BOSTON – A proposed system for staging adrenocortical carcinomas appears to more accurately predict prognoses across all age and stage groups, but the system is not quite ready for prime time, investigators say.
The system combines information on patient age, tumor stage, and nodal and metastatic (TNM) status. In a retrospective study, the TNM-age system was better at predicting 5-year overall survival than was the European Network for the Study of Adrenal Tumors (ENSAT) staging system, which was in turn a modification of another system, said Dr. Elliot Asare, a research resident in the department of surgical education at the Medical College of Wisconsin in Milwaukee.
Dr. Elliott Asare
The improved predictive power of the TNM-age system may be due to differences in tumor biology between older and younger patients, Dr. Asare noted.
"Improved staging allows for a more accurate assessment of the natural history of the disease," he said at the annual meeting of the American Association of Endocrine Surgeons.
The two main staging systems currently used for adrenocortical carcinoma (ACC) are the American Joint Committee on Cancer/International Union Against Cancer (AJCC/UICC) system, and ENSAT, which was proposed in 2009 as a modified version of the AJCC/UICC staging system. Under the ENSAT modification, stage IV disease would be limited to patients with distant metastases.
However, the ENSAT criteria were not good at discriminating between stage I and stage II disease, and failed to show a significant survival difference, Dr. Asare noted.
The investigators undertook to see whether the ENSAT’s prognostic accuracy might improve with a larger data set, and to determine whether adding age as a variable to staging ACC could improve the accuracy of survival predictions.
They drew data on patients with a histologic diagnosis of ACC from 1985 through 2006 in the National Cancer Database, and used Surveillance, Epidemiology and End Results (SEER) summary stage information to establish TNM stage according to ENSAT criteria.
They considered tumor size, resection margin status, histologic grade, lymph node status, SEER summary stage, vital status, and age of diagnosis.
Out of a total of 3,263 patients with ACC, sufficient data were available on 1,597.
When they applied the staging criteria, they were able to validate the ENSAT system for stage III vs. stage IV (P less than .0001), and for stage II vs. state III (P less than .0001), but no significant differences between I and II (P =.04). The 5-year overall survival rates under ENSAT were 68% for stage I and 61% for stage II.
They then developed their alternative staging system by adding age to the mix, as follows:
• Stage 1: T1-T2, N0, M0, age 55 or younger.
• Stage II: T1-T2, N0, M0, age over 55.
• Stage III: T1-T2, N1, M0, any age, or T3-T4, any N, M0, any age.
• Stage IV: any T, any N, M1.
By using this system applied to the same cohort, they found that the respective 5-year overall survival rates (stage I-IV) were 70%, 53%, 37%, and 9.7%, respectively. In addition, the survival rates were significantly different between stages I and II (P less than .0001), stages II and III (P = .0004), and stages III and IV (P less than .0001).
Significant predictors of death under the TNM-age staging were stage II and above, positive tumor resection margins, and grade.
Dr. Asare noted that the study was limited by the lack of some variables in the database and an absence of information on cause-specific mortality, and by the fact that age cannot be used as a continuous variable in a classification system.
The staging system needs to be tested in a validation study with information from an independent database, he added.
Dr. Asare disclosed receiving support from the American College of Surgeons Clinical Scholars in Residence fellowship, partially supported by an unrestricted education grant from Genentech.