To revise the FIGO staging for carcinoma of the vulva using a new approach that involves analyses of prospectively collected data. The FIGO Committee for Gynecologic Oncology reviewed the recent literature to gain an insight into the impact of the 2009 vulvar cancer staging revision. The Committee resolved to revise the staging with a goal of simplification and actively collaborated with the United States National Cancer Database to analyze prospectively collected data on carcinoma of the vulva. Many tumor characteristics were collected for all stages of vulvar cancer treated between 2010 and 2017. Statistical analysis was performed with SAS software. Overall survival was estimated based on tumor characteristics. Log-rank and Wilcoxon tests were used to analyze overall survival similarities between and within groups of tumor characteristics. Characteristics with similar survivals were then grouped into the same stages and substages. Kaplan-Meier overall survival curves were generated for the resulting stages and substages. There were 12 063 cases with available data. The resulting new staging for carcinoma of the vulva has two substages in Stage I, no substage in Stage II, three substages in Stage III, and two substages in Stage IV. The Kaplan-Meier overall survival curves showed clear separation between stages and substages. The 2021 vulvar cancer staging is the first from the FIGO Committee for Gynecologic Oncology to be derived from data analyses. This revision has a new definition for depth of invasion, uses the same definition for lymph node metastases utilized in cervical cancer, and allows findings from cross-sectional imaging to be incorporated into vulvar cancer staging. The 2021 FIGO staging for carcinoma of the vulva is data-derived, validated, and much simpler than earlier revisions.

Background and Objectives Cancer registries must focus on data capture which returns value while reducing resource burden with minimal loss of data. Identifying the optimum length of follow‐up data collection for patients with cancer achieves this goal. Methods A two‐step analysis using entropy calculations to assess information gain for each follow‐up year, and second‐order differences to compare survival outcomes between the defined follow‐up periods and lifetime follow‐up. A total of 391 567 adult cases, deidentified in the National Cancer Database and diagnosed in 1989. Comparisons examined a subset of 61 908 lung cancer cases, 48 387 colon and rectal cancer cases, and 64 134 breast cancer cases in adults. A total of 4133 pediatric cases were diagnosed in 1989 examining 1065 leukemia cases and 494 lymphoma cases. Results Annual increases in information gain fell below 1% after 16 years of follow‐up for adult cases and 9 years for pediatric cases. Comparison of second‐order differences showed 62% of the comparisons were similar between 15 years and lifetime follow‐up when examining restricted mean survival time. In addition, 90% of the comparisons were statistically similar when comparing hazard ratios. Conclusions Survival analysis using 15 years postdiagnosis follow‐up showed minimal differences in information gain compared to lifetime follow‐up.

256 Background: Current staging for colon cancer utilizing the American Joint Committee on Cancer TNM framework stratifies disease into groups according to how advanced the disease is and provides important standardized, individual prognostic information. Ideally, staging systems should exhibit hierarchical logic, with increasing stage reflecting worse prognosis. We sought to evaluate hierarchy within the 8th edition AJCC staging system for colon cancer. Methods: All patients with primary colon (including rectosigmoid) cancer diagnosed 2010-2017 were identified from the National Cancer Database (NCDB). Patients were only included if they carried a histologic diagnosis of invasive primary (non-recurrent) cancer, underwent radical surgery, and had complete pathologic staging data (T, N, M category data). Patients with missing follow-up data were excluded. Kaplan-Meier curves were used to assess overall survival by AJCC 8th edition pathologic summary stage group to determine whether staging was hierarchical for colon cancer. Multivariable modeling and log likelihood ratio test was then used to identify relative contributions of T and N variables to overall survival. Results: Among 270,584 colon cancer patients, pathologic staging was noted to be hierarchical by T category alone and by N category alone. However, summary staging was non-hierarchical in both instances. Multivariable modeling confirmed the AJCC non-hierarchy issue (colon HRs: 1, 1.54, 2.62, 2.60, 1.02, 2.15, and 4.38 for stage 1 as reference, 2a, 2b, 2c, 3a, 3b, 3c respectively). Multivariable modeling demonstrated that high T category (T4a, T4b) conferred the greatest risk of mortality based on hazard ratio (T4a HR 2.76, T4b HR 3.04), while high T category as well as high N category (N2a, N2b) contributed substantially to the survival model based on z-score. Further modeling assessing the predictive power of T and N category with respect to overall survival demonstrated that T category more strongly predicted overall survival than N category in colon cancer (likelihood chi-square test statistic pT:7009 vs pN:3586). Conclusions: Hierarchy in survival based on AJCC summary stage was not observed for patients with stage II-III colon cancer resulting in a paradoxical better observed survival with some stage III cancer subgroups compared to stage II cancer subgroups. High T category appears to more significantly impact survival than N category for patients with N0-N1 disease, while high N category appears more important for patients with T1-3 disease. Future revisions of the AJCC staging system should account for the differential impact of T and N category identified herein to more accurately confer prognostic information.

e18811 Background: The Covid-19 pandemic caused unprecedented challenges in the diagnosis and evaluation of cancer. At the same time, cancer treatment was potentially impacted by significant constraints on patients and hospitals; however, the extent and differential influence on different hospital types is unknown. Our objective was to assess the patterns of treatment utilization to better characterize the impact of the first year of the Covid-19 pandemic on the US healthcare system. Methods: The National Cancer Database (NCDB) was queried for patients treated for any type of malignancy diagnosed from 2018-2020. Autoregressive models were used to forecast expected findings for 2020 based on observations from the prior two years. Descriptive univariate statistics using chi-squared tests were performed to compare observed-to-expected findings for treatment utilization and losses in provided care in 2020. Results: Overall, 1,229,654 patients underwent treatment for any newly diagnosed cancer in the NCDB in 2020, representing a 16.8% reduction compared to what was expected. Stratified by treatment modality, 146,805 fewer patients than expected underwent surgery, 80,480 fewer received radiation and 68,014 fewer received chemotherapy. Reductions in treatment were examined by hospital type. Academic hospitals experienced the greatest reduction in provided care (-105,093 patients, -19%) compared to community programs (-72,432 patients, -14%) and integrated networks (-40,827 patients, -13%). However, there were fewer hospitals in the academic cohort which exaggerated the impact on each hospital. Thus, academic hospitals lost approximately 484 patients per hospital while community hospitals lost 99 patients and integrated networks lost 110 patients per hospital. The losses in provided care were most dramatic in terms of surgical care, as academic hospitals operated on 314 fewer patients per hospital (-20%) than expected while community hospitals on average operated on 69 fewer patients (-16%) and integrated networks 71 fewer patients (-14%). Conclusions: The impact of the first year of the Covid-19 pandemic on cancer treatment was heterogenous, resulting in nearly twice the number of missed surgical patients, compared to other treatment modalities. While all hospital types were affected by the pandemic, cancer care at academic hospitals experienced disproportionate reductions, with each hospital losing more than 4 times the number of treated patients than other hospital types. Continued efforts to recover from the strain of the pandemic on the US healthcare system will need to consider the complex influence of treatment declines across hospital types and different cancer service lines.