EP694 - Concordance rate of preoperative radiologic staging by CT and histopathology in the assessment of colon cancer at Philippine General Hospital

Purpose/Background: Colorectal cancer is one of the most commonly diagnosed malignancies worldwide. In the Philippines, it ranks third as the most prevalent cancer, with over 20,000 reported cases in 2022. Computed tomography (CT) is the primary imaging modality for colon cancer staging, providing information on tumor resectability, local invasion, and distant metastases. However, there is limited local data on how accurately radiologic staging matches histopathologic staging. Evaluating this concordance in the Philippine context may yield valuable insights into current diagnostic practices and guide efforts in enhancing healthcare delivery.

Methods/Interventions: This retrospective cohort study included colon cancer patients who underwent oncologic resection at the Philippine General Hospital (PGH) from 2020-2024, with preoperative imaging and histopathologic reading of the resected specimen performed in the same institution. Patients with non-colonic malignancies (e.g., appendiceal tumors, rectal cancers) who underwent colectomy and whose preoperative CT was not done at PGH were excluded. The radiologic and histopathologic reports were reviewed, comparing tissue invasion and lymph node assessment using the American Joint Committee on Cancer (AJCC) 2017 guidelines. Data were analyzed using IBM SPSS software through the Cohen's kappa coefficient to determine the level of concordance.

Results/Outcomes: A total of 26 patients with complete records were evaluated. (The limited sample size may be attributed to PGH being a tertiary referral center, where most patients are transferred after initial management at other healthcare institutions. To minimize cost, the imaging patients have on hand-often without a radiologist's report - are not repeated unless necessary.) The overall concordance rate was 31.8% with a Cohen's kappa coefficient (κ) of 0.17. This indicates a low level of agreement between radiologic and pathologic staging methods. Although almost a third of cases had congruent staging, the kappa value suggests that much of this concordance may be attributable to chance rather than true diagnostic agreement.

Conclusion/Discussion: This study demonstrated a low concordance rate between preoperative radiologic and postoperative tissue analysis in colon cancer patients treated at PGH. These variations in staging highlight limitations in the preoperative interpretation of disease extent in a considerable number of cases. Potential causes of discrepancy include: limitations in imaging resolution, difficulty in appreciating tumor microinvasion radiographically, and disease progression between imaging and definitive surgery. While radiologic staging remains valuable for initial assessment and surgical planning, this should always be correlated with intraoperative and pathologic findings for definitive clinical decision-making. Improving diagnostic accuracy may be achieved through standardizing imaging protocols, strengthening multidisciplinary collaboration, and providing continued training in radiologic-pathologic correlation.

This study recognizes its limited sample size due to incomplete preoperative records (i.e., radiographic reports)-reflective of realities of the current status of health care in the country -and future efforts will focus on larger cohorts and expanding data collection to include preoperative imaging done in other institutions. Further studies are needed to validate these results and advance diagnostic accuracy and quality of care for colon cancer patients in the Philippines.