Pain after transrectal ultrasonography-guided prostate biopsy: the advantages of periprostatic local anaesthesia

Authors


U. Patel, Department of Radiology, St George's Hospital, London SW17 0QT, UK.
e-mail: Uday.Patel@ccmail.stgh-tr.sthames.nhs.uk

Abstract

Objective To prospectively evaluate the efficacy and safety of periprostatic local anaesthesia (LA) during prostatic biopsy guided by transrectal ultrasonography (TRUS), as 20–65% of men report moderate to severe pain, and there is anecdotal and published evidence that periprostatic anaesthesia improves patients' tolerance.

Patients and methods In all, 157 patients were prospectively recruited and sequentially randomized to receive either LA or no anaesthesia. Sextant biopsies were taken in all men but some had more than six biopsies. All were asked to complete questionnaires immediately after TRUS-guided biopsy and for the subsequent week, giving pain scores and recording any morbidity, including symptoms of infection; analgesic use was also surveyed.

Results Patients given LA had significantly lower pain scores at the time of biopsy than those given no anaesthesia, with median (sd) pain scores of 1.53 (0.7) and 1.95 (0.65) (P < 0.001), respectively. In addition, there was a trend towards less analgesic use by those given LA, although this was not statistically significant. There was no difference in the amount of haematuria, haematochezia or haematospermia, or infection rate, between the groups. The additional cost and time of the procedure was minimal (£3.00 and 3 min/per patient, respectively).

Conclusion Periprostatic LA infiltration is a quick and simple procedure which significantly improves immediate pain with no added morbidity; we strongly advocate its use to improve patient tolerance of TRUS-guided prostate biopsy.

Introduction

Most prostate cancers are now diagnosed by TRUS- guided prostate biopsy. Originally a transperineal needle puncture was used after a local anaesthetic (LA), but even then it was a painful procedure. With the development of transrectal needle insertion in the late 1980s the LA was discontinued as rectal wall puncture causes no great discomfort and transrectal biopsies were considered to be painless. However, 5–90% of patients report discomfort or pain after TRUS biopsy [1–4]. This may arise from puncture of the prostatic capsule and stroma, although some men also find the introduction and maintenance of the ultrasound probe within the rectum uncomfortable [5,6]. Previous studies have evaluated the role of anaesthesia (rectal lidocaine gel [1] or periprostatic nerve blockade [5,7]) in improving patient comfort. Of the last two studies only one was a prospective report, but it was a small series [5].

Thus we prospectively evaluated the role of periprostatic LA injection in improving pain after TRUS biopsies and its effect on the rate of bleeding. Finally and importantly, the issue of excess infection was specifically addressed, as this procedure involves the injection of up to 10 mL of potentially bacteria-contaminated fluid.

Patients and methods

All patients referred to our department for outpatient TRUS-guided prostatic biopsy over a 7-month period were eligible for the study. All had an abnormal DRE and/or a raised PSA level. Several patients had been referred after previous biopsy had shown prostatic intraepithelial neoplasia (PIN). Those with a bleeding diathesis or on anticoagulation therapy, or with a history of radical prostatectomy or radiotherapy, were excluded. All patients received a standard antibiotic prophylaxis, i.e. one metronidazole suppository 1 g (Hawgreen Ltd, Dublin, Ireland) and then ciprofloxacin 500 mg twice daily for 3 days, both started 1 h before the procedure.

Patients were sequentially randomized to receive either LA or no injection. A consultant radiologist or specialist registrar in radiology performed all procedures. The method of anaesthesia is given below. All patients had sextant biopsies, but some also had additional seminal vesicle or focal prostate biopsies, and these were recorded.

