Adjuvant radiation therapy for recurrent PSA after radical prostatectomy in T1±T2 prostate cancer

Transcription

1 Adjuvant radiation therapy for recurrent after radical prostatectomy in T1±T2 prostate cancer Prostate Cancer and Prostatic Diseases (1998) 1, 321±325 ß 1998 Stockton Press All rights reserved 1365±7852/98 $ V Ravery 1, F Lamotte 1, CH Hennequin 3, M Toublanc 2, Li Boccon-Gibod 4, JF Hermieu 1, V Delmas 1 and L Boccon-Gibod Departments of 1 Urology and 2 Pathology, Bichat Hospital, 3 Department of Therapeutic Radiology, Saint-Louis Hospital and 4 Department of Pathology, Trousseau Hospital, Paris, France To evaluate retrospectively the ef cacy of adjuvant radiation therapy (ART) in patients with T1 ± T2 prostate cancer (CaP) in whom extracapsular cancer (pt3) was detected after radical prostatectomy (RP), together with biochemical failure characterized by a recurrent level of serum prostate-speci c antigen () > 0.1 ng=ml. Twenty-two patients with T1 ± T2 CaP treated by RP who subsequently were found to have pt3 CaP with (13) or without (9) positive surgical margins and=or seminal vesicle invasion, exhibited biochemical failure characterized by a recurrent level of serum, 2 ± 40 (mean: 25) months after RP and were treated with ART (65 Gy). Bone and CT scans were negative in every patient, 15 of whom were submitted to TRUS biopsy (Bx) of the anastomosis (resection site), which was positive in 8. Patients were followed up for between 6 and 60 (mean: 32.5) months. Transient side effects (urgency, proctitis, diarrhea) were experienced by 9 patients after ART. A decrease in serum was observed in 19 patients; however, only 14 of these achieved an undetectable level (< 0.1 ng=ml) on one or more occasions after completion of ART (in 12 cases this was after 3 months). Of the 14 patients, 8 achieved a persistently unmeasurable level at a mean follow-up of 20.4 (range: 9 ± 48) months. There was no difference between patients in whom an undetectable level of serum was attained and those in whom it was not, with regard to specimen pathology, doubling time, timing of ART, and the result of Bx. Patients who achieved an undetectable had a lower mean at the time of ART (1.1 vs 2.9 ng=ml, P < 0.05) and a lower preoperative mean. Although ART for biochemical failure after RP may lead to undetectable levels in a signi cant proportion of patients for a signi cant period of time, a longer follow-up shows that such unmeasurable levels persist in only 36.4% of such patients. Keywords: prostate cancer; radiation therapy; treatment Introduction Detectable serum levels of prostate-speci c antigen () 1 month after radical prostatectomy (RP) indicate residual=recurrent disease. External beam radiation therapy Correspondence: Dr V Ravery, Department of Urology, Bichat Hospital, 46 rue Henri Huchard, Paris, France. is therefore most often applied if local recurrence is suspected. However, the modalities of adjuvant radiation therapy (ART) after RP are still debated. Should it be initiated immediately, where pt3 disease is noted in the pathology report, or should it be delayed until recurs? Furthermore, it is dif cult to pinpoint which patients, among those who show such a recurrence, actually need adjuvant treatment and therefore are more likely to bene t from ART.

