3 Clinical evidence

3 Clinical evidence

Summary of clinical evidence

3.1 The key clinical outcomes for the UroLift system presented in the decision problem were:

  • length of hospital stay

  • need for, or duration of, catheterisation

  • number of follow‑on consultations after discharge, both in primary and secondary care

  • re‑operation rates and time to re‑operation

  • symptoms of benign prostatic hyperplasia (using International Prostate Symptom Score [IPSS])

  • reduction in ejaculatory or sexual function

  • time to return to normal activities

  • quality of life

  • healthcare‑associated infection

  • device‑related adverse events.

3.2 The company's submission of clinical evidence was based on a recently published systematic review, Perera et al. (2014). It did not carry out additional literature searches or evidence synthesis. The company also submitted 2‑year follow‑up results (Roehrborn et al. 2014) for a study included in the review (Roehrborn et al. 2013).

3.3 The External Assessment Centre carried out an independent search of the literature and did not identify any relevant studies other than the 9 used in the systematic review (Cantwell et al. 2013, Chin et al. 2013, Delongchamps et al. 2012, Roehrborn et al. 2013, McVary et al. 2014, McNicholas et al. 2013, Shore et al. 2014, Woo et al. 2011, Woo et al. 2012). It found 1 study, Abad et al. (2013), which had been excluded from the systematic review. However, having reviewed its English abstract, the External Assessment Centre considered the findings (which were published in Spanish) to be potentially relevant and obtained a translation. It excluded the study by Delongchamps et al. (2012) because it was published in French and included only 4 patients. The External Assessment Centre noted that 2 studies, Chin et al. (2012) and Woo et al. (2012), reported results from the same case series, and that 3 studies, Roehrborn et al. (2013, 2014) and McVary et al. (2014), reported results from the LIFT study.

3.4 Perera et al. (2014) reported a systematic review and meta‑analysis of studies which used the UroLift system. The authors conducted a literature search for studies involving the use of the UroLift system and identified 61, of which 23 conference proceedings and 28 editorials were excluded from further analysis. The remaining 10 studies included 2 published papers on a randomised controlled trial (the LIFT study; McVary et al. 2014 and Roehrborn et al. 2013) and 8 uncontrolled before‑and‑after studies (Abad et al. 2013, Cantwell et al. 2014, Chin et al. 2012, Delongchamps et al. 2012, McNicholas et al. 2013, Shore et al. 2014, Woo et al. 2011 and 2012). These 10 studies were included in the meta‑analysis.

3.5 The meta‑analysis used data from 88–1298 responses (depending on the score) obtained from 452–680 patients. The results presented the outcome as a compound 'prostate symptom score', which comprised the combined values of IPSS and Benign Prostatic Hyperplasia Impact Index (BPHII). This made it difficult to identify how much IPSS had changed. In the abstract, Perera et al. presented an improvement in IPSS of −8.0 points (95% confidence interval [CI] −8.8 to −7.2) at 12‑month follow‑up. However the External Assessment Centre noted this value was obtained from the pooled prostate symptom score and represents a smaller IPSS improvement than any individual publication included in the meta‑analysis. The External Assessment Centre stated that this method of calculation was unwarranted when actual IPSS values, means and overall changes could have been reported instead. The External Assessment Centre's own calculations of weighted mean IPSS indicated an improvement in IPSS of −11 points. The meta‑analysis results also showed quality of life measurements improved by 2.2–2.4 points, and sexual health scores showed a small improvement of 0.3–0.4 points.

3.6 The LIFT study was a randomised controlled trial designed to evaluate the safety and effectiveness of the UroLift system when used in men with symptomatic benign prostatic hyperplasia. It was patient‑blinded and the comparator used was sham cystoscopy. The primary end point was IPSS reduction in the active arm at least 25% more than that in the control arm. The trial was conducted at 19 centres across the USA (14), Canada (2) and Australia (3) in men aged 50 years or older, with prostate volumes of 30–80 ml and an IPSS greater than 12. Patients were randomised 2:1 in favour of the intervention group, resulting in 140 men having the UroLift system and 66 having sham cystoscopy.

