corresponding author: Benjamin C. Greenspun, MD, Department of Surgery, Weill Cornell Medicine, 525 E 68th St, K-836, New York, NY 10065.
E-mail address: bcg9006@med.cornell.edu (B.C. Greenspun).
adenomatous polyposis syndrome (APC), MEN1 (MEN1), neurofi- bromatosis type 1 (NF1), SDH-syndromes (SDHx), and Carney complex (PRKAR1A), as well as their corresponding somatic vari- ants.1 More recent efforts aimed at characterizing genomic signa- tures of ACC tumorigenesis have focused on somatic driver mutations. In a multicenter study by Zheng et al in 2016, whole- exome sequencing of ACCs identified ZNFR4, CTN, NB1, CCNE1, and TERF2 as novel driver mutations, as well as the phenomenon of whole genome doubling being characteristic of ACC.3 Other geno- types have been cited, including deletions of WNT repressors, such as KREMEN1 and ANRF3, acknowledging WNT as the most frequently mutated pathway.4 Furthermore, other studies have looked beyond synonymous nucleotide polymorphisms or struc- tural variants, to expression level anomalies including copy number alterations and epigenetics to identify genetic signatures.5 Despite these efforts, the association of these variants with recurrence after surgery for curative intent (R0 resection) remains unknown.
Given that ACC is a rare disease only occurring in <0.5% of ad- renal nodules <4 cm, investigating genomic signatures using a multiinstitutional data set can be of benefit.1 The Cancer Genome Atlas (TCGA) is a National Cancer Institute-sponsored publicly available database, requiring rigorous standardization of data sets from published works studying oncologic genomics. At the time of this writing, it reflects the sequencing of 9125 samples from 33 distinct types of cancer. It specifically annotates oncologic canonical signaling pathways involved in the tumorigenesis and therapeutic resistance of the 33 malignancies. These pathways include WNT, Hippo, PI3K, transforming growth factor beta (TGF-B), RTK/RAS, NRF2, cell cycle, P53, Myc, and NOTCH and reflect both oncogenes and tumor suppressors. As of 2018, 89% of cancers studied had at least 1 driver alteration in these pathways, and 57% of all tumors had a potentially actionable alteration with existing therapies6 Consequently, this pathway analysis provides a unique way to identify potentially actionable genomic drivers of rare malignancies while omitting passenger events detected on whole-exome screening. Thus, given this as well as the paucity of data regarding genomic signatures of recurrence after curative-intent surgery, we aimed to identify pathways associated with ACC recurrence after R0 resection.
Methods
Institutional review board exemption
The study was exempt from institutional review board approval of the Weill Cornell Medicine Review Board because of the publicly available and deidentified nature of the data in cBioPortal.
Data source
ACC patients treated with adrenalectomy from TCGA were analyzed using cBioPortal. The publicly available cBioPortal website is a software developed by a multiinstitutional team hosted by the Center for Molecular Oncology at Memorial Sloan Kettering Cancer Center. It is a portal for analyzing TCGA and other studies made available by authors that combines nonsynonymous mutations, DNA copy number data, messenger RNA and microRNA expression data, protein-level data, epigenetic (methylation) data, and dei- dentified clinical data when available.7-9
Patient and variable selection
The study’s primary aim was to identify genomic pathways associated with recurrence after surgery for curative intent (R0 resection). Secondary aims were to identify genomic drivers
(pathways) associated with early recurrence. To accomplish this, all ACC patients within the TCGA with available recurrence status and canonical oncologic pathway data were included. Canonical onco- logic pathways included WNT, Hippo, PI3K, TGF-ß, RTK/RAS, NRF2, cell cycle, P53, Myc, and NOTCH. These pathways are standardized across TCGA studies. Notably, data on nononcologic pathways are absent from this collection of studies. These clinicopathologic var- iables, genomics including canonical oncologic pathways, treat- ment patterns, and outcomes were retrospectively analyzed. Patients who experienced a recurrence and those who did not were divided into separate cohorts. Of note, the site of recurrence regarding locoregional vs distant metastases was absent from cBioPortal. Subanalyses of R0 resected patients with documented recurrences and those with early recurrences-defined as preced- ing the median time to recurrence for the whole cohort-were performed. Patients who experienced an early recurrence before this median underwent the same pathway analysis. In addition to pathway analysis, the tumor mutational burden (TMB) was evalu- ated. TMB is measured in mutations per megabase of sequenced tumor and has emerged as a marker of tumor immunogenicity-a value critical in many solid organ tumors to predicting response to immunotherapy, specifically existing checkpoint inhibitors. Efforts have been made to establish a generalizable threshold for which tumors will benefit from immunotherapy that typically falls be- tween 8 and 10 mutations/Mb.10-12 Consequently, TMB was inter- preted as a binary value of >10 muts/Mb as TMB-high and <10 muts/Mb as TMB-low.