Patients were examined in the left lateral decubitus position with a 9–5 MHz curved-array transrectal probe, and the prostatic volume and ultrasonographic appearances in the longitudinal and transverse planes recorded. The probe was then fitted with the needle guide and repositioned in the rectum. Local anaesthetic was then injected if appropriate but the probe was not removed again until the end of the procedure. Standard-protocol sextant biopsies were taken in sequence through the left and right peripheral zone, at the base, middle and apex of the gland. Additional biopsies were then taken if necessary. An 18 G 20 cm automatic cutting needle (MDTech, Gainsville, Florida, USA) and automated biopsy gun (Manan Medical Products, Northbrook, Illinois, USA) were used to obtain the biopsies.

For the periprostatic LA injection, 10 mL of 1% lignocaine hydrochloride (Braun Melsungen AG, Germany) was used. A 22 G 13.97 cm Chiba needle (Becton Dickinson, Franklin Lakes, NJ, USA) or spinal needle (Steriseal, Worcester, UK) was inserted through the needle guide under TRUS guidance. The tip of the needle, seen as an echogenic focus, was advanced into the gland just to the right of the gland apex, in the longitudinal plane. There is resistance to injection of lignocaine when the needle is within prostatic tissue. The needle was then slowly withdrawn under direct vision and once immediately outside the prostate ‘capsule’, with the needle in Denonvillier's fascia, the resistance dissipates and 5 mL of lignocaine can be injected easily (Fig. 1a– c). The lignocaine separates and spreads along the tissue planes of Denonvillier's fascia, bathing the entire posterior surface of the gland, from the apex to the seminal vesicles (Fig. 1d). A second injection of 5 mL was then given at the same site on the left side of the gland. The spinal needle was then withdrawn and biopsies taken after an interval of ≈ 2 min.

Figure 1.

A series of transrectal images of the prostate in the sagittal plane, showing periprostatic LA injection. With a single injection the entire prostate gland can be anaesthetized. a, The gland base is on the right of the image (B); the echogenic focus of the LA needle tip is visible adjacent to the marker (N), along the line of the needle guide (dotted line), and has been inserted under direct vision to abut the prostate ‘capsule’. In this instance, LA will be injected initially in the region of the gland apex. b, Early LA injection; the needle tip (N) is clearer in this image as the LA pools around the needle tip before spreading along the posterior aspect of the prostate gland towards the gland base. The LA pool is seen as an anechoic collection and a small amount of the LA is shown separating the tissue planes posterior to the prostate gland. c, Late LA injection; near the end of the LA injection the anechoic LA has clearly spread from the initial pool, around the gland apex (1) along the posterior aspect of the prostate gland to pool up to the gland base (2). The needle tip is still visible near the gland apex. Thus, the entire posterior surface of the gland is bathed by the LA. d, Late LA injection; the transrectal probe has been directed more cranially and clearly shows the anechoic pool of LA lying posterior to the seminal vesicle (SV) around the base of the gland.

After the biopsy all patients were given a questionnaire to complete; the questionnaire was devised and validated in our department, using a pilot study of 20 patients biopsied with no LA. Patients were asked to score the severity of their pain at the time of biopsy and at various intervals for 1 week afterward, using a four-point scale (1 = no pain, 2=mild pain, 3=moderate pain and 4=severe pain). Patients were asked about other additional symptoms, e.g. haematuria, haematochezia, haematospermia, pyrexia and any visit to their GP. Subsequent analgesic use and whether they would consent to another TRUS-guided biopsy was also recorded. All patients were given a stamped addressed envelope to return their questionnaire. Returned questionnaires were analysed using the nonparametric Mann–Whitney U-test and chi-squared test, where appropriate.

Results

Of the 157 patients assessed 73 received no LA and 84 had an LA injection; the mean ages of the two groups were similar. The LA group had a higher questionnaire return rate (74, 88%) than the other group (59, 81%), but the difference was not significant (P = 0.21; Table 1). Those who did not return the questionnaires were excluded from further analysis, but none of these patients required subsequent hospitalization for any biopsy-related complication. Some patients (eight in the no-LA group and five in the LA group) did not complete the pain scores for the whole week and were excluded from the analysis of pain scores.

