Complications and postoperative events after cryosurgery for prostate cancer

Authors


Thomas J. Polascik, Duke University Medical Center, Box 2804, Yellow Zone, Durham, NC 27710, USA. e-mail: polas001@mc.duke.edu

Abstract

Study Type – Harm (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

There is limited data available to determine the rate of complications following prostate cryosurgery. This is particularly true when limiting the experience to third generation cryoprobes. Herein we thoroughly report all complications and postoperative events, defined as any deviation from an ideal postoperative course, with extended follow up.

OBJECTIVES

  • • There has been a call in the urological literature for standardized reporting of complications.
  • • To use strict criteria aiming to report our complications and other postoperative events in a cohort of men undergoing third-generation prostate cryosurgery.

PATIENTS AND METHODS

  • • Demographic, clinical and pathological data were collected on men undergoing primary cryosurgery from 2002–2010, excluding those who had received neoadjuvant or adjuvant radiotherapy.
  • • Complications and events were broadly defined as any deviation from the expected postoperative course and any subjective complaint expressed during a follow-up visit.
  • • Descriptive statistics were generated and compared between groups using chi-squared and rank sum tests as appropriate.
  • • Logistic regression analyses were performed to assess the potential predictors of any complication or event.

RESULTS

  • • In total, 106 consecutive patients were identified.
  • • There were no intra-operative complications or instances of equipment failure. We observed 42 early complications, with 48 delayed and 10 late postoperative events.
  • • A total of five (4.7%) patients had persistent urge and/or stress incontinence.
  • • Thirty-one patients had International Prostate Symptom Score (IPSS) and bother index scores available before and after surgery; median scores decreased from 7 and 2 to 6 and 1, for IPSS and bother index, respectively. Twenty-four patients had Sexual Health Inventory for Men scores available before surgery and at 2-year follow-up; median scores decreased from 11 to 2.
  • • On multivariate analysis, there were no significant associations.

CONCLUSIONS

  • • Modern cryosurgery is safe, and most of the complications and postoperative events are transient.
  • • Erectile function, however, has marked deterioration.
  • • We were unable to identify significant risk factors for complications or postoperative events.
Abbreviations
BI

bother index

IRE

incontinence-related event

OR

odds ratio

SHIM

Sexual Health Inventory for Men

INTRODUCTION

Prostate cryosurgery has become an increasingly employed treatment modality for prostate cancer, prompting the release of a best practice policy statement by the AUA in 2008 [1]. The compiled complications in that review originated mainly from series with patients treated before the introduction of modern, third-generation cryoprobe systems that utilize argon and helium gases, monitor the iceball more precisely under TRUS guidance, protect the urethra with a warming catheter, and incorporate needle-sized cryoprobes that do not require tract dilation [2].

There has been a call for standardized reporting of surgical complications. Martin et al. [3] defined criteria for elements that should be included, including how the data were collected, the duration of follow-up, the inclusion of outpatient information, definitions of complications, mobidity and mortality rates, procedure-specific complications, severity grades, length of stay data, and risk factors used in the analysis [3]. For grading of complications, the modified Clavien system has been commonly used [4]. In the present study, we utilize the Martin criteria to thoroughly report any possible complication and postoperative event, either subjective or objective, in our cohort of men undergoing primary cryosurgery with third-generation cryoprobes.

PATIENTS AND METHODS

Approval was obtained from our institutional review board for the retrieval of demographic, clinical and pathological data on patients undergoing primary cryosurgery from our prospectively maintained database between 2002 and 2010. All patients were treated using third-generation technology. We excluded patients who had received neoadjuvant or adjuvant radiotherapy as part of salvage therapy or a clinical trial. The surgical procedure has been described previously in detail [2]. Neurovascular preservation was attempted in certain potent men before surgery by placing an active cryoneedle outside the prostate and keeping the area warm with helium. Hormonal ablation was used in some men with prostate volumes estimated at >40 g and was stopped immediately after surgery.

The variables collected were age, race, family history of prostate cancer, use of tobacco, hormonal ablation, prebiopsy PSA level, prostate volume and sagittal length (TRUS estimated at time of biopsy), biopsy Gleason score, comorbidities using the Charlson index, American Society of Anesthesiology scores, pre-existing LUTS, clinical stage, presence of a median lobe, hospital stay and follow-up, as well as IPSS, bother index (BI) and Sexual Health Inventory for Men (SHIM) scores before surgery and at 2 years after surgery.