3.7 Roehrborn et al. (2013) reported the results of the LIFT study and noted that the primary end point was met at 3 months. After 12 months, IPSS, quality of life, peak urinary flow rate (Qmax) and BPHII score were all improved in patients who had the UroLift system, compared with their baseline measurements. McVary et al. (2014) reported specifically on the preservation of sexual function in patients in the LIFT study, and recorded sexual health outcome scores using the sexual health inventory for men (SHIM) and male sexual health questionnaire for ejaculatory dysfunction (MSHQ‑EjD). The results showed that using the UroLift system improved lower urinary tract symptoms and urinary flow without compromising sexual function. There was no evidence of erectile or ejaculatory dysfunction in patients treated using the UroLift system. There was no difference in SHIM or MSHQ‑EjD scores at 3 months compared with their baseline values, but these scores improved and were statistically significantly different from baseline after 1 year. An average of 4.9 UroLift implants was used per patient.

3.8 At 2‑year follow‑up, Roehrborn et al. (2014) reported a mean 42%±7.6% decrease in IPSS (95% CI −48.5% to −35.4%) compared with baseline measurements. Similar improvements were also reported in BPHII score and quality of life. Sexual health outcomes measured by SHIM and MSHQ‑EjD scores indicated that improvements in sexual function were preserved throughout the second post‑operative year. Within 2 years of first having the UroLift system, 7.5% of patients had a further procedure to treat lower urinary tract symptoms; 5 had further UroLift implants and 5 had transurethral resection of the prostate (TURP) or holmium laser enucleation (HoLEP). All second procedures were done with no complications from the initial UroLift procedure. Cantwell et al. (2013) reported the subsequent use of the UroLift system in men who had been assigned to the control arm of the LIFT study. Of the 66 men who first had the sham procedure, 53 (80%) chose further treatment with the UroLift system after unblinding (mean age=68 years, mean prostate volume=40.3 cm3). Results showed that the UroLift system was statistically significantly more effective than the sham procedure, producing a mean reduction in IPSS at 12 months of 37% (95% CI −46% to −27%). The authors reported that sexual function was maintained, with no notable deterioration after use of the UroLift system: in fact, the UroLift system caused a statistically significant improvement in ejaculatory function at 3 months compared with sham procedure. Adverse events were reported as mild to moderate and no blood transfusions were needed. The authors noted the symptomatic relief, low morbidity and preservation of sexual function associated with the UroLift system.

3.9 During consultation, the company submitted a conference abstract (Roerhborn et al. 2015) reporting 3‑year follow‑up results from the LIFT study. IPSS improvement was 43% after 3 years compared with patients who had sham (p<0.0001). Fewer patients were reported at 3‑year follow‑up (n=62) than at 2 years (n=104) but the abstract did not describe reasons for drop‑outs at any time point. However, the results and adverse events remained consistent with those collected for the 2‑year follow‑up. Twelve subjects (8.6%) had a secondary procedure over the 3‑year period. Sexual function outcomes were not reported in detail in this abstract.

3.10 Abad et al. (2013) was an uncontrolled case series of 20 men (mean age=74.3 years, mean prostate volume=42.6 cm3) treated in Spain. At 1‑month follow‑up, IPSS fell by 37.5% and peak Qmax increased from 8.6 ml/s to 13.2 ml/s. The authors noted no cases of urinary incontinence or sexual dysfunction. Minor complications included transient dysuria (70%) and urgency (40%), and slight haematuria (30%). Two patients (10%) needed post‑operative catheterisation. The authors stated that longer follow‑up times and larger patient numbers were needed before conclusions could be made on the safety and efficacy of the technology. An average of 3.8 UroLift implants was used per patient.