2 322 At our institution, ART is offered to pt3 patients who have elevated levels of serum after RP, but no other evidence of metastatic disease. We report here the results of ART in 22 patients with pt3 cancer and detectable postoperative. Materials and methods Between September 1988 and September 1996, 215 RP operations were performed at our hospital, for clinically localized prostate cancer. Of these patients, 22 had ART for persistent=recurrent with no evidence of distant metastasis and were studied retrospectively. All had pt3 disease noted in the initial pathology report. The mean preoperative serum (modi ed polyclonal Yang assay) was 17.9 (4.5 ± 35) ng=ml. Postoperative levels were measured using an ultrasensitive (Yang modi ed) assay with a detection threshold of 0.1 ng=ml. Postoperative biochemical failure was de ned as a persistent=recurrent postoperative level in excess of 0.1 ng=ml. Low preoperative level was arbitrarily de ned as less than 10 ng=ml. RP specimens were processed according to the Stanford method and positive surgical margins were de ned as the presence of carcinoma at the linked limits of the specimen, suggesting that removal had been incomplete. A total of 15 patients underwent transrectal ultrasound-guided biopsy of the anastomosis (resection site). Patients underwent bone=ct scanning and chest radiography before ART; normal ndings with these procedures were a prerequisite for ART. A total dose of 65 Gy was delivered to the prostatic bed (the area from which the prostate had been removed), in fractions each of 2.25 Gy, four fractions per week (9 Gy=week). All the patients were treated using a four- elds box technique and using a 18 MV photons linear accelerator. No conformal radiation therapy was done. The pelvic lymph nodes were not included in the initial elds. The mean interval between RP and ART was 25.6 (6 ± 72) months. After ART, was determined at least once every six months with a mean follow-up of 32.5 (6 ± 60) months. After ART, two groups of patients were de ned, according to the biochemical results: group I comprised those patients in whom was undetectable on one or more occasions after completion of ART; Patients in group II showed only a partial decrease, if any, in serum level after completion of ART. It was postulated that post-art biological failure, was due to either metastasis undetectable prior to radiotherapy or to local tumoral recurrence. However, no further conventional radiology exam was performed since the volume of distant=local tumoral recurrence is at this time so small that radiology would not make a worthwhile contribution to the diagnosis and to the therapeutic strategy either. Preoperative, pathology features (Gleason score, positive surgical margins, invasion of seminal vesicles), doubling time, timing of ART, serum level prior to ART, and the results of biopsy of the resection site (anastomosis), were recorded and compared with the biochemical results after ART. Urinary and rectal morbidity were graded in accordance with Radiation Therapy Oncology Group (RTOG) guidelines. The chi-squared test was used to evaluate statistical differences between groups (with correction for low numbers) and the Wilcoxon nonparametric test was used for comparison of means. Results The tolerance of the treatment was usually good. We noted, according to RTOG morbidity grading system, grade 1 or 2 side effects (urgency, proctitis, diarrhea) in nine patients (40.9%). One patient only experienced a grade 3 complication. These symptoms were always transient and disappeared between 1 and 6 months after ART. In this series, nine patients had positive seminal vesicles and 13 had positive surgical margins (8 out of 14 in group I: patients with post-art undetectable ; 5 out of 8 in group II: patients with uncomplete post- ART response). The mean preoperative was 10.7 (6.2 ± 16) ng=ml in group I and 19.3 (4.5 ± 35) ng=ml in group II (P < 0.05). The mean prior to ART was 1.1 (0.14 ± 2.5) ng=ml in group I and 2.9 (0.3 ± 7.2) ng=ml in group II. In 19 patients (19 out of 22, 86.4%) a decrease in serum levels was noted; in 14 of these (14 out of 22, 63.6%; group I) a complete biochemical response (de ned as an undetectable level at any time (that is, on one or more occasions) after completion of ART) was achieved, in 12 patients this being noted within three months of the completion of ART. In eight patients (8 out of 22, 36.4%) this complete biochemical response was durable (persistent) over a mean follow-up period of 20.4 (9 ± 48) months and in one patient a durable but incomplete response (0.15 ng=ml at 12 months) was noted (Table 1). Nevertheless, as shown in Figure 1, the percentage of patients with an undetectable level of after ART showed a continual decrease with increasing length of follow-up. Of the eight patients who, prior to ART, had serum levels lower than 1 ng=ml, a complete biochemical response was noted in seven, and this was durable in 6. In all seven, was undetectable one month after ART. The only two factors found to have a statistically signi cant correlation with complete biochemical response to ART were low preoperative and low prior to ART (Table 2). No difference in biochemical response after ART was observed between patients with recurrent elevated postoperative and those with persistently elevated postoperative.

