A Systematic Review of the Role of Robotics in Plastic and Reconstructive Surgery—From Inception to the Future

Background The use of robots in surgery has become commonplace in many specialties. In this systematic review, we report on the current uses of robotics in plastic and reconstructive surgery and looks to future roles for robotics in this arena. Methods A systematic literature search of Medline, EMBASE, and Scopus was performed using appropriate search terms in order to identify all applications of robot-assistance in plastic and reconstructive surgery. All articles were reviewed by two authors and a qualitative synthesis performed of those articles that met the inclusion criteria. The systematic review and results were conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta Analysis (PRISMA) guidelines. Results A total of 7,904 articles were identified for title and abstract review. Sixty-eight studies met the inclusion criteria. Robotic assistance in plastic and reconstructive surgery is still in its infancy, with areas such as trans-oral robotic surgery and microvascular procedures the dominant areas of interest currently. A number of benefits have been shown over conventional open surgery, such as improved access and greater dexterity; however, these must be balanced against disadvantages such as the lack of haptic feedback and cost implications. Conclusion The feasibility of robotic plastic surgery has been demonstrated in several specific indications. As technology, knowledge, and skills in this area improve, these techniques have the potential to contribute positively to patient and provider experience and outcomes.

FiGURe 1 | A 15-year literature review of the number of publications relating to robotic surgery demonstrating a highly significant exponential increase. Each column represents the number of papers published in that year, rising from 168 in 2000 to over 2,000 in the year 2014 (Source; Pubmed, searched using the terms "robot" and "surgery" from January 2000 to December 2014).
Since the first reported use of the daVinci ® Surgical Robotic System (Intuitive Surgical, Sunnyvale, CA, USA) in a roboticassisted laparoscopic cholecystectomy (1), Intuitive Surgical has become the leading force in surgical robotics. The daVinci ® robot has been widely implemented in many surgical specialties, from cardiac surgery (2,3) to gynecology (4,5). In the USA, 80% of radical prostatectomies are now being performed robotically (6). With updates to the daVinci ® robot including a fourth instrument arm, its application is broadening to other specialties such as colorectal surgery (7). The dominance of the daVinci ® system is, however, beginning to be challenged with new competitors entering the market.
Plastic and reconstructive surgery is an innovative specialty, often at the forefront of technical innovation within surgery. It is also a specialty that works collaboratively with many other surgical disciplines and, therefore, those practicing it will likely come across advances in robotic surgery in these other specialties. It is, therefore, important for plastic surgeons to embrace this new surgical platform, explore potential uses for it, and learn from those who have already incorporated robotics into their surgical armamentarium.
This systematic review aimed to identify all current reported uses of robotic assistance in plastic and reconstructive surgery, from cadaveric to clinical examples. We have provided and upto-date list of all areas of interest to the plastic and reconstructive surgeon, evaluating the relevant advantages and disadvantages of the use of robotics in these areas. metHODS A database search was performed to identify all articles describing the use of robotic assistance in plastic and reconstructive surgery. The search strategy was constructed in line with the Preferred Reporting Items for Systematic Reviews and Meta Analysis (PRISMA) guidelines (8) and the Cochrane handbook (9). Key words and Medical Subject Heading terms were combined using Boolean logic and refined with the help of an information specialist (see Figure 2 for an example of the full search strategy).
Medline (1946-present), EMBASE (1980-present), and Scopus electronic databases were all searched using the developed search strategy up to May 2017.
All studies identified were downloaded into EndNote V8 for Mac (Clarivate Analytics) and duplicates removed. De-duplicated results were then uploaded to Covidence (www. covidence.org) for screening. Titles and abstracts were reviewed by two independent reviewers (OC and HS) against the inclusion and exclusion criteria and discrepancies resolved through discussion with a third, independent reviewer (TD). Studies were considered eligible for qualitative synthesis if they met the following inclusion criteria: • the study was published in English • the study design was one of the following: case reports, case cohorts, case-control and randomized controlled studies. Both prospective and retrospectively designed studies were included. • the study reports the use of a robotic surgical system for a potential plastic surgery-related operation, with both preclinical and clinical applications included.
Full-text articles of those included studies were subsequently reviewed independently for final inclusion. References were checked for further, un-identified articles, and these were added in if appropriate.
A data extraction sheet was developed to extract the following data from studies: Author, date of publication, location of study, study design, number of operations performed, operations/ techniques, outcomes measured. This was piloted on a random sample of papers and subsequently refined. All data were extracted and tabulated using Microsoft Word and Excel (Redmond, Washington, DC, USA). Figure 3 illustrates the PRISMA flow diagram demonstrating the process of article retrieval and screening. A total of 7,904 articles were identified after de-duplication for screening. Of these 213 made it to full-text review. A total of 68 studies met the inclusion criteria and were eligible for inclusion in this systematic review. Included papers were divided into groups based on operative type or body location and a qualitative synthesis of the outcomes reported performed. microsurgery A total of 13 studies were identified discussing the use of robotics for a microsurgery application ( Table 1). Eleven of these were preclinical studies in synthetic, animal, and cadaveric models (10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) while two were clinical studies (21,22). Katz et al. performed the first daVinci ® system assisted anastomosis in a porcine model in 2005 (10), closely followed by work in canine tarsal and femoral vessels (13). In these studies, they concluded significant advantages such as the elimination of tremor at a microsurgical level, but that the lack of purpose-built microsurgical instruments was an important limitation. Further animal and human cadaveric work cemented the idea that robotically assisted microvascular surgery is both feasible and in some instances potentially beneficial, such as when working at depth and for surgeon comfort (20).