Statistical analysis
Two-tailed hypothesis tests were employed (P < . 05 considered significant). StataSE was used for all statistical analysis. x2 and Fisher exact tests were employed for categorical data, including sex, ethnicity, race, vascular invasion, venous invasion, capsular inva- sion, adjuvant radiation, surgical margin status, and oncologic pathway mutation data. Unpaired t tests were used for continuous variables (Mann-Whitney for nonnormally distributed data) including age, TMB, tumor size, and median follow-up. Univariate logistic regression and multivariable logistic regression modeling was used for clinically and statistically significant variables.
Results
Whole-cohort analysis
Review of the TCGA within cBioPortal yielded 92 unique, sequenced ACC tumors submitted to TCGA under a multicenter study. Eighty-four of 92 had available recurrence and oncologic canonical pathway data documented. Forty-four of 84 patients (52%) had a documented recurrence after resection. Median follow- up differed significantly between recurrent and nonrecurrent co- horts at 35 months and 54.9 months, respectively (P = . 003), attributed to mortality specific to the recurrent cohort. For general cohort demographic characteristics (Table I), sex (27% vs 35% male, P = . 251) and mean age (46 vs 47 years, P = . 917) did not differ significantly between recurrent and nonrecurrent cohorts. Simi- larly, ethnicity (74.1% vs 95.2%, not Hispanic or Latino, P =. 064) and race (97.6% vs 94.7% White, P = . 606) did not differ significantly.
For clinicopathologic results (Table I), median tumor size (10.5 cm vs 9.2 cm, P =. 129), incidence of capsular invasion (65% vs 44.4%, P =. 105) or venous invasion (50% vs 33.3%, P =. 164), and rate of receiving adjuvant radiation (25% vs 16.2%, P =. 416) did not differ significantly between recurrent and nonrecurrent cohorts. R0 sur- gical margin status was significantly more frequent in the nonre- current cohort (67.6% vs 94.9%, P =. 002) whereas R2 margins were
| Characteristic | Recurrent cohort (n = 44) | Nonrecurrent cohort (n = 40) | P value |
|---|---|---|---|
| Sex (% male) | 27 | 35 | .251 |
| Mean age, yr | 46 | 47 | .917 |
| Ethnicity | |||
| Not Hispanic or Latino | 20 (74.1) | 20 (95.2) | .064 |
| Hispanic or Latino | 7 (25.9) | 1 (4.8) | .064 |
| Race | |||
| Asian | 1 (2.4) | 1 (2.6) | >.999 |
| Black | 0.0 | 1 (2.6) | .481 |
| White | 40 (97.6) | 36 (94.7) | .606 |
| Median tumor size, cm | 10.5 | 9.2 | .129 |
| Capsular invasion | 26 (65.0) | 16 (44.4) | .105 |
| Venous invasion | 20 (50.0) | 11 (33.3) | .164 |
| R0 | 25 (67.6) | 37 (94.9) | .002* |
| R1 | 3 (8.3) | 1 (2.6) | .351 |
| R2 | 9 (24.3) | 1 (2.6) | .006* |
| Adjuvant radiation | 11 (25.0) | 6 (16.2) | .416 |
Unless otherwise noted, values are n (%).
