Purpose: Touch preparation cytology has been used in oncology as a technique to assist in predicting local tumor recurrence. We prospectively investigated the relationship between this cytological evaluation and the standard histological method of assessing specimens, measuring the distance from the tumor to the various anatomical boundaries and disease recurrence in radical retropubic prostatectomy patients.
Materials and methods: In a prospective study of 91 consecutive clinical stages T1c and T2 cancer cases radical retropubic prostatectomy touch preparation cytology was performed intraoperatively in an anatomical fashion (apex, posterior, lateral right and left, and base). A single blinded cytopathologist reviewed all prostate touch preparation specimens and categorized them as malignant, benign or atypical cells. Benign or atypical cells were classified as negative cytology. Detailed histological margin analysis of the surgical specimens was also done in which distances between the tumor front, and prostate capsule (inner and outer edge) and surgical margins (apex, posterior, right and left lateral, and base) were measured. All specimens were re-staged by the same pathologist. Median followup was 38 months. Disease recurrence was determined biochemically (prostate specific antigen), and with bone scans, prostatic fossa biopsies and digital rectal examinations.
Results: Of the 91 specimens 25 were excluded from study because distance measurements could not be made for technical reasons. Multivariate analysis was performed on the remaining 66 patients based on the variables of stage, age, cytology status, distance from tumor to the inner prostatic capsule, distance from tumor to the surgical margin and postoperative Gleason sum. The only variable with independent prognostic value was postoperative Gleason sum (p = 0.04). Cytology status was not statistically significant (p = 0.07) nor were distance data to the inner capsule (p >0.05) and surgical margin (p >0.05).
Conclusions: Although touch preparation cytology does not enhance prognostic information already provided by Gleason sum, it does correlate highly with postoperative Gleason sum. Other gross macroscopic variables, that is pathological stage, margin status and distance measurements, although lacking in independent predictive value, correlated with postoperative Gleason sum. The constancy of Gleason sum leads us to believe that the key to predicting prostatic cancer behavior lies not on the macroscopic but on the molecular or cellular level. Of the various factors analyzed in this study postoperative Gleason sum remains the most powerful predictor of recurrence risk.