The Modified S-GRAS Scoring System for Prognosis in Korean with Adrenocortical Carcinoma
Article information
Abstract
Background
Adrenocortical carcinomas (ACCs) are rare tumors with aggressive but varied prognosis. Stage, Grade, Resection status, Age, Symptoms (S-GRAS) score, based on clinical and pathological factors, was found to best stratify the prognosis of European ACC patients. This study assessed the prognostic performance of modified S-GRAS (mS-GRAS) scores including modified grade (mG) by integrating mitotic counts into the Ki67 index (original grade), in Korean ACC patients.
Methods
Patients who underwent surgery for ACC between January 1996 and December 2022 at three medical centers in Korea were retrospectively analyzed. mS-GRAS scores were calculated based on tumor stage, mG (Ki67 index or mitotic counts), resection status, age, and symptoms. Patients were divided into four groups (0–1, 2–3, 4–5, and 6–9 points) based on total mS-GRAS score. The associations of each variable and mS-GRAS score with recurrence and survival were evaluated using Cox regression analysis, Harrell’s concordance index (C-index), and the Kaplan–Meier method.
Results
Data on mS-GRAS components were available for 114 of the 153 patients who underwent surgery for ACC. These 114 patients had recurrence and death rates of 61.4% and 48.2%, respectively. mS-GRAS score was a significantly better predictor of recurrence (C-index=0.829) and death (C-index=0.747) than each component (P<0.05), except for resection status. mS-GRAS scores correlated with shorter progression-free survival (P=8.34E-24) and overall survival (P=2.72E-13).
Conclusion
mS-GRAS scores showed better prognostic performance than tumor stage and grade in Asian patients who underwent surgery for ACC.
INTRODUCTION
Adrenocortical carcinomas (ACCs) are very rare malignancies with the incidence of 0.7 to 2.0 per million per year [1,2] and about 1.0 per million per year in Korea [3]. Prognosis of ACC vary widely by 5-year overall survival (OS) rates range from 10% to 60% [4,5], despite generally poor prognosis due to the aggressive behavior and tendency to recur of these tumors. A registry-based nationwide survey of 204 patients in Korea showed that the 5-year OS and disease-specific survival (DSS) rates were 64.5% and 70.6%, respectively [6]. Surgery remains the only curative modality in the treatment of ACC. However, more than 50% ACC underwent initial complete resection will still develop disease recurrence or metastasis [7], and 15% die within 2 years [5]. So, reliable determination of patient prognosis after resection is critical to guide the frequency of follow-up examinations and adjuvant treatment, as well as to more accurately counsel patients regarding their long-term outcomes.
Recently, several clinical parameters, histopathological, pathological, and surgical related factors were found to be correlated with the prognosis of ACC patients [8,9]. The European Network for the Study of Adrenal Tumours (ENSAT) staging system is widely used as the standard prognostic factor in ACC [5,8,10]. Five-year survival is 60%–80% for tumors limited to adrenal, 35%–50% for locally advanced disease, and much lower in case of metastasis [8]. Other prognostic factors include resection status [11-13], Ki67 proliferation index [14], and mitotic counts [8,15,16]. The European Society of Endocrinology-ENSAT guidelines have stratified the risk of recurrence based on tumor stage, resection status, and Ki-67 index (or mitotic counts). Although older age and hypercortisolism are associated with poorer OS [17-19], their prognostic value remains uncertain [20,21].
Although individual clinical and histopathological prognostic markers are limited in predicting patient prognosis, their combinations may have better prognostic accuracy in patients with ACC. The prognostic value of the GRAS components, including tumor grade (G; Weiss score >6 and/or Ki67 index ≥20%), resection status (R), age (A), and tumor- or hormone-related symptoms (S), was initially evaluated in 444 patients with advanced ACC [22]. Stage, Grade, Resection status, Age, Symptoms (S-GRAS) scores, a modified form of the GRAS classification that includes ENSAT stage and focuses on Ki67 index for grading, allowed better stratification than its individual clinical and histopathological characteristics in 107 patients [23]. Recently, S-GRAS scores showed a prognostic performance superior to that of tumor stage and Ki67 index in predicting progression-free survival (PFS) and DSS in 942 European patients with ACC [24]. Less is known, however, about the prognostic performance of S-GRAS scores in Asian patients who undergo surgery for ACC. Understanding the clinical characteristics and prognostic stratification of ACC in Asian patients is essential for their proper management. The present study therefore investigated the prognostic utility of modified S-GRAS (mS-GRAS) scores, which incorporate a modified grade by integrating mitotic counts with the Ki67 index, in Korean patients undergoing adrenalectomy for ACC, prompted by the limited availability of Ki67 index data.
