Palliative transurethral resection of the prostate: functional outcome and impact on survival

Authors


Stephan Madersbacher, Department of Urology and Andrology, Donauspital, Langobardenstrasse 122, A-1220 Vienna, Austria. e-mail: stephan.madersbacher@wienkav.at

Abstract

OBJECTIVE

To assess the long-term functional and oncological outcome in a consecutive series of patients undergoing palliative transurethral resection of the prostate (pTURP).

PATIENTS AND METHODS

We retrospectively assessed all patients who had a pTURP between 1992 and 2004 at our institution. Patients with incidental prostate cancer (pT1a/pT1b) were excluded. In all, 89 patients (mean age 75.9 years, sd 0.9, at diagnosis) entered the study.

RESULTS

The median (range) prostate-specific antigen level at diagnosis was 25.7 (0.7–5000) ng/mL and the mean Gleason score was 7. The mean (sd, range) interval between the diagnosis of prostate cancer and pTURP was 1.5 (0.3, 0.5–10.9) years. The indications for pTURP were refractory urinary retention in 30%, severe bladder outlet obstruction with a postvoid residual urine volume of > 100 mL in 43%, and bladder stones, haematuria and hydronephrosis in 9% each. The mean (sd, range) follow-up after pTURP was 2.6 (0.2, 0.1–7.3) years. The peri-operative mortality (<30 days) was 2%, and 22 patients (25%) died during the follow-up. As estimated by Kaplan–Meier analysis, the 1-, 2- and 5-year survival rates were 83%, 70% and 61%, respectively. Patients with prostate cancer in the pTURP specimen had a shorter 3-year survival (52%) than those with a negative histology (89%, P = 0.03). At the last follow-up, 79% of men voided spontaneously and were continent. A repeat pTURP was necessary in 25% of patients, 11% required permanent catheterization and 10% were incontinent.

CONCLUSION

Despite greater peri-operative mortality and morbidity than conventional TURP, pTURP is a fairly safe and effective procedure. Although a potential negative impact of pTURP on survival cannot be excluded, the estimated 5-year survival of 61% in this series seems to justify this intervention.

Abbreviations
pTURP

palliative TURP

PVR

postvoid residual urine volume.

INTRODUCTION

Besides haematuria, BOO, with its complications such as urinary retention, a high postvoid residual urine volume (PVR), bladder stones or hydronephrosis, is the most frequent complication of locally advanced prostate cancer [1]. Under these circumstances treatment options are antiandrogen therapy, various minimally invasive procedures (stents, laser, etc.), catheterization and, most commonly, palliative TURP (pTURP) [2].

Although pTURP has been common under these circumstances for some decades, it is controversial. This uncertainty arises in particular from a potential negative oncological effect due to intravascular dissemination of prostate cancer cells [3,4]. This is thought to take place via the open venous sinus when using high-pressure resection techniques [4]. These oncological concerns contrast with the significant negative impact of acute urinary retention, bladder stones and the need for catheterization on quality of life.

There is controversy about pTURP partly because long-term follow-up studies that assess the oncological and functional outcome after pTURP are largely absent. In particular, reports that excluded patients with incidental prostate cancer, as in the present study, are rare. To address these issues we assessed the functional and oncological outcome after pTURP in 89 consecutive patients who had this procedure at our institution between 1992 and 2004. Patients with incidental prostate cancer (pT1a/b) were excluded. The mean interval between the initial therapy and pTURP was 1.5 years and the mean follow-up after pTURP was 2.6 (0.1–7.3) years.

PATIENTS AND METHODS

This retrospective study included 107 patients with prostate cancer who had a pTURP at our institution between 1992 and 2004; 16 patients with incidental prostate cancer (pT1a/pT1b) were excluded. In all, 89 charts provided sufficient follow-up data, and these patients were included.

The primary diagnosis of prostate cancer was made by TRUS-guided biopsies in all patients, and tumours were graded using the Gleason score. Serum levels of PSA were obtained from the time of diagnosis to the end of the follow-up. At diagnosis a bone scan was usually taken if the serum PSA level was > 20 ng/mL or if histology revealed tumours with a Gleason score of ≥ 8. During the follow-up, bone scans or CT were used only when clinically indicated (e.g. because of symptoms, hydronephrosis, etc.). As a screening tool for upper urinary tract dilatation, transabdominal ultrasonography is the method of choice in our country.

