Matsui Yoshiyuki, Ichioka Kentaro, Terada Naoki, Yoshimura Koji, Terai Akito, Dodo Yoshihiro, Arai Yoichi
Department of Urology, Kurashiki Central Hospital, Okayama, Japan.
J Urol. 2004 Feb;171(2 Pt 1):687-91. doi: 10.1097/01.ju.0000106864.91375.80.
Although salvage radiation therapy (RT) is a potentially curative treatment option for men with biochemical failure after radical prostatectomy (RP), to our knowledge there are no definitive pretreatment factors predicting patients likely to benefit from this treatment. We examined the impact of volume weighted mean nuclear volume (MNV) of biopsy specimens on disease outcomes and describe its usefulness as a new independent predictor.
We analyzed 33 patients who received salvage RT for biochemical failure after RP, including 11 who had received neoadjuvant hormone therapy before RP. Salvage RT was delivered to the prostatic bed at a total dose of 60 Gy with a 4-field contoured technique. Unbiased estimates of MNV were calculated from more than 100 cancer nuclei per patient captured from biopsy specimens based on a stereological method and compared with other clinical and pathological findings, including patient age, pretreatment prostate specific antigen (PSA), PSA density, biopsy Gleason score, neoadjuvant therapy, surgical Gleason score, pathological stage, tumor volume, surgical margin status, biochemical disease-free duration before RT, nadir PSA and PSA doubling time before RT, and pre-RT PSA with regard to predicting the disease outcome after salvage RT.
The median followup after salvage RT was 43.4 months. A total of 17 patients (52%) experienced biochemical failure a median of 6.7 months (range 0 to 48.1) after RT. On univariate analysis MNV and log(pre-RT PSA) were significant predictors of disease outcome in all patients and in the 22 nonneoadjuvant patient subset (p = 0.0124 and 0.0159, respectively). Log(nadir PSA) and PSA doubling time were also significant in the latter subset (p = 0.0287 and 0.0475, respectively). However, dual multivariate analysis revealed that MNV was the only independent predictor in the 2 groups (logistic regression analysis p = 0.00931 and 0.03511, and Cox proportional hazards analysis p = 0.00483 and 0.02277, respectively). There was a statistically significant biochemical disease-free survival advantage for small vs large MNV in each data set (p = 0.0072 and 0.0036, respectively).
Our results suggest that an estimate of MNV contributes significantly to the prediction of biochemical control after salvage RT. However, further investigation in a larger nonneoadjuvant population is needed to confirm its significance in combination with other clinical and pathological findings.