ReSULtS
Two clinical examples were identified, with one cohort study by Boyd et al. including 22-patients where the robot was used for harvesting the internal mammary vessels in free breast reconstruction (21). Van der Hulst et al. used the robot to perform the anastomosis, commenting on the increased time taken for this over traditional methods, as would be expected early on in the learning curve (22). As in preclinical studies, the benefits of using the daVinci ® robot for performing the microvascular anastomosis include elimination of tremor and motion scaling.

muscle Flap Harvest
Traditionally muscle free-flaps are raised through a large incision overlying the muscle belly and are, therefore, a perfect example of where the robot can have marked benefit as minimally invasive harvesting can significantly reduce the size of externally visible scarring. Laparoscopic harvesting has been attempted, but with  poor uptake due to difficulties with visualization of the operative field and the inherent limitations of laparoscopic instruments (23,24). Three human cadaveric studies (25)(26)(27) and five clinical reports (28)(29)(30)(31)(32) were identified describing the use of the robot for muscle flap harvest ( Table 2). In those clinical studies, it is clear that the robot improves visualization, reduces the scar burden and resulted in reduced postoperative pain and hospital stay. The traditional approach to rectus muscle harvest is with a large abdominal skin incision. Not only is this cosmetically unappealing but also, in combination with division of the anterior abdominal wall fascia, can result in incisional hernia formation. As robot-assisted colorectal surgery becomes increasingly routine, with the advantages of minimal scarring, reduced conversion to open procedure, reduced time to intestinal motility, and reduced postoperative sexual dysfunction reported (33), it would seem a retrograde step to then introduce a large abdominal wound when harvesting the rectus abdominis muscle for perineal reconstruction. In a case series by Singh et al. the robot was used in tandem with a robotically performed abdominoperineal resection for adenocarcinoma to raise the rectus abdominis flap for reconstruction (32). This produced satisfactory closure of the defect without the risk of a ventral hernia. In these combined procedures the risks associated with entering the abdominal cavity are already present from the colorectal resection and, therefore, one of the major disadvantages of robotically assisted rectus abdominis muscle harvest is not a risk purely implicated through the use of this novel muscle harvest technique.

Nerve Surgery
A total of eight preclinical studies and five clinical studies were identified, with the majority investigating the role of robotics in brachial plexus work ( Table 3) (34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46). Epineural nerve repair using robotic assistance has been shown to be technically feasible in animal models, with the benefits of reduced physiological tremor and improved vision of the surgical field noted (35). Nerve harvest has also been demonstrated to be feasible in cadaveric and animal models (35,37).
In those clinical studies identified, robotic assistance was successfully used to repair a brachial plexus (45), repair the sympathetic chain to treat Horner's syndrome (42), perform a thoracic sympathectomy for palmar hyperhidrosis (43), repair a peripheral nerve following tumor excision (46), and undertake an Oberlin procedure (44). Table 4 illustrates those articles relating to procedures in the upper limb, with three preclinical (47)(48)(49) and one clinical study identified (50). As with a number of other areas of the body the use of the robot has so far only been for proof of concept and there has yet to be any concrete studies demonstrating a benefit.

trans-Oral Robotic Surgery (tORS)
Trans-oral robotic surgery has allowed head and neck surgeons to treat benign and malignant conditions of the oral A remarkable degree of tremor filtration, but significantly slower operative time. All anastomoses were patent and non-leaking Case controlled Karamanoukian et al. (12) 2006 Animal tissue samples Slit arteriotomy and end-to-end arterial anastomoses in procine hearts The Zeus robotic system is a viable tool for microsurgical vascular reconstruction. It allows for precise movement, lack of hand tremor, enhanced microvascularisation and improved ergonomics, compared to conventional human assistance. The major advantage is the ability of the robot to scale down the surgeon's movements to a microscopic level Generation of learning curves for robot-assisted microvascular anastomosis. Important aspects of learning identified included starting level, learning plateau and learning rate Robert et al. (17) 2013 Human cadaver Radial/ulnar artery dissection and microvascular anastomoses (N = 2 cadavers, 4 anastomoses)