Patients were seen after discharge at 1–2 weeks for a voiding trial and at 3 months for baseline post-ablation PSA level and assessment of voiding. Follow-up was typically carried out every 3–6 months at which time PSA level, IPSS, SHIM score and subjective patient status were determined. Clinic visits were spaced out after 1 year if the PSA level and patient's functional status were stable. Charts were reviewed for complications and other postoperative events during these visits and graded according to the Clavien system [4]. All adverse outcomes from surgery to within 30 days of the first postoperative visit were considered as complications; otherwise, they were classified as delayed (≥1–12 months) or late (>12 months) events. Any deviation from the normal or ideal postoperative course involving the lower urinary tract, genitalia or surgical site was included, in addition to any subjective complaint expressed during a follow-up visit, regardless of the time from surgery.

Worsening LUTS were determined clinically by subjective patient complaint (classified as having primarily an urgency or frequency component), as well as objectively by IPSS and BI. Dysuria was defined as painful urination. Haematuria was considered when the patient had to undergo a full urological evaluation. An incontinence-related event (IRE) was broadly defined as the use of any wetness protection or perceived threatened leakage, even if no protection was used. IREs included a range of experiences such as urgency, urge-incontinence, stress incontinence, post-void dribbling, etc. In contrast, true incontinence was defined as leakage requiring the use of any pad or protective garment. Incontinence was classified clinically as having primarily an urge, stress, or mixed component.

Descriptive statistics were generated and compared between the two groups (with and without a complication/event) using chi-squared and rank sum tests as appropriate. Logistic regression analyses were performed to assess potential predictors of any complication and any complication or event, adjusted for age, race (African American, non-African American), Charlson index (classified as 0, 1–2 and ≥3), PSA level, prostate volume, tobacco use, hormonal ablation, pre-existing LUTS (yes/no), median lobe (yes/no), clinical stage (T1/T2) and biopsy Gleason score. PSA level and prostate volume were log-transformed [log PSA, 10(log Volume)] to almost normal distribution. American Society of Anesthesiology, age-adjusted Charlson score and sagittal lengths were not included as a result of significant intercorrelations (data not shown). Statistical analyses were performed using SPSS, version 17 (SPSS Inc., Chicago, IL, USA) Values are given as the median (interquartile range) and n (%) unless otherwise indicated. P < 0.05 was considered statistically significant, whereas values in the range 0.05–0.1 were defined as trends toward significance.

RESULTS

We identified 106 consecutive patients. Biochemical recurrence and biopsy data post cryosurgery on the first 97 patients have been reported in detail elsewhere [5]. For the current series, 30 of 106 patients underwent a post-procedure biopsy for a PSA level ≥0.5 ng/mL; 14 were found to have a biopsy proven cancer. Demographic, clinical and pathological characteristics of the cohort are shown in Table 1. Bilateral and unilateral nerve preserving procedures were each attempted in six (5.7%) cases. Hormonal manipulation was used before surgery in 28 (26.4%) patients for a mean of 6 months. A median lobe was present in 18 (17.0%) patients. Median lobes were not routinely treated during cryoablation.

Table 1.  Demographic, clinical, and pathological characteristics of the cohort
CharacteristicCohort (n= 106)With any event (n= 62)Without event (n= 44)P
  1. ASA, American Society of Anesthesiologists.

Age (years)69.6 (63.1–73.9)70.3 (62.8–75.2)69.4 (64.7–72.7)0.530
Family history   1.000
 Yes21129 
 No855035 
Race   0.306
 African American371918 
 Non-African American694326 
Tobacco use   0.779
 Yes1587 
 No915437 
Hormonal ablation   0.180
 Yes281315 
 No784929 
Preoperative Impotence   1.000
 Yes714229 
 No352015 
Preoperative LUTS   0.325
 Yes553524 
 No512720 
Age-adjusted Charlson score   0.927
 1963 
 21798 
 3301812 
 4–5362016 
 6–91495 
Charlson score   0.437
 0462818 
 125178 
 219811 
 3–41165 
 5–7532 
ASA score   0.649
 2432419 
 3513219 
 41266 
TRUS volume (mL)27.2 (21.4–34.1)27.6 (22.2–34.4)26.6 (20.0–34.0)0.534
Sagital length3.7 (3.5–4.1)3.8 (3.5–4.3)3.7 (3.4–4.0)0.349
Presence of median lobe   0.443
 Yes19136 
 No874938 
Gleason score   0.293
 ≤6633330 
 7332211 
 ≥81073 
Clinical stage    
 T18048320.650
 T2261412
Biopsy PSA level5.1 (4.2–8.5)5.0 (4.0–7.1)6.4 (4.5–8.8)0.054
Follow-up (months)25.7 (9.4–43.2)26.8 (10.8–45.4)22.9 (6.9–41.3)0.224