METHODS
Patients and data collection
This retrospective, multicenter study included patients who underwent surgery for ACC at three tertiary hospitals in Korea, Asan Medical Center (AMC), Seoul National University Hospital (SNUH), and Seoul National University Bundang Hospital (SNUBH), and was part of the Korean Adrenal Disorder Study (KADS; clinicaltrial.gov No. 06229405). Patients aged ≥18 years diagnosed with or treated for histologically confirmed ACC at these three tertiary hospitals between January 1, 1996 and December 31, 2022 and available for follow-up were recruited. Only conventional subtype of ACC included [25]. The study protocol conformed to the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Boards of AMC (No. 2022-1496), SNUH (No. 2204-155-1320), and SNUBH (No. B-2207-769-401), which waived requirements for written informed consent owing to the retrospective nature of this study.
Data recorded from patient records included age, sex, age at diagnosis, body mass index, tumor size, modality of tumor diagnosis (e.g., incidental; symptom-related, either to the tumor mass or hormone secretions; or other/unknown), ENSAT stage, treatment (surgery, adjuvant mitotane, or palliative mitotane), and date of adrenalectomy. Other factors included resection status (with R0 indicating complete resection, RX indicating unknown resection status, R1 indicating the removal of all macroscopic disease, accompanied by microscopic residual disease, and R2 indicating macroscopic residual disease); mitotic counts; Ki67 index; date of disease recurrence after primary surgery; and date of last visit or death. Symptoms were defined as hormone-related if they were associated with excess adrenal hormone(s) (e.g., glucocorticoids, mineralocorticoids, or androgen), tumor-related if they were associated with a mass effect (e.g., abdominal pain), or systemic cancer-related (e.g., fatigue or weight loss).
S-GRAS scores of patients undergoing adrenalectomy were calculated as described [23,24]: ENSAT stage (S; stage 1 or 2=0 points; stage 3=1 point; stage 4=2 points), grade based on Ki67 index (G; 0%–9%=0 point; 10%–19%=1 point; ≥20%=2 points), resection status of the primary tumor (R0=0 points; RX=1 point; R1=2 points; R2=3 points), age at diagnosis (A; <50 years=0 point; ≥50 years=1 point), and hormone, tumor, or systemic cancer-related symptoms at presentation (S; no=0 point; yes=1 point). If Ki67 index was not available, it could be replaced by mitotic count [8]. Modified G based on mitotic counts in patients without information on Ki67 index was scored as 0 points for ≤5 mitoses/50 high powered fields (HPF), 1 point for 6–20 mitoses/50 HPF, and 2 points for >20 mitoses/50 HPF [25]. mS-GRAS scores ranged from 0 to 9 and were categorized into four groups, consisting of patients with mS-GRAS scores of 0–1, 2–3, 4–5, and 6–9 points [23,24].
Endpoints
PFS and OS were determined by retrospective review of medical records at each institution. PFS was defined as the time from primary tumor resection to the first radiological evidence of relapse or death from any cause, and OS was defined as the time from diagnosis of ACC until death from any cause.
Statistical analysis
Continuous variables are presented as the median (interquartile range [IQR]) and categorical variables are presented as the number (percentage). The prognostic effects of total mS-GRAS score, its individual components, and groups of mS-GRAS scores for PFS and OS were evaluated by univariable Cox regression analysis, with results reported as hazard ratios (HRs), 95% confidence intervals (CIs), and P values. The discriminative performances of total mS-GRAS score and groups of mSGRAS scores were compared with the performances of single components of mS-GRAS scores using Harrell’s Concordance index (C-index) [26]. Harrell’s C-index is defined as the proportion of patient pairs in which the predicted and observed survival outcomes are in agreement with respect to rank, with a higher C-index indicative of better model discrimination. PFS and OS for each of the groups of mS-GRAS score groups were determined by the Kaplan–Meier method and compared by log-rank tests. All statistical analyses were performed using R version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria), with P values <0.05 defined as statistically significant.