* Statistically significant.
| Pathway | Recurrent cohort, n (%) (n = 44) | Nonrecurrent cohort, n (%) (n = 40) | P value |
|---|---|---|---|
| WNT | 20 (45.5) | 7 (17.5) | .009* |
| Hippo | 22 (50.0) | 13 (32.5) | .124 |
| PI3K | 20 (45.5) | 3 (8.1) | <. 0001* |
| Cell cycle | 22 (50.0) | 7 (17.5) | .002* |
| MYC | 11 (25.0) | 5 (12.5) | .172 |
| TP53 | 24 (54.6) | 8 (20.0) | .001* |
| TGF-B | 5 (11.4) | 3 (7.5) | .715 |
| RTK-RAS | 22 (50.0) | 11 (27.5) | .045* |
| NRF2 | 4 (9.1) | 0 | .117 |
| NOTCH | 20 (45.5) | 13 (32.5) | .267 |
* Statistically significant.
more frequently observed in the recurrent cohort (24.3% vs 2.6%, P = . 006). R1 status did not differ significantly (8.3% vs 2.6%, P = .351). Twelve patients (14.3%) presented with distant metastases at the time of resection. However, those who underwent R0 resection with metastases (n = 2/12) were surgically cleared of disease, with no evidence of disease at 3-month follow-up. Similarly, 3 R0 resected patients required en bloc resection of T4 tumors.
Genomic analysis began with TMB interpretation. Among pa- tients with a TMB <10, 51% (38 of 75) recurred, whereas 71.4% (5 of 7) of patients with TMB >10 recurred (P =. 4363). Genomic pathway alterations were then reviewed for the whole cohort (Table II).
Alterations in the WNT (45.5% vs 17.5%, P = . 009), PI3K (45.5% vs 8.1%, P =. 0001), cell cycle (50% vs 17.5%, P =. 002), TP53 (54.6% vs 20%, P =. 002), and RTK-RAS (50% vs 27.5%, P =. 045) pathways were significantly more frequent in the recurrent cohort, whereas Hippo, MYC, TGF-ß, NRF2, and NOTCH did not differ significantly.
Analysis of R0 resected cohort
There were 62 patients with R0 resection (Table III), with 40.3% (n = 25) with documented recurrence. WNT (52% vs 16.2%, P = .004), PI3K (36% vs 8.1%, P =. 009), cell cycle (40% vs 16.2%, P =. 043), and TP53 (44% vs 18,9%, P =. 046) remained significantly associated with recurrence, whereas RTK-RAS was no longer significant (44% vs 24.3%, P = . 165). For the RO cohort, multivariable logistic regression was performed for the outcome of recurrence. When controlling for the clinically significant variables of capsular inva- sion, venous invasion, and rates of receiving adjuvant radiation, WNT (OR 4.43 [1.09-18.0], P = . 038), PI3K (OR 7.80 [1.33-45.65], P = . 023), and cell cycle (OR 6.81 [1.43-32.3], P = . 016) remained significantly associated with recurrence. No clinical variable, including capsular invasion (OR 1.76 [0.49-6.32], P = . 386), venous invasion (OR 1.47 [0.39-5.59], P = . 573), or adjuvant radiation (OR 3.72 [0.76-18.12], P = . 103) achieved statistical significance on multivariable analysis. Further, these pathways remained signifi- cantly associated with increased odds of recurrence, even after exclusion of R0 resected patients with distant metastases or en block resections of T4 tumors. Finally, of 59 R0 resected tumors with available biochemical status, 34 (57.6%) were functional. However, on univariate analysis, steroid hormone producing tu- mors were not significantly associated with recurrence (2.81 [0.90-8.79], P =. 075).
Early recurrence subanalysis
Lastly, time to recurrence was assessed for the whole cohort, yielding a median time to recurrence of 7.9 months (IQR 3.7, 20). Among patients who recurred early (n = 22) compared to late (n = 22), only MYC pathway alterations were significantly more frequent (40.9% vs 9.1%, P = . 033) (Table IV).