3 Table 1 Pathological features and biological results after adjuvant radiation therapy of patients in groups I and II 323 Group Patient no. Preop. (ng/ml) Result of NBA prior to RT Gleason score Surgical margins RP ± ART interval (months) prior to ART (ng/ml) < 0.1 at 3 months Durable complete biochemical response Follow-up period (months) I 3 12 ND no no Ð 7 Ð yes no ND yes yes Ð 6 Ð yes no yes no Ð yes no ND 8 Ð yes yes ND yes yes Ð yes yes Ð 8 Ð yes yes ND yes yes Ð yes no ND 9 Ð yes yes no yes 40 II Ð Ð 7 Ð ND Ð yes a Ð Group I ˆ patients in whom was undetectable on at least one occasion after ART; Group II ˆ patients in whom only a partial (or no) decrease in was noted after ART. a Detectable but low durable result (0.15 ng/ml at 12 months); NBA ˆ needle biopsy of anastomosis; RP ˆ radical prostatectomy; ART ˆadjuvant radiation therapy; ND ˆ not done; ˆ positive; Ð ˆ negative. Discussion It has been postulated that biochemical failure (recurrence of serum > 0.1 ng=ml) after RP may be attributable to local residual=recurrent disease, where resection site biopsy is positive in conjunction with low doubling time and no other evidence of distant metastasis. CT scan and bone scan must be performed prior to ART for tumor widespread assessment; however, most often they are negative since tumor volume is very low at this time. Previously reported rates of response to ART have varied widely. 1 ± 3 These differences may re ect differences in response de nition, patient selection, time of follow-up, and detection threshold, making it dif cult to compare results between studies. However, in considering the most recent series, 4 ± 6 of studies that focus on the biochemical response obtained in pt3 patients, the proportion of patients who are free of disease as de ned according to biochemical criteria is constantly less than that of patients clinically free of disease. In our series, the rate of durable complete biochemical response is low (36.4%) at a mean follow-up of 20.4 months, whereas previously published rates of pt2=pt3 patients with undetectable at last follow-up range from 50 ± 59%. 7 ± 10 These con icting results are explained by the use, in our study, of an ultrasensitive assay (detection threshold 0.1 ng=ml) which enables biochemical failure to be detected at an earlier stage than in other series, despite a shorter follow-up period. We have noted a more durable response to ART if prior to ART is 1 ng=ml or lower. Schild et al 11 reported similar results. These ndings probably re ect a lesser degree of residual disease in these patients, before ART. Link et al 7 underlined the poor results of radiation therapy in post-radical prostatectomy patients whose does not decrease to undetectable levels after surgery. Their dates suggested that the lower the prior to ART, the better the results. This opinion is widely shared in literature. 8,9 Preoperative also appears to be predictive of good results after ART, possibly because, in the context of patients who show biochemical progression after RP, high levels of re ect micrometastatic disease at the time of surgery. Indeed, if ART fails to cure a patient it is either because local residual=recurrent tumor is too bulky or because metastasis has already occurred. With regard to the postoperative velocity, most authors 6,9,10 have underlined the fact that an increasing level of 0.75 ng=ml per year is related to metastatic disease in over 50% of cases, and that these patients are probably not the best target for ART. In our series we have not noted any correlation between postoperative velocity and response to ART, probably because the calculation of velocity depends on sampling and, in this retrospective study, has not necessarily been sampled at the same time for every patient. We previously reported signi cantly worse results in patients with seminal vesicle involvement. However, with

4 324 Table 2 Potential prognostic factors and their relationship to biochemical response Biochemical response Complete Incomplete Factor n ˆ 14 n ˆ 8 P Gleason score 7 7 NS Positive seminal vesicle (%) 5/14 (36) 4/8 (50) NS Positive surgical margins (%) 8/14 (57) 5/8 (62.5) NS Preoperative (ng/ml) < 0.05 prior to ART (ng/ml) < 0.05 doubling time (months) NS RP ± ART interval (months) NS Positive NBA (%) 3/8 (37.5) 5/7 (71.4) NS NS ˆ nonsigni cant; NBA ˆ needle biopsy of the anastomosis; RP ˆ radical prostatectomy; ART ˆadjuvant radiation therapy. Figure 1 Kaplan Meier curve showing percentage of patients with a complete biochemical response, according to length of follow-up. [JCM1] further follow-up we have been unable to reproduce this nding. Schild et al 11 reported similar data. Most previous reports 1,4,12 have noted that an undoubted advantage is conferred by surgery plus ART over surgery alone, with regard to speci c survival rate, metastasis-free disease, and local recurrence. However, Freeman et al 13 found no difference in local recurrence when comparing both treatments. The extreme heterogeneity of these studies does not make any comparison easy. It would be interesting to compare the results of surgery alone vs surgery plus ART in randomized pt3 patients with positive=negative surgical margins. Therefore, although our results do not support the use of ART in patients with detectable after RP, they indicate that, where ART is, nevertheless, resorted to, it is better to administer it when serum is still less than 1 ng=ml. The occurrence of rising postoperative is related to tumor recurrence. Therefore, a curative dose (60 ± 70 Gy) must be recommended to treat these patients rather than a prophylactic dose (45 Gy). This attitude is followed by most of the authors using postoperative radiation therapy in case of pt3. 