RESULTS
Clinical characteristics of patients with available mS-GRAS data
Of the 153 patients who underwent adrenalectomy for ACC between January 1996 and December 2022 at three medical centers in Korea, 114 (74.5%) had available mS-GRAS data (Table 1). These patients were followed-up for a median 58.8 months (range, 1.4 to 204.6). Of these 114 patients, 60 (52.6%) were aged ≥50 years, and 54 (47.4%) had symptoms. Thirteen (11.4%), 54 (47.4%), 17 (14.9%), and 30 (26.3%) patients had ENSAT stages 1, 2, 3, and 4 disease, respectively. Thirty-five (30.7%) patients received adjuvant mitotane, at a median dose of 2,000 mg. Blood levels of mitotane were measured in nine (26.5%) of these 35 patients, with six (66.7%) of these nine having mitotane levels over 14 mg/L. Relapse occurred in 70 (60.5%) of the 114 patients, at a median 23.0 months (IQR, 12.1 to 39.1) (data not shown). Locoregional recurrence occurred in nine (7.9%) patients, distant metastases in 45 (39.5%), and both in 16 (14.0%). Fifty-five (48.2%) patients died at a median 120.0 months (IQR, 67.9 to not reached).
The proportions of women, symptomatic patients, and patients with ENSAT stage 1 were higher, whereas the proportion of patients with ENSAT stage 3 were lower in the 114 patients with available mS-GRAS data than in the 39 without available mSGRAS data (Supplemental Table S1). In addition, patients with available mS-GRAS data were younger in age and had larger sized tumors than those without available mS-GRAS data.
Prognostic performance of total mS-GRAS score, its individual components, and categories of mS-GRAS scores on PFS ENSAT stage 4 (vs. ENSAT stage 1–2), modified G2 (vs. modified G0), RX and R2 (vs. R0), and symptoms (vs. no symptoms) were associated with PFS, whereas age was not (Table 2). Evaluation of all 153 patients with ACC who underwent adrenalectomy showed that ENSAT stage 3 and 4 (vs. ENSAT stage 1–2), RX and R2 (vs. R0), and symptoms (vs. no symptoms) were associated with relapse, whereas age was not (Supplemental Table S2).
Discrimination statistics using Harrell’s C-index showed that total mS-GRAS score (C-index=0.829; 95% CI, 0.759 to 0.880) was superior to each of its components (P<0.05 each), except for R (P=0.333), in predicting relapse (Table 2). Groups of mSGRAS scores (C-index=0.815; 95% CI, 0.751 to 0.864) also showed better prognostic performance for relapse than each of its components (P<0.05 each), except for ENSAT stage (P=0.115) and R (P=0.832) (Table 2). Compared with patients in the mS-GRAS 0–1 group, the HRs of the mS-GRAS 6–9, 4–5, and 2–3 groups for relapse were 69.51 (95% CI, 25.99 to 185.86; P<0.001), 3.18 (95% CI, 1.19 to 8.48; P=0.021), and 1.84 (95% CI, 0.87 to 3.90; P=0.109), respectively. Groups of mS-GRAS score also correlated significantly with shorter PFS (P=8.34E-24) (Fig. 1).
Kaplan–Meier analyses of (A) progression-free survival and (B) overall survival according to modified Stage, Grade, Resection status, Age, Symptom (mS-GRAS) score grouping in patients with adrenocortical carcinoma who underwent adrenalectomy (n=114). Components of the mS-GRAS score included European Network for the Study of Adrenal Tumour (ENSAT) stage (S), modified grading (modified G), resection status (R), age (A), and tumor- or hormone-related symptoms (S).
Prognostic performance of total S-GRAS score, its original grading components, and categories of S-GRAS scores on PFS
In 79 patients with available Ki67 index as original grading component of S-GRAS to investigate the prognostic performance of original S-GRAS score, S-GRAS score (C-index=0.805; 95% CI, 0.720 to 0.870) and groups of S-GRAS scores (C-index=0.803; 95% CI, 0.717 to 0.864) showed better prognostic performance for relapse than grading compartment using Ki67 index of original S-GRAS score (C-index=0.648; 95% CI, 0.552 to 0.731; P<0.001 each) (Supplemental Table S3).
Prognostic performance of total mS-GRAS score, its individual components, and categories of mS-GRAS scores on OS
ENSAT stage 4 (vs. ENSAT stages 1–2), modified G2 (vs. modified G0), and RX and R2 (vs. R0) were all associated with OS, whereas age and symptom were not (Table 3). In all 153 patients with ACC who underwent adrenalectomy showed that ENSAT stages 3 and 4 (vs. ENSAT stages 1–2), R2 (vs. R0), and symptoms (vs. no symptoms) were also associated with death, whereas age was not (Supplemental Table S2).