Discussion
This study’s unique focus on driver events, using a pathway- specific approach, identified PI3K, WNT, and cell cycle as discrete signatures of recurrence after R0 resection. Further, among patients who experienced an early recurrence (before 7.9 months), we identified MYC alterations as an additional unique molecular signature. Jointly, these data point to a subset of R0 resected
| Pathway | R0 recurrent cohort, n (%) (n = 25) | R0 non-recurrent cohort, n (%) (n = 37) | P value | Multivariable logistic regression*, OR [CI], P value |
|---|---|---|---|---|
| WNT | 13 (52.0) | 6 (16.2) | .0041 | 4.43 [1.09-18.0], P = . 0381 |
| Hippo | 12 (48.0) | 12 (32.4) | .289 | 2.57 [0.74-8.98], P = . 138 |
| PI3K | 9 (36.0) | 3 (8.11) | .009+ | 7.80 [1.33-45.65], P = . 0231 |
| Cell cycle | 10 (40.0) | 6 (16.2) | .043 | 6.81 [1.43-32.3], P = . 0161 |
| MYC | 4 (16.0) | 4 (10.8) | .703 | 2.58 [0.34-19.44], P = . 357 |
| TP53 | 11 (44) | 7 (18.9) | .046 | 4.17 [0.98-17.74], P = . 053 |
| TGF-B | 3 (12.0) | 3 (8.1) | .677 | 3.78 [0.31-46.38], P = . 299 |
| RTK-RAS | 11 (44.0) | 9 (24.3) | .165 | 1.99 [0.54-7.37], P = . 301 |
| NRF2 | 1 (4.0) | 0 | .403 | Colinear |
| NOTCH | 10 (40.0) | 11 (29.7) | .425 | 0.86 [0.24-3.15], P = . 823 |
CI, confidence interval; OR, odds ratio.
* Multivariable logistic regression was performed using the clinical variables of capsular invasion, venous invasion, and adjuvant radiation for an outcome of recurrence for each genomic pathway.
+ Statistically significant.
| Pathway | Early recurrence cohort, n (%) (n = 22) | Late recurrence cohort, n (%) (n = 22) | P value |
|---|---|---|---|
| WNT | 11 (50.0) | 9 (40.9) | .762 |
| Hippo | 12 (54.6) | 10 (45.5) | .763 |
| PI3K | 12 (54.6) | 8 (36.4) | .354 |
| Cell cycle | 14 (63.6) | 8 (36.4) | .130 |
| MYC | 9 (40.9) | 2 (9.1) | .033* |
| TP53 | 15 (68.2) | 9 (40.9) | .129 |
| TGF-B | 3 (13.6) | 2 (9.1) | >.999 |
| RTK-RAS | 14 (63.6) | 8 (36.4) | .130 |
| NRF2 | 2 (9.1) | 2 (9.1) | >.999 |
| NOTCH | 13 (59.1) | 7 (31.8) | .129 |
Statistically significant.
patients with a genomic signature associated with high risk for recurrence. This is particularly notable, given that other clinical high-risk features, such as venous or capsular invasion were not significantly associated with recurrence comparatively. Clinically, identifying this population is critical as adjuvant therapy is not uniformly administered for R0 resected ACCs, often being reserved for those who are incompletely resected or in those who are deemed to be poor surgical candidates.
Current 2023 National Comprehensive Cancer Network (NCCN) guidelines minimize the role of genetic sequencing of patients with ACC to testing for familial syndromes, rather than to guide imme- diate therapy.13 Thus, although some germline variants may be detected in patients who meet these criteria, most patients have tumors driven by sporadic mutations and would not undergo testing, despite potentially possessing targetable variants. Howev- er, prior studies have provided a foundation for our knowledge of ACC tumor genomics. At present, our existing knowledge of WNT pathway alterations in ACC reflects a heterogenous cohort.4 Thus, it is difficult to interpret these findings within the subset of patients who undergo R0 resection with intent-to-cure. Consequently, applying routine sequencing to detect these variants may not augment clinical care in the context of varying surgical margins and adjuvant therapies. However, using these same pathways to expand our understanding of risk stratification for recurrence in an R0 resected population may parallel advances in risk assessment seen in other malignancies. For example, in colorectal adenocarcinoma, localized, node-negative disease, absent of metastases may still be advised to receive adjuvant chemotherapy after complete resection based on the presence of BRAF mutations.14 Similarly, R0 resected ACCs with PI3K, WNT, and cell cycle aberrancies detected on routine pathology sequencing may benefit from adjuvant therapy. Current NCCN guidelines only recommend adjuvant external beam radiation after R0 resection for Ki-67 >10%, rupture of capsule, large tumors, and high grade. Mitotane is similarly conditionally rec- ommended for hormonal symptom management over tumor con- trol for R0 resected tumors.13 Notably, European guidelines recommend mitotane for all stage III disease and for those with Ki- 67 >10%, even in completely resected tumors.15 Thus, studies investigating the utility of guiding adjuvant therapy with genomic markers associated with recurrence in this localized R0 cohort are warranted. Beyond existing agents, the detection of the individual somatic drivers within these select pathways may prove to be valuable markers for novel targeted therapies.
