Critical analysis of the literature investigating urogenital function preservation following robotic rectal cancer surgery.

AIM: To analyses the current literature regarding the urogenital functional outcomes of patients receiving robotic rectal cancer surgery.
METHODS: A comprehensive literature search of electronic databases was performed in October 2015. The following search terms were applied: "rectal cancer" or "colorectal cancer" and robot* or "da Vinci" and sexual or urolog* or urinary or erect* or ejaculat* or impot* or incontinence. All original studies examining the urological and/or sexual outcomes of male and/or female patients receiving robotic rectal cancer surgery were included. Reference lists of all retrieved articles were manually searched for further relevant articles. Abstracts were independently searched by two authors.
RESULTS: Fifteen original studies fulfilled the inclusion criteria. A total of 1338 patients were included; 818 received robotic, 498 laparoscopic and 22 open rectal cancer surgery. Only 726 (54%) patients had their urogenital function assessed via means of validated functional questionnaires. From the included studies, three found that robotic rectal cancer surgery leads to quicker recovery of male urological function and five of male sexual function as compared to laparoscopic surgery. It is unclear whether robotic surgery offers favourable urogenital outcomes in the long run for males. In female patients only two studies assessed urological and three sexual function independently to that of males. In these studies there was no difference identified between patients receiving robotic and laparoscopic rectal cancer surgery. However, in females the presented evidence was very limited making it impossible to draw any substantial conclusions.
CONCLUSION: There seems to be a trend towards earlier recovery of male urogenital function following robotic surgery. To evaluate this further, larger well designed studies are required.

Core tip: Urogenital dysfunction is a significant problem following rectal cancer surgery that significantly affects quality of life. Despite laparoscopic total mesorectal excision becoming the standard approach in much of the developed world, the incidence of post-operative urogenital dysfunction remains high. Robotic surgery allows for precision surgery in the pelvis, therefore enabling better preservation of the pelvic autonomic nerves. Current studies examining the urogenital outcomes following robotic rectal cancer surgery have several limitations, but suggest that robotic surgery may offer favourable outcomes when compared to laparoscopic and open surgery. Larger scale prospective studies are required to validate these results.

INTRODUCTION

Colorectal cancer is one of the most common cancers in the developed world[1-3] with rectal cancers making up a third of those cancers[2-4]. The aim of rectal cancer surgery is to radically resect the cancer in order to achieve oncological cure and avoid local recurrence. During the past three decades significant improvements have been made to combat this predicament. These advances include earlier diagnosis, advanced surgical techniques and the improvement of adjuvant and neoadjuvant treatment[4-8]. These developments were not only aimed to improve the patients’ survival but also directed to improve the quality of life after cancer rectal surgery.

Urogenital function is one of the most important aspects of quality of life and rectal cancer may have adverse effects on it[5,9-13]. Although urogenital dysfunction is considered to be multifactorial, intra-operative damage to the pelvic autonomic nerves is the primary cause[14-16]. This is mainly due to the close proximity of the mesorectum to the autonomic nerves, and the difficulty in identifying such small structures such as the autonomic nerves in a narrow operative space such as the pelvis[13,17]. Damage to the sympathetic nerves results in urinary incontinence, ejaculation disorders in men and decreased orgasmic intensity in women[13,18]. Damage to the parasympathetic nerves leads to a lack of detrusor muscle function and subsequent voiding disorder, as well as erectile problems and lubrication dysfunction in men and women respectively[13,18]. These are significant post-operative and life changing events that jeopardise patients quality of life[9].

It is logical to assume that better visualisation of the structures of the pelvis, such as offered from laparoscopic or robotic surgery, can aid preservation of the autonomic nerves. Nevertheless, there is a debate as to whether laparoscopic surgery offers improved urogenital functional outcomes when compared to open surgery[19], as some studies have shown improved outcomes[20] while other advocate the contrary[21]. A probable reason for the disparate results is due to laparoscopic rectal surgery being technically difficult[22], as evident from its long learning curve[23] and the high conversion rate demonstrated in the CLASSICC and COLOR II trials[24,25]. Existing laparoscopic instruments have a restricted range of movement compared with that of the surgeons hand and are difficult to use in confined spaces such as the pelvis[26,27].

