OFFICIAL JOURNAL OF THE SOCIETY OF SURGICAL ONCOLOGY
Check for updates
Cumulative GRAS Score as a Predictor of Survival After Resection for Adrenocortical Carcinoma: Analysis From the U.S. Adrenocortical Carcinoma Database
Jordan J. Baechle, BS1,2, Paula Marincola Smith, MD1, Carmen C. Solórzano, MD1, Thuy B. Tran, MD3, Lauren M. Postlewait, MD4, Shishir K. Maithel, MD4, Jason Prescott, MD, PhD5, Timothy Pawlik, MD5, Tracy S. Wang, MD6, Jason Glenn, MD6, Ioannis Hatzaras, MD7, Rivfka Shenoy, MD7, John E. Phay, MD8, Lawrence A. Shirley, MD8, Ryan C. Fields’, Linda Jin, MD9, Daniel E. Abbott, MD10, Sean Ronnekleiv-Kelly, MD1º, Jason K. Sicklick, MD11, Adam Yopp, MD12, John Mansour, MD12, Quan-Yang Duh, MD13,
Natalie Seiser, MD13, Konstantinos Votanopoulos, MD14, Edward A. Levine, MD14, George Poultsides, MD3, and Colleen M. Kiernan, MD, MPH1
1Department of Surgery, Vanderbilt University Medical Center, Nashville, TN; 2School of Medicine, Meharry Medical College, Nashville, TN; 3Department of Surgery, Stanford Medical Center, Stanford, CA; 4Department of Surgery, Winship Cancer Institute, Emory University, Atlanta, GA; 5Department of Surgery, The Johns Hopkins Medical Center, Baltimore, MD; 6Department of Surgery, Medical College of Wisconsin, Milwaukee, WI; 7Department of Surgery, New York University Langone Health, New York, NY; 8Department of Surgery, The Ohio State University, Columbus, OH; 9Department of Surgery, Washington University School of Medicine, St Louis, MO; 10Department of General Surgery, School of Medicine and Public Health, University of Wisconsin, Madison, WI; 11Department of Surgery, University of California San Diego, San Diego, CA; 12Department of Surgery, University of Texas Southwestern Medical Center, Dallas, TX; 13Department of Surgery, University of California San Francisco, San Francisco, CA; 14Department of Surgery, Wake Forest School of Medicine, Winston-Salem, NC
ABSTRACT
Background. Adrenocortical carcinoma (ACC) is a rare but aggressive malignancy, and many prognostic factors that influence survival remain undefined. Individually, the GRAS (Grade, Resection status, Age, and Symptoms of hormone hypersecretion) parameters have demonstrated their prognostic value in ACC. This study aimed to assess
Jordan J. Baechle and Paula Marincola Smith contributed equally.
Supplementary information accompanies this paper at https://doi. org/10.1245/s10434-020-09562-8.
@ Society of Surgical Oncology 2021
First Received: 21 April 2020
Accepted: 21 December 2020; Published Online: 14 February 2021
C. M. Kiernan, MD, MPH e-mail: colleen.m.kiernan@vumc.org
the value of a cumulative GRAS score as a prognostic indicator after ACC resection.
Methods. A retrospective cohort study of adult patients who underwent surgical resection for ACC between 1993 and 2014 was performed using the United States Adreno- cortical Carcinoma Group (US-ACCG) database. A sum GRAS score was calculated for each patient by adding one point each when the criteria were met for tumor grade (Weiss criteria ≥ 3 or Ki67 ≥ 20%), resection status (mi- cro- or macroscopically positive margin), age (≥ 50 years), and preoperative symptoms of hormone hypersecretion (present). Overall survival (OS) and disease-free survival (DFS) by cumulative GRAS score were analyzed by the Kaplan-Meier method and log-rank test.
Results. Of the 265 patients in the US-ACCG database, 243 (92%) had sufficient data available to calculate a cumulative GRAS score and were included in this analysis. The 265 patients comprised 23 patients (10%) with a GRAS of 0, 52 patients (21%) with a GRAS of 1, 92 patients (38%) with a GRAS of 2, 63 patients (26%) with a GRAS of 3, and 13 patients (5%) with a GRAS of 4. An
increasing GRAS score was associated with shortened OS (p <0.01) and DFS (p <0.01) after index resection.
Conclusion. In this retrospective analysis, the cumulative GRAS score effectively stratified OS and DFS after index resection for ACC. Further prospective analysis is required to validate the cumulative GRAS score as a prognostic indicator for clinical use.
Adrenocortical carcinoma (ACC) is among the rarest and most aggressive cancers, with an incidence of approximately one per million and 21% to 46% of patients presenting with advanced unresectable disease.1-5 For those with resectable disease, adrenalectomy with or without en bloc resection of adjacent organs to achieve negative surgical margins remains the only possibility of cure.68 Patients with fully resectable disease have a reported 5-year survival rate of approximately 50%, whereas patients with unresectable disease have a 5-year survival rate near 0% and a median survival of less than 12 months. 1,9
The rarity and aggressive nature of ACC make prospective examination difficult. Therefore, much of the clinical management and prognostication for ACC depends on retrospective analyses and/or expert opinion. The cur- rent prognostication for patients with ACC primarily hinges on the presence or absence of metastases and tumor resectability.1,4,6,9 However, for better prognostication and identification of patients at high risk for recurrence and shortened survival further understanding of how various histologic, operative, demographic, and clinical risk factors may influence ACC patient survival is needed.