Although their efficacy within ACC remains to be determined, there are several existing targeted therapies in both the preclinical and FDA-approved phases that act on the WNT, PI3K, and cell cycle pathways. CTNNB1 mutants are among the most common in ACC and represent an aberration in the WNT pathway.4 Early evidence
of other tumors with this genotype have exhibited high sensitivity to threonine tyrosine kinase (TTK) inhibitors.16 Similarly, FDA- approved cyclin-dependent kinase inhibitors (CDK inhibitors), including palbociclib, ribociclib, and abemaciclib, and ATM inhib- itors-both of which act on the cell cycle pathway-have become routinely used for node-positive and metastatic breast cancer.17,18 Finally, although not yet implemented in ACC trials, the FDA has approved several agents targeting the PI3K pathway. Alpelisib, among several others, acts to inhibit PI3KCA, which has been commonly employed for metastatic breast cancer.19
Identifying targeted therapies is critical, as existing immuno- therapies have not shown significant promise with ACC. Guidelines propose consideration of tumor microsatellite instability, mismatch repair deficiency, and TMB testing to conditionally recommend pembrolizumab in those with TMB >10.13 However, ACCs are characteristically immunologically “cold,” concurrent with our TMB analysis, exhibiting a poor response to pembrolizumab in existing literature.20 Whether other pharmacologic means of manipulating the tumor microenvironment can potentiate an immunologic response remains to be determined.
Study limitations
Although the oncogenic pathway model permits detection of high-yield genomic aberrancies in rare malignancies, using this approach only addresses mutations in aggregate, rather than in- dividual events. Consequently, this study cannot comment on specific somatic and germline driver mutations that manifested in tumor recurrence. Additionally, these data were collected retro- spectively, and analysis is limited to the existing variables found within the primary source. Although rates of patients receiving adjuvant radiation were consistently available, systemic therapies were inconsistent or not routinely reported, likely differing signif- icantly given the multiple institutions’ protocols from which these treatment outcomes were reported. Further, given the subset of R0 patients with completely resected locally invasive and metastatic disease, the extent or necessity of en bloc resection for adjacent organs is also unknown. Similarly, the association between tumor grade, recurrence, and pathway alterations remains unknown given lack of available Ki-67 data. Thus, biases may exist that cannot be excluded in this analysis. Finally, although the TCGA is the largest publicly available database for ACC, it is still limited and relatively small given the rare nature of this malignancy.
In conclusion, this study has identified novel genomic signa- tures of recurrence after R0 resection of ACCs. Future studies evaluating the clinical utility of WNT, cell cycle, and PI3K pathway alterations to guide adjuvant therapy after R0 resection are war- ranted. Further, as larger cohorts are studied, the individual and combination molecular events within these pathways driving recurrence after intended curative surgery may become evident. Similarly, as study populations grow, the role of MYC pathway al- terations in early recurrence may be elucidated in an R0 resected cohort.
Funding/Support
The authors received no funding for this work.
Conflict of Interest/Disclosure
The authors have no relevant financial disclosures.
Acknowledgments
Paul Christos, DrPH, MS, served as the biostatistician advisor.
CRediT authorship contribution statement
Benjamin C. Greenspun: Writing - review & editing, Writing - original draft, Visualization, Validation, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Dawn Chirko: Writing - review & editing, Writing - original draft, Formal analysis, Data curation. Rajbir Toor: Writing - review & editing, Writing - original draft, Formal analysis, Data curation. Kyle Wierzbicki: Writing - review & editing, Writing - original draft, Formal analysis, Data curation. Teagan E. Marshall: Writing - review & editing, Writing - original draft, Methodology, Formal analysis, Data curation. Abhinay Tumati: Writing - review & editing, Writing - original draft, Methodology, Formal analysis, Data curation. Rasa Zarnegar: Writing - review & editing, Writing - original draft, Supervision, Formal analysis, Data cura- tion, Conceptualization. Thomas J. Fahey: Writing - review & editing, Writing - original draft, Supervision, Formal analysis, Data curation, Conceptualization. Brendan M. Finnerty: Writing - re- view & editing, Writing - original draft, Visualization, Validation, Supervision, Resources, Methodology, Investigation, Formal anal- ysis, Data curation, Conceptualization.