Robotic surgical systems were introduced to overcome the technical limitations of laparoscopic surgery[28]. They provide a superior three dimensional view, tremor filtering and superior ergonomic instrumentation[26,29]. These chattels enable precise dissection in narrow surgical fields such as the pelvis and help preserve the autonomic nerves. Even though multiple studies have examined the pathological, oncological and postoperative outcomes of robotic rectal surgery, there are only a few studies that have investigated the urological and sexual outcomes of robotic rectal cancer surgery and these tend to be predominantly about male patients.

Therefore the aim of this systematic review is to examine the available literature on the postoperative urogenital outcomes of robotic rectal cancer surgery on both male and female patients.

MATERIALS AND METHODS

A comprehensive literature search of electronic databases was performed in October 2015 by using the Discovery search engine tool (for more info refer to: http://www.port.ac.uk/library/infores/discovery/). Discovery is Portsmouth University’s search engine tool and it simultaneously searches over 200 scientific electronic databases including MEDLINE (PubMed), Google Scholar and Science Direct. The following search terms were applied: “rectal cancer” or “colorectal cancer” and robot* or “da Vinci” and sexual or urolog* or urinary or erect* or ejaculat* or impot* or incontinence. All original studies that reported the urological and/or sexual outcomes of patients having robotic rectal cancer surgery were included. Reference lists of all retrieved articles were manually searched for further relevant articles. A flow diagram of the selection process is given in Figure 1. Abstracts were independently searched by two authors. Fifteen full text articles fulfilled the inclusion criteria.

Figure 1

Selection process flow diagram.

RESULTS

Original studies

A total of 1338 patients were included in the reviewed studies (818 received robotic, 498 laparoscopic and 22 open rectal cancer surgery). The characteristics of all the original studies reporting either urinary or sexual outcomes are outlined in Tables 1 and 2. Of the 15 studies that met the inclusion criteria, 14 were cohort studies[5,6,9,18,30-39] and one a randomised control trial[40]. Nine of the cohort studies were comparing robotic rectal cancer surgery to either laparoscopic[9,30-33,35,38,40] or open[18] rectal cancer surgery.

Table 1

Characteristics of original studies

Ref.	Country	Study design	Control group	No. of cases for urogenital outcomes	Study specifically examines urogenital outcomes
Hellan et al[34]	United States	Retrospective	No control group	39	No
Patriti et al[40]	Italy	RCT	Robot vs lap	29 rob vs 37 lap	No
Luca et al[6]	Italy	Prospective	No control group	74	Yes
Kim et al[31]	South Korea	Prospective	Robot vs lap	30 rob vs 39 lap	Yes
Park et al[39]	United States	Prospective	No control group	30	No
Leung et al[5]	Hong Kong	Prospective	No control group	33	Yes
Park et al[32]	South Korea	Retrospective	Robot vs lap	14 rob vs 15 lap	No
D'Annibale et al[33]	Italy	Retrospective	Robot vs lap	30 vs 30	No
Stănciulea et al[37]	Romania	Retrospective	No control group	78	No
Erguner et al[38]	Turkey	Prospective	Robot vs lap	27 rob vs 37 lap	No
Park et al[9]	South Korea	Retrospective	Robot vs lap	32 vs 32	Yes
Ozeki et al[18]	Japan	Prospective	Robot vs open	15 rob vs 22 open	Yes
Cho et al[35]	South Korea	Retrospective	Robot vs lap	278 vs 278	No
Alecu et al[36]	Romania	Retrospective	No control group	79	No
Morelli et al[30]	Italy	Retrospective	Robot vs lap	30 vs 30	Yes

These include: (1) the studies country of origin; (2) the study design (prospective, retrospective or randomised control trial); (3) the control group (if present) used to compare with the robotic rectal surgery, this was either laparoscopic or open rectal surgery cases; (4) the number of cases included in each study whose urogenital outcomes were evaluated; and (5) whether the study was specifically designed to investigate the urogenital outcomes of robotic surgery or not. RCT: Randomised control trial; Robot: Robotic; lap: Laparoscopic.