Before pTURP the prostate volume was determined by TRUS and serum PSA levels measured. pTURP was performed with a conventional 26 F resectoscope (Richard Wolf GmbH, Knittlingen, Germany). The surgical technique was a ‘channel’ rather than a ‘radical’ TURP. After pTURP patients had an irrigation catheter for 1–3 days.

For this retrospective study the hospital, surgical, pathological and outpatient records were reviewed. In addition, all patients were invited to a follow-up examination at our institution to update past and current treatments for prostate cancer and to assess the current voiding status. During this investigation, serum PSA levels and the PVR were measured. Patients were further categorized in those with a positive functional outcome (PVR <100 mL, no need to use pads after pTURP, no need for repeated or continuous catheterization) and those with a negative outcome (PVR > 100 mL, incontinent after pTURP or need for repeated or continuous catheterization).

The significance of changes in Gleason score between diagnosis and pTURP were assessed using the Wilcoxon test; the Spearman test was used to calculate correlations between the Gleason score and PSA levels. The Kaplan–Meier method was used to assess survival duration and the log-rank method to determine survival differences between groups. Cox regression analysis was used to assess the impact of clinical variables on the overall survival, and logistic regression analysis to evaluate the effect on the need for repeat TURP, and the functional outcome. All statistical tests were two-sided and deemed significant at P < 0.05.

RESULTS

In all, 89 patients (mean age 75.9 years, sd 0.9, range 57–92) entered this retrospective study. The mean (sd, median, range) PSA level at diagnosis was 273 (112, 26, 0.7–5000) ng/mL; 21 patients (24%) had a serum PSA level of <10, 20 (22%) of 10–20, 27 (30%) of 20–100 and 21 (24%) of > 100 ng/mL. The mean (sd, range) Gleason score was 7.3 (1.7, 3–9). The prostate volume as measured by TRUS was 54.6 (4.6, 19–141) mL. The initial therapy is shown in Table 1. The mean (sd, range) interval between the diagnosis and pTURP was 1.5 (0.3, 0.5–10.9) years. The indications for pTURP are also shown in Table 1; overall, there was refractory urinary retention in 30%, severe bladder outlet obstruction with a PVR of > 100 mL in 43%, and bladder stones, haematuria and hydronephrosis in 9% each. In general, these patients had no LUTS-specific management before pTURP (e.g. intermittent catheterization), except for α1-blockers, that were given temporarily to some patients with LUTS and a high PVR.

Table 1. 
The association of initial therapy with symptoms triggering pTURP, and histology and functional outcome
VariableType of initial therapyP
AntiandrogenEBRTBrachytherapy
  1. EBRT, external beam radiotherapy.

N (%)51 (57)28 (32)10 (11) 
Symptoms triggering pTURP, n (%)
PVR > 100 mL16 (31)16 (57) 6/100.37
Refractory  retention15 (29) 8 (29) 4/100.58
Bladder stones 4 (8) 4 (14) 00.56
Haematuria 8 (16) 0 0
Hydronephrosis 8 (16) 0 0
Outcome
Histology at pTURP
 malignant39 (77)19 (68) 00.06
 benign12 (23) 9 (32)10/100.87
Functional
 positive37 (73)23 (82) 9/100.15
 negative14 (27) 5 (18) 1/100.10

The mean (sd, range) prostate volume at pTURP was 62 (7, 20–141) mL and the mean (sd, median) PSA level 101 (25, 16) ng/mL. The hospital stay was 6.0 (0.5) days and the resection weight 16 (9, 1–43) g. In 81 patients (91%) no blood transfusions were necessary during or after pTURP, and the remaining eight required 2–4 units of blood. The catheter was removed after a median of 2 days; 96% of men were discharged with no catheter, with a mean (sd) PVR of 30 (6) mL. There was positive histology at pTURP in 56 patients (63%) (Table 2). The Gleason score at pTURP was higher, at 8.1 (0.3), than at diagnosis (P = 0.035). The association between Gleason score and PSA at diagnosis (P = 0.008) was not evident at pTURP (P > 0.05). Patients with a positive histology at pTURP were younger (P = 0.03), had a higher mean PSA level (P = 0.014), a higher Gleason score at diagnosis (P = 0.01) and more positive biopsies (P = 0.02). The resection weight in both groups was comparable (Table 2).