In total, 87 (82.1%) patients had outpatient procedures and were discharged after surgery, 17 (16.0%) patients stayed overnight and one (0.9%) patient was discharged on the second postoperative day. The majority of admissions were overnight observations early in our experience as a result of hospital policy. All complications and postoperative events for the cohort are shown in Table 2. There were no intra-operative complications or instances of equipment failure. We noted 42 complications, 48 delayed events and 10 late events. A single death of a suspected cerebrovascular accident (Clavien grade V) was recorded 6 days after surgery.

Table 2.  Complications and postoperative events
CharacteristicComplications (<1 months)Delayed events (≥1–12 months)Late events (>12 months)Total
  1. IRE, Incontinence-related event.

Dysuria21 3
Worsening LUTS   14
 Urgency18  
 Frequency1   
 Unspecified 22 
Bruised perineum2  2
Scrotal oedema8  8
Epididymitis1214
Prostatitis1 12
UTI24 6
Balanitis21 3
Prostatic abcess 112
Perineal abcess 1 1
Pericatheter bleed8  8
Haematuria2 24
Clogged catheter3  3
Pain   4
 Rectal2   
 Testicular1   
 Urethral 1  
Splayed stream12 3
Phimosis 1 1
Meatal stenosis 1 1
Blood per rectum1  1
Leg paresthesia1  1
Retention/suprapubic catheter 4 4
IRE   24
 Urge1112 
 Stress13  
 Mixed 2  
 Unspecified 31 
Death1  1
Clavien grade   100
 I30163 
 II8203 
 III3133 
 V1   

Foley catheters remained in place for a median (range) of 8 (6.2–10.0) days, after which the patients underwent a voiding trial. In total, 19 (17.9%) patients failed the first voiding trial by being unable to void or having consistent post-void residuals >150 mL. Of those, four (21.1%) patients failed two voiding trials and subsequently had a suprapubic catheter (SPC) placed (Clavien grade IIIa). All patients had the suprapubic catheter removed within 30 days after placement.

There were 24 (22.6%) patients who experienced a postoperative IRE. This was usually transient in nature, resolving after an mean of 9 months. Of these, five had threatened incontinence and were using no pads. There were five (4.7%) patients who had persistent incontinence, mostly urge-related and in patients with a high IPSS or BI score. The clinical characteristics of these patients are shown in Table 3. In total, 18 patients underwent cystoscopy for indications including worsening LUTS, haematuria or as follow-up for bladder cancer. Some degree of mild, incomplete urethral slough was noted in three (2.8%) patients, although this was not felt to be clinically significant or to require therapy. Overall SHIM, IPSS and BI scores are shown in Table 4. There were 31 patients who had IPSS and BI scores available before and after surgery; median (range) scores decreased from 7 (3–12) and 2 (1–3) to 6 (2–9) and 1 (0–2), respectively. There were 35 (33%) patients who identified themselves as ‘potent’ before cryosurgery; despite this, the median (range) SHIM for these patients was 7 (2–15). Among 24 patients for whom SHIM scores were available before surgery and at 2-year follow-up, the median (range) score decreased from 11 (2–17) to 2 (1–6). SHIM scores for patients with attempted bilateral nerve preservation decreased from a median (range) of 21 (18–23) to 7 (5–10). There were 48 patients who sought post-treatment therapy for their erectile dysfunction; 22 used two or more erectile aids during the course of follow-up. Outside of the aforementioned cases of persistent incontinence and erectile dysfunction, all complications and events resolved.

Table 3.  Characteristics of patients with persistent incontinence
PatientPreoperative IPSS/BITypeProtectionFollow-up (months)Comments
  1. *Follow-up time limited, resolution unknown. BI, bother index; PPD, pads per day.

