Laparoscopic Surgery for Crohn’s Disease: Reasons for Conversion

Abstract

Objective

To examine factors influencing conversion from a laparoscopic to an open procedure in patients requiring surgery for Crohn’s disease.

Summary Background Data

Laparoscopic management of patients with complications of Crohn’s produces better outcomes than traditional open approaches, but it is difficult to determine before surgery who will be amenable to laparoscopic management. In this series, a laparoscopic approach was offered to virtually all patients to determine reasons for laparoscopic failure.

Methods

Data regarding patients who underwent attempted laparoscopic procedures for Crohn’s (January 1993 to June 2000) were collected prospectively. The bowel was mobilized laparoscopically and extracorporeal anastomoses were performed. Conversion to open surgery was defined as creation of an incision of more than 5 cm.

Results

One hundred ten patients (age 37 ± 1.1 years, 58% female) underwent 113 attempted laparoscopic interventions. Indications for surgery included obstruction (77%), failure of medical management (35%), fistula (27%), and perineal sepsis (4%). Sixty-eight procedures (60%) were completed laparoscopically. Procedures completed laparoscopically included ileocecectomy (n = 46), small bowel resection (n = 22), fecal diversion (n = 7), intestinal stricturoplasty (n = 7), resection of prior ileocolonic anastomosis (n = 5), segmental colectomy (n = 1), and lysis of adhesions (n = 1). Forty-five procedures (40%) were converted as a result of adhesions (n = 21), extent of inflammation or disease (n = 9), size of the inflammatory mass (n = 7), inability to dissect a fistula (n = 5), or inability to assess anatomy (n = 3). Factors associated with conversion were internal fistula as an indication for surgery, smoking, steroid administration, extracecal colonic disease, and preoperative malnutrition. In laparoscopic patients, mean times to passage of flatus and first bowel movement were 3.6 ± 0.2 days and 4.4 ± 0.2 days, respectively. Mean time to discharge was 6 ± 0.2 days.

Conclusions

Attempted laparoscopic management is safe and effective if there is an appropriate threshold for conversion to an open procedure. Conversion factors identified in this study largely reflect technical challenge and severity of disease. Patients taking steroids and those with known fistulas or colonic involvement threaten laparoscopic failure, but many of these patients can be managed laparoscopically and have better outcomes. By understanding the reasons for conversion, it is hoped that the chances of laparoscopic success can be improved by modifying standard preoperative medical management or using additional technological capabilities (e.g., robotics).

In patients with Crohn’s disease, complications often develop that require surgical intervention. Such complications include obstruction, fistulization, bleeding, infection, and intractability. In most situations the surgical remedy involves some form of enteric resection. Traditionally, these patients have required celiotomy, but more recently a few centers have been selecting patients suitable for laparoscopically assisted approaches to enteric resection. Early results of these series have been promising. Indeed, our group and others ^1–3 have shown that laparoscopic management of patients with complications of Crohn’s disease produces better outcomes than traditional open approaches. Specifically, in our retrospective cohort analysis comparing laparoscopic and traditional open approaches, we have shown significant improvements in blood loss, length of stay, return of bowel function, and functional recovery indices. ¹

Not all patients with complications of Crohn’s disease may be candidates for laparoscopic management. Laparoscopic approaches are contraindicated in patients with a known frozen abdomen and may be contraindicated in patients needing emergent surgery (e.g., active hemorrhage, peritonitis, complete bowel obstruction). These situations aside, it is difficult to determine before surgery which patients with complications of Crohn’s disease are amenable to laparoscopic management. Crohn’s disease is a transmural inflammatory condition that can involve not only the bowel but adjacent organs as well. The degree of inflammation can be variable and unpredictable, which in turn affects the technical complexity and safety of surgery and the likelihood of laparoscopic success.