Discrimination statistics using Harrell’s C-index showed that total mS-GRAS score (C-index=0.747; 95% CI, 0.684 to 0.805) was superior to each of its components (P<0.05 each), except R (P=0.148), in predicting death (Table 3). Group of mS-GRAS score (C-index=0.746; 95% CI, 0.673 to 0.795) also showed better prognostic performance than each of its components (P<0.05 each), except R (P=0.141), in predicting death (Table 3). Compared with patients in the mS-GRAS 0–1 group, the HRs of the mS-GRAS 6–9, 4–5, and 2–3 groups for death were 13.14 (95% CI, 5.01 to 34.50; P<0.001), 3.22 (95% CI, 0.93 to 11.15; P=0.065), and 2.32 (95% CI, 0.85 to 6.28; P=0.099), respectively. Groups of mS-GRAS scores correlated significantly with shorter OS (P=2.72E-13) (Fig. 1).
Prognostic performance of total S-GRAS score, its original grading components, and categories of S-GRAS scores on OS
Compared with grading (C-index=0.629; 95% CI, 0.534 to 0.714), S-GRAS score (C-index=0.727; 95% CI, 0.633 to 0.798; P=0.024) and S-GRAS group (C-index=0.740; 95% CI, 0.640 to 0.802; P=0.035) showed better prognostic performance for death in the 79 patients with available K67 index data (Supplemental Table S4).
DISCUSSION
To our knowledge, this study is the first to validate the prognostic performance of mS-GRAS score-a minimally modified version of the S-GRAS scoring system, based on clinical and histopathological characteristics, for both PFS and OS in Asian patients with ACC. The mS-GRAS score was found to be superior to ENSAT staging, grading, age, and symptoms, but not to resection status, in predicting OS and PFS in Asian patients with ACC. Grouping of mS-GRAS scores was also superior to grading, age, and symptoms in the prognosis discrimination of both PFS and OS. These findings indicated that the mS-GRAS scoring system might be useful in predicting post-surgical prognosis in patients with ACC.
Several clinical and histopathological characteristics have been associated with the prognosis of ACC patients following surgery. ENSAT stage, reflecting the extent of the tumor, was found to be the most important prognostic factor [5,8,10]. Other major prognostic factors include resection status [11-13], Ki67 index [8,14-16], and mitotic counts [8,27]. Guidelines have therefore stratified the risk of recurrence based on tumor stage, resection status, and Ki-67 index (or mitotic count) [8], with a recent study reporting that the prognostic performance of S-GRAS was superior to that of tumor stage, resection status, and Ki67 index in predicting PFS and DSS in European patients with ACC [24]. Furthermore, a single-center study involving 51 patients in China showed that groups of S-GRAS scores were predictive of cancer-specific survival (CSS) [28]. That study, however, found that of the parameters used to calculate S-GRAS score, only ENSAT stage was significantly associated with CSS. Therefore, to our knowledge, the present study is the first to validate the superior prognostic performance of the mS-GRAS score, compared with ENSAT staging and grading (Ki67 index or mitotic counts), for both PFS and OS in Asian patients with ACC.
In agreement with the results of previous studies showing that adequate resection status, including locoregional lymph node dissection and tumor thrombus embolectomy, improved both diagnostic accuracy and therapeutic outcome [8,29,30], resection status was significantly associated with PFS and OS in the present study. In addition, S-GRAS score was superior to resection status in predicting PFS and DSS [24]. The present study showed that Harrell’s C-index of mS-GRAS score was numerically higher than that of resection status, but the difference was not statistically significant. Although R1 status has been associated with a poorer prognosis than RX [24], R1 status was not significantly associated with PFS and OS in the present study. It is difficult to exactly quantify the impact of resection status on the risk of relapse and death due to lack of accurate reporting and frequent RX reports [9]. In the present study, the small number of patients with R1 status (seven of 114, 6.1%) and the potential misclassification of R2 as RX status due to the inclusion of patients with ACC beginning in 1996 may explain the lack of statistical significance between the prognostic performances of mS-GRAS score and resection status.