Table 2. 
Comparison of patients with and with no positive histology at pTURP
VariableHistologyP
positivenegative
N (%) patients 56 (63) 33 (37) 
Mean (sd):
At diagnosis;
 age, years 73.7 (1.2) 78.0 (1.2)0.017
 PSA level, ng/mL296 (160)126 (103)0.004
 Number of +ve biopsies  5.9 (0.7)  3.1 (0.7)0.005
 Gleason score  8.0 (0.3)  6.3 (0.5)0.011
PSA at pTURP, ng/mL126 (27) 14 (5)0.014
Resection weight, g 14.1 (1.4) 15.4 (2.0)0.780
3-year survival after pTURP, % 52 890.030

The peri-operative mortality (<30 days) was 2.2%, with nine patients dead (10%) within a year, 15 (17%) within 2 years and 19 (21%) within 3 years. In all, 22 patients (25%) died after a mean (sd, range) of 1.5 (0.3, 0.1–5.6) years.

The mean (sd, median, range) follow-up from the diagnosis of prostate cancer was 4.4 (0.4, 4.0, 0.1–14.1) years and from pTURP was 2.6 (0.2, 2.3, 0.1–7.3) years. As estimated by Kaplan–Meier analysis the 1-, 2- and 5-year overall survival rates were 83%, 70% and 61%, respectively (Fig. 1a). Overall survival correlated with PSA level at diagnosis (P = 0.037), the number of positive biopsies at diagnosis (P = 0.03), the PSA level at pTURP (P = 0.03), the Gleason score at pTURP (P = 0.04) and a positive histology at pTURP (P = 0.03).

Figure 1.

Overall survival of: a, all 89 patients; and b, depending on the histology at pTURP.

As shown in Table 2 and Fig. 1b, patients with prostate cancer in the pTURP specimen had a shorter overall survival than those with no positive histology (P = 0.03). Furthermore, survival correlated with the functional outcome; those with a good functional outcome had better survival, although this difference was not quite significant (P = 0.08).

A repeat pTURP was necessary in 22 (25%) patients after a mean of 11 months, four of whom required more than one re-intervention. Hence, three-quarters of patients were treated with just one intervention. At the last follow-up 79% voided spontaneously with a PVR of <100 mL (mean 22, sd 8) and reported no urinary incontinence; these patients were considered as having a successful functional outcome. Failure was defined as requiring permanent transurethral (four men) or suprapubic catheterization (six) or reporting incontinence after pTURP (nine; one pad/day in three, and more than one pad/day in six; Table 1). None of the following variables correlated with the functional outcome after pTURP: age, number of positive biopsies, prostate volume at diagnosis/pTURP, PSA at diagnosis/pTURP, time to pTURP, histology at pTURP, PVR after pTURP, resection weight, duration of surgery, or distant metastases.

DISCUSSION

Any urologist counselling a patient with severe BOO due to locally advanced prostate cancer and deemed unsuitable for, or after, curative treatment, is in a dilemma. On the one hand there is the theoretical possibility of accelerating tumour growth by surgery, and on the other patients with severe BOO usually requiring catheterization have a significantly impaired quality of life. This dilemma is in part due to the paucity of long-term functional and oncological data after pTURP.

When assessing the role of pTURP three aspects need to be considered: (i) the safety of the procedure; (ii) the functional outcome; and (iii), probably the most important, oncological aspects. We tried to address these three aspects in a contemporary, consecutive series of 89 patients. Notably, patients with incidental prostate cancer were excluded, and all patients had surgery after primary treatment, with a mean interval of 1.5 years. This group is considered to be that with the worst prognosis [2].

We do not advocate pTURP at the initial diagnosis of prostate cancer even in the presence of BOO, but rather await the effect of primary treatment. Furthermore, only patients with a strong indication (recurrent urinary retention, high PVR, bladder stones) were candidates for pTURP in the present series. Finally, it is our policy to use a ‘channel’ and not a radical TURP, to minimize the risk of tumour spillage. This is also underlined by the relatively small resection volume.