Our objective was to determine factors associated with intraoperative conversion from laparoscopic to the open approach. We offered a laparoscopic attempt to virtually all patients (with the exceptions noted above). By determining such factors, we hope to inform our patients before surgery regarding their chances of laparoscopic success. In addition, by understanding the reasons for conversion, we hope to improve our chances of laparoscopic success by modifying standard preoperative medical management or using additional technological capabilities (e.g., robotics).

PATIENTS AND METHODS

Patient Population

One hundred ten consecutive patients referred between January 1993 and June 2000 requiring surgical procedures for Crohn’s disease were offered a laparoscopically assisted procedure. The only excluded patients were those with a known frozen abdomen or those with urgent or emergent complications of Crohn’s (active hemorrhage, peritonitis, or complete bowel obstruction). During the informed consent discussion, patients were clearly apprised about the early stages of development of this approach and were offered the traditional alternative. Because conversion to an open technique is readily accomplished in the same setting, all patients agreed to an initial attempt at a laparoscopically assisted procedure.

Preoperative Evaluation

In addition to evaluation by history and physical examination, patients were evaluated before surgery with small bowel contrast radiography, computed tomography scanning, and in most instances colonoscopy. These studies take on added importance in the absence of the ability to palpate the bowel exhaustively. Some patients had undergone previous bowel resection and others had known internal fistulas.

Surgical Technique

Patients received standard bowel preparation. In the operating room, under general anesthesia, a pneumoperitoneum was established by Veress needle insufflation or direct fascial incision (Hassan technique) just above the umbilicus. An initial laparoscopic survey of the abdomen was accomplished using the video telescope. Based on this examination, at least two auxiliary trocar locations were chosen, usually just above the symphysis pubis and in the upper abdomen in the midline. When helpful, a third auxiliary port was placed in the lower midline. All auxiliary trocars were usually 5 mm.

In most instances, the terminal ileum and cecum was the primary disease site. In these situations, the cecum was grasped and retracted medially toward the midline. The line of Toldt was then carefully incised and the retroperitoneum entered. The ureter was identified in the retroperitoneal space where it crosses the internal iliac artery and was protected. If necessary, the hepatic flexure region was also dissected free using an ultrasonic dissector, electrocautery, or clips. The dissection was continued until the mobilized bowel and mesentery was free enough to reach the abdominal wall. In some instances, the mesentery was divided laparoscopically. A small incision was then created in the midline, the right lower quadrant, or a suprapubic area (Pfannenstiel). The bowel targeted for resection was then delivered through this incision onto the abdominal wall. The actual resection and anastomosis, including division of mesenteric vessels in some instances, was then accomplished on the abdominal wall in a standardized fashion. The mesenteric defect was then closed and the bowel placed back into the abdominal cavity. Using retractors and the laparoscopic telescope, the abdomen was irrigated through the small incision and evaluated for hemostasis. The incision was closed in a standard fashion. Laparoscopically assisted diversion was accomplished by simply mobilizing the target bowel, delivering it through a premarked ostomy site, and tacking it to the external rectus sheath.

A procedure was defined as a successful laparoscopically assisted procedure when the incision used to deliver the bowel was 5 cm or less.

Data Collection and Analysis

Data were extracted from a prospectively maintained database and from medical records. Subtypes of Crohn’s disease were analyzed based on medical records. To analyze surgical outcomes, each procedure was assigned a symptomatic resolution score from 0 to 3 (0 = worse, 1 = little or no change, 2 = marked improvement, 3 = complete resolution) based on evaluation at postoperative clinic visits. For all data, differences between groups were analyzed using the unpaired, two-tailed Student t test or chi-square analysis as appropriate.

RESULTS

The demographic data and laboratory values are summarized in Table 1. Of the 110 patients, 66 (60%) underwent 68 surgical procedures that were completed laparoscopically. The remaining 44 (40%) underwent 45 attempted laparoscopic procedures that were converted to open procedures. The mean age at diagnosis was 26 ± 12 years, and the mean age at the time of surgery was 37 ± 12 years. Age differences were not statistically different between the laparoscopic and converted groups, nor was the interval from diagnosis to surgery. The sex distribution was predominantly female in both laparoscopic and converted groups. There was a greater proportion of men in the converted group, but this was not statistically significant.

