Different lasers in the treatment of benign prostatic hyperplasia: a network meta-analysis

All available surgical treatments for benign prostatic hyperplasia (BPH) have their individual advantages or disadvantages. However, the lack of head-to-head studies comparing different surgeries makes it unavailable to conduct direct analysis. To compare the efficacy and safety among different lasers and transurethral resection of prostate (TURP) for BPH, randomized controlled trials were searched in MEDLINE, EMBASE, Cochrane library, WHO International Clinical Trial Registration Platform, and Clinical Trial.gov by 2015.5; and the effectiveness-, perioperation- and complication-related outcomes were assessed by network meta-analysis. 36 studies involving 3831 patients were included. Holmium laser through resection and enucleation had the best efficacy in maximum flow rate. Thulium laser through vapo-resection was superior in improving international prostate symptom score and holmium laser through enucleation was the best for post-voiding residual volume improvement. Diode laser through vaporization was the rapidest in removing postoperative indwelling catheter, while TURP was the longest. TURP required the longest hospitalization and thulium laser through vapo-resection was relatively shorter. Holmium and thulium lasers seem to be relatively better in surgical efficacy and safety, so that these two lasers might be preferred in selection of optimal laser surgery. Actually, more large-scale and high quality head-to-head RCTs are suggested to validate the conclusions.

Among reported lasers, Nd:YAG was the earliest laser used in the treatment of BPH 3 . Almost all studies investigated laser as monotherapy compared with standard treatment (TURP), and only two studies applied combined strategy-KTP plus Nd:YAG. Green light laser technique was the most commonly used technology and was reported in 13 studies, and seven studies exactly recorded the wavelength (532 nm). Studies varied in ways of treatment, surgical techniques and publication years (from 1995 to 2015); however, the baseline characteristics of all patients were basically similar among interventions. The main surgical technique of lasers was vaporization; enucleation and resection were mainly conducted by Holmium laser. One study stopped early due to the need for prolonged catheterization and a high rate of urinary tract infection 4 .
Risk of bias in included studies was summarized graphically in Fig.2B. One study had high risk of bias for sequence allocation and concealment, as it was reported as an open-label clinical trial 5 . Another trial was considered with high risk of other bias due to early discontinuation of study 4 . Since it was sometimes difficult to blind surgeons and patients, we did not include the blinding items of risk of bias in our analysis.
Short-term complications-related outcomes. Short-term complications-related outcomes, including dysuria, urinary retention, re-catheterization, clot retention, transfusion, incontinence, TURS, and UTI were analyzed. Total of nine studies reported dysuria, which was availably compared in five interventions through different techniques (holmium laser, green light laser, KTP/Nd:YAG, Nd:YAG laser and TURP).The network analysis showed that dysuria was the most common short-term complication in patients who underwent green light laser(vapo-enucleation) and holmium laser(enucleation), and the least common in Nd:YAG laser with vaporization. The relative effect estimates of Nd:YAG(vaporization) versus green light laser (vapo-enucleation) and holmium (enucleation) were 0.00 (0.00, 0.20) and 0.06 (0.00, 16.93), respectively. The Nd:YAG laser through vaporization technique was the most common technique inducing urinary retention. Besides, re-catheterization was always seen in patients underwent green light laser (vaporization) and diode laser (vaporization); and it was barely occurred in Nd:YAG laser (vaporization), basically corresponding with the outcome of dysuria but not urinary retention ( Fig. 5A-C). KTP/Nd:YAG (vaporization) and green light (vaporization) were mostly related with post-operation incontinence (Fig. 5D).
Transfusion was mostly seen in TURP but rarely in diode laser(vaporization) and thulium laser (vapo-resection); Nd:YAG laser (vaporization) and TURP had much higher rates of occurrence of clot retention, while holmium laser (enucleation) had the least rate (Fig. 6A,B). In addition, TURS was most seen in patients treated with TURP (Fig. 6C). While considering UTI, the most related technique was Nd:YAG laser through vaporization (Fig. 6D).
Long-term and other complications-related outcomes. Long-term complications-related outcomes, including bladder neck contracture or stenosis and urethral stricture or meatal stenosis, were analyzed in our network meta-analysis. Bladder neck contracture or stenosis was frequently occurred in KTP/Nd:YAG (vaporization) and holmium laser (enucleation); while green light (vaporization) and diode laser (vaporization)were on the contrary (Fig. 7A). The relative effect estimate of green light (vaporization) versus KTP/Nd:YAG (vaporization) was 0.47 (0.02, 5.05). Intriguingly, green light through vaporization plus enucleation had fewer bladder neck contracture or stenosis than vaporization alone. TURP was mostly associated with urethral stricture or meatal stenosis, and diode laser (vaporization) had the least incidence (Fig. 7B). Analysis of green light through different techniques showed that combination of vaporization and enucleation had less urethral stricture or meatal stenosis, compared to vaporization alone.