Table 2

Further characteristics of original studies

Ref.	Fully or hybrid robotic procedure	Functional scores applied	Follow up in months	No. of surgeons performing cases	SIGN score
Hellan et al[34]	Hybrid	No	Median f/u 13 mo	Not stated	+
Patriti et al[40]	Hybrid	No	Mean f/u 12 mo	Not stated	+
Luca et al[6]	Fully	Yes	1, 6, 12	2 surgeons	++
Kim et al[31]	Hybrid	Yes	1, 3, 6, 12	1 surgeon	++
Park et al[39]	Reverse hybrid	No	Not stated	Not stated	+
Leung et al[5]	Mixture	Yes	3	Not stated	++
Park et al[32]	Hybrid	Yes	3, 6, 12	1 surgeon	++
D'Annibale et al[33]	Fully	Yes	1, 12	1 surgeon	++
Stănciulea et al[37]	93% fully	Yes	Once b/n 6 and 12 mo	3 surgeons	+
Erguner et al[38]	Mixture	No	Not stated	Not stated	+
Park et al[9]	Hybrid	Yes	3, 6, 12	1 surgeon	++
Ozeki et al[18]	Fully	Yes	3, 6, 12	2 for robot cases	++
Cho et al[35]	Fully	No	1	3 surgeons did 97.1% cases	++
Alecu et al[36]	Hybrid	Yes	Not stated	Not stated	+
Morelli et al[30]	Not stated	Yes	1, 6, 12	1 surgeon	++

These include: (1) whether the surgeons used the hybrid or robotic approach for their study; (2) whether urogenital function was assessed by means of functional scores or not; (3) the follow up period during which data for urogenital outcomes was collected; (4) the number of surgeons performing the cases in each study; and (5) the studies SIGN score. f/u: Follow up; SIGN: Scottish Intercollegiate Guidelines Network.

Out of the 15 studies only six[5,6,9,18,30,31] were specific to urogenital outcomes; the rest reported urogenital outcomes amongst a multitude of outcomes examined in those studies.

Outcome assessment

Functional questionnaire scores were used in ten[5,6,9,18,30-33,36,37] of these studies to access the urological and sexual function of patients. These questionnaires are validated tools that have been used in a multitude of previous studies to access urinary and sexual function in males and females[41-45]. Out of the 1338 patients included in this review, only 726 (54%; 442 robotic, 262 laparoscopic, 22 open) had their urogenital function assessed via functional questionnaires.

To assess male urological function the majority of studies used the International Prostatic Symptoms Score (IPSS) or a slight modification of it. This is a subjective scoring system examining seven categories[41]. These include incomplete bladder emptying, frequency, intermittency, urgency, weak stream, straining and nocturia. Patients score each category and assign a higher score for increasing severity of symptoms. Alternative questionnaires used to assess urological function were the the International Consultation on Incontinence Questionnaire - Male Lower Urinary Tract Symptoms[44], and the International Consultation on Incontinence Questionnaire - Female Lower Urinary Tract Symptoms[45] questionnaire.

Male sexual function was assessed in ten studies by the international index of erectile function (IIEF)[42] score. The IIEF is a 15-item score that analyses five factors: Erectile function, orgasmic function, libido, intercourse satisfaction and overall satisfaction. Unlike the IPSS score for urinary function, a high IIEF score is associated with good sexual function and the lower the IIEF score the greater the degree of sexual dysfunction.

Female sexual function was assessed in three studies[6,30,37] via the Female Sexual Function Index (FSFI)[43]. This is a validated questionnaire that is in many ways the female version of the IIEF questionnaire.

The studies that did not use validated scoring tools to assess functional outcomes simply reported the incidence of dysfunction. The limitations present in this method of reporting are the inability to quantify dysfunction and the difficulty in defining what makes a case.

Finally, one study[31] assessed urological function by performing urodynamic studies as well as using a validated functional questionnaire, making it the only study to report urinary outcomes with both subjective and objective measurement tools.

Pre-operative assessment and follow up

The studies assessing functional outcomes via validated questionnaires asked their participants to fill the questionnaires pre-operatively in order to establish their baseline urogenital function. In this way post-operative scores were assessed against the pre-operative scores for each patient, allowing the change of function from baseline to be assessed. Reporting the change of function from baseline is a more accurate way of assessing the impact of the intervention, rather than reporting the postoperative functional scores alone.

It was unclear across several of the studies[6,18,30,32] how many patients were sexually inactive pre-operatively and whether they were included in the analysis. Adding sexually inactive patients in the analysis will result in skewing of the data and it is therefore important to report how many patients were sexually inactive and whether they were included in the analysis or not.

In contrast to the studies applying validated functional scores, most of the studies that simply reported the incidence of urogenital dysfunction did not mention the pre-operative state of their participants. This makes it difficult to assess whether any cases of dysfunction became cases because of the intervention or not.