Patients were categorized by subtype of Crohn’s disease. ¹ All patients were categorized into at least one of three main subtypes based on both current presentation and past medical history: fistulizing, obstructive, and colonic (involvement of the colon other than cecal disease continuous with terminal ileal disease). The most common of the main subtypes was obstructive. There were no differences in the prevalence of the obstructive subtype in the laparoscopic (81%) versus the converted (80%) groups. There was a smaller percentage of fistulizing disease in the laparoscopic group (37%) versus the converted group (47%), but this was not statistically significant. Patients with colonic Crohn’s disease were less common in the laparoscopic group (13% vs. 29%, P < .05).

A history of perianal disease, duodenal disease, extraintestinal manifestations, and irritable bowel syndrome was also recorded. Patients with perianal involvement (16% vs. 32%, P < .06) and irritable bowel syndrome (4% vs. 15%, P < .10) were less common in the converted group, but these differences did not reach statistical significance. There were no differences between the groups in terms of duodenal involvement or extraintestinal manifestations of Crohn’s disease.

Hematocrit, white blood cell count, blood urea nitrogen, creatinine, albumin, and erythrocyte sedimentation rate were all evaluated. The means of these laboratory values all fell within the normal range. There were no statistically significant differences in laboratory values between the groups. However, the converted group’s mean white blood cell count (9.8) compared with the laparoscopic group (8.1) approached significance (P = .07).

Pain was the most common complaint from patients (86% overall, 84% in the laparoscopic group, 89% in the converted group), followed by diarrhea (40% overall, 38% in the laparoscopic group, 42% in the converted group). Several other symptoms, including nausea, vomiting, weight loss, and bloating, occurred in approximately one third of patients. Fever, fatigue, bleeding, flatulence, and heartburn occurred in less than 20%. There were no statistically significant differences in presenting symptoms between the laparoscopic and converted groups. Time from onset of the patient’s chief complaint to surgical management was also recorded. This interval tended to be longer in the laparoscopic group (2.1 vs. 1.4 months, P = .20), although this was not statistically significant.

Mesalamine was the most common preoperative medication used, followed by prednisone (67%), metronidazole (38%), azathioprine (20%), and cyanocobalamin (7%). There were no statistically significant differences in use of these medications between the groups, with the notable exception of prednisone: prednisone was used in 59% of patients in the laparoscopic group and 80% of patients in the conversion group. This result was statistically significant (P < .05) (Table 2).

Obstructive symptoms (77%) were the most common indication for surgery in both groups. This was followed by “failure of medical management” (35%). There was no statistically significant difference between the groups for either of these two common indications. The presence of a fistula by preoperative contrast radiography was the determinant for surgery in 27% overall, 42% of patients who required conversion, but only 18% of patients whose procedures were completed laparoscopically (P = .004). Gastrointestinal bleeding and perineal sepsis were indications for surgery only in the laparoscopic group and accounted for less than 5% of these patients (Table 3).

Ileocecectomy (n = 76) was the most common procedure performed in both the laparoscopic (n = 46) and converted (n = 30) groups. Small bowel resection (n = 31) was the next most common procedure: 22 were completed in a laparoscopically assisted manner and 9 required conversion to an open procedure. The next most common procedures, in order of frequency, were stricturoplasty (n = 19), fistula takedown (n = 12), resection of previous ileocolonic anastomosis (n = 12), fecal diversion (n = 8), segmental colectomy (n = 5), and lysis of adhesions (n = 4). Conversion rates varied depending on the procedure. The lowest conversion rates occurred during fecal diversion (12%), stricturoplasty (21%), enterectomy (29%), and ileocecectomy (39%). The highest conversion rates were for segmental colon resection (80%) and lysis of adhesions (75%). The conversion rates that approached parity were resection of prior anastomosis (58%) and fistula takedown (58%) (Table 4).

