Clinically and Statistically Significant Changes Seen in Sham Surgery Arms of Randomized, Controlled Benign Prostatic Hyperplasia Surgery Trials

Transcription

1 Clinically and Statistically Significant Changes Seen in Sham Surgery Arms of Randomized, Controlled Benign Prostatic Hyperplasia Surgery Trials Charles Welliver,*, Michael Kottwitz, Paul Feustel and Kevin McVary From the Division of Urology, Department of Surgery (CW) and Center for Neuropharmacology and Neuroscience (PF), Albany Medical College, Albany, New York, and Division of Urology, Department of Surgery, Southern Illinois University School of Medicine (MK, KM), Springfield, Illinois Abbreviations and Acronyms BPH ¼ benign prostatic hyperplasia LUTS ¼ lower urinary tract symptoms Qmax ¼ maximum urinary flow rate RCT ¼ randomized controlled trial SS ¼ symptom score Accepted for publication June 11, Study received institutional review board approval. * Correspondence: Albany Medical College, 23 Hackett Blvd., Albany, New York (telephone: , FAX: ). Financial interest and/or other relationship with AMS and Coloplast. Financial interest and/or other relationship with Lilly/ICOS, NxThera, Watson Pharmaceuticals, National Institute of Diabetes and Digestive and Kidney Diseases, Astellas, AMS and Sophing. Editor s Note: This article is the fourth of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 1834 and Purpose: Medication trials frequently involve a placebo arm to more fairly assess the efficacy of the study drug. However, benign prostatic hyperplasia surgery trials rarely include a sham surgery group due to the inherent risks associated with simulating treatment in these patients. As a result the placebo response to sham surgery for benign prostatic hyperplasia is largely unknown. Materials and Methods: We systematically reviewed the available literature to look for randomized, controlled trials involving endoscopic or intraprostatic injection benign prostatic hyperplasia treatments that included a sham surgical arm from January 1990 to February Studies that included an objective symptom questionnaire and maximum urinary flow at 3 months were included. Results were analyzed together with weighting based on study sample size. Results: The initial search yielded a total of 1,998 potential studies. After reviewing abstracts and full text articles 14 randomized, controlled trials were included in some part. An average decrease from 22.3 to 16.7 (e27%) was seen in studies of the AUASS (American Urological Association symptom score) 3 months after a sham endoscopic procedure (p ¼ ) with an increase in maximum urinary flow of 1.3 ml per second (14%, p ¼ 0.001) at 3 months. Prostate injection based studies at 3 months were similar with a decrease from 21.3 to 15.7 (e26%, p <0.001). Maximum urinary flow increased by 2.0 ml per second (23%, p ¼ 0.043). Conclusions: Sham controlled endoscopic and injection benign prostatic hyperplasia interventions demonstrate a considerable and statistically significant change in symptom scores and maximum urinary flow, which is comparable to the response seen in medication trials. Future uncontrolled benign prostatic hyperplasia surgical trials should consider these findings when interpreting outcomes. Key Words: prostatic hyperplasia, lower urinary tract symptoms, endoscopy, injections, randomized controlled trials as topic RANDOMIZED, placebo controlled clinical trials are considered the gold standard in research and yield level 1 evidence. While these trials are common in studies examining medications, they are less frequently seen in studies of patients undergoing surgical procedures. In medication trials a placebo medication is easy to manufacture and administer, creating an excellent comparative group with minimal additional risk to the 1682 j /15/ /0 THE JOURNAL OF UROLOGY Ó 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH,INC. Vol. 194, , December 2015 Printed in U.S.A.