Follow up was fairly variable between the different studies and the follow up intervals for each study are summarised in Table 2. The majority of the studies followed up their patients in more than one occasion following surgery. The commonest follow up intervals were 3, 6 and 12 mo post-operatively.

Quality of included original studies

The Scottish Intercollegiate Guidelines Network critical appraisal tool for cohort studies was used to evaluate the original studies included in this review. However, none of the studies met the majority of the criteria for a high quality study. Most of the studies fell between the acceptable and low quality bracket (Table 2). The majority of studies were retrospective in nature, included a small number of patients, were subject to selection bias in terms of patient selection and made no adjustments for confounding factors.

The studies included in this review have significant differences in terms of outcome reporting and methodology. In addition, almost all of them are non-randomised in nature. Considering this and because of the heterogeneity of the data in these studies it was not appropriate to perform a meta-analysis. There are only a few studies whose data were homogeneous enough to permit a meta-analysis. However, this has already been performed by two previous systematic reviews[46,47] which combined the data of three studies. We discuss these systematic reviews in our discussion.

Male urological function

Out of the 15 original studies included, 12 studies reported male urological functional outcomes. The characteristics of these studies plus a summary of their results are present in Table 3.

Table 3

Original studies reporting male urological function

Ref.	Males assessed independently of females	Functional scores applied	Control group	No. of cases examining male urological function	Follow up in months	Outcome summary
Kim et al[31]	No	Yes	Robot vs lap	30 rob vs 39 lap	1, 3, 6, 12	Urological function recovered faster in robotic group (3 mo vs 6 mo) IPSS change from baseline lower in robotic group at 3 mo (P = 0.036) Mean voiding volume deterioration lower in 3 and 6 mo in robotic group (P = 0.007, P = 0.049) Similar outcomes at 12 mo in both groups
Park et al[9]	Yes	Yes	Robot vs lap	32 vs 32	3, 6, 12	IPSS scores elevated post-operatively in both groups At 12 mo IPSS change from baseline lower in robotic group but non-significant (P = 0.051)
Park et al[32]	Yes	Yes	Robot vs lap	14 rob vs 15 lap	3, 6, 12	Deterioration of IPSS scores in 3 mo which recovered by 6 mo in both groups
D'Annibale et al[33]	Yes	Yes	Robot vs lap	30 vs 30	1, 12	Deterioration of IPSS scores in 3 mo which recovered by 12 mo in both groups
Ozeki et al[18]	Yes	Yes	Robot vs open	15 rob vs 22 open	3, 6, 12	No statistical deterioration of IPSS scores in either group
Morelli et al[30]	Yes	Yes	Robot vs lap	Not available	1, 6, 12	Voiding and incontinence worse 1 mo in both groups, incontinence recovered by 6-12 mo in both groups
Leung et al[5]	Yes	Yes	No control group	33	3	No significant male urological function deterioration
Luca et al[6]	Yes	Yes	No control group	38	1, 6, 12	No significant male urological function deterioration
Stănciulea et al[37]	No	Yes	No control group	78	Once b/n 6 and 12	No deterioration in IPSS scores but no data presentation in results
Hellan et al[34]	No	No	No control group	39	median F/U 13 mo	One patient (2.56%) developed bladder dysfunction post operatively
Park et al[39]	No	No	No control group	30	Not stated	No patients developed bladder dysfunction post operatively
Cho et al[35]	No	No	Robot vs lap	278 vs 278	1	Voiding dysfunction rate higher in the laparoscopic group (4.3% lap vs 0.7% rob; P = 0.012)

The following study characteristics are described: (1) whether male patients were assessed independently of female patients or not, in studies that this was not the case data from male and female patients was combined; (2) whether functional scores were used to assess urogenital outcomes or not; (3) the control group used in the study if applicable; (4) the number of cases examining male urological function; (4) the follow up periods in months; and (5) a brief summary of the study’s findings regarding male urological function. Robot: Robotic; lap: Laparoscopic; f/u: Follow up; IPSS: International Prostatic Symptoms Score.

Validated functional scores were used in nine of the above studies. Six of those compared the scores of patients undergoing robotic surgery with those undergoing laparoscopic or open surgery. Most studies[18,30,32,33] showed that urological function tended to deteriorate in the early postoperative phase (1-3 mo) but later recovered with time (6-12 mo) irrespective of surgical modality. One study[9] found that IPSS score change from baseline was less in the robotic group at 12 mo after surgery, but failed to reach statistical significance (P = 0.051).