120/3UrgeDiaper4* 
224/4Mixed1–2 PPD3*History of detrusor instability
39/3UrgeDiaper24History of urinary retention
     Incontinence 2 years post procedure
40/0Unk2 PPD3* 
5UnknownUrge6 PPD69Incontinence 5 years post procedure
Table 4.  Preoperative and postoperative (2 year) IPSS, bother index (BI) and Sexual Health Inventory for Men (SHIM) score
TimingSHIMIPSSBI
Preoperative6.5 (2–15) (n= 54)10 (4–13.5) (n= 63)2 (1–3)
Postoperative2 (1–6) (n= 58)6 (2–10) (n= 58)1 (0–2)

On multivariate analysis for all complications and postoperative events (Table 5), we found no significant associations, albeit a trend toward significance was noted for hormonal ablation (odds ratio, OR, 0.340; P= 0.052), PSA level (OR, 0.187; P= 0.068) and pathological Gleason score ≥8 (OR, 4.695; P= 0.092). On multivariate analysis for complications, no significant associations were found (data not shown).

Table 5.  Multivariate analysis of risk factors for any complication or postoperative event
CharacteristicOdds ratio95% CIP
Age (years)1.0000.940–1.0640.998
Race (non-African American reference)0.6540.249–1.7150.388
Tobacco use1.0040.266–3.7850.995
Hormonal ablation0.3400.114–1.0100.052
LUTS1.7410.695–4.3630.237
Presence of median lobe2.0510.605–6.9600.249
Clinical stage   
 T1 (reference)
 T20.7240.269–1.9500.523
PSA level0.1870.031–1.1350.068
Prostate volume (mL)1.0820.807–1.4500.600
Gleason score  0.109
 ≤6 (reference)
 72.1000.787–5.6040.138
 ≥84.6950.778–28.3120.092
Charlson score  0.931
 0
 1–21.0950.298–4.0290.891
 >31.2560.349–4.5220.728

DISCUSSION

Although the original Clavien grading system was created to rate complication severity based on the length of hospital stay, with the increasing number of outpatient surgeries performed, additional emphasis has been placed on the patient's perspective [4]. A patient who undergoes cystoscopy in the clinic to evaluate urethral pain or an anatomical abnormality (classified as grade IIIa) that subsequently resolves may be considered as having more mild complication than a patient who has to take a medication such as an anticholinergic to control urge incontinence (grade II). Despite this, use of the Clavien system has been increasing in the urological literature. Recently, two studies addressed the issue of standardized reporting of complications with traditional extirpative surgery for prostate cancer [6,7]. Novara et al. [6] considered complications within 90 days of surgery and Rabbani et al. [7] considered all complications regardless of time after surgery. The number of cases, prostate size and type of surgery were among the independent risk factors that they identified in their analyses [6,7].

Previous reporting of complications after cryosurgery has not been standardized. Early in the experience, rectal fistulae were occasionally reported, a complication seldom seen with modern methods [8–10]. In the mid-1990s (coincident with removal of the uretheral warming catheter from clinical practice), one of the more common complaints was bladder outlet obstruction, which was present in just under 25% of patients undergoing therapy [10,11]. Treatment for refractory cases included TURP, which left close to half the patients with stress urinary incontinence post resection [10]. The recognition of the safety features of the urethral warming catheter and its subsequent reinstatement decreased this complication. Urethral slough was only incidentally observed to a limited degree in three of our patients, similar to the rate of ≈5% noted in other modern series [8,12], with rates that vary depending on whether cystoscopic evaluation is performed in response to complaints of worsening LUTS. Transient urinary retention immediately after the procedure is expected as a result of temporary prostate swelling. Some cryosurgeons have addressed this issue by placing a suprapubic catheter at the time of surgery, although this adds morbidity to the procedure. Retention rates vary depending on the criteria and timing of voiding trials chosen by cryosurgeons. In total, 18% of our patients failed their initial voiding trial (after a median of 8 days catheterization) of which four of 19 later required a suprapubic catheter after failing a second voiding trial. The duration of foley catheterization in other series before a voiding trial or how prolonged catheter drainage needed to be before it was deemed ‘retention’ is unclear, although, as in our patients, the inability to void was transient, did not require TURP, and resolved within ≈1 month [12–14].