Several studies have determined that tumor grade, resection status, age, and preoperative symptoms of hor- mone hypersecretion can have an impact on ACC prognosis.10-16 These factors (grade [G], resection status [R], patient age [A], and preoperative symptoms [S] of hormone hypersecretion) have been demonstrated retro- spectively to have individual prognostic value in stages III- IV ACC, and in 2015 these factors were collectively referred to as the “GRAS” parameters by the European Network for the Study of Adrenal Tumors (ENSAT). 10-16
The utility of combining the individual GRAS parame- ters into a cumulative GRAS score for further stratification of ACC patient prognosis was recently assessed by Liang et al.11 for 65 patients with stages I-III ACC in a single- center study using a three-tiered system (GRAS score ≤ 1, 2, ≥ 3). These early findings require validation with a larger multi-institutional cohort for patients with all stages (I-IV) of surgically resectable ACC. This study aimed to examine the value of using a cumulative GRAS score to stratify ACC patient survival after primary resection of ACC stages I-IV in the United States (U.S.).
METHODS
The U.S. Adrenocortical Carcinoma Group (US-ACCG) consists of 13 academic medical centers including Van- derbilt University, Emory University, Stanford University, The Johns Hopkins University, Medical College of Wis- consin, New York University, The Ohio State University, Washington University in St. Louis, University of Wis- consin, University of California San Diego, University of Texas Southwestern, University of California San Fran- cisco, and Wake Forest University. This multi-institutional collaboration retrospectively identified adult patients (age ≥ 18 years) who underwent resection for ACC between 1993 and 2014. The institutional review boards at each participating institution approved this study. Demographic, clinical, preoperative, intraoperative, pathologic, perioper- ative morbidity, and survival data were collected through review of the medical records at each institution.
Of the 265 patients in the US-ACCG database, 243 (91.7%) had sufficient data available to calculate a GRAS score and were included in our analysis. A sum GRAS score was calculated for each patient meeting the criteria by adding one point for each of the following GRAS cat- egories: G (grade: ≥ 3 Weiss criteria met12 (Table 1) or Ki67 ≥ 20% on the final pathology report), R (resection status: a microscopically [R1] or macroscopically [R2] positive final surgical margin at the index operation), A (age of ≥ 50 years at the index operation), and S (symp- toms: preoperative symptoms of hormone hypersecretion) (Table 2).
The American Joint Commission on Cancer (AJCC) Staging Manual, seventh edition, was used to determine tumor-node-metastasis (TNM) classification.17 The ENSAT staging system was used to define clinical stage of disease.5 Overall survival (OS) was defined as time from
| Weiss system | Score |
|---|---|
| High nuclear grade | 1 |
| Mitotic rate > 5/50 high power | 1 |
| Atypical mitotic figures | 1 |
| Eosinophilic tumor cell cytoplasm (> 75%) | 1 |
| Diffuse architecture (> 33%) | 1 |
| Necrosis | 1 |
| Venous invasion | 1 |
| Sinusoidal invasion | 1 |
| Capsular invasion | 1 |
| Totalª | 9 |
ªTotal ≥ 3 correlates with subsequent malignnt behavior12
| Points | |
|---|---|
| Grade | |
| Weiss criteria < 3 and Ki67 < 20% | 0 |
| Weiss criteria ≥ 3 | |
| or Ki67 ≥ 20% | 1 |
| Resection | |
| R0 | 0 |
| R1/2 | 1 |
| Age (years) | |
| < 50 | 0 |
| ≥ 50 | 1 |
| Symptoms | |
| Absent | 0 |
| Present | 1 |
| Cumulative GRAS score: |
the date of the index operation to the date of death, and disease-free survival (DFS) was defined as time from the date of the index operation to documented disease recur- rence or death.
Statistical Analysis
Categorical variables are presented as frequency and percentage and compared using the chi-square test or Fisher’s exact test as appropriate. Continuous variables were reported as median value with interquartile ranges (IQR) and compared using the Kruskal-Wallis test. Both OS and DFS were calculated using the Kaplan-Meier method and compared using the log-rank test. Individual and concurrent GRAS components were analyzed using univariant Cox regression methods. Significance was set at a p value lower than 0.05. All statistical analyses were performed using the 1.1.383 R statistics software (R Core Team, Vienna, Austria).
RESULTS
For the entire study cohort of 243 patients, the median follow-up time was 17.6 months (range 0.33-220 months). The GRAS score was 0 for 23 patients (9.5%), 1 for 52 patients (21.4%) 2 for 92 patients (37.9%), 3 for 63 patients (25.9%), and 4 for 13 patients (5.3%). The GRAS score groups differed significantly in age at the time of the index operation (p<0.01), ENSAT stage (p<0.01), T stage (p< 0.01), presence of metastases (p < 0.01), rate of postop- erative chemotherapy (p < 0.03), and rate of postoperative mitotane therapy (p < 0.04). The groups also differed significantly in rate of preoperative abdominal pain (p <
0.01), palpable abdominal mass (p <0.04), leg edema (p < 0.01), and hormone hypersecretion (p < 0.01). The demographic, clinical, and pathologic features of the study cohort by cumulative GRAS score are summarized in Table 3. A breakdown of cumulative GRAS score com- ponents for each group is summarized in Table 4.
