Reversal of the Hartmann's procedure: A comparative study of laparoscopic versus open surgery.

BACKGROUND: The Hartmann's operation, although less frequently performed today, is still used when initial colonic anastomosis is too risky in the short term. However, the subsequent procedure to restore gastrointestinal continuity is associated with significant morbidity and mortality. PATIENTS AND METHODS: The review of an institutional review board (IRB)-approved prospectively maintained database provided data on the Hartmann's reversal procedure performed by either laparoscopic or open technique at our institution. The data collected included: demographic data, operative approach, conversion for laparoscopic cases and perioperative morbidity and mortality.
RESULTS: Over a 14-year period from January 1997 to August 2011, 74 Hartmann's reversal procedures were performed (laparoscopic surgery-49, open surgery-25). The average age was 55 years for the laparoscopic and 57 years for the open surgery group, respectively. Male patients represent 61% of both groups. There was no significant difference in operative time between the two groups (149 min vs 151 min; P = 0.95), and there was a tendency to lower morbidity (3/49-7.3% vs 4/25-16%; P = 0.24) in the laparoscopic surgery group. In the laparoscopic group, eight patients (16.3%) were converted to open surgery, mostly due to severe adhesions. The length of hospital stay was significantly shorter for the laparoscopic group (5 days vs 7 days; P = 0.44).
CONCLUSIONS: The Hartmann's reversal procedure can be safely performed in the majority of the cases using a laparoscopic approach with a low morbidity rate and achieving a shorter hospital stay.

INTRODUCTION

The most serious cause of morbidity and mortality in emergency colonic resection is anastomotic dehiscence and subsequent sepsis. Left colonic resection associated with terminal colostomy and rectal closure, also known as the Hartmann's procedure, has been the tried and true approach to avoid this risk when this anastomosis is risky.[12] Common indications for the Hartmann's procedure include perforated tumours, complicated diverticulitis, trauma, or patients with serious comorbidities where primary anastomosis is considered tenuous.

Although the Hartmann's procedure can reduce the risks associated with the index operation, there is a significant volume of data suggesting that the subsequent reversal surgery is associated with significant morbidity and even mortality.[123] The majority of these data have been associated with a repeat laparotomy. The most common complications include surgical site infection, wound complications and unfortunately anastomotic leak and death. The increasing adoption of laparoscopic colon surgery has been associated with a global reduction in operative morbidity with colon resection. Despite these significant benefits for index colectomy, there are limited data regarding the role of laparoscopic techniques for laparoscopic Hartmann's reversal.[4] The primary purpose of this study is to compare the outcomes of laparoscopic and open surgery approaches to the Hartmann's reversal surgery and to describe the experience and characteristics of the patients undergoing this procedure at our centre.

PATIENTS AND METHODS

After obtaining approval from our institutional ethics committee and review board, a prospective database was prepared compiling the data on the patients treated under the Hartmann's procedure reversal between January 1997 and August 2011. All the patients underwent their original surgery by open surgery. After obtained informed consent all the patients, the following data were recorded: Patient's demographics; ASA score; causes of the Hartmann's procedure, reversal technique and conversions (in the laparoscopic group patients who need to do laparotomy to complete the surgery); complications in the first 30 days; start of liquid oral intake; and length of hospital stay. The open-surgery group included patients with associated ventral hernias, and the surgeries were performed preferentially by surgeons without experience in laparoscopic colonic surgery. Results and group characteristics for each surgical approach were analysed using t-test or Chi-square test accordingly, with significance set at P < 0.05.

Surgical technique

Laparoscopic approach: The intervention is performed in modified lithotomy position. The first trocar (12 mm) is introduced into the right lower quadrant using open technique. The second and third trocars are introduced under direct vision and into the right hypochondrium (5 mm) and at the umbilical level (10 mm). Some cases required a 5-mm fourth trocar in the left lower quadrant. The rectal stump is identified and freed from adhesions. The head of a circular stapler (No. 29) is installed in the entrance of the proximal colon. Subsequently, the colostomy opening in the abdominal wall is closed. A circular stapler is introduced transanally and the colorectal intracorporeal anastomosis is performed under laparoscopic vision. An air leak test is performed routinely by insufflating air via the anus with the anastomosis immersed under isotonic saline solution. If an air leak is identified intracorporeal 3/0 polyglactin sutures are used for reinforcement. Trocar sites are closed using polyglactin 0 suture. Skin incisions are sutured with 3/0 Nylon. Open-surgery approach: The patient position is the same as previously described and a midline infraumbilical laparotomy incision is used for abdominal access. The anastomotic segment is freed and the anastomosis is fashioned in a similar manner as described above for the laparoscopic approach. Laparotomy and colostomy openings are closed with uninterrupted suturing of the aponeurosis using Polydioxanone suture 1/0, and the skin is closed with 3/0 Nylon. Drains are not routinely used in both open and laparoscopic surgery.