13,14 In the opinion of Forman, 15 ART may have better results if the total dose delivered to the prostate bed is up to 74 Gy; however, the use of conformal radiation techniques are mandatory if a dose higher than 70 Gy is delivered. It has been suggested that positive biopsy of the resection site (anastomosis) re ects recurrent=residual disease, rather than metastatic disease that ART cannot be expected to eradicate. In our series, only a few patients have undergone such a biopsy; however, in this setting, the number of positive biopsies was found to be even higher in the group of patients showing an incomplete biochemical response. It is not known, therefore, whether this nding is due to random distribution or whether it is because a complete biochemical response is achieved chie y in cases of low-volume tumors, in which there is some dif culty in obtaining a biopsy specimen of cancerous tissue from the anastomatic area prior to ART. Conclusions The results of our series indicate that a short interval between RP and ART is not predictive of a good biochemical result. These ndings are similar to those of Partin, 10 but do not corroborate those of others. 11,16 A persistently elevated postoperative has been described 5,7,12 as predictive of an unfavourable response to ART, but, again, we did not nd such a correlation. These con icting reports are probably explicable in terms of the wide variations in response de nitions and patient settings in these different series. In this group of high risk pt3 patients, ART for biochemical failure after RP may lead to undetectable levels in a signi cant (63.6%) proportion of patients, however a longer follow-up shows that such unmeasurable levels persist in roughly one third of such patients. Since 20 ± 30% of pt3 patients do not experience any postoperative rising at time of last follow-up, and would be irradiated for nothing, we strongly recommend that only those patients who are most likely to bene t from ART are selected. In our experience, such patients are those with low preoperative, and with serum prior to ART of less than 1 ng=ml. References 1 Anscher MS, Prosnitz LR. Postoperative radiotherapy for patients with carcinoma of the prostate undergoing radical prostatectomy with positive surgical margins, seminal vesicle involvement and=or penetration through the capsule. J Urol 1987; 138: 1407 ± Forman JD, Wharam WD, Lee DJ. De nitive radiotherapy following prostatectomy: Results and complication. Int J Radiat Oncol Biol Phys 1986; 12: 184 ± Freeman JA et al. Radical retropubic prostatectomy and postoperative adjuvant radiation for pathological stage C (PcNO) prostate cancer from 1976 to 1989: Intermediate ndings. J Urol 1993; 149: 1029 ± Gibbons RP et al. Adjuvant radiotherapy following radical prostatectomy: Results and complications. J Urol 1986; 135: 65 ± Hudson MA, Catalona WJ. Effect of adjuvant radiation therapy on prostate speci c antigen following radical prostatectomy. J Urol 1990; 143: 1174 ± 1177.

5 6 Lange PH et al. The effect of radiation therapy after radical prostatectomy in patients with elevated prostate speci c antigen levels. J Urol 1990; 144: 927 ± Link P, Freiha FS, Stamey TA. Adjuvant radiation therapy in patients with detectable prostate speci c antigen following radical prostatectomy. J Urol 1991; 145: 532 ± McCarthy J, Catalona W, Hudson M. Treatment outcome of early vs. delayed XRT after radical prostatectomy: correlation with post-operative serum. J Urol 1994; 149: 430A (Abstract). 9 Medini E, Medini I, Reddy PK, Levitt SH. Delayed=salvage radiation therapy in patients with elevated prostate speci c antigen levels after radical prostatectomy. A long term followup. Cancer 1996; 78: 1254 ± Partin AW, Pound CR, Clemens JQ. Serum after anatomic radical prostatectomy: The Johns Hopkins experience after 10 years. Urol Clin North Am 1993; 20: 713 ± Schild SE. The use of radiotherapy for patients with isolated elevation of serum prostate speci c antigen following radical prostatectomy. J Urol 1996; 156: 1725 ± Syndikos I, Pickles T, Kostashok E, Sullivan LD. Postoperative radiotherapy for stage pt3 carcinoma of the prostate. J Urol 1996; 155: 1983 ± Anscher M, Robertson C, Prosnitz L. Adjuvant radiotherapy for pathologic stage T3=4 adenocarcinoma of the prostate: ten-year update. Int J Radiat Oncol Biol Phys 1995; 33: 37 ± Zietman A, Coen J, Shipley W, Althausen A. Adjuvant irradiation after radical prostatectomy for adenocarcinoma of prostate: analysis of freedom failure. Urology 1993; 42: 292 ± Forman JD. Therapeutic irradiation for patients with an elevated post-prostatectomy level. J Urol 1997; 157: Steckel J, Danella JF, dekernion JB. The detectable prostate speci c antigen after radical prostatectomy: Evaluation, treatment, and results. J Urol 1994; 151: 376A (Abstract).