Table 1.  Patient demographics and details, the number of patients visiting their GP after TRUS biopsy and the reason for the visit, and the median number of days with haematuria, haematochezia or haematospermia reported by patients after biopsy
VariableNo LALA
  • *

    P   =  0.96.

No. of patients7384
Number replied (%)59 (81)74 (88)
Mean (range) age, years67.6 (53–88)68.2 (45–88)
Previous biopsy, n716
Patients with multiple biopsies1018
No. patients/no. biopsies2/7, 8/813/8, 1/10,
4/12
Would you have another biopsy?
Yes/No52/366/5
Seen GP since procedure?79*
Fever01
Pain11
Bleeding55
Other25
Median (sd) no. of days after biopsy with:
Haematuria2.3 (2.4)2.6 (2.4)
Haematochezia0.36 (0.89)0.55 (1.1)
Haematospermia0.69 (1.2)0.65 (1.3)

A similar number of men visited their GP after biopsy (P < 0.96, Table 1). Of these, only one in the LA group attended because of pyrexia, and was treated successfully with further oral antibiotics as an outpatient. Similar numbers of patients in each group attended their GP for additional symptoms (Table 1), and the episodes of haematuria, haematochezia and haematospermia (median number of days experienced) were comparable in the two groups (Table 1). This was also true for those who had multiple biopsies taken (8–12), although there were few such patients, i.e. 10 in the no-LA arm and 18 in the LA arm (Table 1). A similar number of patients from each group reported that they would consent to a repeat biopsy, i.e. 95% of the no-LA group and 93% of the LA group (Table 1).

Seven men in the LA group (9%) used analgesia for 1–3 days, compared with eight (14%) of those in the no-LA group, who used analgesia for 1–6 days. Whilst this was not a significant difference (P = 0.39), there was a trend towards less frequent analgesic use in the LA group, which is shown in Figure 2.

Figure 2.

The number of days of analgesic use after TRUS-guided biopsy in the two groups (green, no LA, red, LA).

However, the pain scores in the two groups at the time of biopsy were very significantly different (P < 0.001) between the no-LA and LA groups, with median (sd) pain scores of 1.95 (0.65) and 1.53 (0.7), respectively. At 1 h after biopsy the median pain scores were similar, at 1.59 (0.76) and 1.58 (0.76), respectively (P = 0.93). However, the overall pain rated by the LA group was less, with a significantly lower median pain score for the whole week of 1.3 (0.33) in the no-LA and 1.2 (0.28) in the LA group (P = 0.0185). This supports the noted trend towards less analgesic use in the LA group.

In all, 322 TRUS-guided biopsies were taken in 1999–2000, at a cost of £203.04 per procedure (£65 366 per year). The additional needles used to administer the LA cost £2.80 per patient. The LA used was of minimal additional cost, at £3.00 per patient (£966 per annum). Furthermore, the total additional time for the periprostatic LA injection was estimated to be 2–3 min. No patient had any adverse effects at the time of LA injection. Patients occasionally complained of a mild ‘stinging’ sensation when the LA was injected in the position described, but injection in the midline (into the external sphincter or urethra) was painful.

Discussion

The hope that TRUS-guided biopsy would be painless and well tolerated has been not been realized [1–3,6,8]. In our department, patient surveys [9] have shown that up to 30% of men report pain of any severity, and procedural pain is recognized as a significant limitation to increasing the number of biopsies taken to minimize sampling error [10]. Pain also discourages those patients with high-grade PIN or persistent clinical suspicion of unidentified prostate cancer from agreeing to repeat biopsies.