Discussion
TURP is always considered as the gold standard surgical treatment for patients with BPH; however, it is still associated with significant morbidity and mortality, such as TURS and transfusion. In the past two decades, both researchers and surgeons devoted to developing novel surgical treatments for pursuing a much better efficacy, as well as much more improvement of safety. Since Nd:YAG laser was firstly reported being used in the treatment of BPH by Costello and colleagues in 1992 43 , more lasers were introduced to treat BPH with improved surgical safety and efficacy. Afterwards, the improvement of technology led the appearance of innovative lasers applying in surgical practice of BPH. At present, the most often used lasers are holmium laser, green light laser and thulium laser with different surgical techniques, both in clinical practice and academic research 44 .
However, among diverse kinds of laser techniques, there is no completely or absolutely perfect intervention. Each laser has its advantages or disadvantages with different clinical outcomes. How to select the best surgical treatment for BPH is difficult but be of importance. Thus, it is worthy of comparing different surgical techniques, no matter direct or indirect studies. And to our knowledge, this is the first study, applying Bayesian analytical method of network meta-analysis, to indirectly compare the efficacy and safety of different lasers with TURP in the treatment of BPH. Based on our network meta-analysis, holmium laser achieved a better Q max and less PVR, and thulium laser achieved a lower IPSS than other lasers with different techniques and TURP. In the aspect of peri-operation-related outcomes, holmium laser through resection technique took the longest operating time and thulium laser (vapo-resection) ranked in the middle. It was also proven by a direct analysis that thulium laser required a longer operating time than holmium laser (72.4 vs 61.5 minutes, p = 0.034) 42 . However, both the two techniques had shorter time of indwelling catheter and time of stay in hospital resulting in a reduced incidence of complications. So, holmium laser and thulium laser showed a better surgical efficacy and had a higher safety with the least incidence of complications.
The resection and enucleation techniques of thulium and holmium lasers were seen as similar as open prostatectomy in the complete removal of the prostatic lobes 31,39,42 . The thulium laser could perform a smooth incision or vaporization by continuous mode 42 . This might decrease the occurrence of LUTS and improve IPSS after operation. The holmium laser always was characterized as scar-free disruption and its incision could reach the surgical capsule of prostate 31,39 . And this might increase the outflow of bladder and reduce the PVR. But all these advantages were still not enough to explain the better effectiveness of holmium laser and thulium laser techniques in improving patients' obstructive symptoms.
For peri-operation-related outcomes and the incidences of complications, green light laser through vapo-enucleation was more frequent associated with dysuria. Urinary retention was frequently seen in Nd:YAG laser alone or combined with KTP technique. Also, re-placement of catheter was mainly associated with green light laser and diode laser techniques. It was very likely that these two techniques could easily induce tissue edema and cause deficient effect when resecting the apex of prostate. Nd:YAG/KTP, diode laser, green light laser and TURP were observed with more complications (particularly green light laser) only with medium efficacies. As shown in the analysis, we compared six kinds of laser techniques and TURP. Nowadays, since the numerous postoperative complications of the initial laser procedures and the improvement of new equipment, some lasers had to be eliminated. Nd:YAG laser technique was firstly applied in the surgical treatment of BPH and is now abandoned in clinical practice, for its incapability in ablating prostate tissue immediately, postoperative frequency and urgency led by deep tissue necrosis and high rate of re-operation 45 . As show in the study, Nd:YAG technique achieved the worst surgical effect. In contrast to obsolete laser procedures, the current higher powered lasers-holmium laser, thulium laser, green light laser and diode laser-are able to ablate prostatic tissue rapidly. Holmium laser is optimized for incision, and green light laser is optimized for vaporization. Previously researches  showed that these two techniques had the superiorities over other lasers in the respect of functional outcomes, which were inconsistent with the results of our analysis 44 .
According to the results of our network meta-analysis, it seemed that the choice of laser techniques for BPH should be depended on what kinds of aims the patient wanted to get benefit, such as TURP may not be the best choice for patients with previous documented or suspected urethral stricture or meatal stenosis, as its incidence of postoperative urethral stricture or meatal stenosis was the highest.
There were some limitations in this network meta-analysis, such as included RCTs studied from 1995 to 2015, evident differences in sample sizes and significant transformation in techniques of the same lasers (green light laser improved form initial 80-W to 120-W, and subsequently to the current laser at 180-W) 5,7,8,9,10,13,18,22,24,25,26,35,40 . Although all authors were contacted to provide un-reported data for the integrity of analysis, the results of this study were restricted by incomplete data reported by included studies. Also, time-points of analyzed outcome measures were restrained to 12 months, for the limited data reported. But this review still enjoyed several advantages-rigidly analytical process by pre-published protocol, analyses of both lasers and TURP, and indirect comparisons in predicting efficacies of one technique to another due to the absence of head-to-head RCTs.
Furthermore, it might be encouraged for us to offer referable ideas for future researches and valuable advice for clinical surgeons by using this first comprehensive systematic review and network meta-analysis to assess the efficacy and safety of different laser techniques and TURP for BPH.
In conclusion, this is the first time to indirectly compare the efficacy and safety of different lasers with TURP for BPH by applying network meta-analysis. To date, no completely or absolutely perfect laser technique could be found to take the place of TURP in the surgical treatment of BPH. Holmium laser and thulium laser may seem (C) rank probability of re-operation. Thulium laser was not availably compared for outcome bladder neck contracture or meatal stenosis. All six kinds of lasers through different surgical techniquesand TURP were availably compared in the network meta-analysis of outcomes urethral stricture or meatal stenosis and re-operation.
to be relatively better in terms of surgical efficacy and safety, so that these two lasers might be preferred in the selection of optimal laser surgery. Actually, much longer-term, larger-scale and higher quality head-to-head RCTs are needed to validate the conclusions.