Kim et al[31] reported IPSS scores in favour of the robotic group. They found that IPSS scores significantly increased 1 mo after surgery; but then recovered in 3 mo in the robotic group and 6 mo in the laparoscopic group with a statistically significant lesser deterioration of scores from baseline in the 3 mo follow up period in the robotic group (P = 0.036). It is worth noting that Kim et al[31]’s study was the only one to assess urinary function by means of urodynamic studies in conjunction with a functional score. He reported that the deterioration in mean voiding volume from baseline was statistically less in 3 and 6 mo post-op in favour of the robotic group (P = 0.007, P = 0.049). The only other study to report urological outcomes in favour of the robotic group was Cho et al[35]’s study; reporting a higher voiding dysfunction rate in the laparoscopic group (4.3% vs 0.7%; P = 0.012). However, this study did not use any functional scores to assess urological function.

Female urological function

Seven studies reported female urological functional outcomes (Table 4). However, there are only two studies that report female urological dysfunction independently to that of males.

Table 4

Original studies reporting female urological function

Ref.	Females assessed independently of males	Functional scores applied	Control group	No. of cases examining female urological function	Follow up in months	Outcome summary
Morelli et al[30]	Yes	Yes	Robot vs lap	Not available	1, 6, 12	No difference between the pre- and post-operative scores in both groups
Luca et al[6]	Yes	Yes	No control group	36	1, 6, 12	Worse female urological function at 1 mo with full recovery by 12 mo in both groups
Kim et al[31]	No	Yes	Robot vs lap	30 rob vs 39 lap	1, 3, 6, 12	As in Table 3
Stănciulea et al[37]	No	Yes	No control group	78	Once b/n 6 and 12	As in Table 3
Hellan et al[34]	No	No	No control group	39	Median f/u 13 mo	As in Table 3
Park et al[39]	No	No	No control group	30	Not stated	As in Table 3
Cho et al[35]	No	No	Robot vs lap	278 vs 278	1	As in Table 3

This table describes the same study characteristics included in Table 3 but for female instead of male patients. Robot: Robotic; lap: Laparoscopic; f/u: Follow up.

Both studies used approved functional scores to assess urinary function and both studies compared robotic surgery patients with laparoscopic surgery patients. Morelli et al[30] found no difference between the pre-operative and post-operative scores concerning voiding and filling symptoms in both groups. Conversely, Luca et al[6] reported worsening of symptoms one month post operatively with full recovery by 12 mo in both robotic and laparoscopic groups.

Male sexual function

Fourteen original studies reported male sexual functional outcomes (Table 5). Ten of those assessed male sexual function via the IIEF[42] questionnaire.

Table 5

Original studies reporting male sexual function

Ref.	Males assessed independently of females	Functional scores applied	Control group	No. of cases examining male sexual function	Follow up in months	Outcome summary
Kim et al[31]	Yes	Yes	Robot vs lap	18 rob vs 20 lap	1, 3, 6, 12	Quicker recovery of male sexual function in robotic group (6 mo vs 12 mo) No difference in IIEF change from baseline between two groups at any stage Erectile function and libido deteriorated significantly more in lap group at 3 mo
Park et al[9]	Yes	Yes	Robot vs lap	20 vs 20	3, 6, 12	Quicker recovery of male sexual function in robotic group (6 mo vs 12 mo) IIEF deterioration significantly higher in lap group at 6 mo (P = 0.03)
Park et al[32]	Yes	Yes	Robot vs lap	14 rob vs 15 lap	3, 6, 12	Better male sexual function scores at 3 and 6 mo in robotic group No difference in IIEF change from baseline between two groups at any stage
D'Annibale et al[33]	Yes	Yes	Robot vs lap	18 rob vs 23 lap	1, 12	Erectile function restored 1 yr post-operatively in robotic group (P = 0.066) and partially in lap group (P = 0.048) No statistical comparison of IIEF change from baseline b/n 2 groups at any stage
Ozeki et al[18]	Yes	Yes	Robot vs open	15 rob vs 22 open	3, 6, 12	IIEF scores unchanged at 3, 6 and 12 mo in both groups
Morelli et al[30]	Yes	Yes	Robot vs lap	Not available	1, 6, 12	Quicker recovery of erectile and orgasmic function in robotic group (6 mo vs 12 mo) No difference in IIEF change from baseline between two groups at any stage
Leung et al[5]	Yes	Yes	No control group	15	3	No significant difference between post- and pre-operative IIEF scores
Luca et al[6]	Yes	Yes	No control group	38	1, 6, 12	Male sexual function scores decreased at 1 and 6 mo, recovered at 12 mo
Stănciulea et al[37]	Yes	Yes	No control group	31	Once b/n 6 and 12	No difference of pre- and post-op IIEF scores with exception of 3 patients (9.68%) with severe erectile dysfunction
Alecu et al[36]	No	Yes	No control group	79	Not stated	3 patients (3.79%) developed important sexual dysfunction. No mention of IIEF scores in results
Patriti et al[40]	Yes	No	Robot vs lap	11 rob vs 12 lap	Mean f/u 12 mo	No difference in the incidence of sexual dysfunction between the 2 groups
Erguner et al[38]	No	No	Robot vs lap	27 rob vs 37 lap	Not stated	No difference in the incidence of sexual dysfunction between the 2 groups
Cho et al[35]	No	No	Robot vs lap	278 vs 278	1	No difference in the incidence of sexual dysfunction between the 2 groups
Park et al[39]	Yes	No	No control group	16	Not stated	1 patient (6.25%) developed ejaculatory dysfunction, no patients developed erectile dysfunction