Incontinence after cryosurgery has a markedly deleterious effect on the patient's quality of life. Rates of incontinence vary from 2–32% in the literature and such a wide range reflects differences in the definition of incontinence and the length of follow-up [9–18]. The IRE rate of 22.6% in the present study was based on a very broad definition that included any patient complaint, including threatened incontinence where no pad was used. Of note, only five (4.7%) patients had persistent incontinence; 80% with predominantly urge-related symptoms. Some 60% of these men had relatively short follow-up, with the possibility of later resolution; the others developed incontinence after several years. For these late developing cases, the aetiology may have been a result of factors other than or contributing to changes from their previous cryosurgery. The time to resolution of incontinence was 9 months in the present study and is similar to that reported by Bahn et al. [9], where 84.1% of previously continent men regained continence within ≈6 months. Interestingly, a quality of life survey found that only ≈50% of patients treated with primary cryosurgery were ‘not at all’ troubled by leakage of urine [19].

We have previously reported on urinary function after primary cryotherapy [20]. Initially, IPSS and BI scores worsened at 6 months of follow-up, although improvement was noted at 2 years, and stabilized with longer follow-up. Similarly, we noted better IPSS and BI scores in our larger cohort after 2 years. However, 14 (13.2%) patients complained of worsening LUTS, particularly urgency as a delayed event, at a time point where the IPSS increased. These symptoms resolved in all patients, and may indicate the natural history of voiding complaints after cryosurgery. Other studies have noted worsening urgency and frequency in 32–65% of patients [15–17], with improved or stable symptoms in a subset of their patients [13,16]. Erectile dysfunction occurs commonly after cryosurgery; some advocate not offering this therapy to patients who wish to preserve potency [8]. Impotence rates are in the range 29–94% [8,9,12,14–16,19,21]. Because several men in the present study cohort were not sexually active before and did not attempt sexual activity after their cryosurgical procedure, and therefore did not mention this as a complaint, the rate of true worsening erectile function (i.e. erectile dysfunction) was unknown. An often cited timepoint for the recovery of erectile function after surgery is 2 years [21]. We noted a high rate of impotence, with median 2 years postoperative SHIM scores of 2 among all patients (i.e. with severe erectile dysfunction) [22]. Despite this, the desire to maintain potency is shown by the almost 50% rate of using erectogenic aids after surgery, with almost 25% of the patients trying more than one therapy.

In total, eight of our patients had scrotal oedema, four had haematuria and three complained of genital or rectal pain, which complications noted in other studies after cryosurgery [12,13,15,17,18]. We have previously reported on a case of delayed prostatic abscess [23], a rare postoperative occurrence, although others have commented on a ‘contained cavity’ within the prostate seen on cystoscopy [13]. Other reported complications and events in the present study, such as epididymitis or prostatitis, may not have been related directly to the surgery and may represent normal background urological complaints in ageing males. Also, complaints such as dysuria after Foley catheter removal or bruising in the perineum may be considered as normal occurrences before surgery, although we considered them in the present analysis because they caused patient concern. We did not note other less commonly reported cryosurgical complications, such as penile numbness [10–12] and stricture [11,16,24].

The results of our multivariate analysis showed trends to significance for a protective effect of previous hormonal ablation and increasing biopsy PSA level. Both of these parameters indicate larger prostates that would have had the most dramatically improved voiding after therapy either with androgen deprivation therapy and cryosurgery or cryosurgery alone [20]. However, prostate size alone was not independently significant, indicating that the cause may be multifactorial. Finally, higher Gleason grades showed a trend to higher risk for complications. These cases would have required the most aggressive therapy. Given the variety of complications and postoperative events reported, involving multiple sites and at varying time points, it is not unexpected that no single variable accounted significantly for increased risk.

In conclusion, modern cryosurgery is a safe treatment modality for prostate cancer. The majority of complications and other postoperative events are mild and transient, with the notable exception of erectile dysfunction. Because there is a great deal of variety of events and affected organ systems among these, we were unable to identify significant risk factors for complications or postoperative events.

ACKNOWLEDGEMENTS

The authors would like to acknowledge Lori Elison, CRNA, for her help in preparing the manuscript.

CONFLICT OF INTEREST

Jorge R. Caso and Vladimir Mouraviev are supported by research funds from the Committee for Urologic Research, Education and Development (CURED) of Duke University. Thomas J. Polascik is supported by research funds from Galil Medical.

Ancillary