The results of published series vary but, depending very much on the cases selected, survival rates at 10 years have been reported to exceed 90% (338, 452, 453) with a comparable percentage of continent and many men potent post-operatively (338). Lower preoperative PSA, pathologic stage, and Gleason grade are associated with high rates of cure after RRP. For patients with preoperative PSA values <10 ng/ml, Catalona demonstrated that 7-year disease-free survival was 76-93% (417). If PSA was >10 ng/ml, disease-free survival decreased to 51%. Similar results have been shown by others (247, 454, 455).
Not unexpectedly, the probability of cure is considerably greater for patients with localized, non-palpable, and well to moderately differentiated tumors. The risk of treatment failure rises in relation to increases in clinical and pathological stage and correlates with increasingly aggressive histologic patterns. Survival rates has been reported to decrease with increasing pathologic stage, from 81% for pT2 lesions, 76% for pT3 tumors and 19% for N(+) disease. As indicated above, survival rates also vary with grade. In a multi-institutional study, Gerber et al (1996) (456) demonstrated that 10 year progression-free rates correlated with grade with 87% of low-grade tumors, 68% of moderate and 52% of high-grade cancers progressing, with other authors reporting comparable results (339, 340, 455-457).
A positive surgical margins is present if the tumour extends to the edge of the specimen. Patients with positive surgical margins, whether surgically induced or not, have approximately a 2 fold higher recurrence rate compared with patients with negative margins. The progression-free probability at 5 years for men with negative margins at RRP has been reported to be ~80% compared with 42-64% for men with positive margins (297, 458)
A number of molecular markers have been evaluated for their prognostic value but none has been adopted for routine use in practice. Amongst others, the potential of CD151, c-myc and calveolin-1 overexpression in predicting outcome has been discussed earlier in this chapter. Overexpression of p21 protein has been reported to be a predictor of response to salvage radiotherapy after RRP (459) and Fizazi et al (2002) (460) correlated p21 expression with progression to androgen independence. In addition, expression of p53 and the combined loss of PTEN and p27 have been shown to be predictive in the identification of the likelihood of disease recurrence (461, 462). Fradet et al (2004) (463) demonstrated that expression of NF-κB in prostate cancer cells at the sites of positive margins was highly indicative of biochemical recurrence (463).
Although clearly in the minority, some men with microscopic N(+) disease have undetectable PSA levels at 5-10 years. Progression in N(+) disease appears to be related to the volume of tumour in the lymph nodes. As well, not all men with seminal vesicle involvement have poor outcomes (464).
Despite the observation that seminal vesicle and lymph node involvement are associated with poor results post-prostatectomy, in the Catalona series, at seven years the disease-free survival rates were 26% for men with seminal vesicle involvement and 9% for men with N(+) disease (337). There also appears to be a difference if seminal vesicle involvement is a direct extension from the primary tumour, in which case it behaves more like an extracapsular extension rather than a separate metastatic focus which implies systemic involvement.
PSA serology is very useful in monitoring for disease recurrence post-RRP. If all cancer tissue is removed, PSA values should drop to the undetectable range (between 0-0.02 ng/ml). Understandably, patients become strongly focussed on their post-operative PSA values with raised levels following RRP indicating persisting prostatic tissue at a distant and/or local site (444, 452, 465). Most PSA recurrences occur in the first few years after surgery and precede clinical evidence of disease by as much as 5 years. It is rare for the PSA to become detectable after levels have been undetectable for 5 years. The rate of PSA doubling time is an important predictor of poor prognosis (452, 466, 467).
Using the definition of PSA failure of the American Society of Therapeutic Radiation and Oncology’s (ASTRO’s) 1996 consensus statement (ie 3 consecutively rising PSA values, each obtained at least 3 months apart), D’Amico et al analysed 888 patients followed for a median time of 38 months (8-100) after RRP. Based on the medical literature, they established 3 groups using known prognostic factors (viz. PSA level, biopsy Gleason score and the American Joint Commission on Cancer [AJCC] staging system). Over 95% of the PSA failures were evident for the intermediate and high-risk groups by 4 years. Questioning the ASTRO guidelines, Amling et al reviewed 2,782 RPP patients followed for a median of 6.3 years. After examining four points of “PSA cut-off”, these authors concluded that a PSA >0.4ng/ml was most appropriate since a significant number of patients with a lower PSA do not continue to increase or proceed to demonstrable metastases (468).
Freedland et al (2003) (469) undertook a retrospective review of 358 men who underwent RRP between 1991 and 2001 to determine the most relevant cutpoint for determining biochemical recurrence. For patients with a detectable postoperative PSA value from 0.11 to 0.2 ng/ml, the 1 and 3-year risk of PSA progression was 64% and 93%, respectively. For men with a PSA value from 0.21-0.3 ng/ml, the 1 and 3-year risk of PSA progression was 86% and 100%, respectively. The 5-year risk of PSA recurrence using a greater than 0.1 ng/ml cutpoint resulted in a 43% risk of recurrence compared with only 23% for a greater than 0.5 ng/mL cutpoint (469). Contemporarily, it has become common practice to regard a PSA of >0.2ng/ml as the cutpoint to indicate BCR.
Reporting on 213 men who had biochemical failure on the basis of a PSA of >0.2 ng/ml following RRP for a mean follow-up of 56 months (range 1 to 125), 99 of whom were treated with androgen ablation and/or radiation therapy at the time of detection of BF, Jhaveri et al (1999) (470) cited 10-year overall survival rates of 88% compared with 93% for those who did not have BF. They found no difference in survival rates in relation to age (>65 years), preoperative PSA >10 ng/ml, biopsy or specimen Gleason score >7, clinical Stage including the presence of extracapsular extension, positive surgical margins, and seminal vesicle invasion. (470).
Pound et al (1999) (452) reported in a retrospective study on the fate of 304 men with biochemical escape (PSA >0.2 ng/ml) following RRP, 103 of whom progressed to metastatic disease, not having had androgen suppressive therapy. The median actuarial time to metastases was 8 years from the time of PSA elevation and once patients developed metastases, the median actuarial time to death was 5 years. The time interval from surgery to the appearance of metastatic disease was predictive of the time until death (452).
However, it is clear that BCR in itself is not a predictor of imminent demise from prostate cancer and, as alluded to above, PSA doubling time (PSADT) is becoming accepted increasingly as a preferred prognostic indicator (471). Albertsen et al (2004) reported that men who had a PSADT of <1 year had a high risk of dying of prostate cancer within 10 years of diagnosis, and conversely (472). D’Amico et al (2003) (473) examined the results of men who experienced BCR subsequent to RRP and radiation therapy. From their analyses of 5,918 men following RRP and 2,751 patients after radiation therapy, they concluded that a PSADT of <3months met the criteria of surrogacy in predicting death from prostate cancer at a median survival time of 6 years (473).