3.11 Chin et al. (2012) and Woo et al. (2012) both reported on the same Australian multicentre study of 64 men (mean age=66.9±7.3 years, mean prostate volume=51±23 cm3) with moderate to severe lower urinary tract symptoms. The authors reported improvements following use of the UroLift system, using an average of 4 implants per procedure. At 2‑year follow‑up, IPSS had decreased by 42% in the entire population (95% CI −54% to −31%); at 3 years, some patients continued to show a 34% symptomatic improvement. Similar improvements were shown in BPHII and quality of life. Results were statistically significant for all of these outcomes at all time intervals. No decrease in sexual function was observed, and the MSHQ‑EjD showed significant improvements at some intervals. Adverse events were minor, such as dysuria and haematuria, and typically resolved within 1 week. No blood transfusions were needed. Cystoscopic follow‑up at 6 months (n=22) showed no evidence of encrustation or infection. Post‑operative catheterisation rate was 53% (for a median of 20 hours). After 2 years, reoperation rate was 20% using TURP, a repeat prostatic urethral lift or photoselective vaporisation of the prostate.

3.12 McNicholas et al. (2013) reported an uncontrolled study of 102 men with symptomatic benign prostatic hyperplasia (mean age=68 years, mean prostate size=48 cm3, mean IPSS=23.2), which was done in 7 centres across 5 countries. IPSS, quality of life, BPHII, Qmax and adverse event reports, including sexual function, were used as outcome measures. The authors reported that all procedures were completed successfully with a mean of 4.5 implants per patient. Mean statistically significant improvements at 12 months were noted in IPSS (52%), quality of life (53%) and Qmax (51%). Adverse events were mild and transient, with no reported loss of antegrade ejaculation. During the follow‑up period 6.5% of patients progressed to having TURP without complication. The authors noted the potential advantages of the UroLift system, including its minimally invasive nature, the avoidance of retrograde ejaculation, symptomatic improvement, and the fact that it can be performed under local anaesthesia.

3.13 Shore et al. (2014) reported a prospective non‑randomised study of 51 patients having the UroLift system, with particular emphasis on their experiences (mean age=66±7.6 years, mean prostate volume=41.3±11.6 cm3, mean IPSS=21.5±5.4). Average procedure time was 52±22 minutes, with an average of 3.7 implants per procedure. All procedures were done as day procedures. Post‑operative catheterisation was needed in 20% of patients and mean duration was 16 hours. Follow‑up was 1 month. Outcomes included IPSS, quality of life, BPHII, Qmax, post‑void residual, SHIM and MSHQ‑EjD. In addition, the study looked at quality of recovery, work productivity and activity impairment. IPSS improved by an average of −47.5% at 1‑month follow‑up (95% CI −56.4% to −38.5%). Average number of days before return to work was 2.8±3.7; 73% of patients did not miss any work days. There were no serious adverse effects and no reported cases of sexual dysfunction. Ejaculatory function, measured by MSHQ-EjD score, showed statistically significant improvement at 1 month (10.3±2.6 at baseline to 11.9±3.1) after using the UroLift system. There was no statistically significant change in erectile dysfunction after 1 month.

3.14 Woo et al. (2011) reported a case series of 19 patients in Australia with benign prostatic hyperplasia (mean prostate volume=49 cm3) who had the UroLift system. The objective of the study was to assess the safety and efficacy of the technology. All procedures were performed successfully, with a post‑operative catheterisation rate of 58%. Some minor side effects were reported (haematuria, dysuria and irritation) but all resolved within a month. No retrograde ejaculation was reported. At 12‑month follow‑up, 4 patients had had TURP. The authors noted that IPSS improvement was highest at 3 months after the UroLift procedure (57% reduction). They also noted that there was no statistically significant change in Qmax or post‑void residual.

3.15 During consultation, the results of the BPH6 trial (Sønksen et al. 2015) became available as an in‑process document (Clinicaltrials.gov identifier: NCT01533038). This study most closely matches the scope for this evaluation because it directly compares UroLift with TURP as part of a randomised, multicentre clinical trial. The report described outcomes in patients using the composite BPH6 end point at 12 months; the External Assessment Centre noted that these end points are well justified and supported by published sources. There were no statistically significant differences in baseline parameters except for the MSHQ‑EjD function score. The UroLift arm experienced a significant improvement in MSHQ‑EjD from baseline (p=0.03) whereas the TURP arm experienced a significant deterioration (p<0.0001). The UroLift system did not cause any adverse events that needed surgical intervention or revision but further intervention was needed in 2 patients (6%) in the TURP group. patients having the UroLift system also experienced fewer treatment‑related infections (7%) than patients having TURP (14%; p=0.46).