2 CHANGES IN SHAM SURGERY ARMS OF BENIGN PROSTATIC HYPERPLASIA SURGERY TRIALS 1683 patients taking placebo. The inclusion of this group is important as it serves as a reference point above which the study drug must exceed results to be considered effective. However, assessing surgical outcomes is more challenging since creating a reasonable comparative placebo group is more difficult. As most surgical procedures are performed using some form of anesthesia and with a surgical incision, there is more risk associated with inclusion in a sham surgery group than with administration of a placebo medication. Uniform failure to account for the placebo effect leads one to assume that improvements in LUTS were related to treatment while the intervention effect remains unknown. However, because endoscopic and intraprostatic injection BPH surgeries are done transurethrally or percutaneously and do not require a skin incision, some trials exist that include a sham surgery group that was evaluated with subjective and objective outcomes. In this study we sought to quantify the expected response to sham surgery for endoscopic procedures or intraprostatic injections for the treatment of LUTS related to BPH. METHODS A systematic review was done based on a literature search through the PubMedÒ-MEDLINEÒ database. The key words used to search were ( placebo or randomized controlled or sham ) and ( prostatic hyperplasia or benign prostatic hyperplasia ). Search terms were chosen to be overly inclusive so that any potential relevant reports would be identified. The literature search aimed to identify all full-length articles reporting the results of RCTs of surgery for LUTS due to BPH that were published between 1990 and February After screening titles and abstracts a first selection for eligibility was made (fig. 1). The reference list of applicable RCTs was screened for additional references. Review articles and book chapters on BPH were also searched to look for additional studies. Only RCTs including a sham procedure (ie catheter insertion, antenna insertion to simulate microwave therapy, or intraprostatic injection of placebo or vehicle) for LUTS due to BPH were considered. Once selected the articles were studied in full text to gather information about study design, baseline assessments of the study metric (questionnaire based SSs and Qmax) and reassessment of these outcomes at some later interval. To create standardization studies were only analyzed if they included outcome data on 3 months after the sham procedure or intervention. When data were missing from articles that could otherwise be included, investigators 1,2 or device manufacturers 3 were contacted to obtain this missing information. In some cases missing results were pulled from the NIH (National Institutes of Health) website (ClinicalTrials.gov). 4e6 Studies were excluded from data analysis on SS if they did not report relevant data about main functional outcomes, including well recognized SSs such as AUASS, Madsen-Iversen SS or Boyarksy SS. If studies did not include accepted SS outcomes, they were still included in Qmax calculations if these data were available. In a separate calculation for SS articles that reported outcomes only with a study that used a 0 to 27 scale (Boyarsky and Madsen-Iversen SSs) were scaled to a 0 to 35 scale similar to AUASS and analyzed together. Studies were then separated based on whether they were for intraprostatic injection or mimicked endoscopic treatment of BPH. These different groups were analyzed separately as intraprostatic injections may have had a different placebo effect (mass effect from drug, or trauma to prostate parenchyma or capsule from needle insertion), which could cause local changes in the prostate that would differ from those of an endoscopic procedure. Data organization and statistical analysis were completed with ExcelÒ. Weighted means of SS and Qmax were calculated using reported means with weights determined by the number of patients in each study. The p values of changes were determined by the paired t-test comparing before and after differences, which were again weighted by the number of patients. Outcomes were considered statistically significant at p <0.05. Forest-type plots were constructed based on these findings. In these plots the relative size of the diamond depicted was scaled to the number of patients examined in the respective study with a standard effect size added to the right side of the figure. Figure 1. RESULTS A PubMed-MEDLINE database search using the selected key words yielded 1,997 results (fig. 1). One additional and potentially relevant record was found based on a review of book chapters. After reviewing the titles and abstracts of these studies 1,978 were excluded as they did not meet initial study inclusion criteria due to a variety of reasons (ie not sham controlled or trials were for medical treatment of BPH). Full text articles of the remaining 20 studies were then reviewed. Four articles were excluded