In the univariate Cox regression analysis examining the impact of each GRAS parameter on OS individually, worse OS was significantly associated with positive (R1 or R2) resection margins (hazard ratio [HR], 3.32; 95% confidence interval [CI], 2.23- 4.95; p < 0.01) and preoperative symptoms of hormone hypersecretion (HR 1.60; 95% CI 1.09-2.35; p = 0.02) but not tumor grade or patient age (Fig. 1).
For the patients of all ENSAT stages, an increasing cumulative GRAS score was significantly associated with worse OS and DFS after the index resection for ACC (OS: HR 1.57; CI 1.28-1.92; p < 0.01; DFS: HR 1.32; CI 1.13-1.54; p < 0.01), with the median OS decreasing with each additional GRAS point. The median OS was 120.7 months for a GRAS score of 0, 60.7 months for a GRAS score of 1, 38.2 months for a GRAS score of 2, 31.7 months for a GRAS score of 3, and 9 months for a GRAS score of 4 (Fig. 2a). The median DFS was 36.1 months for a GRAS score of 0, 14.3 months for a GRAS score of 1, 15.6 months for a GRAS score of 2, 10.1 months for a GRAS score of 3, and 5.6 months for a GRAS score of 4 (Fig. 2b).
Because the Kaplain-Meier curves for OS and DFS demonstrated some overlap for the patients with interme- diate GRAS scores (1-3), an additional analysis was performed to examine the utility of using a three-tiered GRAS scoring system that grouped the patients with intermediate GRAS scores (1, 2, and 3) into a single arm. When considered together in a single study arm, the patients with intermediate cumulative GRAS scores (GRAS 1-3) had a median OS of 44.5 months and a median DFS of 13 months (Fig. 3a, b).
DISCUSSION
The cumulative GRAS score comprising a function of tumor grade, operative resection status, patient age, and presence of preoperative hormone hypersecretion symp- toms was found to be useful in stratifying OS and DFS for the patients who underwent primary resection of stages I-IV ACC. Although some of the GRAS parameters (re- section status and preoperative symptoms of hormone hypersecretion) were individual predictors of shortened OS, others (tumor grade and patient age) were not,
| Cumulative GRAS score | 0 n (%) | 1 n (%) | 2 n (%) | 3 n (%) | 4 n (%) | p Value |
|---|---|---|---|---|---|---|
| Total | 23 | 52 | 92 | 63 | 13 | |
| Gender | ||||||
| Female | 16 (69.6) | 34 (65.4) | 48 (52.2) | 47 (74.6) | 9 (69.2) | 0.062 |
| Male | 7 (30.4) | 18 (34.6) | 44 (47.8) | 16 (25.4) | 4 (30.8) | |
| Race | ||||||
| White | 20 (90.9) | 42 (82.4) | 72 (80.9) | 45 (76.3) | 11 (84.6) | 0.901 |
| Hispanic | 0 (0.0) | 3 (5.9) | 5 (5.6) | 7 (11.9) | 0 (0.0) | |
| Black | 1 (4.6) | 2 (3.9) | 6 (6.7) | 2 (3.4) | 1 (7.7) | |
| Asian | 0 (0.0) | 2 (3.9) | 4 (4.5) | 2 (3.4) | 1 (7.7) | |
| Other | 1 (4.6) | 2 (3.9) | 2 (2.6) | 3 (5.1) | 0 (0.0) | |
| Age: years (IQR) | 36.0 (31.0-47.0) | 46.0 (38.1-51.0) | 54.0 (44.5-63.0) | 58.5 (53.2-67.0) | 59.3 (57.0-67.0) | <0.001 |
| ASA class | ||||||
| 1 | 3 (16.7) | 9 (22.5) | 13 (20.0) | 7 (18.9) | 0 (0.0) | 0.057 |
| 2 | 8 (44.4) | 13 (32.5) | 21 (32.3) | 4 (10.8) | 0 (0.0) | |
| 3 | 6 (33.3) | 16 (40.0) | 24 (36.9) | 22 (59.5) | 7 (100.0) | |
| 4 | 1 (5.6) | 2 (5.0) | 7 (10.8) | 4 (10.8) | 0 (0.0) | |
| BMI: kg/m2 (IQR) | 27.0 (23.5-32.0) | 27.0 (25.0-31.2) | 26.0 (23.1-32.8) | 28.3 (24.0-33.0) | 32.9 (27.8-37.8) | 0.510 |
| ENSAT stage | ||||||
| I | 4 (22.2) | 1 (2.0) | 3 (3.5) | 4 (6.6) | 1 (8.3) | 0.001 |