External assessment centre synthesis and results

3.16 The External Assessment Centre considered that the meta‑analysis in Perera et al. (2014) demonstrated the clinical effectiveness of the UroLift system. However, it stated that the reporting of the results from the meta‑analysis was not clear, particularly in the reporting of the methodologies used, patient numbers, and how pooled effect sizes based on multiple outcomes were used.

3.17 In order to address this, the External Assessment Centre presented data for each reported outcome measure as changes from baseline and used weighted means to account for cohort sizes. Outcomes were reported at 1, 3, 12 and 24 months follow‑up, where available, from the 7 studies identified (LIFT [Roehrborn et al. 2013, 2014; McVary 2014], Cantwell et al. 2013, Abad et al. 2013, combined cohort study [Chin et al. 2012, Woo et al. 2012], McNicholas et al. 2013, Shore et al. 2014 and Woo et al. 2011). The External Assessment Centre also provided context for the results by identifying clinically important effect sizes for each outcome measure. It obtained this information using published and validated differences where available, such as for IPSS; others were based on expert advice.

3.18 The External Assessment Centre noted that at the time there was no published evidence directly comparing the UroLift system with the specified comparators. Because of this, it selected a recent systematic review, Li et al. (2014), which reviewed studies comparing TURP with HoLEP from which to derive effect sizes for the comparator interventions. The External Assessment Centre considered the patient populations in these studies to be broadly similar to those found in studies of the UroLift system, with patient ages and IPSS baselines falling within the same range. Prostate volumes were more varied in the studies of TURP and HoLEP, but skewed towards larger prostates and slower flow rates than the studies on the UroLift system.

3.19 The External Assessment Centre conducted an evidence synthesis of the outcomes in these studies. It presented changes in baseline outcomes for the comparators similar to the calculations that were done for the UroLift system (table 1). Results showed that both TURP and HoLEP were associated with greater improvements in IPSS than the UroLift system at all time points (−17.34 to −19.7 with TURP and −17.68 to −20.88 with HoLEP, compared with −9.22 to −11.82 with the UroLift system). Qmax and post‑void residual improvements were also higher with TURP and HoLEP.

3.20 The External Assessment Centre noted that sexual function was poorly reported on in the publications on TURP and HoLEP. This is likely to be because impaired sexual function is a well‑recognised complication of these procedures and so may not be explicitly reported. Clinical advice suggested that 5% of patients having TURP will develop erectile dysfunction, and that 70–80% will experience retrograde ejaculation. One clinical expert stated that the GOLIATH study (Bachmann et al. 2015) reported reliable data on erectile function for TURP. Results of the study (n=119) reported an erectile dysfunction score (based on the international index for erectile dysfunction) of 13.7±7.2 at baseline, rising to 14.1±8.2 at 12 months after having TURP (indicating a higher level of erectile dysfunction). This change was not significant. Another study reporting 6‑year follow‑up of HoLEP reported a retrograde ejaculation rate of 76% (Gilling et al. 2008).

3.21 The External Assessment Centre emphasised that the results of its evidence synthesis did not represent a direct comparison of the UroLift system with either TURP or HoLEP, and that patient populations may vary and outcome measures are dependent on original baseline scores. Nevertheless, it considered that this approach may give an idea of improvements from baseline and complications after TURP and HoLEP, presented in the same format as the UroLift system data.

3.22 During consultation, the External Assessment Centre reviewed the additional evidence (see sections 3.9 and 3.15) on the 3‑year LIFT study outcomes and on the in‑process publication of the BPH6 trial. It concluded that both were supportive of, and consistent with, the findings presented in the UroLift assessment report which formed part of the evidence presented to the Committee.