Successful anastomoses
The assembling and disassembling of the vascular clamp were time consuming In both cases (radial and ulnar arteries), the 10/0 needle was bent and a second suture had to be used  cavity and oropharynx avoiding more traditional jaw and lip split approaches, facilitated by the improved access and visualization afforded by the robotic instruments (51,52). If there is no communication between the oral cavity or oropharynx and neck dissection then the defect could be left to heal by secondary intention; however, in more complex or advanced stages of disease, reconstruction using local flaps or free tissue transfer is required (53). If a jaw split has not been performed, access for satisfactory reconstruction can be almost impossible and thus developing reconstructive techniques using the robot in order to capitalize on the minimized morbidity associated with a TORS resection is of paramount importance. Trans-oral robotic surgery has become the biggest area for robotic-assisted plastic surgery procedures, with 2 preclinical (54, 55) and 21 clinical studies identified (56-76) ( Table 5). Local reconstructive options include the use of the Facial Artery Musculomucosal flap, commonly used in reconstruction of the floor of the mouth and soft palate. Bonawitz and Duvvuri have described using the robot for raising and in-setting the flap with good results (64,65). Others demonstrated that the use of the robot to perform a musculomucosal advancement flap pharyngoplasty gives good results, both in terms of orocutaneous fistula risk and functional outcomes (60,61).
In larger or more complex composite defects there is often the requirement for free-flap reconstruction, with specific indications including exposure of the carotid artery, large baseof-tongue defects and defects of the soft palate and tonsillar fossa which cannot be closed with local flap options. The commonest reported free-flap used following TORS resection is the radial forearm flap; however, others such as the anterolateral thigh flap are also described. In the majority of cases the robot was used for flap inset, with authors reporting good access and visualization that allowed a water-tight inset to be achieved and no flap complications despite the lack of a traditional jaw spilled. The robot was also used in a number of studies to perform the vascular anastomosis (58,62,63,69).

trans-Oral Robotic cleft Surgery (tORcS)
Trans-oral robotic cleft surgery is still in its infancy with only three articles identified (77-79) ( Table 6); however, it builds upon the same benefit profile achieved by TORS that has been outlined previously for access to the oral cavity and oropharynx in cleft lip and palate patients. Table 7 demonstrates four other studies identified in the systematic review, which do not fit into the categories above (80)(81)(82)(83). Of these indications, it is likely that only lymph node based procedures are likely to progress in the future, with some benefits such as the ability to perform supermicro-surgery an obvious advantage in lymph node transfer. The robot allowed microsurgery to be performed in a very small space with telemanipulation and minimally invasive techniques In postoperative surveys, neuropathic pain halved from 6/10 to 3/10 postop, with no worsening of sensory deficits