| II | 9 (50.0) | 25 (50.0) | 31 (36.0) | 13 (21.3) | 0 (0.0) | |
| III | 5 (27.8) | 19 (38.0) | 37 (43.0) | 29 (47.5) | 5 (41.7) | |
| IV | 0 (0.0) | 5 (10.0) | 15 (17.4) | 15 (24.6) | 6 (50.0) | |
| Incidentaloma | ||||||
| Yes | 10 (43.5) | 18 (34.6) | 44 (47.8) | 26 (41.3) | 3 (23.1) | 0.356 |
| No | 13 (56.5) | 34 (65.4) | 48 (52.2) | 37 (58.7) | 10 (76.9) | |
| Weight loss | ||||||
| Yes | 1 (4.4) | 5 (9.6) | 17 (18.5) | 8 (12.7) | 0 (0.0) | 0.219 |
| No | 22 (95.7) | 47 (90.4) | 75 (81.5) | 55 (87.3) | 13 (100.0) | |
| Abdominal pain | ||||||
| Yes | 13 (56.5) | 31 (59.6) | 43 (46.7) | 16 (25.4) | 3 (23.1) | 0.001 |
| No | 10 (43.5) | 21 (40.4) | 49 (53.3) | 47 (74.6) | 10 (76.9) | |
| Palpable mass | ||||||
| Yes | 2 (8.7) | 12 (23.1) | 6 (6.5) | 7 (11.1) | 0 (0.0) | 0.038 |
| No | 21 (91.3) | 40 (76.9) | 86 (93.5) | 56 (88.9) | 13 (100.0) | |
| Leg edema | ||||||
| Yes | 1 (4.4) | 6 (11.5) | 8 (8.7) | 17 (27.0) | 8 (61.5) | < 0.001 |
| No | 22 (95.7) | 46 (88.5) | 84 (91.3) | 46 (73.0) | 5 (38.5) | |
| Tumor diameter: cm (IQR) | 10.2 (7.65-14.0) | 11.1 (9.00-15.8) | 12.0 (8.00-14.5) | 12.0 (8.12-15.0) | 11.0 (9.50-12.0) | 0.799 |
| Tumor weight: grams (IQR) | 280 (135-670) | 731 (371-1871) | 318 (131-1012) | 321 (145-1001) | 251 (201-331) | 0.052 |
| T stage | ||||||
| T1 | 4 (22.2) | 2 (4.1) | 2 (2.4) | 3 (4.9) | 1 (7.7) | 0.002 |
| T2 | 9 (50.0) | 26 (53.1) | 37 (44.0) | 20 (32.8) | 1 (7.7) | |
| T3 | 5 (27.8) | 12 (24.5) | 37 (44.0) | 25 (41.0) | 9 (69.2) | |
| T4 | 0 (0.0) | 9 (18.4) | 8 (9.5) | 13 (21.3) | 2 (15.4) | |
| N stage | ||||||
| N0 | 1 (50.0) | 14 (82.4) | 22 (64.7) | 12 (57.1) | 2 (40.0) | 0.297 |
| N1 | 1 (50.0) | 3 (17.6) | 12 (35.3) | 9 (42.9) | 3 (60.0) |
| Cumulative GRAS score | 0 n (%) | 1 n (%) | 2 n (%) | 3 n (%) | 4 n (%) | p Value |
|---|---|---|---|---|---|---|
| M stage | ||||||
| M0 | 20 (100.0) | 47 (90.4) | 75 (84.3) | 48 (76.2) | 6 (50.0) | 0.002 |
| M1 | 0 (0.0) | 5 (9.6) | 14 (15.7) | 15 (23.8) | 6 (50.0) | |
| Tumor function | ||||||
| Nonfunctional | 23 (100.0) | 43 (82.7) | 64 (69.6) | 19 (30.2) | 0 (0.0) | < 0.001 |
| Glucocorticoid | 0 (0.0) | 6 (11.5) | 9 (9.8) | 28 (44.4) | 10 (76.9) | |
| Virilizing/feminizing | 0 (0.0) | 2 (3.9) | 13 (14.1) | 11 (17.5) | 3 (23.1) | |
| Mineralocorticoid | 0 (0.0) | 1 (1.9) | 6 (6.5) | 5 (7.9) | 0 (0.0) | |
| Preoperative chemotherapy | ||||||
| Yes | 0 (0.0) | 2 (3.9) | 1 (1.2) | 2 (32) | 0 (0.0) | 0.781 |
| No | 20 (100.0) | 49 (96.1) | 82 (98.8) | 61 (96.8) | 13 (100.0) | |
| Surgical approach | ||||||
| Open | 17 (77.3) | 42 (84.0) | 70 (79.5) | 53 (84.1) | 10 (76.9) | 0.856 |
| Minimally invasive | 5 (22.7) | 8 (16.0) | 18 (20.5) | 10 (15.9) | 3 (23.1) | |
| EBL: ml (IQR) | 475 (125-900) | 775 (300-2425) | 500 (200-1500) | 800 (275-1900) | 800 (162-1150) | 0.445 |
| Operative time: min (IQR) | 175 (139-256) | 199 (158-311) | 240 (159-316) | 298 (214-352) | 186 (152-252) | 0.154 |
| Postoperative chemotherapy | ||||||
| Yes | 0 (0.0) | 5 (11.1) | 17 (20.0) | 16 (26.2) | 3 (25.0) | 0.028 |
| No | 21 (100.0) | 40 (88.9) | 68 (80.0) | 45 (73.8) | 9 (75.0) | |
| Postoperative mitotane | ||||||
| Yes | 3 (16.7) | 15 (39.5) | 31 (43.1) | 24 (44.4) | 9 (75.0) | 0.035 |
| No | 15 (83.3) | 23 (60.5) | 41 (56.9) | 30 (55.6) | 3 (25.0) | |
| Postoperative radiation | ||||||
| Yes | 0 (0.0) | 2 (5.0) | 62 (89.9) | 51 (89.5) | 11 (84.6) | 0.463 |