This table describes the same study characteristics included in Tables 3 and 4 but for studies assessing male sexual function. Robot: Robotic; lap: Laparoscopic; f/u: Follow up; IIEF: International Index of Erectile Function score.

Six of the ten studies using the IIEF scores compared the scores of patients receiving robotic rectal cancer surgery with that of a control. Park et al[9]’s study showed that sexual function recovers faster in the robotic group. At 6 mo the IIEF scores in the robotic group were higher than in the laparoscopic group and showed a significantly smaller decrease from baseline (P = 0.03). Kim et al[31] also found that sexual function recovered quicker in the robotic group (6 mo vs 12 mo), but unlike Park et al[9]’s study, when comparing the change of total IIEF scores from baseline no significant difference was detected. However, erectile function and libido had deteriorated significantly more in the laparoscopic group 3 mo post op. Park et al[32] showed similar results, with significantly higher mean IIEF scores at 3 and 6 mo post op in favour of the robotic group. Like Kim et al[31]’s study, the change of scores from baseline did not statistically favour either intervention. In Morelli et al[30]’s study erectile and orgasmic function was significantly worse 1 mo after RobTME while it was significantly worse after 1 and 6 mo after LapTME, with erectile and orgasmic function normal at 12 mo in both groups. The other components of the IIEF score deteriorated 1 and 6 mo following surgery in both groups, with normalisation of the scores at 12 mo. D’Annibale et al[33] reported better restoration of erectile function 1 year after surgery in the robotic group; however, there is no mention of the actual IIEF scores or their change from baseline in the study so any results need to be interpreted with caution. Overall, the above comparative studies seem to report a trend towards quicker recovery of sexual function in the robotic group. However, Park et al[9]’s study was the only one to reveal an interval change in IIEF scores in favour of the robotic group that was statistically significant.

Female sexual function

In contrast to male sexual function, only a few studies have investigated sexual function in females (Table 6). Only three studies have examined female sexual dysfunction independently with that of males[6,30,37] and only one of those compared robotic outcomes to those of a control group[30]. All three studies assessed female sexual function via the FSFI.

Table 6

Original studies reporting female sexual function

Ref.	Females assessed independently of males	Functional scores applied	Control group	No. of cases examining female sexual function	Follow up in months	Outcome summary
Morelli et al[30]	Yes	Yes	Robot vs lap	not available	1, 6, 12	Female sexual function worse at 1 and 6 mo and restored by 12 mo, in both groups
Luca et al[6]	Yes	Yes	No control group	36	1, 6, 12	Female sexual function worse at 1 and 6 mo and restored by 12 mo
Stănciulea et al[37]	Yes	Yes	No control group	13	Once b/n 6 and 12	No difference between pre- and post-operative FSFI scores (but data not provided in results section)
Alecu et al[36]	No	Yes	No control group	79 pts	Not stated	As in Table 5
Erguner et al[38]	No	No	Robot vs lap	27 rob vs 37 lap	Not stated	As in Table 5
Cho et al[35]	No	No	Robot vs lap	278 vs 278	1	As in Table 5

This table describes the same study characteristics included in Tables 3-5 but for studies assessing female sexual function. Robot: Robotic; lap: Laparoscopic; FSFI: Female Sexual Function Index.