DiScUSSiON
In the 30 years since the first robot was used in a surgical procedure the arena of robotic surgery has changed at a breathtaking pace, with the use of the daVinci ® robot now common place in some specialties. This revolution has taken longer to impact on the plastic surgery community. It is, therefore, somewhat ironic that it was a plastic surgeon who was at the forefront of robotic and tele-surgery at its inception (84). However, this systematic review has shown that significant developments have been made in the last few years. The benefits of robotic surgery have been well documented, albeit with no large scale studies, and include reduced blood loss, reduced postoperative pain, faster recovery, and improved cosmesis (85). In relation to plastic and reconstructive surgery the elimination of tremor, greater degree of freedom of the instrument and motion scaling all have the potential to improve the accuracy and reproducibility of microsurgery. The evidence suggests that while the initial learning curve is steep, proficiency in microsurgical skills using the robot can be gained in a short number of sessions (18).
Of the areas identified in this systematic review there are some that are further down the development road than others and some, where the advantages of robotic assistance are greater. For example, with the recent uptake of free-perforator flaps in the field of reconstructive surgery we are beginning to approach the limits of human dexterity at which point the robot may prove to    (64)   be advantageous. However, to fully exploit this there needs to be focused development in the field of robotic instrument design, expanding the portfolio of micro-instruments. It is our opinion that the potential for robotic head and neck reconstruction is huge and is one of the areas that will most definitely see growth due to the obvious benefits it offers. This will be especially true as the indications for TORS resection continue to widen, resulting in larger and more complex defects. The current limitation to more widespread utilization is instrument design in order to perform microvascular anastomoses and easier inset and it is this area that research should focus. This may also be the case with TORCS. As with cancer resection, there are many circumstances where adequate access to the pharynx and palate in the pediatric cleft patient can pose a significant challenge. The space in which to operate, as well as access for instrumentation, can be severely restricted, especially in cases with abnormal anatomy, poor jaw opening, or anomalies of the nasopharyngeal space. Adequate illumination and visualization can also be difficult. Early work has shown that performing posterior pharyngeal wall surgery using the daVinci ® robot is feasible, with benefits such as an improved view, easier dissection, reduced secondary surgical insult and preferential ergonomics for the operating surgeon (77). Its use may also open up avenues of new surgical interventions to areas of the oropharynx that were previously inaccessible.
There is currently less convincing evidence for the use of robotics in areas such as nerve and upper limb surgery. In brachial plexus reconstruction nerve harvest is often required and, therefore, reduced donor site morbidity through robotic harvest, such as with trans-thoracic harvest of intercostal and phrenic nerves, is an area that has the potential for future advancement. It will be important, however, to also demonstrate its safety and cost-effectiveness in order to justify the marginal reductions in scarring when compared to more traditional harvest sites. To date all of the preclinical and clinical studies investigating robotic nerve surgery have demonstrated that it is technically feasible. However, it is still mostly at a proof of concept level and while does have benefits in terms of reduced tremor, it is most likely to be of benefit in difficult to assess areas or when the robot is already being used to perform other parts of the procedure. Finally, at present the indications for the use of robotics in hand surgery are probably more limited than other areas discussed, especially as access is not normally a problem in hand surgery. However, the benefits as discussed for microsurgical anastomosis may prove to  be useful in specific indications such as traumatic replantation or congenital reconstruction. Robotic surgery's main disadvantage remains the high cost of purchasing and maintaining the equipment. This will undoubtedly improve with time as a greater number of procedures are performed using the robot and the unit cost per operation reduces. A recent comparison of the cost of TORS compared to radiotherapy demonstrated that TORS is currently more expensive; however, this is likely to reduce through the creation of high-volume centers performing TORS (86). It has also been shown that in a center where the learning curve had already been overcome, robotic surgery was cheaper than equivalent open surgery for the surgical treatment of endometrial cancer (87).
Lack of haptic feedback is also often cited as another disadvantage of robotic surgery, with studies demonstrating that operators of augmented robotic surgical systems prefer those with haptic feedback (88). However, other studies such as by Hagen and colleagues who looked at 52 individuals and their perception of haptic feedback while performing robotic surgery demonstrated that visual cues are able to give the perception of haptic feedback, even when true haptic feedback is not present (89). Despite this evidence there is still a tremendous amount of working looking at ways to incorporate haptic feedback into robotic systems, summarized in a review by Okamura (90).
Finally, robotic surgery often results in longer operative times, although this improves with proficiency and in some cases is now comparable to traditional techniques.
The future of robotics in plastic surgery is clearly exciting. Over the last 5 years the range of procedures using the daVinci ® robot being attempted by the plastic surgery community has increased significantly and, as technology continues to improve, this will gain further momentum. Of the 68 studies included in this review, only three used a robotic system other than the daVinci ® . This dominance is beginning to be challenged and while equipment additions such as a micro-forcep is currently available for the daVinci ® robot and external companies have developed micro-doppler probes and hydrojet dissectors (91) it will be the development of further microsurgical instruments that will allow greater use of the robot in the field of plastic and reconstructive surgery. The combination of motion scaling and tremor-free instrument manipulation with new instrument design will also allow new avenues in microsurgery that have to date been too technically demanding to be explored. Furthermore, the introduction of a new single port addition to the daVinci ® system will allow greater access in trans-oral surgery, improving instrument maneuverability within the tight confines of the intra-oral cavity.

cONcLUSiON
The potential value of robotic plastic surgery has already been investigated in several specific indications. It is still early days for the field and only time will tell if the use of robotics in plastic surgery is truly of benefit. As the technology, knowledge, and skills in this area improve, it is likely that in specific indications the use of robotic surgery will further contribute positively to patient and provider experience and outcomes. It is, therefore, imperative that the plastic surgery community embraces this new technology platform, but in doing so conducts well-designed, patient-focused research to ensure that it is only being used when there is true benefit to our patients. aUtHOR cONtRiBUtiONS TD, KK, and IW developed the idea for this paper. TD, OC, and HS developed the search strategy and performed the systematic review and data extraction. TD, KK, IW, OC, and HS wrote the manuscript and all authors edited and agreed the final version.

FUNDiNG
This work has received no specific funding. TD is funded by the Welsh Clinical Academic Training Fellowship.