| No | 18 (100.0) | 38 (95.0) | 7 (10.1) | 6 (10.5) | 2 (15.4) | |
| Postoperative complication (Clavien-Dindo classification) | ||||||
| None | 8 (61.5) | 26 (70.3) | 42 (63.6) | 26 (52.0) | 5 (41.7) | 0.875 |
| 1 | 1 (7.67) | 1 (2.7) | 5 (7.6) | 4 (8.0) | 1 (8.3) | |
| 2 | 2 (15.4) | 5 (13.5) | 11 (16.7) | 12 (24.0) | 3 (25.0) | |
| 3 | 1 (7.7) | 3 (8.1) | 4 (6.1) | 5 (10.0) | 3 (25.0) | |
| 4 | 1 (7.7) | 2 (5.4) | 3 (4.6) | 1 (2.0) | 0 (0.0) | |
| 5 | 0 (0.0) | 0 (0.0) | 1 (1.5) | 2 (4.0) | 0 (0.0) | |
| In-hospital mortality | ||||||
| Yes | 0 (0.0) | 0 (0.0) | 3 (3.6) | 2 (3.4) | 0 (0.0) | 0.733 |
| No | 17 (100.0) | 49 (100) | 80 (96.4) | 57 (96.6) | 13 (100.0) | |
| Hospital stay: days (IQR) | 5.00 (4.00-6.50) | 6.00 (4.00-7.00) | 6.00 (4.00-8.00) | 6.00 (5.00-9.25) | 8.00 (5.50-16.0) | 0.269 |
| 90-day readmission | ||||||
| Yes | 3 (20.0) | 5 (11.4) | 9 (12.5) | 8 (15.4) | 6 (46.2) | 0.056 |
| No | 12 (80.0) | 39 (88.6) | 63 (87.5) | 44 (84.6) | 7 (53.8) | |
Bold values are p < 0.05
IQR, interquartile range; ASA, American Society of Anesthesiology; BMI, body mass index; ENSAT, European Network for the Study of Adrenal Tumors; EBL, estimated blood loss
suggesting that a cumulative GRAS score may be greater than the sum of its parts in stratifying patient prognosis after index resection for stages I-IV ACC.
In current clinical practice, ACC resectability remains the primary prognostic indicator for patients with ACC. However, several groups in this study demonstrated that other factors including tumor grade, hormone
| GRAS score | 0 n (%) | 1 n (%) | 2 n (%) | 3 n (%) | 4 n (%) |
|---|---|---|---|---|---|
| Total | 23 | 52 | 92 | 63 | 13 |
| Grade | |||||
| Weiss criteria <3 | 23 (100.0) | 24 (46.2) | 16 (17.4) | 3 (4.76) | 0 (0.0) |
| Weiss criteria ≥ 3 or Ki67 ≥ 20% | 0 (0.0) | 28 (53.8) | 76 (82.6) | 60 (95.2) | 13 (100.0) |
| Resection status | |||||
| R0 | 23 (100.0) | 52 (100.0) | 70 (76.1) | 31 (49.2) | 0 (0.0) |
| R1/2 | 0 (0.0) | 0 (0.0) | 22 (23.9) | 32 (50.8) | 13 (100.0) |
| Age (years) | |||||
| < 50 | 23 (100.0) | 37 (71.2) | 34 (37.0) | 10 (15.9) | 0 (0.0) |
| ≥ 50 | 0 (0.0) | 15 (28.8) | 58 (63.0) | 53 (84.1) | 13 (100.0) |
| Symptomsª | |||||
| Absent | 23 (100.0) | 43 (82.7) | 64 (69.6) | 19 (30.2) | 0 (0.0) |
| Present | 0 (0.0) | 9 (17.3) | 28 (30.4) | 44 (69.8) | 13 (100.0) |
ªSymptoms defined as presence of symptoms of preoperative hormone hypersecretion
Hazards Ratios of GRAS Components
| Hazard Ratio | 95% CI | |
|---|---|---|
| Grade Weiss Criteria ≥3 | 1.32 | (0.84-2.06) |
| Resection R1/2 | 3.32 | (2.23-4.95) |
| Age ≥50 years old | 1.15 | (0.78-1.70) |
| Symptoms Present | 1.60 | (1.09-2.35) |
p-value
0.225
<0.001
0.486
0.016
1
2
3
4
5
Hazard Ratio (95% CI)
hypersecretion, and age also were individually helpful in predicting survival for the ACC patients with stage III-IV disease. 10-16
Our study demonstrated that calculation of a cumulative GRAS score, which is a sum of a patient’s individual GRAS risk factors, can stratify OS and DFS for ACC patients with ENSAT stages I-IV disease after the index resection. Our findings are consistent with those recently published by Liang et al.11 who demonstrated with a small cohort of stages I-III ACC patients (n = 65) at a single center that a cumulative GRAS score was beneficial in stratifying patient survival.