Adhesions, defined as tissues too adherent to one another for safe dissection, were cited as the most common reason for conversion from laparoscopy to open technique (47%). Although adhesions were the most common reason for conversion, neither the number of prior abdominal procedures nor the presence or absence of previous abdominal surgery predicted conversion. The next most common reasons for conversion included extensive inflammation or disease (18%) and the size of the inflammatory mass (16%). The inability to dissect a fistula resulted in conversion in five patients (11%), and inability to assess anatomy was responsible in three patients (7%) (Table 5).

Mean operative time for the laparoscopic group was less than for the converted group (207 vs. 219 minutes), but this difference was not statistically significant. Estimated blood loss was significantly less in the laparoscopic group (121 vs. 202 mL, P = .05). Intravenous fluid administration was greater in the converted group (3.29 vs. 2.82 L), but this difference was not statistically significant. Correspondingly, urine output was significantly greater in the converted group (579 vs. 311 mL, P = .05). The surgeon, assistant, or experience of the assistant did not significantly correlate with the conversion rate. There were no intraoperative complications.

Mean time to passage of flatus and bowel motion was less for the laparoscopic group (3.4 and 4.2 days vs. 3.7 and 4.6 days), but this difference was not statistically significant (P = .2). Mean time to liquid diet and solid food tolerance was also less for the laparoscopic group (3.1 and 4.9 days vs. 4.1 and 5.1 days). The difference in days to liquid diet tolerance was statistically significant (P = .01). Finally, length of stay was significantly reduced in the laparoscopic group (5.7 vs. 6.5 days, P = .05).

There were three major complications in our series. The combined major and minor complication rate was 13% for the laparoscopic group and 16% for the converted group. Two patients in the laparoscopic group had major complications: cerebrovascular accident and pelvic abscess. There was one complication in the converted group, a postoperative large bowel obstruction that required reoperation. Three patients were readmitted after initial hospital discharge, two in the laparoscopic group and one in the converted group. The symptomatic resolution score was 2.65 in the laparoscopic group and 2.59 in the converted group (difference not significant) (Table 6).

DISCUSSION

Patients with complications of Crohn’s disease often require enteric resection. Traditionally, these patients have required celiotomy, but a few centers are selecting patients suitable for laparoscopically assisted approaches. Several retrospective series have reported that laparoscopic management of patients with complications of Crohn’s disease produces better outcomes than traditional open approaches. Specifically, in our retrospective cohort analysis comparing laparoscopic and traditional open approaches, we have shown significant improvements in blood loss, length of stay, return of bowel function, and functional recovery indices. ¹ Bemelman et al ² compared patients who underwent open versus laparoscopically assisted ileocecectomy and showed a significantly shorter length of stay in the laparoscopic group. Alabaz et al ³ also compared patients who underwent ileocecectomy using open or laparoscopically assisted approaches. Patients in the laparoscopic group had a significantly shorter length of stay and a quicker return to work. In addition, at a mean follow-up of nearly 3 years, significantly more patients in the open group had bowel obstruction than in the laparoscopic group.

Because the results of these series were promising, we wished to offer this laparoscopically assisted option to as many patients as possible. We have not offered it to patients with a known frozen abdomen or patients who need emergent surgery. Several studies have addressed appropriate indications for laparoscopic surgery in Crohn’s disease in more detail. ^4–6 With these few exceptions, we have offered a laparoscopic attempt to all patients with complications of Crohn’s disease needing enterectomy. In so doing, we hoped to understand the reasons for laparoscopic failure (i.e., intraoperative conversion from the laparoscopic-assisted to the open approach).

