Safety, Feasibility, and Short-term Outcomes of Laparoscopic Ileal-Pouch-Anal Anastomosis

Abstract

Objective:

To compare safety and short-term outcomes of 100 laparoscopic ileal pouch-anal anastomosis (IPAA) versus 200 conventional open IPAA patients.

Summary Background Data:

Outcomes of laparoscopic IPAA (LAP-IPAA) have been incompletely characterized. Previous reports are characterized by small numbers of patients and rarely include case-matched or randomized trial methodology. This report describes 100 LAP-IPAA patients case matched to 200 open IPAA patients.

Methods:

Between 1998 and 2004, 100 consecutive LAP-IPAA patients (75 laparoscopic assisted, 25 hand assisted) were identified and case matched to 200 open IPAA control patients by age, operation, gender, date of operation, and body mass index. Operative and postoperative outcomes at 90 days were compared.

Results:

A total of 300 patients (180 female) with a median age of 32 years (range, 17–66 years), and a median body mass index of 23 kg/m² (range, 16–34 kg/m²) underwent IPAA (100 LAP-IPAA, 200 open IPAA). Diagnosis (chronic ulcerative colitis 97%, familial adenomatous polyposis 3%) and previous operative history were equivalent between groups. One intraoperative complication occurred in each group. Overall, the laparoscopic conversion rate was 6%. Median operative time was longer for the LAP-IPAA group (333 minutes versus 230 minutes, P < 0.0001). LAP-IPAA patients had shorter median time to regular diet (3 versus 5 days), time to ileostomy output (2 versus 3 days), length of stay (4 versus 7 days), and decreased IV narcotic use (all P < 0.05.Postoperative morbidity was equivalent (LAP-IPAA = 33%, open IPAA = 37%), mortality was nil, and readmission rates were equal (LAP-IPAA = 21%, open IPAA = 22%). Reoperation was required in 3% of LAP-IPAA and 6.5% of open IPAA patients (P < 0.2) during the first 3 months.

Conclusion:

LAP-IPAA is equivalent to open IPAA in terms of safety and feasibility. In addition, LAP-IPAA provides significant improvements in short-term recovery outcomes.

Short-term outcomes of the first 100 laparoscopic IPAA procedures case matched with 200 open IPAA controls are reported. We documented enhanced postoperative recovery among the laparoscopic patients with no differences compared with open controls in terms of postoperative morbidity and mortality. Laparoscopic IPAA is feasible and safe and confers early outcome advantages after ileo-anal anastomosis.

Ileal pouch-anal anastomosis (IPAA) is accepted as the operative procedure of choice for patients with ulcerative colitis and familial adenomatous polyposis. Since its introduction,¹ the technique of pouch construction has evolved, but a number of technical points continue to be debated.^2–6 The most recent iteration of the debate is the adaptation of laparoscopic methodology to accomplish the proctocolectomy and ileo-anal anastomosis. Laparoscopic approaches such as segmental colectomy for benign and malignant disease^2,7–9 as well as for more complex procedures2,7,10–29 have consistently demonstrated a number of benefits.

However, there are few reports of proctocolectomy and ileo-anal anastomosis performed either laparoscopically assisted (LA) or by hand-assisted techniques (HAL). Recently, several authors have reported short-term outcomes after LAP-IPAA, which were equivalent to those after open IPAA.11,13,16,22 However, conclusions were based on small numbers of patients. Our aim, therefore, was to compare outcomes in a large number of patients undergoing laparoscopic IPAA for chronic ulcerative colitis and familial adenomatous polyposis to a case-matched cohort of patients undergoing open IPAA.

METHODS

All patients who underwent a laparoscopic IPAA (LA and HALS) with a diverting ileostomy between 1998 and 2004 at the Mayo Clinic, Rochester, Minnesota were identified from a prospective laparoscopic database. This group of 100 laparoscopic IPAA was matched against a group of 200 open IPAA patients using the prospective Mayo clinic surgical registries. Each laparoscopic patient was matched in a 1 to 2 fashion by age (±2 years), gender, date of surgery (±3 years), BMI (±3), and procedure. The study was approved by the institutional IRB. LA and HALAP-IPAA were performed by 5 surgeons, while open-IPAA was performed by 7 surgeons in the Division of Colon and Rectal Surgery Mayo Clinic, Rochester, Minnesota. Postoperative recovery pathways were not standardized but left to the individual surgeon.