References
1. Yip L, Duh QY, Wachtel H, et al. American association of endocrine surgeons guidelines for adrenalectomy: executive summary. JAMA Surg. 2022;157: 870-877.
2. Bilimoria KY, Shen WT, Elaraj D, et al. Adrenocortical carcinoma in the United States: treatment utilization and prognostic factors. Cancer. 2008;113: 3130-3136.
3. Zheng S, Cherniack AD, Dewal N, et al. Comprehensive pan-genomic charac- terization of adrenocortical carcinoma. Cancer Cell. 2016;29:723-736.
4. Juhlin CC, Goh G, Healy JM, et al. Whole-exome sequencing characterizes the landscape of somatic mutations and copy number alterations in adrenocortical carcinoma. J Clin Endocrinol Metab. 2015;100:E493-E502.
5. Gupta S, Won H, Chadalavada K, et al. TERT copy number alterations, promoter mutations and rearrangements in adrenocortical carcinomas. Endocr Pathol. 2022;33:304-314.
6. Sanchez-Vega F, Mina M, Armenia J, et al. Oncogenic signaling pathways in the cancer genome Atlas. Cell. 2018;173:321-337.e10.
7. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data [published correction appears in Cancer Discov. 2012 Oct;2(10):960] Cancer Discov. 2012;2:401-404.
8. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer ge- nomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6, pl1.
9. de Bruijn I, Kundra R, Mastrogiacomo B, et al. Analysis and visualization of longitudinal genomic and clinical data from the AACR project GENIE biopharma collaborative in cBioPortal. Cancer Res. 2023;83:3861-3867.
10. Aggarwal C, Ben-Shachar R, Gao Y, et al. Assessment of tumor mutational burden and outcomes in patients with diverse advanced cancers treated with immunotherapy. JAMA Netw Open. 2023;6:e2311181.
11. Raj N, Zheng Y, Kelly V, et al. PD-1 blockade in advanced adrenocortical car- cinoma. J Clin Oncol. 2020;38:71-80.
12. Strickler JH, Hanks BA, Khasraw M. Tumor mutational burden as a predictor of immunotherapy response: is more always better? Clin Cancer Res. 2021;27: 1236-1241.
13. Shah MH, Goldner WS, Benson AB, et al. Neuroendocrine and adrenal tumors, Version 2.2021, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw. 2021;19:839-868.
14. Formica V, Sera F, Cremolini C, et al. KRAS and BRAF mutations in stage II and III colon cancer: a systematic review and meta-analysis. J Natl Cancer Inst. 2022;114:517-527.
15. Fassnacht M, Dekkers OM, Else T, et al. European Society of Endocrinology clinical practice guidelines on the management of adrenocortical carcinoma in adults, in collaboration with the European Network for the study of adrenal tumors. Eur J Endocrinol. 2018; 179:G1-G46.
16. Zaman GJR, de Roos JADM, Libouban MAA, et al. TTK inhibitors as a targeted therapy for CTNNB1 (B-catenin) mutant cancers. Mol Cancer Ther. 2017;16: 2609-2617.
17. Zimmermann A, Zenke FT, Chiu LY, et al. A new class of selective ATM in- hibitors as combination partners of DNA double-strand break inducing cancer therapies. Mol Cancer Ther. 2022;21:859-870.
18. Sheikh MS, Satti SA. The emerging CDK4/6 inhibitor for breast cancer treat- ment. Mol Cell Pharmacol. 2021;13:9-12.
19. Mishra R, Patel H, Alanazi S, Kilroy MK, Garrett JT. PI3K inhibitors in cancer: clinical Implications and adverse effects. Int J Mol Sci. 2021;22:3464.
20. Araujo-Castro M, Pascual-Corrales E, Molina-Cerrillo J, Alonso-Gordoa T. Immunotherapy in adrenocortical carcinoma: predictors of response, efficacy, Safety, and Mechanisms of resistance. Biomedicines. 2021;9:304.