Conversion rates did not appear to be significantly influenced by age at surgery, age at diagnosis, or gender. Conversion rates did, however, vary by disease subtype. Patients with the colonic (extracecal) subtype were less common in the laparoscopic group (13% vs. 29%). We believe this is explained by two factors. First, colon resection is technically challenging, particularly in patients with Crohn’s disease, because the transmural inflammation and foreshortened mesentery make things precarious. ⁷ Second, patients with the colonic subtype appear in general to have a higher severity of disease, making them less amenable to laparoscopic approaches.

There was also a smaller percentage of fistulizing subtype disease in the laparoscopic group (37%) compared with the converted group (47%), although this did not reach statistical significance. This trend is consistent with our finding that internal fistula as an indication for surgery was significantly associated with conversion to an open procedure. It may also be the case that patients with this disease subtype (as with colonic involvement) may in general have a higher severity of disease, making them less amenable to laparoscopic approaches.

Patients with perianal disease and irritable bowel syndrome showed trends toward being amenable to the laparoscopic approach. Few patients in this series underwent surgery secondary to their perianal disease. This once again suggests that patients with these manifestations of Crohn’s disease are more amenable to laparoscopic approaches. Finally, smoking was significantly associated with conversion, most likely reflecting severity of disease, given that smoking is known to exacerbate Crohn’s disease.

As is common with Crohn’s disease, many of these patients (nearly 50%) had undergone one or more previous surgical procedures. Wu et al ⁸ have also investigated the importance of recurrent disease in laparoscopic Crohn’s surgery. They concluded that the recurrent disease is not a contraindication to a laparoscopic approach. In the current series, we reached a stronger conclusion; namely, there were no statistically significant differences between the laparoscopic and converted groups in terms of the presence of prior abdominal surgery, or the number of prior abdominal procedures.

There were no statistically significant differences in presenting symptoms between the laparoscopic and converted groups, although the presence of a mass on physical examination nearly reached significance (more common in the converted group, P = .08). Our study also examined the duration of the patient’s chief complaint before surgical intervention. There was a trend suggesting that patients who endured their symptoms longer before surgical intervention were more likely to have a procedure that was completed laparoscopically (2.1 vs. 1.4 months, P = .2). It is possible that optimization of medical management (by means of bowel rest, parenteral nutrition, and time), allowing inflammation to subside before surgery, may allow more patients to undergo successful laparoscopic procedures.

Several different procedures were performed during the operations in this series, with some patients having multiple tasks accomplished during one surgical event. Most common was bowel resection with extracorporeal anastomoses. Conversion rates varied depending on the procedure attempted and in general reflected disease severity and technical difficulty. These aspects of laparoscopic surgery for Crohn’s disease likely will be improved by technologic advancements such as surgical robotics and improved imaging.

The presence of a fistula (excluding enterocutaneous or perirectal fistulas) as the indication for surgery was associated with conversion to an open procedure. A fistula is more likely to present a technically difficult situation at surgery. Some fistulas were entirely within the resected specimen and did not require fistula division. Others involved adjacent organs, such that success with the laparoscopic approach required division of the fistula and repair of the “innocent” involved tissue. Such an endeavor is more difficult and more concerning in terms of safety.

Adhesions, such that the surgeon could not safely separate the target organ from surrounding tissues, were the most common intraoperative reason given for conversion to an open procedure. Despite this (as stated earlier), there was no trend of greater conversion rates in patients who had undergone prior abdominal surgery. Thus, these adhesions were not just from prior surgery, but also from active inflammation. Adhesions, as well as the second most common reason for conversion (extent of disease or inflammation and size of inflammatory mass), theoretically might also be improved by a period of bowel rest before surgery.