Ninety-day laparoscopic postoperative data were captured using a prospective database, while postoperative outcomes in the open IPAA patients were recorded by using retrospective chart review.

Operative Technique

Open Approach

Standard open IPAA was performed as described in a previous report from this institution.³⁰ Briefly, abdominal colectomy and rectal resection with preservation of the nervi erigentes and presacral nerves was performed. A J-pouch pouch was constructed from the terminal 24 to 30 cm of ileum and either anastomosed by hand after endoanal mucosal resection or double-stapled after preserving the anal transition zone.³¹ A diverting loop ileostomy was constructed in all patients and closed 2 to 3 months later.

Laparoscopic Approach

Seventy-five of the 100 LAP-IPAA patients underwent LA using a standardized 4-port technique and a 4- to 5-cm periumbilical, low midline, or Pfannenstiel incision as previously described.²⁹ Beginning in August 2003, intracorporeal vessel ligation using either laparoscopic stapling or vessel sealing devices was performed in all patients. The remaining 25 laparoscopic patients had a HALAP-IPAA performed using a hand assist device. The placement of the hand assist device was either in the lower midline or Pfannenstiel position and would involve an incision of 6 to 8 cm. Unlike LA, in the HALAP-IPAA approach, the rectal resection is performed in an open fashion through the hand access site.

Statistics

The SAS statistical analysis system V9.1.3 (SAS Institute Inc., Cary, NC) was used for all data analysis. All tests are two-tailed unless otherwise stated, and P values of <0.05 were considered statistically significant. All outcomes were compared between LAP versus open using Mantel-Haenszel for categorical variables and Friedman's test for quantitative variables. Outcomes were also compared within the laparoscopic group between LA and HAL patients. For these comparisons, χ² tests were used for categorical variables and Wilcoxon Rank Sum tests were used for qualitative variables. The laparoscopic group was analyzed under the premise of intention to treat; ie, converted cases were included in the analysis of the laparoscopic cases. Results are presented as median (range) for quantitative outcomes and as frequencies and percentages for categorical outcomes.

RESULTS

Patient demographics, previous operative history, and indications for surgery were equivalent (Table 1). Groups were comparable for all matched factors, including age, date of operation, operation, body mass index, and gender. No differences were seen in the factors above between LA and HAL except for date of operation, as all HAL operation occurred after October 2003, and a higher male predominance in HAL (LA, n = 24, 32%; versus HAL, n = 16, 64%; P = 0.0047).

TABLE 1. Patient Demographics of Laparoscopic (LAP) Patients Versus Matched Open Controls

Operative outcomes are reported in Table 2. Operative times were significantly longer in the laparoscopic group (333 minutes; range, 134–546 minutes; versus 230 minutes; range, 134–421 minutes). Interestingly, HAL procedures took longer to perform than LA (LA = 320 minutes; range, 137–546 minutes; HAL = 372 minutes; range, 255–475 minutes; P = 0.0015). Conversion occurred in 6 patients (6%) with no differences between LA and HAL. One conversion was due to intraoperative bleeding; the remaining were secondary to the inability to mobilize the colon or rectum laparoscopically.

TABLE 2. Operative Outcomes for Laparoscopic (LAP) Versus Open Patients

Mortality was zero. Morbidity occurred in (37%) of patients after open IPAA and (33%) after LAP-IPAA (Table 3). Readmission was equivalent between groups (open IPAA, n = 44, 22%; versus LAP-IPAA, n = 21, 21%). Reasons for readmission are presented in Table 3.

TABLE 3. Postoperative Outcomes After 90-Day Follow-up in Laparoscopic (LAP) Versus Open Patients

Significant improvement in early postoperative recovery parameters were found in the LAP-IPAA patients (Table 4). Furthermore, comparing LA and HAL, there were no significant differences between time to diet, time to first bowel movement, or use of IV narcotics. Median hospital stay was longer for the HAL patients when compared with LA (LA = 4 days; range, 3–22 days; versus HAL = 5 days; range, 4–14 days; P < 0.0229). However, both laparoscopic groups had significantly shorter length of stay than the 7 days found in the open group.