Efforts have been reported to control biopsy-related pain; rectal anaesthetic gel was ineffective [1] and the value of an analgesic suppository has not been systematically evaluated, although some use it routinely. Periprostatic LA (both the method of introduction and its value) have not been thoroughly evaluated either. There is anecdotal evidence of improved tolerance of TRUS biopsy with LA [7] and Nash et al.[5] found that periprostatic lidocaine influenced pain scores at the time of TRUS biopsy. However, both studies were small, involving 50 and 64 patients, respectively. Thus quantitative controlled data on this subject are sparse, particularly for safety, as there is a theoretically increased risk of infection from the introduction of a pool of bacteria-contaminated fluid.

In the present prospective study there was a significant improvement in pain and a trend towards less analgesic use after LA. There was no apparent difference in pain between those patients having multiple and sextant biopsy. This is encouraging, as an earlier study showed that pain was proportional to the number of biopsies taken [7], although Naughton et al.[11] found otherwise. This may become increasingly relevant as patients return for repeat biopsy, and there is a trend towards taking more biopsies to maximize the diagnostic yield, particularly in large glands. The similarity of pain scores at 1 h after biopsy may reflect the short action of lignocaine; we are evaluating a longer-acting LA.

Most importantly, LA infiltration appears to be a safe procedure with no extra morbidity. There was no significant added infection risk, while the incidence of haematuria, haematochezia and haematospermia was also similar and compared favourably with previous studies [4,10,11]. There has been concern that the administration of periprostatic LA may result in fibrosis and interfere with nerve-sparing radical prostatectomy [12]; this has not been our experience (personal communication, Mr R. Kirby), but further data are required. Neither are there any prohibitive cost or time implications to the routine use of LA.

Questionnaires completed by the patients were used to assess pain and morbidity because in our experience it is a simple and cost-effective means of assessing events after TRUS biopsy [9]. Our previous surveys also produced a high return rate, as in the present study. More detailed studies, e.g. patient interviews at return outpatient visits, are prone to difficulties with recall and the possible introduction of interviewer bias. Assessing infection rates is more difficult; the ideal method should be serial blood or urine cultures, neither of which is logistically simple or reliable in a large group of outpatients. We consider the present method of assessment, using pyrexia and GP visits, to be an adequate alternative, as no patient required hospital admission for biopsy-related infection or complications.

The method of periprostatic infiltration used here was developed empirically, considering the prostatic anatomy. The gland receives sympathetic and parasympathetic nerve supplies to both the glandular and stromal elements. The nerves join the neurovascular bundles, which pass along the posterolateral margins of the gland between the capsule and Denonvillier's fascia, and pierce the capsule, particularly at the base and apex in the 4 and 8 o'clock positions. Thus, anaesthetic introduced at these points will numb the entire gland. Additionally, anaesthetic at this point dissects and spreads between what were taken to be the planes of the Denonvillier's fascia. Apparently, pain control is most effective when the anaesthetic spreads widely and freely (Fig. 1). In our experience many sites of small-volume infiltration, as used by some, is not as beneficial. However, in a proportion of patients, the LA does not flow freely and in these cases we inject further small volumes at the base of the gland. None of the present patients had undergone previous prostatectomy or radiotherapy. We recently used LA with equal benefit in these groups, but always ensured that the patients had an empty bladder, as there is a theoretical risk of temporarily relaxing the external sphincter, with loss of bladder control. Predictably, these patients do not always show free flow between the tissue planes, and require many small-volume infiltrates. In yet other patients we noted that the seminal vesicle biopsies remain painful despite otherwise adequate prostatic anaesthesia.

In conclusion, periprostatic LA significantly improved patient tolerance of TRUS-guided biopsies with no added morbidity. Using such anaesthesia, multiple and repeated biopsies become feasible, and it is now part of our routine practice.

Acknowledgements

We thank Mr R. Kirby, Mr M.J. Bailey, Mr K. Anson and Mr C. Anderson, Consultant Urologists at St George's Hospital, for referring and allowing us to study their patients. The assistance of Professor Bland (Professor of Medical Statistics, St George's Hospital Medical School, London) is equally gratefully acknowledged.

Ancillary