Among medications taken, only steroids were significantly associated with conversion to an open procedure. This finding likely reflects severity of disease rather than any specific effects of the steroids themselves. Studies have shown that steroids are indeed a risk factor for postoperative complications. ⁹

In addition to steroid administration and smoking, patients with the clinical evaluation of being malnourished in the preoperative period (as determined by gastroenterologist or surgeon) were more likely to require conversion to an open procedure (P = .04). In some patients designated as malnourished, surgery was postponed to improve their status. However, actual preoperative weight (as determined at the preoperative anesthesia clearance visit), weight loss (>10 lb as determined by patients), and total parenteral nutrition use were not significantly associated with conversion. Reasons for this may be the lack of quantification of weight loss into severity categories and the few numbers of patients receiving TPN in our series.

Surgical outcomes in this series of patients with complications of Crohn’s disease yielded expected results. Patients with completed laparoscopic procedures had less blood loss, received less intravenous fluids, and made less urine. They also had swifter return of bowel function and a shorter hospital stay. Outpatient variables were comparable in terms of reoperation, readmission to the hospital, complications, and symptom resolution. Symptom resolution as measured on our symptom resolution scoring system was equivalent between the groups. Thus, in this nonrandomized comparison, both modes of surgery were equally effective in improving symptoms. The paradigm of offering a laparoscopic approach to “all comers,” with conversion being the key judgment, resulted in excellent symptom resolution for the entire series of patients.

SUMMARY

In this series of patients requiring surgery for Crohn’s disease, we attempted a laparoscopic approach in all, hoping to provide benefit to the maximum number of patients. Our results support the safety and efficacy of such an approach, as reflected in a low complication incidence and excellent symptom resolution scores. There is no harm in an initial laparoscopic attempt if the surgeon has a safe conversion threshold, and the most sensitive determinant of laparoscopic success is an attempt at laparoscopic management. Clearly, patients taking steroids and those with known fistulas or colonic involvement threaten laparoscopic failure, but many of these patients can be managed laparoscopically and have better outcomes.

Analysis of our results shows that fistula as an indication for surgery, smoking, steroid administration, colonic involvement subtype, and preoperative malnutrition were significantly associated with conversion to an open procedure. Three of the factors (smoking, steroid administration, and malnutrition) probably are associated with conversion because they indicate severity of disease. The other factors (colonic involvement and fistula as an indication for surgery) reflect severity of disease subtype and present more of a technical challenge. Measures that reduce severity of disease at the time of surgery such as bowel rest, total parenteral nutrition, and time may increase the likelihood of success with the laparoscopic approach. Conversely, it is possible that surgery at an earlier, less severe stage of the disease may increase the likelihood of a successful laparoscopic procedure, although data from this series do not directly support such a position. A randomized prospective trial may help determine these endpoints. Future improvements in minimally invasive technology will likely address the technical challenges that resulted in conversion in some of these patients. Knowledge of the factors associated with success or failure of the laparoscopic approach will be useful in preoperative preparation and counseling of patients.

Discussion

Dr. Susan Galandiuk (Louisville, Kentucky): Dr. Lawrence, Dr. Townsend, members, and guests. I have some questions for Dr. Talamini. I think laparoscopic surgery has now been accepted as safe in a number of conditions, including Crohn’s disease. I have some questions for Dr. Talamini regarding his patient population. The age of the population is much older than in most series. These patients were an average of 37 years, whether they were converted or done entirely laparoscopically. What do they attribute this to?

There is a natural progression of Crohn’s disease. With increasing duration of disease, there is an evolution of the very acutely edematous disease to much more fibrotic disease. Is that the reason why more of these patients had obstruction as their indication for surgery? In my practice, there is a much, much larger group of patients who have failed medical management. Here only 68% of patients were on steroids.

There is an increase in reporting in the GI literature of patients being treated with infliximab (anti-tumor necrosis factor antibody) who have the treatment with this closure of fistulas, but develop obstruction. Was true in Dr. Talamini’s patients?

Increasing the length of medical management does lead to an increasing complexity of disease, with a much larger proportion of patients having enteroenteric fistulae phlegmons. Was this also noted in his group of patients?