Morelli et al[30] reported worsening of sexual outcomes in both groups 1 and 6 mo following surgery, but sexual outcomes were restored by 12 mo. There were no differences between the robotic and laparoscopic groups. Luca et al[6] demonstrated similar results in their robotic group as in Morelli et al[30]’s study, whereas Stănciulea et al[37] reported no difference between pre- and post-operative FSFI scores.

DISCUSSION

This literature review highlights the fact that the impact of robotic rectal surgery on urogenital functional outcomes is yet to be established. There are number of limitations in the current studies. These include poor study design, small number of participants, lack of stringent follow up and limitations to the methods and types of data collected.

The main limitations of the primary studies were the lack of randomisation, retrospective design and small number of cases in the majority of studies (Tables 1 and 2). As for the prospective studies, most of them failed to mention the number of patients excluded during recruitment, the number of patients refusing to participate and the number of drop outs. There was one RCT but randomisation was abandoned early on as the operating surgeon quickly favoured the robotic approach for low rectal tumours. In terms of participant selection only nine studies reported their outcomes against those of a control, with the other studies essentially only describing their case series rather than comparing them to alternative treatment methods.

Case matching was performed in 2 of the comparative studies[9,35], but in the remaining studies patient selection was susceptible to selection bias due to the method of patient selection and allocation. In a number of studies patients were only able to receive robotic surgery if they covered the extra costs themselves, leaving the patients that couldn’t afford it opting for laparoscopic or open surgery instead. Therefore the validity of the data may be skewed since patients that opted for robotic surgery were more likely to be from a higher socio-economic background, which is a potential confounding factor. Moreover, two studies compared their robotic cases with an equivalent number of their first laparoscopic cases[30,33]. This selection method was done to eliminate the confounding factor of a learning curve from either method. However, the learning curve for each method is not equal[48] and since in both studies all cases were performed by one surgeon only, it is possible that many of the skills acquired from the laparoscopic method were transferable to the robotic one. This way, results in favour of the robotic group could simply represent advancement in the surgeon’s operative technique rather than superiority for the robot.

Patients in the robotic cohort either had a fully robotic procedure or a hybrid procedure (Table 2). The main difference between the two approaches is that in the hybrid approach robotic rectal dissection is preceded by laparoscopic mobilisation of the left colon and ligation of the inferior mesenteric vessels. It is possible that the difference in approach could influence urogenital outcomes. Supporters of the fully robotic approach would advocate that robotic dissection around the inferior mesenteric artery pedicle is an essential step of the procedure for identification and preservation of the periaortic nerves[49], which is where the superior hypogastric plexus lies. Moreover, the paired hypogastric nerves are susceptible to injury during mobilisation of the rectosigmoid colon from the gonadals and the ureter[13]; a step performed laparoscopically during the hybrid approach. Since injury to those nerves can lead to urogenital dysfunction, the hybrid approach might not exploit the full potential of the robotic system.

Five studies did not use functional scores to assess urogenital outcomes. The challenge with only reporting the incidence of urological or sexual dysfunction is not only the inability to quantify the level of dysfunction but also to define what makes a case. Furthermore, where studies fail to report how many of the patients were sexually active pre-operatively, observational bias may be present.

It is important to mention that even though iatrogenic nerve injury is the primary cause of urogenital dysfunction[14-16], this group of symptoms is probably multifactorial in origin. Ozeki et al[18] utilised univariate analysis and found that age and post-operative complications significantly affected urinary function and sexual function respectively at 12 mo follow up. Sexual function in comparison to urological function is reported as being influenced by psychological factors and this is the case more so in women[4,6]. Luca et al[6] showed that whereas the presence of an ileostomy in men did not influence sexual function, it deeply affected it in women. Furthermore, poor body image, fatigue, depression, loss of independence and changes in relationships have all been identified as important factors in women’s sexual dysfunction[4]. In addition, radiation induced ovarian failure in premenopausal women can further worsen sexual symptoms[4]. Since the above are potentially important confounding factors, it is important for the control group to be as similar to the experimental group as possible or control for these confounders in the analysis, something absent in the studies examined in this review.