TABLE 4. Postoperative Recovery: Laparoscopic (LAP) Versus Open Patients

DISCUSSION

We found that IPAA performed laparoscopically confers several significant benefits when compared with open IPAA. Patient experience a significantly smaller incision with less postoperative pain, earlier return of bowel function, early resumption of diet, and shorter length of stay. Although not summarized in this group of 100 LAP-IPAA patients, previous studies have shown tha improved short term outcomes lead to equivalent quality of life scores.²⁹ Interestingly, LAP-IPAA was not less morbid than open-IPAA. However, with the exception of pelvic sepsis, complications whether occurring after LAP-IPAA or open-IPAA were short-lived and of little long-term significance.

Initial reports of laparoscopic techniques for the surgical management of patients with chronic ulcerative colitis and familial adenomatous polyposis provided little evidence of benefit over the standard open operative approach.^32–34 Moreover, several actually detailed increased morbidity and longer hospital stays.^32,34,35 More recently, laparoscopic IPAA has been shown to be safe and feasible2,7,10–22,24–27,36 but the number of patients within each of these studies were small. One of the studies documenting short-term outcomes was a matched series by Dunker et al¹³; short-term outcomes were significantly improved but only 18 patients were studied. Marcello et al²¹ published the first case matched series of 20 patients in the United states and reported improved short-term outcomes and length of stay (LA = 6 days versus open = 8 days). Hasegawa et al¹⁶ published a series of 18 patients undergoing laparoscopic assisted IPAA with good short-term and long-term outcomes, but numbers were small and no control group of open-IPAA were studied. The largest LAP-IPAA series to date was published by Kienle et al¹⁹ and included 50 consecutive patients who underwent a LA-IPAA. They concluded that LA-IPAA was both safe and feasible; however, increasing BMI and immunosuppression increased the risks for conversion and complications. Overall, morbidity in this series was 30% and wound infections occurred in fully 16% of patients. Median length of stay was 12 days for the laparoscopic group and short-term results were not reported. The results reported in the present study compare favorably indeed with these series.

The best laparoscopic approach to use, laparoscopic-assisted (LA) or hand-assisted (HAL), is vigorously debated currently. Several authors have used a LA technique with good results.16,19,21,22,27–29 but others prefer the hand-assisted approach.²⁰ Importantly, few authors have enough experience with both techniques to compare them accurately. Ky et al published a series of 32 patients undergoing single stage HALAP-IPAA with good results, but the lack of controls limited the usefulness of this report.²⁰ The study from Maartense and Dunker et al¹³ is the only randomized controlled trial comparing 30 HAL to 30 open patients. Unlike our study, this series documented no advantage to HAL with regards to improved perioperative recovery nor quality of life. Moreover, they found morbidity and mortality to be equal among groups and the cost of HAL was significantly more than open IPAA. Very long hospital stays, and a 10% reoperation rate base makes comparing these outcomes to ours difficult.

The current study is the largest case matched series demonstrating a significant short term benefit to LAP-IPAA, and thus, provides the best data available to establish the potential benefits of laparoscopic IPAA. These data confirm and expand the findings of others that regardless of surgical approach, LAP-A or HALS provide patients with improved short-term outcomes and no increased risk of morbidity or mortality. Long-term benefits in terms of quality of life improvements and functional outcomes of L IPAA are unknown. The primary benefit identified from our study is over the short-term and future studies should investigate not only long-term outcomes of LAP IPAA specifically, but also the more general effect of laparoscopic surgery on pelvic function.

CONCLUSION

LAP-IPAA, accomplished either using the LA or HAL technique, is safe, feasible, and results in postoperative recovery benefits compared with open IPAA. IPAA continues to evolve. Technical improvements will continue to occur, and the goal of removing disease, preserving anal function, and eliminating the stoma, performed by increasingly minimal surgical techniques, is the goal of our future work.

Discussions

Dr. Eugene F. Foley (Charlottesville, Virginia): The authors have presented their experience with their first 100 laparoscopic ileal pouch anal anastomosis procedures done primarily for ulcerative colitis, providing data suggesting that this complex operation can be safely done by a laparoscopic approach. They further present data comparing this series with a case-controlled cohort of ileal pouch anal anastomoses that their group has done by laparotomy, concluding that the laparoscopic approach may hold advantages over laparotomy in terms of more rapid return of GI function and decreased length of hospital stay.

I would like to comment on three areas, and have some specific questions for the authors regarding each of them.