With respect to the duration of the disease of Crohn’s, did that differ in the patients that were undergoing laparoscopic procedures versus those that had to be converted?

In my experience, the hand-assisted laparoscopic (HAL) technique has been very useful in Crohn’s patients because there are many cases where the current laparoscopic instruments don’t permit you to assess the degree of fibrosis in this. Were any of these patients done by the HALS technique, or was that used prior to converting them?

Thank you very much for the privilege of the floor.

Dr. John H. Pemberton (Rochester, New York): Thank you for the privilege of asking a couple of questions. This was a great paper, and I think it is going to really add to the surgical literature on laparoscopic techniques. As laparoscopic techniques have definitely arrived, there is little doubt that full laparoscopic approaches and laparoscopic-assisted approaches for straightforward Crohn’s disease is a technique that is important now and will continue to become more important as expertise is gained throughout the United States in different programs and learning curves, both on the part of faculty and students, is negotiated.

In our experience of about 100 patients with Crohn’s disease, which is primarily the series done by Tonya Fadoc, one of my junior colleagues, the conversion rate for all comers is about 20%. Now when phlegmon, abscess, fistula, and previous surgery have not been a part of the patient’s history, then the conversion rate is only 4%, which tends to support what you have already said. Could you make a comment about that?

With your data set now how have you changed your approach to the patient who needs an operation for Crohn’s disease? Are you more specific in the type of patient that you operate on now or offer that operation to? Or, interestingly, because the conversion time was no different than the full laparoscopy time, do you just give it to everybody?

And, finally, something that we have been looking at, given our area of the country, we think that body mass index may have something to do with the ability of surgeons to do laparoscopic operations in the first place and to get them done without converting in the second place. Has that been a part of your data set? And if not, perhaps you might look at that again in the future.

Thank you for the opportunity to ask these questions.

Dr. Bruce V. MacFadyen (Houston, Texas): Dr. Lawrence, Dr. Townsend, members, and guests. Thank you for the opportunity of discussing this paper. This is a very interesting paper because it is addressing the issue of complicated bowel management laparoscopically, and there are very few papers in the literature that are really addressing this particular problem. Particularly in Crohn’s disease, it is a disease which challenges us as surgeons for our judgment and surgical skills. Mobilizing thickened mesentery and dealing with the inflammatory mass, the possibility of ureteral injury, fistulas, underlying malnutrition, numerous adhesions, and oftentimes abscesses will be a great challenge to all of us.

Laparoscopically, there can be disadvantages, potential disadvantages, with visualization; instrumentation is not perfect and needs to certainly improve. But there are advantages, if it can be done laparoscopically, with earlier return to work.

In most series that have been reported thus far, the complication rate has been reported between 10% and 18% or 20% for small bowel disease and up to 30% for those with colonic disease. This series emphasizes a very low complication rate of approximately 2%, although the conversion rate is much higher than reported in other series, of 60%. It should be noted that these patients did have a very low complication rate, which would mean that the overall result certainly has been very satisfactory.

Now you had mentioned before that all comers were taken for this procedure, although in your paper you had mentioned the fact that dense adhesions were an issue that would contraindicate laparoscopic approach. And I would ask you how you determine dense adhesions preoperatively. But other series have also talked about the potential for intraabdominal abscess, intraabdominal infection, or complete bowel obstruction as being a potential contraindication.

Adhesions were certainly one of the major reasons for conversion, as also was inflammatory disease, inflammatory mass, and the other issue that would come up is what about skip lesions and how do you accurately assess those intraoperatively? Or was that all made by preoperative determination with an upper GI small bowel series, etc.?

I think that overall this is an excellent paper because a good result was obtained with a very low complication rate. Although the conversion rate is higher than normal, I think it should not be considered a complication that one did convert, and I would ask you what you defined as a true conversion. You mentioned something about an incision greater than 5 cm, but as was mentioned previously by one of the discussants, the idea of a hand-assisted approach would put my conversion rate higher, with my hand being approximately 7.5 cm in size.