Methods
Protocol and registration. We developed a protocol defining the search strategy and a systematic review was performed to identify those randomized controlled trials (RCTs) investigating the efficacy and safety of different lasers or TURP for BPH. The review was registered on PROSPERO of the Centre for Reviews and Dissemination (CRD) (CRD42015024227). The data searching, study selection, quality assessment of included studies and data extraction were performed independently by two researchers (Z.X.M. and S.P.F.). Disagreements were resolved by discussing or with the help of a third investigator to reach the final consensus.
Eligibility and exclusion criteria. The eligibility criteria included: 1) Patients with BPH needed surgery intervention; 2) RCTs comparing different lasers and TURP (comparing different laser with one another), either laser alone or in combination with others, for the treatment of BPH; 3) Each laser may be through different surgical techniques:vaporization, resection of tissue pieces, enucleation, or combinations; 4) Published in English, German, French, Italian, Russian, Dutch, and Japanese.
Exclusion criteria: 1) Patients with neurogenic bladder disorder, urethral strictures, history of prostate adenocarcinoma or any previous prostatic, and bladder neck or urethral surgery. 2) Non-RCTs, reviews, reports only focusing on laboratory findings, trials published only as abstracts. 3) Not published in English, German, French, Italian, Russian, Dutch, and Japanese.
Outcome measures included: 1) Effectiveness-related outcomes: maximum flow rate (Q max), International Prostate Symptom Score (IPSS), post-void residual volume (PVR); 2) Perioperation-related outcomes: operating time, hospitalization and catheterization; 3) Complications-related outcomes: ① short-term complications-related outcomes (always occurred within one month after operation): dysuria, re-catherization, urine retention, clot retention, transfusion, transurethral resection syndrome (TURS), urine incontinence, and urinary tract infection (UTI); ② long-term complications-related outcomes (always occurred after one month of operation): bladder neck contracture or stenosis, urethral stricture or meatal stenosis; ③ other complication-related outcome: re-operation because of any other complications. Study selection and data extraction. Two investigators independently assessed the titles and abstracts of the searched results. The full text versions of those studies, which were potentially eligible, were then assessed. For the study design, selection criteria, participant's characteristics, interventions, outcome measures, study duration, results and other data of each included study were extracted. Extracted contents were recorded on data extraction forms, which were designed according to the advice given in the Cochrane Handbook 46 . We contacted authors to seek additional information where data were not reported or not clear.

Assessment of risk.
Two independent reviewers (Z.X.M. and S.P.F.) used RevMan 5 software to assess the risk of bias of all included studies according to the Cochrane Handbook 47 , as follows: 1). adequate sequence generation; 2). allocation concealment; 3). incomplete outcome data; 4). free of selective reporting; 5). free of other bias. The judgments were categorized as 'yes' (low risk' of bias), 'no' (high risk of bias) or 'unclear' (unclear risk of bias). Data synthesis. ADDIS software (version 1.16.6) was applied to conduct indirect comparison analysis (network meta-analysis to compare different interventions not directly matched). Network meta-analysis led us to predict the likely comparable estimates between indirect comparisons based on two or more studies with one common intervention. Using Bayesian approach, the relative effect estimates were calculated and could be used to estimate the probability that which one was the best. Node-split analysis was utilized to check inconsistency among comparisons, and p < 0.05 was set as significant inconsistency.