First, I would like to emphasize that this is an impressive series. It is by far the largest of its kind presented, and if nothing else, I think it stands as a testimony to the creativity, determination, and fortitude of the surgeons who completed these operations.

For those of you in the room who do not do complex GI surgery or complex laparoscopy, I can assure you that this can be a hard operation to do even with a big open incision, and I guarantee to you that this is a demanding procedure done with a laparoscope. The mean operative time of 5 and a half hours in the laparoscopy group highlights this fact.

We should take note that this series was completed by one of the most experienced groups in the nation doing ileal pouch anal surgery. As they mentioned, they have done at least several thousand ileal anal procedures prior to the study they have shown you here today. Similarly, the Mayo group has been an early champion and leader in the field of colon laparoscopy.

These observations lead me to my first question to the authors: Can and should this technique be extrapolated to other hospitals without the same institutional or surgeon experience? Or will it remain a novelty item only for the most elite centers? Will all ileal anal procedures be done this way in the future?

Secondly, a few technical questions. Did you measure incisional length? The manuscript suggests that the incisions in the hand-assisted group were somewhere between 6 and 8 cm. Do you know your average open incision length? What were the lengths of the incisions made in the cases that were done by straight laparoscopy? You suggest in your length of stay data that the hand ports had a somewhat intermediate 5-day length of stay compared to 7 for the open group and 4 for the straight laparoscopy group. In your mind, is incisional length the main issue or are there other features about laparoscopy which account for the short recovery?

Finally, and most importantly, although I am convinced that this operation can be done safely by this experienced group of laparoscopic colorectal surgeons, we must ask the question: Does the presented data convince us of the paper's final conclusion that the laparoscopic approach is clearly better than laparotomy?

Patients in this study got one of three operations: straight laparoscopy, hand-assisted laparoscopy, or open incision. It is unclear to me how patients got into these arms. Although the case-control feature of the study reveals that the basic preoperative characteristics of the patients in the different treatment arms were the same, the lack of a formal randomization potentially allows for all kinds of biases to occur, and variations in many aspects of perioperative care other than the operative approach may account for differences in the outcomes of the groups.

One example of this: It appears from the manuscript that different surgeons contributed the open versus the laparoscopic cases, and there was no standardization of perioperative care including postoperative feeding protocols. Is it possible that the patients taken care of by the non-laparoscopic surgeons were also subjected to a more traditional slower postoperative feeding regimen delaying early discharge and biasing your results in favor of the laparoscopy group?

Although the study convinces me that these expert surgeons can do this operation laparoscopically safely and feasibly, this limitation makes me hesitant to fully embrace the more technically difficult and time-consuming laparoscopic approach as the clear procedure of choice for patients requiring ileal pouch anal surgery. I look forward to the continued leadership of the Mayo group to help refine the role of laparoscopy in the treatment of colorectal disease.

Dr. David W. Larson (Rochester, Minnesota): Your first question was whether or not this is indeed generalizable, and certainly this is a very complex operation, as you alluded to. We also must point out that in a single institution series it is often difficult to generalize to the surgical population at large.

However, one of the interesting thoughts remains that utilizing new technology such as vessel sealing devices, hand-assisted devices, have rapidly changed our current practice at the Mayo Clinic. As an example, over the last 5 years, ileal pouch anal anastomosis has rapidly adapted to over 80% of our cases being performed in this manner versus less than 1% just 5 years ago. Likewise, only one surgeon 5 years ago performed this operation and now five of the eight of us. So we feel that, overall, given the changes in technology, this will be adopted generally in the surgical community at large and will not remain solely in the hands of institutions like the Mayo Clinic.

Your second question regarding length of incision and whether or not this matters in terms of postoperative recovery. Overall, we did not measure specifically the length of incision. However, it was in general somewhere between 4 and 8 cm, generally slightly smaller in the lap-assisted group. We, however, did not feel there was any significant difference in length of stay depending on the incision site specifically. We feel that the technique of laparoscopy intrinsically changes the dynamics of postoperative care as opposed to incision size alone.

You also questioned the difference in length of stay in the hand-assisted technique. Overall, you are absolutely correct; there were no standardized postoperative protocols. However, when we looked at all the postoperative findings, they were exactly equal between the hand-assisted and the lap-assisted group, with the exception, as you pointed out, length of stay.