The other question I would ask is, what is your recommendation for colonic disease? It appears that segmental colonic resection did have a high conversion rate in your series. Are you recommending that at this stage in our development of the management laparoscopically of this problem that we not treat the colonic problems laparoscopically?

What about external fistulas? Is that differently managed than internal fistulas? And what about all that matted bowel together when you get in there and you see all these complications? Should we take time to lyse those adhesions or should we convert immediately to an open operation?

I think that the emphasis that Dr. Talamini mentioned about a prospective randomized trial is extremely important. I think, though, that we need to deal also with cost/benefit issues, although this paper did not address those issues. Because when you think about preoperative bowel rest and considering TPN preoperatively, you may be talking in the range of $500 to $700 a day, and that may mean that it would be much more cost-effective to proceed with an open procedure rather than going ahead and giving preoperative TPN unless they were significantly malnourished.

Again, I want to congratulate Dr. Talamini on the excellent paper and the presentation, and the Society for the opportunity to discuss it.

Dr. Mark A. Talamini (Baltimore, Maryland): I thank Dr. Galandiuk, Dr. Pemberton, and Dr. MacFadyen for their excellent and insightful questions. I will do my best to answer them.

In terms of the age of our series, I think that you are exactly right. Our patients are a little bit older, and I would attribute this to the excellent and persistent care of our gastroenterologists. One of my coauthors, Dr. Bayless, is certainly one of the world’s finest clinicians when it comes to the care of Crohn’s disease. And I think he simply sends to me a set of patients that are very well cared for and therefore get farther along in age before they require surgery. As we all know, most patients with Crohn’s disease eventually come to the operating room sooner or later.

In terms of Inflexomed, it is a bit early for that in our series or in our patients at Hopkins, I think, to affect this sort of surgery—although I have had a few patients where Inflexomed has enabled us to take a very difficult enterocutaneous fistula patient, close that fistula, and then allow us to get them to the operating room, which also brings up the point of enterocutaneous fistulas themselves. Those we did not attempt to do anything laparoscopically with, recognizing that that would be a technical feat well above our heads.

In terms of hand-assisted laparoscopic surgery, we define conversion as greater than 5 cm, so we did not pursue the hand-assisted technology. We found that even in most of the patients who required conversion, that often meant going from 5 cm to 8 or 9 cm or maybe 10 cm, and with an incision the large, by simply retracting aggressively, we could get what we needed to do accomplished.

In terms of Dr. Pemberton’s questions, body mass index in our series has not been a problem. We did look at weight, preoperative weight, and it was not a factor affecting whether patients needed to be converted, although there were not many patients who were overweight in this series.

Our patient selection really has not changed. The theme of our series was that we would take virtually all comers and be willing to convert and make conversion the key judgment. So we still pretty much will take all comers, the exclusions being pancolonic disease with a contracted mesentery which, quite frankly, is simply above my laparoscopic skill currently, and those who we know from our Hopkins experience have had five, six, seven laparotomies and the previous one was horrible and we knew there just would be no way to actually get into the abdomen.

In many of these patients the key judgment about whether we would be successful was within the first 30 seconds, when we attempted to see if we had open space within the abdomen to enable a pneumoperitoneum in a visual field. If we could get that, in most cases we could work to free up the bowel and eventually succeed. So in most cases, that was what decided things.

In terms of a prospective trial, early on in the series I tried to convince our gastrointestinal practicing colleagues to do a prospective randomized trial. Unfortunately, to them it was not an unanswered question. There was simply no issue: they wanted their patients to be approached in this way. Their sense was, from early on, that there would be no disadvantage to at least trying. So in our referral base we never really had an opportunity to do a randomized prospective trial early on. However, this issue of early or late surgery, I think, may be something that we might be able to approach that way.

I’d like to, once again, thank the Society for being able to present.