The present series indicates that pTURP is a fairly safe procedure, although the peri-operative mortality was significantly higher (2%) than for contemporary series of BPH (<0.25%) [5]. However, the need for a blood transfusion (9%), mean hospital stay (6 days) and duration of catheterization (median 2 days) were comparable with current series of men with BPH [5]. Similar data were recently reported by Crain et al.[6], who compared the outcome after TURP for benign prostatic obstruction (520 men) to that after pTURP (19 men). Compared with patients with BPH, those undergoing pTURP were more likely to have a failure of the initial voiding trial, to require a re-operation and to need chronic drainage, as well as re-catheterization for bleeding [6]. Nevertheless, these authors concluded that pTURP can be safe and is effective in relieving obstruction [6].

The long-term functional outcome after pTURP in the present series was satisfying; 75% of patients were successfully treated with just one intervention. At the last follow-up 79% voided spontaneously with a PVR of <100 mL and reported no urinary incontinence. Similar data were reported by others [6]. In the present series all patients had what is currently considered a strong indication for BPH surgery, i.e. refractory urinary retention, a high PVR (>100 mL), bladder stones or hydronephrosis due to a high PVR.

Probably the most important issue is the potential negative effect of pTURP on the oncological outcome and survival. Clearly, cutting through a malignant tumour violates any oncological principle. Moreno et al.[3] detected prostate cancer cells in the peripheral blood circulation in all cases undergoing pTURP, although the clinical relevance of this finding remains unclear. There is a long-standing debate on the potential negative effect of pTURP on the oncological outcome [4,7–18]. A series of studies analysed the outcome of patients whose cancer was diagnosed either by TURP (incidental) or by biopsy, and who subsequently had external beam radiation [4,7–17]. These studies showed conflicting results, with six positive [8–10,12,15,16] and seven negative studies [4,7,11,13,14,17,18]. Although many studies have tried to correct for confounding factors, the differences in some of these might have been due to higher stages and larger volumes of disease in patients undergoing pTURP. None of these studies was randomized. In a surgical series of 490 stage C patients, Nativ et al.[16] reported no unfavourable outcome in those who were diagnosed by pTURP, compared to those diagnosed by needle biopsy. Similar data from another surgical series were reported by Paulson and Cox [12] analysing 139 patients.

In a retrospective, comparative study Hübner et al.[2] compared the survival of patients who had orchidectomy only (group A, 35 men) to those who had orchidectomy and pTURP at diagnosis (group B, 32) and to those who had orchidectomy and pTURP later in the disease course (group C, 17). Within 3 years, 8% died in group A, 36% in group B and 60% in group C [2]. The authors concluded that pTURP had a negative impact on survival. The problem with this study is in its design (comparative and retrospective), the few patients and that patients requiring pTURP in the disease course probably have at least locally more advanced stages [2]. The present patients were comparable to group C, but we could not duplicate these poor survival data; in the present series of 89 patients only 21% died within 3 years.

Patients with a positive histology at pTURP have a significantly shorter survival than those with a negative histology, despite comparable resection weights. Baseline data indicated that these patients had more advanced tumours at diagnosis and that this worse prognosis cannot be attributed to the surgery itself. However, some authors suggested that pTURP disseminates prostate cancer mainly in patients with T3 and T4 stage tumours [4] and documented tumour cell spread into the circulation [3], yet it remains unclear whether this tumour cell spillage leads to more rapid progression of the disease.

The present study has several limitations, which are largely because it was retrospective and had a long study period. The functional assessment before pTURP was not uniform, and in particular pressure-flow studies were used rarely, mainly because most patients had strong indications for surgery (refractory urinary retention, bladder stones, haematuria, significant PVR). Furthermore, the IPSS was not regularly obtained for the same reason. In addition, we were unable to provide reliable data on disease-specific survival, because substantially many patients died at home. In these cases the cause of death could not be identified. However, we emphasize that the overall survival is the most stringent endpoint.

In conclusion, pTURP is a fairly safe procedure with a satisfying long-term functional outcome. Although a potential negative impact of pTURP on survival cannot be excluded, the estimated 5-year survival in the present series of 61% seems to justify this intervention. In the future minimally invasive treatment options enabling bloodless tissue ablation, such as the KTP- or holmium laser, will hopefully provide effective treatment without jeopardizing the oncological outcome.

CONFLICT OF INTEREST

None declared.

Ancillary