One of my senior colleagues, John Pemberton, was the highest contributor to the hand-assisted technique. It is quite possible that his historical criteria for discharge are longer than those of his younger colleagues and this may have contributed to this. However, that question remains unsolved, frankly.

The last question is in regards to the randomized control trial, is that needed? Are we biased because of our surgeons or the number of surgeons that we have in our group in terms of postoperative recovery? This is certainly a valid point. I think our next step will be to identify standardized protocols for our postoperative recovery.

Looking at length of stay in the literature, I do not know of any paper out there that produces results of length of stay of under 7 days for their open patients, let alone lengths of stay of under 4 days for a laparoscopic patients. Therefore, we feel that this is a very well-controlled case match study in terms of age, gender, BMI, operation. And more importantly, all of these patients were treated recently from 1998 through 2004. During this time, our protocols and our postoperative recovery were relatively unchanged. Therefore, the improvements that are seen in the laparoscopic arm are not arbitrary. In fact, lengths of stay are nearly half that of our open cohort. So we feel this is indeed relevant to the current literature.

Dr. Thomas R. Russell (Chicago, Illinois): As Dr. Foley has pointed out, this is the largest series reported, with other series only being significantly smaller in numbers. This specific series is contributed to by seven surgeons who performed colectomies in an open fashion and five who have done it laparoscopically.

The patients operated upon are, generally speaking, younger patients and mostly female. There is obviously a great spectrum of the degree of their illness, the severity of illness, and certainly the degree of immunosuppression by steroids, cyclosporine, Infliximab, etc. Nowhere in your paper do you talk at all about treatment dependent upon how sick these patients are. For example, all patients received an ileostomy, and everyone in the initial series was done in an open fashion. Now you have this hybrid model of the laparoscopic methodology with some being hand assisted and the others being laparoscopically assisted. My question is whether there has been any modification over the years of the treatment based on the severity of illness of a specific patient?

Secondly, how do you consent your patients today now that you have gone through the open and you are now into the laparoscopic era?

The third question deals with the hybrid model that you have of the laparoscopic approach where some you use what is termed “laparoscopically assisted” versus those that are “hand-held assisted.” I noted that the time to do this laparoscopic procedure is fairly long, 330 minutes, and whether it would be better to adapt one way of doing it laparoscopically rather than having this so-called “hybrid model.”

Finally, a question about the dissection of the rectum and preservation of the nervi erigentes and the pre-sacral nerves. I understand laparoscopically that you dissect the rectum with a laparoscopic technique but with the hand-held technique, do you do it in the old-fashioned way that was done with the open technique, visually by hand, or do you use the laparoscope? This is obviously a critical part of the operation, and I would like to know your experience of whether or not there is one technique that you favor over the other?

Dr. David W. Larson (Rochester, Minnesota): The first question regarded how did we identify these patients, how did we identify what operation to provide them, either laparoscopic or hand assisted. And this really was a decision made by the surgeon. Some of the surgeons in the group of five that perform the laparoscopic only do hand-assisted technique, and that was a decision made by the surgeons themselves. However, in general, patients who are male or who have high BMIs receive a hand-assisted operation versus a lap assisted.

The issue of consent. At this point, nearly 80% of our patients who require pouch surgery receive a laparoscopic approach to their operation. Consent is in general given for the laparoscopic procedure. Patients are notified, of course, about conversion, and if needed, based upon decision; at the time of surgery, we will obviously convert to an open approach. Overall, our conversion rate is around 6%.

In terms of would it be better to do one or the other, in our institution it would be probably impossible to confine the operation to one simple approach just from the sheer volume in terms of numbers of operations that are required to be done on a yearly basis. And frankly, I think the approach that should be taken is the one that is best suited to the surgeon. If surgeons feel they can accomplish this with a hand-assisted approach, I think this is completely reasonable and should be done. We have shown really no differences in either approach. I do not believe any one is better than the other. What I would say is that it is significantly better than the open alternative.

In terms of rectal dissection, in my practice, the majority of cases I do personally are lap-assisted approach. I feel that I can see the nerves and the pelvic structures extremely well using the laparoscopic approach. But we have no data. No one has really looked at the role of sexual function, the role of bladder dysfunction, from pelvic operations performed laparoscopically. We are currently utilizing the same cohort to investigate just those very issues.