3 1684 CHANGES IN SHAM SURGERY ARMS OF BENIGN PROSTATIC HYPERPLASIA SURGERY TRIALS Table 1. Study summary of SS References SS Name Type No. Pts SS Initial/Final (difference) Scaled SS Initial/Final (difference) Albala et al 3 AUASS Endoscopic /17.0 (e5.7) e Blute et al 13 AUASS Endoscopic /16.3 (e4.5) e de la Rosette et al 2 Madsen-Iversen Endoscopic / /10.6 (e5.1) de Wildt et al 14 Madsen-Iversen Endoscopic / /13.5 (e3.2) Elhilali et al 5 AUASS Injection /16.7 (e5.8) e Larson et al 15 AUASS Endoscopic /14.5 (e6.8) e Lepor et al 16 Boyarsky Endoscopic / /11.4 (e4.0) Marberger et al 6 AUASS Injection /15.2 (e5.5) e McVary et al 4 AUASS Injection /15.9 (e5.6) e Ogden et al 17 Madsen-Iversen Endoscopic / /16.6 (e1.8) Roehrborn et al 1 AUASS Endoscopic /16.2 (e7.7) e Roehrborn et al 18 AUASS Endoscopic /18.5 (e5.9) e since the interval between initial treatment and followup assessment was not 3 months. 7e10 One study focused on changes in sexual function with voiding symptoms presented in a previously published report 11 and another was not randomized. 12 Three of the 14 remaining studies assessed outcomes of intraprostatic injections and 11 assessed an endoscopic procedure (tables 1 and 2). 1e6,13e20 Five endoscopy based studies evaluated AUASS 3 months after sham treatment in 267 patients. 1,3,13,15,18 In these reports a 27% decrease (6.3 points) in AUASS was noted from 23.0 to 16.7 (p <0.001, table 3 and fig. 2). 1,3,13,15,18 In 1 study that used only the Boyarsky SS 16 and 3 that used the Madsen-Iversen SS 2,14,17 at 3 months the initial and final scores were scaled from the usual limits of these scores (0 to 27) to the AUASS scale (0 to 35) and then analyzed with studies that used AUASS. This analysis yielded 9 studies with a total of 374 patients (table 1). 1e6,13e18 In this analysis a change of 21.2 to 15.7 (e26%) was seen (p <0.0001, table 3 and fig. 3). 2,3,11e20 A total of 11 endoscopic studies evaluated Qmax 3 months after a sham procedure in 449 patients. An increase of 1.3 ml per second (14%) was noted in Table 2. Study summary of Qmax References No. Pts Initial/Final ml/sec Qmax (difference) Albala et al /11.1 (2.8) Bdesha et al /9.8 (e1.0) Blute et al /9.4 (2.0) de la Rosette et al /9.5 (e0.2) de Wildt et al /9.7 (0.1) Elhilali et al /10.2 (1.3) Larson et al /9.2 (1.4) Lepor et al /12.9 (2.4) Marberger et al /12.9 (2.8) McVary et al /9.7 (1.7) Ogden et al /9.2 (0.6) Roehrborn et al /9.6 (1.5) Roehrborn et al /9.9 (2.0) Zerbib et al /10.8 (0.2) these patients (range 8.8 to 10.0 ml per second, p ¼ 0.001, table 3 and fig. 4). 4e6 This value was not the absolute sum between these values due to rounding at the tenths place. Three intraprostatic injection studies including a total of 268 patients were included in analysis (table 3 and fig. 5). 4e6 All were published in the last 2 years and used AUASS as the defined SS. A decrease of 5.6 (26%) was seen in AUASS (from 21.7 to 15.7, p ¼ ). The change in Qmax was even larger than that in endoscopic studies since there was an increase of 2.0 ml per second (range 8.8 to 10.8, p ¼ 0.043). DISCUSSION To our knowledge this is the first study examining and summating the sham response to endoscopic and injection based therapies for LUTS due to BPH. Our data show a clinically and statistically meaningful change in SS and Qmax in patients who undergo sham procedures. The changes found were surprising and exceed what are commonly referred to as clinically meaningful changes when discussing BPH treatments. A change in AUASS of at least 3 is considered clinically meaningful 21 and the decrease in AUASS was 6.3 in the endoscopic group and 5.6 in the injection group. The changes seen with a sham endoscopic procedure are actually similar to the expected AUASS response to the commonly used modern a-blockers such as alfuzosin, tamsulosin and silodosin. In placebo based RCTs of changes at 12 weeks of treatment with the standard dose of a-blockers AUASS decreased in the range of to ,24 However, a significant placebo response was noted in these studies and the change in SS that could be attributed only to medication may actually be lower than these values. When subtracting the SS change noted for placebo, the response attributed only to

