The importance of precise pT diagnosis for prognostic prediction of uterine cervical cancer--a single institutional report at a Japanese comprehensive cancer hospital.

We previously reported that the majority of Japanese pathologists misunderstand the International Union against Cancer-pT2 criteria for uterine cervical cancer (UCC). We compared the prognosis of originally diagnosed pT2 (ori-pT) UCC cases at our hospital with reclassified pT2 (re-pT) cases to assess the importance of making a correct pT diagnosis. There were 43 International Federation of Gynecology and Obstetrics (FIGO) II (i.e., cT2) and/or ori-pT2 UCC cases who received surgery without neoadjuvant chemotherapy at Shikoku Cancer Center from 1991 to 2003. The cases (seven ori-pT1 and 36 ori-pT2; 43 cN0 with six pN1) were reclassified as 22 re-pT1 and 21 re-pT2. Fifteen of the 23 ori-pT2a cases (65%) were re-pT1 because their vaginal extension had only been intraepithelial. The difference in the 5-year survival rate (5Y-SR) was not significant between the ori-pT1 and ori-pT2 cases using Fisher's exact test (F test): P = 0.236 > 0.05, whereas 5Y-SR of re-pT1 cases was significantly higher than re-pT2, including pN1 cases and excluding them (F test: P = 0.00164 < 0.01 and P = 0.0108 < 0.05, respectively). The 5Y-SR of ori-pT2-re-pT1 (overdiagnosed pT2) was significantly higher that of ori-pT2-re-pT2 (true pT2) including pN1 cases and excluding them (F test: P = 0.00694 < 0.01 and P = 0.0305 < 0.05, respectively). These results indicated that pT2 of UCC could be frequently misdiagnosed at an institutional level, and that misdiagnosed pT2 might impair the evidence-based medicine of UCC. Multi-institutional assessment of the accuracy of pTNM is recommended, because it is not likely that this is an endemic problem to our hospital.