Interestingly, although each of the individual GRAS parameters have been associated previously with worse OS for ACC patients,10-16 in our cohort of 243 patients, only positive resection margin and preoperative symptoms of hormone hypersecretion were associated with worse OS in the univariate analysis, whereas neither elevated tumor grade nor increased patient age were significantly associ- ated with worse OS. This disparity may have been related to inherent differences in the US-ACCG and ENSAT patient populations or to the fact that the Libé et al.10 study investigated stages I-III patients versus our study, which examined all stages of disease. Nonetheless, tumor grade
(a)
Overall Survival by GRAS Score from Resection
1.00
p <0.001
0.75
Survival Probability
0.50
0.25
0.00
0
20
40
60
80
100
120
Time (Months)
GRAS Score + 0 + 1 +2 +3 + 4
Number at risk
0
22
12
11
9
7
5
4
1
52
27
16
10
6
3
2
2
89
45
21
15
12
6
5
3
61
23
11
9
5
2
1
4
13
2
1
1
0
0
0
0
20
40
60
80
100
120
Time (Months)
(b) Disease-Free Survival by GRAS Score from Resection
1.00
p =0.002
Disease-Free Survival Probability
0.75
0.50
0.25
0.00
0
10
20
30
40
50
60
Time (Months)
GRAS Score + 0 + 1 + 2+ 3 + 4
0
21
12
10
8
7
6
5
1
52
22
14
10
7
6
5
2
88
42
28
18
9
5
5
3
58
26
15
10
8
6
5
4
13
3
1
0
0
0
0
0
10
20
30
40
50
60
Time (Months)
GRAS Score
GRAS Score
(a)
Overall Survival by GRAS Score
1.00
p < 0.001
0.75
Survival Probability
0.50
0.25
0.00
0
20
40
60
80
100
120
Time (Months)
GRAS Score + 0 + 1-3 + 4
Number at risk
GRAS Score
0
22
12
11
9
7
5
4
1-3
202
95
48
34
23
11
8
4
13
2
1
1
0
0
0
0
20
40
60
80
100
120
Time (Months)
(b) Disease-Free Survival by GRAS Score
1.00
p<0.001
Disease-Free Survival Probability
0.75
0.50
0.25
I
+
0.00
0
10
20
30
40
50
60
Time (Months)
GRAS Score + 0 ++ 1-3 + 4
Number at risk
GRAS Score
0
21
12
10
8
7
6
5
1-3
198
90
57
38
24
17
15
4
13
3
1
0
0
0
0
0
10
20
30
40
50
60
Time (Months)
and patient age did seem to add value in the context of a cumulative GRAS score, suggesting that synergism may exist between the GRAS variables that improves their prognostic value when they are considered simultaneously as a five-tier scoring system.
Additionally, although the median OS did decrease sequentially with each additional GRAS “point,” we noted that the separation of survival was not robust for the intermediate GRAS scores (GRAS scores 1-3), as indi- cated by the overlapping Kaplan-Meier curves, particularly in the case of DFS. Although our study possibly was underpowered to detect a true survival difference between the patients in this intermediate range, it also is possible that differentiating between GRAS scores 1, 2, and 3 is not clinically meaningful. Future investigations examining the utility of a cumulative GRAS score in stratifying postop- erative ACC patient survival should consider both five- and three-tier scoring systems to find the most clinically valu- able means of prognostication.
We acknowledge several limitations of this study. First, the analysis was performed in a retrospective fashion and thus was inherently limited in its ability to make conclu- sions about causality. Second, although we think that the multi-institutional nature of this study was a strength allowing us to assemble a sufficiently robust database to derive meaningful associations about the natural history of these rare tumors, the collaboration was limited to large, academic referral centers, which may have led to selection bias, limiting the generalizability of our conclusions.
Additionally, multi-institutional collaborations are sub- ject to inter-observer variability, both within and between institutions. Our study depended on Ki-67 and mitotic rate data from pathology reports to assign a tumor grade. Whereas some have advocated incorporation of tumor grade into the ENSAT staging system for ACC,18 others have shown that both Ki-67 and mitotic rate are subject to inter-observer variability due to differences in both quan- tification method and tumor heterogeneity.19 Such variability may have limited our ability to detect the prognostic significance of tumor grade in ACC, particularly in the context of a multi-institutional collaboration.
Also due to the multi-institutional nature of our study and its long study period, the possibility of variability in practice patterns between institutions and even within institutions during the study period was likely and even expected. To the latter point, we observed variability in the ACC patient GRAS scores and management during the duration of the study period, as illustrated in Fig. S1a and d.
Additionally, we observed that the rate of mitotane therapy was 32% late in the study period compared with 19% early in the study period (p = 0.07; Fig. S1b), and similarly, that the use of adjuvant chemotherapy (21% vs
10%; p = 0.07; Fig. S1c) and adjuvant radiation therapy (11% vs 1%; p = 0.02; Fig. S1d) appeared to increase slightly over time. However, despite these incremental changes in patient management, OS an DFS did not change significantly between the early and late study cohort (Fig. S2a and b). Importantly, however, the US-ACCG database fails to fully capture other differences in diagno- sis, staging, and medical management. Most notably, although tumor functionality was reported for all patients in the form of diagnoses including non-functional or glu- cocorticoid-, virilizing/feminizing-, or mineralocorticoid- secreting tumors, complete data on preoperative biochem- ical evaluation were not uniformly reported, limiting the authors’ ability to independently validate these diagnoses or compare hormone levels between patients. Similarly, although the US-ACCG database does report on the pre- operative imaging method (CT vs MRI), it does not include the region of imaging (e.g., chest vs abdomen/pelvic CT scan) which may have implications for the accuracy of preoperative staging and, as a result, postoperative calcu- lation of DFS. Furthermore, with regard to medical management, systemic therapies were reported as either mitotane or “chemotherapy” and did not differentiate between types of chemotherapy, nor did the US-ACCG database specify the duration of treatment or whether the recommended therapy regimen was completed as planned or terminated early. All of limitations of the US-ACCG database limited our ability to determine whether delivery of these non-surgical therapies differed between groups or whether they impinged on our findings.