RESULTS

Seventy-four patients were included in this study. Forty-nine patients underwent laparoscopic surgery and 25 patients had open surgery. Patient characteristics are described in Table 1. The average age and gender distribution were similar between the two groups [Table 1]. The majority of the patients in both groups had an ASA score ≤2 [Table 1]. The most frequent indications for the Hartmann's procedure in the open-surgery group were as follows: Obstructive rectosigmoid cancer in 11 cases (44%) and perforated diverticular disease in 5 cases (20%). In the laparoscopic group, the most frequent indications were perforated diverticular disease that occurred in 31 patients (63%) and colorectal carcinoma that occurred in 6 patients (12%) [Table 1]. Two patients only in the open-surgery group had an associated midline ventral hernia. There was no significant difference in operative time or time to resume oral intake between the two groups [Table 1]. Eight patients were converted to open surgery (16.3%), among them five due to severe multiple adhesions and one each due to difficulties to find the rectal stump, poor visualization due to obesity and intraoperative anastomosis failure [Table 2]. In the laparoscopic group, three complications were noted during the first 30 days (7.3%): One unnoticed jejunal lesion in a patient with severe adhesions who was re-operated on the second day, one haemorrhage at the suture line treated endoscopically and one colostomy site infection. In the open-surgery group, four complications (16%) were noted: One pelvic hematoma, one rectovesical fistula, one evisceration (all of these patients were re-operated) and one case of colostomy site infection [Table 3]. The patient with evisceration did not have previous ventral hernia. The length of hospitalization was lower in the laparoscopic group with a median of 5 days versus 7 days in the open-surgery group. This difference is statistically significant (P = 0.044). We reported no mortality in any of the two groups. The average follow-up time of the patients was 3 months for the laparoscopic and 5 months for the open-surgery group.

Table 1

Demographics and clinical data

Table 2

Conversion

Table 3

Thirty-day morbidity

DISCUSSION

The Hartmann's operation described by Henry Hartmann[56] in 1923 for rectal cancer is a technique that is still used in emergency colon surgery because of its safeness in patients at higher risk of anastomosis dehiscence although not without morbidity and mortality, especially in patients with others comorbidities.[7] These patients require a two-stage surgical procedure to restore normal intestinal transit. This second stage also has shown important morbidity rates, which in some cases have given reason to doubt the possibility of restoration.[3689] With the development of laparoscopic colon surgery in the past years, restoration has become one of the procedures performed laparoscopically, in order to use the advantages of this less invasive method to reduce these complication rates. However, enthusiasm for using the laparoscopic approach for this condition has been limited, probably due to technical difficulties.[1011121314] A recent review performed by van de Wall showed a tendency of the laparoscopic approach having lower rates of complications and also lower length of hospitalization.[8] However, some of the reviewed data have only a small series and a few comparative studies.[815] Our comparative study is until now the largest single centre experience in laparoscopic Hartmann's reversal. Some studies show that postoperative morbidity rates in the laparoscopic approach are similar or lower to open surgery, being recorded at 30-50% in open cases[16] and approximately 15% in the laparoscopic ones.[1112131415] These numbers match up with our series with an early morbidity of 7.3%. Haughn et al. has shown that the most frequent morbidity cause in the series has been colostomy wound infection. Interestingly, in this last series, the morbidity after 6 months has also been higher in the open-surgery group than in the laparoscopic group, with incisional hernia being the main complication.[14] In our open study group, there was one case of evisceration, a complication that could have been avoided using the laparoscopy approach. On the other hand, there was no mortality either in the laparoscopic group or in the open-surgery group. This result matches up with most laparoscopic series,[81112131415] which confirms this procedure as safe. However, open-surgery series still show mortality rates between 0.6% and 1.7%.[369] Our study has some inherent limitations because of the nature and specific indications for the procedure and randomization was not possible. On the same line, our open group present more patients with malignant disease in the original surgery than the laparoscopic group, a fact that is also present in some series. Before the COST study, the available evidence about laparoscopic approach and cancer dissemination made us to prefer an open reconstitution for those cases.[17] This fact explains the higher incidence of malignant disease in the original surgery in the open reconstitution group. Operating time in the laparoscopic group (149 min) was not greater than that in the open-surgery group, and these results were similar to that of other series[14161819] and differ from the usually longer time of laparoscopic colon surgery, although the open-surgery group included two patients with incisional hernia who required a laparotomy to repair it, which could have extended the operating time. It is likely that laparoscopy results can be associated with similar operating times, due to the reductions in time for lysis of adhesions and closure of the laparotomy wound. In our experience, the hospital length of stay was significantly lower and this seems to be one of the greatest advantages of the laparoscopic technique used in our series. Similar findings appear in other previous comparative series.[111415161819] With respect to the initial approach, there is no consensus about the best technique. Most series make an open transumbilical Hassan's approach or start with the colostomy liberation.[20] Our technique starting with an open approach in the right inferior quadrant offers the advantage of starting far from the usual adherences at the left side and the greater omentum adherent to the original midline laparotomy. Our conversion rate of 16.3% is comparable to that of other series (7-22%) predominantly due to the management of adhesions to the abdominal wall or rectal stump.[1113] We believe that the case of delayed recognition of the jejunal enterotomy may have been avoided with a conversion. Apart from that, the 83.7% successful laparoscopic completion rate suggests that a prior laparotomy is not a contraindication to laparoscopic re-operative colon surgery, even in some cases with previous peritonitis. It is difficult to predict the tenacity of adhesions and many can be managed laparoscopically, suggesting that an initial laparoscopic approach should be considered routine. In our practice, laparoscopy is the technique of choice for the Hartmann's reversal, which is why, during the period under analysis, there were more cases of laparoscopic than open surgery.

Laparoscopic Hartmann's reversal procedure is a technique that can avoid relaparotomy and is associated with the advantages of this less-invasive approach, including low morbidity and faster recovery. We therefore recommend laparoscopy as the technique of choice for the Hartmann's reversal procedure.

Financial Support and Sponsorship

Nil.

Conflicts of Interest

There are no conflicts of interest.