Table 1 Weighted means of outcome measures for the UroLift system and comparators

Minimally important change

Month

UroLift

(based on trials of the UroLift system alone)

TURP

(based on trials comparing TURP with HoLEP)

HoLEP

(based on trials comparing TURP with HoLEP)

IPSS

(Negative is improvement)

Minimum: 3

Moderate: 5.1

Marked: 8.8 1

1

−10.35

−17.34

−17.68

3

−11.82

−19.7

−20.88

12

−10.49

−18.13

−19.29

24

−9.22

−17.5

−20.4

IPSS QoL

(Negative is improvement)

Minimum: 1−3 1

1

−2.27

−2.99

−2.64

3

−2.48

−2.8

−3

12

−2.31

−3.18

−3.24

24

−2.22

N/A

N/A

BPHII

(Negative is improvement)

Minimum: 0.5

Moderate: 1.1

Marked: 2.2 1

1

−3.29

N/A

N/A

3

−3.96

12

−3.95

24

−3.76

IIEF

(Positive is improvement)

Minimum: 4 2

1

+0.52

N/A

N/A

3

+1.34

12

+0.8

24

N/A

MSHQ‑EjD

(Negative is improvement)

Minimum: 1.5 2

1

+1.82

N/A

N/A

3

+1.47

12

+0.83

24

N/A

MSHQ‑Bother

(Negative is improvement)

Minimum: 1 2

1

−0.67

N/A

N/A

3

−0.79

12

−0.91

24

N/A

Q max

(Positive is improvement)

Minimum: 2 ml/s 3

1

+4.16

+14.58

+15.29

3

+3.78

+14.11

+18.25

12

+3.52

+16.69

+17.78

24

+4.15

+23.20

+23.1

PVR

(Negative is improvement)

Minimum: 50 ml 2

1

−7.0

−137.43

−160.23

3

−10.34

−89.34

−78.0

12

−5.72

−127.29

−161.47

24

N/A

−196.1

−231.4

1 Barry et al. 1995

2 Clinical expert opinion

3 NICE guidance on lower urinary tract symptoms

Abbreviations: BPHII, benign prostatic hyperplasia impact index; HoLEP, holmium laser enucleation of the prostate; IIEF, international index of erectile dysfunction; IPSS, international prostate symptom scores; MSHQ‑Bother, male sexual health questionnaire for bother; MSHQ‑EjD, male sexual health questionnaire for ejaculatory dysfunction; PVR, post‑void residual; Qmax, peak urinary flow rate; QoL, quality of life; RCT, randomised controlled trial; TURP, transurethral resection of the prostate.

Adverse events and complications

3.23 The company searched the Medicines and Healthcare Products Regulatory Agency (MHRA) website and the US Food and Drug Administration Manufacturer and User Facility Device Experience (MAUDE) database to identify reports of adverse events relating to the UroLift system. None were found and the External Assessment Centre confirmed this.

3.24 The External Assessment Centre emphasised the difference in reporting complications between studies using the UroLift system and those using the 2 comparators. For example, mild adverse events such as transient dysuria and haematuria are commonly reported with the UroLift system. However, because these events are recognised as common with TURP and HoLEP, they are typically not reported in the literature.

3.25 Nevertheless, the External Assessment Centre considered that some comparisons can be made between the procedures in terms of complications and adverse events. The External Assessment Centre reported that incontinence was less prevalent with the UroLift system (5%) compared with TURP (11%) and HoLEP (14%). Reoperation rates were higher with the UroLift system (8%, weighted mean of all studies, 95% CI 3% to 14%) than with TURP (6%) and HoLEP (4%). Follow‑up intervals varied, up to a maximum of 2 years.

3.26 The External Assessment Centre stated that it was difficult to compare catheterisation rates between the different procedures because policies vary between centres; for example, some hospitals are reported to catheterise after the UroLift procedure as a matter of course. Mean post‑procedure catheterisation times were shorter for the UroLift system (22.3 hours) than for TURP (62.7 hours) and HoLEP (44.2 hours), based on Li et al. (2014) and Perera et al. (2014). A clinical expert advised the External Assessment Centre that after both TURP and HoLEP, a urinary catheter is inserted in part to irrigate the bladder, which can increase the patient's recovery time. Another expert stated that the UroLift system removes the need for both catheterisation and the use of irrigation fluid after the operation.