Additionally, in our study’s scoring system, “symp- toms” were defined as symptoms of hormone hypersecretion and did not include other symptoms, including those related to tumor mass effect. We chose to include only symptoms of hormone hypersecretion for two primary reasons. First, multiple previous studies have demonstrated that symptoms of hormone hypersecretion have independent prognostic value in ACC,10,20,21 whereas symptoms of tumor mass effect have no independent prognostic value. Second, when symptoms of tumor mass effect were included in the cumulative GRAS scoring system using our study cohort, no additional prognostic benefit was observed (data not shown).
Importantly, there is a known differential impact on prognostication between types of hormone-secreting ACCs, and our cumulative GRAS score failed to account for this differential impact. Consistent with previously published literature,20,21 cortisol-secreting tumors in our patient cohort were significantly associated with worse OS and DFS than non-functional tumors and mineralocorti- coid- or virilizing/feminizing-secreting tumors (Table S1a and b). We chose not to distinguish between subtypes of hormone-secreting tumors in our primary analysis despite
these known differences in order to remain consistent with previous descriptions of the GRAS scoring system and to maintain the simplicity and potential clinical utility of the cumulative GRAS score. Similarly, although it might be argued that a more appropriate scoring system would weigh each GRAS variable proportionally to its prognostic value rather than give each category one equal point, we believe the simplicity and overall effectiveness of the scoring system as outlined outweighs the potential benefits of a more complex scoring system.
Finally, a cumulative GRAS score may simply be co- linear with disease stage and not provide additional prog- nostic benefit. Unfortunately, our study was underpowered to perform subanalyses examining the ability of the GRAS score to stratify survival for each ENSAT stage individu- ally, which would have allowed us to examine this possibility effectively. However, each of the GRAS parameters have previously demonstrated individual prog- nostic value in both early- and late-stage ACC,10-16 and given that these variables are not directly accounted for in the ENSAT staging system, we predict that a cumulative GRAS score may be helpful in supplementing the already well-validated ENSAT staging system. Future studies are needed to determine whether a cumulative GRAS score can supplement the ENSAT staging system by providing additional granularity.
Our study demonstrated the effectiveness of a cumula- tive GRAS score in stratifying prognoses for postoperative ACC patients with stages I-IV disease. Although we pre- sented both five- and three-tier scoring options, it remains unclear which scoring system is superior in stratifying cancer-related outcomes. The five-tier scoring system (GRAS 0-4) may be able to predict differences in OS accurately after resection for ACC, but our three-tier scoring system (GRAS 0, 1-3, 4) may be more meaningful for predicting differences in DFS. Whereas our European colleagues who originally coined the GRAS parameters did not examine the utility of a cumulative score in stratifying survival,10 Liang et al.11 in early 2020 published the utility of a cumulative GRAS score for stratifying survival in a small cohort (n = 65) of stages I-III ACC patients. Inter- estingly, Liang et al.11 suggested a three-tier approach, with cumulative GRAS scores grouped as ≤ 1, 2, and ≥ 3. Future studies should consider all these approaches.
Importantly, the prognostication for patients with ACC continues to evolve over time with the benefit of new research and the advent of new scoring systems.22 How- ever, given the consistency with which the GRAS parameters predict survival in multiple datasets and studies, it is likely that these factors will maintain their clinical benefit and serve as a foundation onto which additional factors can be added as the understanding of disease biol- ogy progresses.
In conclusion, the cumulative GRAS score appears to stratify OS outcomes in ACC effectively for patients of all disease stages who undergo primary surgical resection. The findings of this retrospective study would benefit from validation in a prospective manner, although we acknowledge the challenges of conducting such prospec- tive analyses in the setting of rare tumors such as ACCs.
DISCLOSURES Dr. Jason Sicklick reports research funding from Foundation Medicine, Inc and Amgen and provision of consultant services at Deciphera. The other authors have no conflicts of interest.
REFERENCES
1. Bilimoria KY, Shen WT, Elaraj D, et al. Adrenocortical carci- noma in the United States. Cancer. 2008;113:3130-6. https://doi. org/10.1002/cncr.23886.
2. Golden SH, Robinson KA, Saldanha I, Anton B, Ladenson PW. Prevalence and incidence of endocrine and metabolic disorders in the United States: a comprehensive review. J Clin Endocrinol Metab. 2009;94:1853-78. https://doi.org/10.1210/jc.2008-2291.
3. Fassnacht M, Johanssen S, Quinkler M, et al. Limited prognostic value of the 2004 International Union Against Cancer staging classification for adrenocortical carcinoma. Cancer. 2008;115:243-50. https://doi.org/10.1002/cncr.24030.