4 CHANGES IN SHAM SURGERY ARMS OF BENIGN PROSTATIC HYPERPLASIA SURGERY TRIALS 1685 Table 3. Endoscopic and intraprostatic injection procedures Mean SD Score No. Studies No. Pts Initial Final Mean SD Change (%) Change p Value Endoscopy: AUASS (e27) Scaled SS (e26) Qmax (ml/sec) (14) Intraprostatic injection: AUASS (e26) Qmax (ml/sec) (23) a-blockers would be in the range of to A recent meta-analysis showed that the true drug effect (effect above placebo) was e2.01 (95% CI e1.57 to e2.44, p <0.01). 26 Surprisingly the purely objective Qmax was also influenced by sham treatment with an increase of 1.3 and 2.0 ml per second in the endoscopic and injection groups, respectively. This also approximately simulates the expected response to a-blockers beyond that of placebo with studies demonstrating a e to 2.3 ml per second range 23,28 and a meta-analysis revealing a true drug effect of 1.11 ml per second above placebo (95% CI 0.79 to 1.43, p <0.01). 26 While medication trials frequently include a placebo group to compare study drug efficacy, it is much more difficult to create this comparison group in surgical trials. The reasons for this are multiple, including the unwillingness of patients to be randomized to a nonsurgical group and the inherent risk associated with general anesthesia. However, as BPH procedures became less invasive many procedures could be completed without general anesthesia. Thus, RCTs involving noninvasive technologies such as balloon dilation and transurethral microwave thermotherapy started to compare results to those of a sham procedure in an effort to demonstrate that the technique had results above an expected placebo response. Controlling for these findings in future BPH surgical trials may be challenging but not impossible. In fact one of the analyzed intraprostatic injection trials included in our study used a sham injection procedure and 4-week run-in before randomization to minimize the sham treatment response. 4 Design models similar to such trials in which a patient undergoes a sham procedure that is believed to be at least part of the overall treatment could help minimize the sham response to eventual active treatment. While this approach may be more challenging with procedures that involve the need for general anesthesia, patients could be potentially started on placebo medication in the weeks before a procedure to generate a generalized response to the treatment of LUTS due to BPH. While our series compares results across different studies, it is not a true meta-analysis but rather a statistically analyzed review. By definition a metaanalysis is a statistical approach to combine compared outcomes from multiple independent studies. These studies must have similar treatment groups to be analyzed together. Because the sham results compiled in our series are from multiple treatment options for LUTS due to BPH (transurethral microwave thermotherapy, prostate urethral lift, different injection approaches, etc), they could not be analyzed in comparison to the sham group as the treatment group was inherently dissimilar. Instead these results are a summation and a Figure 2. Change in AUASS with endoscopic treatments Figure 3. Change in Qmax with endoscopic treatments