3.27 One study (reported by both Chin et al. 2012 and Woo et al. 2012) included data on implant encrustation, which occurs when UroLift implants are placed too close to the bladder and exposed to static urine. The External Assessment Centre sought clinical advice on this issue because of the lack of long‑term data. Three experts described encrustation as a significant issue, 2 considered it to be insignificant and a sixth was unsure. All except 1 of the experts advised that removing encrusted implants is a simple procedure.

3.28 The External Assessment Centre also sought clarification on any potential difficulties the implants may cause if a patient later has TURP. Clinical experts who have done TURP for patients with UroLift implants advised the External Assessment Centre that implants do not impede any later procedure, including TURP and HoLEP, and that no alteration of the standard surgical techniques is needed.

Committee considerations

3.29 When developing its provisional recommendations, the Committee noted that there was no published evidence comparing the UroLift system with the comparators specified in the scope. It considered that the External Assessment Centre's evidence synthesis was useful in showing that UroLift was slightly less effective in terms of improving IPSS compared with TURP and HoLEP, but nevertheless achieved a marked improvement in symptoms. After consultation, the Committee noted that the results of a recent comparative trial (BPH6) of UroLift against TURP were similar to those in the External Assessment Centre's evidence synthesis, and supported its interpretation of the comparative effectiveness of UroLift and TURP.

3.30 The Committee noted the consistent evidence that using the UroLift system does not damage sex function: it considered this to be an important advantage for many men. It noted small improvements in sexual function in some reports. The Committee accepted that damage to sexual function is not commonly reported in studies of TURP and HoLEP, because it is often regarded as an inevitable consequence of the procedures.

3.31 The Committee considered that despite some limitations, the clinical evidence was sufficient to demonstrate that the UroLift system provides clinical and patient advantages as an option for treating symptoms of benign prostatic hyperplasia with a lower risk of important complications.

3.32 The Committee discussed patient selection for treatment using the UroLift system. It noted that the instructions for use state that the device is indicated for men with benign prostatic hyperplasia who are aged 50 years or older and who have a prostate of less than 100 ml. The clinical experts stated that appropriate prostate shape is also important; in particular, a prostate with a hypertrophic median lobe would preclude the use of the UroLift system.

3.33 The Committee was advised that the UroLift system would be appropriate for up to 1 in 4 men needing surgery for lower urinary tract symptoms of benign prostatic hyperplasia. It heard advice from the clinical experts that the UroLift system is most useful for men who wish to preserve sexual function. It also has special advantages for men with blood clotting disorders and for men for whom general anaesthetic would be unsuitable.

3.34 The Committee recognised that there was no clinical evidence which included follow‑up beyond 3 years. The evidence showed that benefit is maintained up to 3 years, but the Committee heard expert advice that the UroLift system may not offer permanent relief of symptoms. Available information on re‑operation rates for the UroLift system suggests that they are similar to those for TURP and HoLEP. The clinical experts also stated that data for re‑operation rates could not be directly compared between the UroLift system and its comparators, because the need to re‑operate often represents the regrowth of prostate tissue, which is cut away as part of TURP and HoLEP. The UroLift system works in a different way, retracting prostatic tissue rather than removing it.

3.35 The Committee noted that no quality of life data were presented, but it considered that the UroLift system was very likely to result in improved quality of life for men in terms of preserving sexual function. The Committee was also advised by the clinical experts that men having UroLift implants typically return to full activity sooner than those having TURP or HoLEP.

3.36 The Committee noted that the NICE interventional procedure guidance on insertion of prostatic urethral lift implants to treat lower urinary tract symptoms secondary to benign prostatic hyperplasia (a procedure that uses the UroLift system) recommends further research and the publication of results from consecutive case series of patients having the procedure. The Committee concurred with this recommendation, in particular with regard to obtaining more evidence on the effects of UroLift implants on symptoms and quality of life, the duration of benefit, and the need for further procedures in the longer term.

  • National Institute for Health and Care Excellence (NICE)