4. Kebebew E, Reiff E, Duh Q-Y, Clark OH, McMillan A. Extent of disease at presentation and outcome for adrenocortical carci- noma: have we made progress? World J Surg. 2006;30:872-8. h ttps://doi.org/10.1007/s00268-005-0329-x.
5. Lughezzani G, Sun M, Perrotte P, et al. The European Network for the Study of Adrenal Tumors staging system is prognostically superior to the international union against cancer-staging system: a North American validation. Eur J Cancer. 2010;46:713-9. h ttps://doi.org/10.1016/j.ejca.2009.12.007.
6. Gaujoux S, Weinandt M, Bonnet S, Reslinger V, Bertherat J, Dousset B. Surgical treatment of adrenal carcinoma. J Visceral Surg. 2017;154:335-43. https://doi.org/10.1016/j.jviscsurg.2017. 06.010.
7. Marincola Smith P, Kiernan CM, Tran TB, et al. Role of addi- tional organ resection in adrenocortical carcinoma: analysis of 167 patients from the U.S. Adrenocortical Carcinoma Database. Ann Surg Oncol. 2018;25:2308-15. https://doi.org/10.1245/s104 34-018-6546-y.
8. Prendergast KM, Marincola Smith P, Tran TB, et al. Features of synchronous versus metachronous metastasectomy in adrenal cortical carcinoma: analysis from the US adrenocortical carci- noma database. Surgery. 2020;167:352-7. https://doi.org/10. 1016/j.surg.2019.05.024.
9. Ayala-Ramirez M, Jasim S, Feng L, et al. Adrenocortical carci- noma: clinical outcomes and prognosis of 330 patients at a tertiary care center. Eur J Endocrinol. 2013;169:891-9. https://d oi.org/10.1530/EJE-13-0519.
10. Libé R, Borget I, Ronchi CL, et al. Prognostic factors in stage III- IV adrenocortical carcinomas (ACC): a European Network for the Study of Adrenal Tumor (ENSAT) study. Ann Oncol. 2015;26:2119-25. https://doi.org/10.1093/annonc/mdv329.
11. Liang J, Liu Z, Zhou L, et al. The clinical utility of “GRAS” parameters in stage I-III adrenocortical carcinomas: long-term data from a high-volume institution. Endocrine. 2020;67(2):449-56. https://doi.org/10.1007/s12020-019-02141-2.
12. Lau SK, Weiss LM. The Weiss system for evaluating adreno- cortical neoplasms: 25 years later. Hum Pathol. 2009;40:1-12. h ttps://doi.org/10.1016/j.humpath.2009.03.010.
13. Abiven G, Coste J, Groussin L, et al. Clinical and biological features in the prognosis of adrenocortical cancer: poor outcome of cortisol-secreting tumors in a series of 202 consecutive patients. J Clin Endocrinol Metab. 2006;91:2650-5. https://doi. org/10.1210/jc.2005-2730.
14. Volante M, Bollito E, Sperone P, et al. Clinicopathological study of a series of 92 adrenocortical carcinomas: from a proposal of simplified diagnostic algorithm to prognostic stratification. Histopathology. 2009;55:535-43. https://doi.org/10.1111/j.1365- 2559.2009.03423.x.
15. Berruti A, Fassnacht M, Haak H, et al. Prognostic role of overt hypercortisolism in completely operated patients with adreno- cortical cancer. Eur Urol. 2014;65(4):832-8. https://doi.org/10. 1016/j.eururo.2013.11.006.
16. Stojadinovic A, Ghossein RA, Hoos A, et al. Adrenocortical carcinoma: clinical, morphologic, and molecular characterization. JCO. 2002;40:941-50.
17. Edge SB, Compton CC 2010 The American Joint Committee on Cancer: the 7th Edition of the AJCC Cancer Staging Manual and the future of TNM. Ann Surg Oncol. 2010;17:1471-4. https://doi. org/10.1245/s10434-010-0985-4.
18. Miller BS, Gauger PG, Hammer GD, Giordano TJ, Doherty GM. Proposal for modification of the ENSAT staging system for adrenocortical carcinoma using tumor grade. Langenbecks Arch
Surg. 2010;395:955-61. https://doi.org/10.1007/s00423-010-069 8-y.
19. Papathomas TG, Pucci E, Giordano TJ, et al. An international Ki67 reproducibility study in adrenal cortical carcinoma. Am J Surg Pathol. 2016;40:569-76. https://doi.org/10.1097/PAS. 0000000000000574.
20. Kim Y, Margonis GA, Prescott JD, et al. Nomograms to predict recurrence-free and overall survival after curative resection of adrenocortical carcinoma. JAMA Surg. 2016;151:365-73. http s://doi.org/10.1001/jamasurg.2015.4516.
21. Margonis GA, Kim Y, Tran TB, et al. Outcomes after resection of cortisol-secreting adrenocortical carcinoma. Am J Surg. 2016;211:1106-13. https://doi.org/10.1016/j.amjsurg.2015.09. 020.
22. Lippert J, Appenzeller S, Liang R, Sbiera S, Kircher S, Altieri B, et al. Targeted molecular analysis in adrenocortical carcinomas: a strategy toward improved personalized prognostication. J Clin Endocrinol Metab. 2018;103:4511-23. https://doi.org/10.1210/jc. 2018-01348.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.