In this review we did not perform a meta-analysis due to the heterogeneity of the included studies. Nevertheless, it should be mentioned that two review articles have performed meta-analyses on male urological and sexual function scores of patients receiving robotic vs laparoscopic rectal surgery[46,47]. For male urological function, the reviews pooled the data from three studies and found that at 3 mo there was a significant difference of IPSS scores in favour of the robotic group. However, this was not the case at 6 mo following surgery and at 12 mo the two meta-analyses reported contradictory results, one showing favourable IPSS scores for the robotic group[46] whilst the other demonstrated no difference between the two groups[47]. Regarding male sexual function, the meta-analyses pooled the data for erectile function only. By including three and two studies respectively[46,47], both reviews demonstrated favourable erectile function scores for the robotic group at 3 and 6 mo following surgery. Weighing these results one should note that as a rule, the overall quality of a meta-analysis is limited to the quality of its primary studies, and since the quality of the evidence available is low, the results of the available meta-analysis are of equally low quality.

There is a degree of inconsistency of results across the research examined in this review and the potential for bias amongst the various studies on the subject. There is a lack of high level evidence supporting any particular approach for preservation of urogenital function following rectal surgery. Nevertheless, the current evidence suggests that robotic surgery might lead to a quicker recovery of male urological and sexual function when compared to alternative methods. It is less clear whether robotic surgery makes any difference in male urogenital outcomes 1 year following surgery. In females the evidence on urogenital function following robotic rectal surgery is further limited. Again functional outcomes seem to improve with time but this is regardless of operative approach.

Larger randomised controlled trials such as the ROLARR trial[50] might provide more insight into this matter. However, even though the ROLARR trial is underway, urogenital outcomes are not one of its primary ends points and urogenital outcomes are only assessed once following surgery, at six months. Therefore, to answer whether robotic rectal cancer surgery truly offers superior urogenital outcomes further randomised control trials specifically designed to evaluate urogenital function with appropriate short and long term follow up are recommended. In addition, urogenital dysfunction should be rigorously assessed through appropriate validated functional scores and males should be analysed separately to females.

COMMENTS

Background

Urological and sexual dysfunctions are unfortunate sequela of rectal cancer surgery. They occur due to iatrogenic injury to the pelvic autonomic nerves during the surgical process and cause significant quality of life limitations for patients. Better visualisation of the pelvis such as during laparoscopy has failed to address this issue due to the stiff, fixed tip instruments used for laparoscopy being hard to use in narrow spaces such as the pelvis. Robotic surgical systems overcome many of the limitations of laparoscopic surgery but whether robotic rectal surgery can lead to superior urological and sexual functional outcomes remains to be determined.

Research frontiers

Robotic surgical systems possess several advantages over conventional laparoscopy such as flexible wristed instruments that mimic the surgeon’s hands. They eliminate the surgeon’s tremor and offer far superior ergonomics and dexterity. In addition, the surgeon, rather than the assistant, controls a 3-D, high definition stable camera, an important aspect for co-ordinated surgery. These advantages allow for precision surgery in narrow spaces such as the pelvis, where other methods have failed and in rectal surgery could enable preservation of the pelvic autonomic nerves and therefore increase the quality of life for these patients.

Innovations and breakthroughs

There are only a few studies that have investigated the urological and sexual outcomes of robotic rectal surgery and these tend to be predominantly about male patients. This study differs by critically reviewing the available literature on the postoperative urological and sexual outcomes of robotic rectal surgery on both male and female patients. As such, this review is unique in that it examines the largest number or relevant studies to date; it focuses solely on the urogenital outcomes of robotic rectal surgery and examines the evidence on both males and females.

Applications

This review critically analyses the literature examining the urogenital outcomes of robotic rectal cancer surgery. Readers will be able to have a concise understanding of the available literature on this subject. Furthermore, this review leads to clear conclusions indicating a paucity of evidence of whether robotic rectal surgery offers favourable urogenital functional outcomes and establishes quality of life differences. Nevertheless, the authors identify that robotic surgery might lead to a quicker recovery of male urological and sexual function when compared to alternative methods of surgery and recommend the direction of further research.

Terminology

Urogenital function is a term referring to the combination of urological and sexual function. Laparoscopic and robotic surgeries are forms of minimally invasive surgery which offer several advantages over open surgery, such as smaller wounds and quicker postoperative recovery.

Peer-review

The manuscript is a comprehensive review addressing pelvic functions (rectal and sexual) after robotic surgery. Content coverage is adequate and focus. Language quality and flow of idea are excellent.