5 1686 CHANGES IN SHAM SURGERY ARMS OF BENIGN PROSTATIC HYPERPLASIA SURGERY TRIALS Figure 4. Change in AUASS with prostatic injection treatments statistical analysis of the changes from any sham procedure for LUTS due to BPH. Injection based techniques were analyzed separately in a different group since they included injection of a placebo, which could potentially act on the prostate by a mass effect. Alternatively needle insertion could lead to a change in the prostate parenchyma or capsule. These results may differentiate this approach from a more benign endoscopic sham procedure. There was large variability in how injections were administered in studies of prostatic injection due to the range of the number of injections. In addition in 1 series 1 needle stick was administered but the needle was injected in a way that created 3 pockets. 5 Injections were also done by a transrectal 4 or transperineal 5 approach, or some combination. 6 For these reasons no additional attempt was made to stratify sham results from these studies based on the number of injections or the injection approach as they were too varied and too few studies existed. Due to the long period included in our review a variety of urinary SSs were reported by investigators. Boyarksy and Madsen-Iversen SSs were frequently used before AUASS became ubiquitous. These 2 scores use a different scale (0 to 27) than AUASS (0 to 35). In an effort to include SS data from these older studies we scaled the high end of these studies (27) to the high end of the AUASS (35). This scaling did little to change SS outcomes that were due to using AUASS alone. Similar decreases were seen when AUASS was used alone (e27%), and when Boyarsky and Madsen-Iversen SSs were scaled and then included with AUASS (e26%). The potentially largest limitation of this study may be the influence of natural regression to the mean for study participants. Briefly regression to the mean is a phenomenon by which the initially observed measurement of a patient is essentially an extreme value when it has been influenced by observation or study enrolment. When this initial value is later compared to more normalized subsequent measurements, there appears to be a decrease in the score that is simply a return to the more normal physiology or symptoms of the patient. In study participants this may be an even larger factor since at study entry men must report a significant enough SS to be included, that is they must have symptoms considerable enough to warrant treatment. Treatment or sham treatment may then simply decrease the reported patient outcomes closer to the actual patient physiology. Additionally we did not use study inclusion criteria to screen studies and studies may have included different thresholds for a minimum SS to qualify for inclusion. The statistical approach that we used assessed only overall changes in the means of these studies. As individual patient data were not universally available, our statistical measures required that we focus on these cohorts as a whole. While this is a subtle difference, these findings should only be used to assess the outcomes of an overall study of surgical outcomes and should not be used to assess the statistical significance of an individual patient outcome. Figure 5. Change in Qmax with prostatic injection treatments CONCLUSIONS Sham controlled, endoscopic and intraprostatic injection BPH related surgeries demonstrate a considerable and statistically significant change in SS and Qmax, which is often as pronounced as the true drug response seen in BPH medication trials. Future uncontrolled BPH surgery trials should incorporate these findings into the assessment of clinical improvements.

6 CHANGES IN SHAM SURGERY ARMS OF BENIGN PROSTATIC HYPERPLASIA SURGERY TRIALS 1687 REFERENCES 1. Roehrborn CG, Preminger G, Newhall P et al: Microwave thermotherapy for benign prostatic hyperplasia with the Dornier Urowave: results of a randomized, double-blind, multicenter, sham-controlled trial. Urology 1998; 51: de la Rosette JJ, de Wildt MJ, Alivizatos G et al: Transurethral microwave thermotherapy (TUMT) in benign prostatic hyperplasia: placebo versus TUMT. Urology 1994; 44: Albala DM, Fulmer BR, Turk TM et al: Officebased transurethral microwave thermotherapy using the TherMatrx TMx J Endourol 2002; 16: McVary KT, Roehrborn CG, Chartier-Kastler E et al: A multicenter, randomized, double-blind, placebo controlled study of onabotulinumtoxina 200 U to treat lower urinary tract symptoms in men with benign prostatic hyperplasia. J Urol 2014; 192: Elhilali MM, Pommerville P, Yocum RC et al: Prospective, randomized, double-blind, vehicle controlled, multicenter phase IIb clinical trial of the pore forming protein PRX302 for targeted treatment of symptomatic benign prostatic hyperplasia. J Urol 2013; 189: Marberger M, Chartier-Kastler E, Egerdie B et al: A randomized double-blind placebo-controlled phase 2 dose-ranging study of onabotulinumtoxina in men with benign prostatic hyperplasia. Eur Urol 2013; 63: Abbou CC, Colombel M, Payan C et al: The efficacy of microwave induced hyperthermia in the treatment of BPH: the Paris public hospitals experience. Prog Clin Biol Res 1994; 386: Brehmer M, Wiksell H and Kinn A: Sham treatment compared with 30 or 60 min of thermotherapy for benign prostatic hyperplasia: a randomized study. BJU Int 1999; 84: 292e Maria G, Brisinda G, Civello IM et al: Relief by botulinum toxin of voiding dysfunction due to benign prostatic hyperplasia: results of a randomized, placebo-controlled study. Urology 2003; 62: Nawrocki JD, Bell TJ, Lawrence WT et al: A randomized controlled trial of transurethral microwave thermotherapy. Br J Urol 1997; 79: McVary KT, Gange SN, Shore ND et al: Treatment of LUTS secondary to BPH while preserving sexual function: randomized controlled study of prostatic urethral lift. J Sex Med 2014; 11: Mulvin D, Creagh T, Kelly D et al: Transurethral microwave thermotherapy versus transurethral catheter therapy for benign prostatic hyperplasia. Eur Urol 1994; 26: Blute ML, Patterson DE, Segura JW et al: Transurethral microwave thermotherapy v sham treatment: double-blind randomized study. J Endourol 1996; 10: de Wildt MJ, Hubregtse M, Ogden C et al: A 12-month study of the placebo effect in transurethral microwave thermotherapy. Br J Urol 1996; 77: Larson TR, Blute ML, Bruskewitz RC et al: A highefficiency microwave thermoablation system for the treatment of benign prostatic hyperplasia: results of a randomized, sham-controlled, prospective, double-blind, multicenter clinical trial. Urology 1998; 51: Lepor H, Sypherd D, Machi G et al: Randomized double-blind study comparing the effectiveness of balloon dilation of the prostate and cystoscopy for the treatment of symptomatic benign prostatic hyperplasia. J Urol 1992; 147: Ogden CW, Reddy P, Johnson H et al: Sham versus transurethral microwave thermotherapy in patients with symptoms of benign prostatic bladder outflow obstruction. Lancet 1993; 341: Roehrborn CG, Gange SN, Shore ND et al: The prostatic urethral lift for the treatment of lower urinary tract symptoms associated with prostate enlargement due to benign prostatic hyperplasia: the L.I.F.T. Study. J Urol 2013; 190: Bdesha AS, Bunce CJ, Kelleher JP et al: Transurethral microwave treatment for benign prostatic hypertrophy: a randomised controlled clinical trial. BMJ 1993; 306: Zerbib M, Steg A, Conquy S et al: Localized hyperthermia versus the sham procedure in obstructive benign hyperplasia of the prostate: a prospective randomized study. J Urol 1992; 147: Barry MJ, Williford WO, Chang Y et al: Benign prostatic hyperplasia specific health status measures in clinical research: how much change in the American Urological Association symptom index and the benign prostatic hyperplasia impact index is perceptible to patients? J Urol 1995; 154: Roehrborn CG: Efficacy and safety of once-daily alfuzosin in the treatment of lower urinary tract symptoms and clinical benign prostatic hyperplasia: a randomized, placebo-controlled trial. Urology 2001; 58: Kawabe K, Yoshida M, Homma Y et al: Silodosin, a new alpha1a-adrenoceptor-selective antagonist for treating benign prostatic hyperplasia: results of a phase III randomized, placebocontrolled, double-blind study in Japanese men. BJU Int 2006; 98: Lepor H: Phase III multicenter placebo-controlled study of tamsulosin in benign prostatic hyperplasia. Tamsulosin Investigator Group. Urology 1998; 51: Oelke M, Giuliano F, Mirone V et al: Monotherapy with tadalafil or tamsulosin similarly improved lower urinary tract symptoms suggestive of benign prostatic hyperplasia in an international, randomised, parallel, placebocontrolled clinical trial. Eur Urol 2012; 61: Wang X, Wang X, Li S et al: Comparative effectiveness of oral drug therapies for lower urinary tract symptoms due to benign prostatic hyperplasia: a systematic review and network meta-analysis. PLoS One 2014; 9: e Kaplan SA, Roehrborn CG, Rovner ES et al: Tolterodine and tamsulosin for treatment of men with lower urinary tract symptoms and overactive bladder: a randomized controlled trial. JAMA 2006; 296: Abrams P, Speakman M, Stott M et al: A dose-ranging study of the efficacy and safety of tamsulosin, the first prostate-selective alpha 1A-adrenoceptor antagonist, in patients with benign prostatic obstruction (symptomatic benign prostatic hyperplasia). Br J Urol 1997; 80: 587.