Hypercortisolism was also reported to be prognostic of clinical outcomes in patients with ACC [17-19]. Because advanced age was reported to be a prognostic factor [31], age was included in the mS-GRAS scoring system. Nevertheless, the present study found that mS-GRAS scores were superior to symptoms and age in predicting both PFS and OS. The presence of symptoms was associated with relapse but not with death in the 114 patients with mS-GRAS scores. By contrast, the presence of symptoms was associated with both relapse and death in the 153 patients who underwent adrenalectomy. In both groups, however, advanced age was not associated with relapse or death, in agreement with a study showing that age was the only factor in S-GRAS scores not associated with PFS and DSS [24]. And a registry-based nationwide survey of 204 patients in Korea showing that age was not associated with mortality [6]. The cut-off value of 50 years was derived from the median age of study subjects in the ENSAT study [22]. Taken together, these findings suggest the need for additional studies assessing whether advanced age is a prognostic factor in patients with ACC.
In this study, approximately 65% of patients who underwent adrenalectomy experienced relapse and 90% of the patients who experienced relapse showed recurrence of distant metastases. Mitotane is the mainstay of adjuvant therapy in ACC, with retrospective studies showing that mitotane treatment was associated with significantly longer PFS and OS [32]. The utility of adjuvant mitotane for all patients with ACC is unclear, as studies have reported conflicting results; mitotane administration is cumbersome, requiring complex hormone replacement, the need for careful monitoring, and the potential toxicity of this reagent; and the lack of reliable markers predicting response to treatment [33,34]. Currently, adjuvant mitotane is proposed for patients considered at high risk of recurrence (Ki67 index ≥ 10%, RX-R1, and/or ENSAT stage 3–4) [8,9,16]. S-GRAS scores of 4–5 have been associated with longer PFS in mitotane-treated patients [24]. Furthermore, an open-label, parallel, randomized, phase 3 (ADIUVO) trial found that adjuvant mitotane might not be indicated in patients with low to intermediate risk of recurrence (R0, ENSAT stage I–III, and Ki-67 ≤10%) [34]. These findings suggested that S-GRAS score can be used to stratify patients more likely to benefit from adjuvant mitotane.
The present study had several limitations, including its retrospective design. In addition, because Ki-67 indices were obtained for only 79 (51.6%) of the 153 patients who underwent adrenalectomy, a modified grading was used, consisting of mitotic counts and Ki-67 indices. Total S-GRAS score and SGRAS score groups by original grading using only the Ki-67 index were associated with PFS and OS. Third, we could not analyze the potential use of mS-GRAS score to stratify patients more likely to benefit from adjuvant mitotane.
In conclusion, the present study found that the prognostic performance of the mS-GRAS scoring system was superior to that of tumor stage and proliferation activity, including Ki67 index or mitotic count, in Korean patients who underwent surgery for ACC. The S-GRAS scoring system may be useful in predicting the prognosis of Asian patients who undergo surgery for ACC.
Supplementary Material
Supplemental Table S1.
Characteristics of ACC Patients witha (n=114) and without (n=39) Available mS-GRASb Data
Supplemental Table S2.
Analyses of PFS and OS Showing the HR of Several Components of mS-GRASa Scores in Patients Who Underwent Adrenalectomy (n=153)
Supplemental Table S3.
Analysis of PFS Showing the HR and Discrimination of the Original G Compartment, the S-GRASa Score, and Groups of S-GRAS Score in Patients with Available Data of Ki67 Index (n=79)b
Supplemental Table S4.
Analysis of OS Showing the HR and Discrimination of the Original G Compartment, the S-GRASa Score, and Groups of S-GRAS Score in Patients with Available Data of Ki67 Index (n=79)b
Notes
CONFLICTS OF INTEREST
Chang Ho Ahn in Lunit as a medical director and has a stock option in the firm. Jung Hee Kim is a deputy editor of the journal. But she was not involved in the peer reviewer selection, evaluation, or decision process of this article. No other potential conflicts of interest relevant to this article were reported.
AUTHOR CONTRIBUTIONS
Conception or design: S.H.L., J.H.K. Acquisition, analysis, or interpretation of data: S.K.B., S.H.L., S.S.P., C.H.A., S.H.K., W.W.K., Y.M.L., S.J.K., D.E.S., T.Y.S., K.E.L., J.H.K., K.C.J., J.M.K. Drafting the work or revising: S.K.B., S.H.L., J.H.K. Final approval of the manuscript: S.K.B., S.H.L., S.S.P., C.H.A., S.H.K., W.W.K., Y.M.L., S.J.K., D.E.S., T.Y.S., K.E.L., J.H.K., K.C.J., J.M.K.
Acknowledgements
We thanked Sookkyeong Seo for the data collection. This study was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare of the Republic of Korea (Project No. HI22C0049, HI21C0032).