World J Surg DOI 10.1007/s00268-014-2507-1
Laparoscopic Versus Open Reversal of a Hartmann Procedure: A Single-Center Study Markus Zimmermann • Martin Hoffmann • Tilman Laubert • Karl-Frederik Meyer • Thomas Jungbluth • Uwe-Johannes Roblick Hans-Peter Bruch • Erik Schlo¨ricke
•
Ó Socie´te´ Internationale de Chirurgie 2014
Abstract Purpose Re-anastomosis after a Hartmann procedure is associated with a higher morbidity and mortality than other elective colorectal operations. The goal of this comparative study was to evaluate whether laparoscopic reversal is a justified operative approach, although the initial operation is most often an emergency laparotomy. Methods A retrospective analysis was conducted on data collected on all 70 patients who underwent laparoscopic and open reversal of a Hartmann procedure at the Department of Surgery, University of Schleswig–Holstein, Campus Lu¨beck, between January 1999 and December 2011. Together with general demographic data, the analysis included the indication for the initial Hartmann procedure, time to reversal, intraoperative findings, the choice of operative method, operating time, postoperative pain control, return of normal bowel function, length of hospital stay, and peri- and postoperative morbidity and mortality.
M. Zimmermann (&) M. Hoffmann T. Laubert H.-P. Bruch Department of Surgery, University of Schleswig–Holstein, Campus Lu¨beck, Ratzeburger Allee 160, 23538 Lu¨beck, Germany e-mail: [email protected] K.-F. Meyer U.-J. Roblick AGAPLESION DIAKONIEKLINIKUM HAMBURG, Hohe Weide 17, 20259 Hamburg, Germany
Results In most patients, the Hartmann procedure was performed after a perforated sigmoid diverticulitis. We were not able to find any statistically significant differences with respect to gender, body mass index (BMI) and American Society of Anesthesiologists classification between the laparoscopic group (LG) (N = 24 patients) and the open group (OG) (N = 46). In the LG, patients were significantly younger (p = 0.019). The median operating time was 210 min (75–245) in the LG, which was significantly longer than in the OG (166 min; 66–230). The statistical analysis of the duration of postoperative analgesic therapy (LG 7 days [6–10]; OG 12 days [6–30] ), return to normal diet (LG 3 days [2–6]; OG 4 days [2–10] ), return of normal bowel function (LG 3 days [2–4]; OG 4 days [2–9] ) and length of hospital stay (LOS) (LG 10 days [8–13]; OG 15 days [8–163]) detected significant differences in advantage for the LG. Unplanned return to theatre during index admission was only necessary in the OG (N = 7, 15.2 %). With a median follow-up of 8 months (range 1–20), we observed a comparable number of minor complications in both groups but a significantly higher number of major complications in the OG (N = 27, 58.7 %) (p = 0.001). Conversion occurred in three cases (12.5 %). There was no mortality in either of the two groups. Conclusions This study was able to demonstrate the feasibility of the laparoscopic approach. In terms of postoperative results it should be seen as equivalent to the open procedure. However, the laparoscopic approach requires profound surgical expertise. The indication should be made after a careful risk/benefit analysis for each individual patient.
T. Jungbluth Klinikum Wolfsburg, Sauerbruchstr. 7, 38440 Wolfsburg, Germany
Introduction
E. Schlo¨ricke Westku¨stenklinikum Heide, Esmarchstraße 50, D-25746 Heide, Germany
Although Hartmann’s procedure [1] was originally introduced for the management of a malignant left-sided tumor, today the
123
World J Surg
most common reason for it to be performed is complicated or perforated diverticular disease [2]. The increasing experience of colorectal surgeons, together with major progress in interventional endoscopic techniques, such as stent implantations and surveillance programs, are causes for the decline in this indication over the past few years [3]. The current most important consideration for the Hartmann procedure is to avoid a postoperative anastomotic leak in order to reduce morbidity and mortality. In particular, emergency operations involving peritonitis and sepsis for perforated sigmoid diverticulitis are an indication for this procedure [4]. In such patients, a primary anastomosis is associated with a high risk of anastomotic leakage due to disturbances of the vascular supply following systemic reactions to peritonitis and vasoconstrictor drugs that are administered to the patients [2]. Despite technical achievements in the management of enterostomas and the advantages of regular bowel irrigation [5], the formation of an artificial anus leads to a considerable reduction in the individual quality of life [6]. Thus, it is understandable that most patients opt for a stoma reversal as soon as possible. The reversal procedure is generally associated with a high morbidity and mortality [7]. The elevated complication rate is a result of the changed anatomy, the presence of interenteric adhesions, the destruction of anatomic layers and compromised vascular perfusion after the resection of the superior rectal artery or the inferior mesenteric artery [8]. The underlying comparative study between the laparoscopic and the open approach for the reversal of the Hartmann procedure is intended to clarify the extent to which morbidity and mortality is influenced by the choice of the operative approach. Additionally, possible difficulties in tissue dissection and possible solutions associated with a laparoscopic reversal procedure after a primary open Hartmann procedure are discussed.
Patients and methods Prospective data of all 70 patients who underwent a reversal of a Hartmann procedure in the Department of Surgery (University Hospital of Schleswig–Holstein, Campus Lu¨beck, Germany) during a period of 12 years (January 1999–December 2011) were collected in an electronic database and analyzed retrospectively. All patients with a terminal sigmoideostoma or descendostoma after partial colectomy were included in the study. On the preoperative day, bowel preparation was carried out in all patients, and preoperative fasting was started 12 h before intervention. Postoperative, low-calorie parenteral nutrition was administered and patients were allowed water in sips 6 h after intervention. On the first postoperative day, oral nutrition was started with tea, soup, and yogurt. Further diet restoration
123
T1
T2
T4: recommended extra trocar to aid mobilization of the splenic flexure
Colostoma T3
Fig. 1 Recommended trocar placement
was carried out according to signs of normal bowel function (bowel sounds, flatulence, and stool). Due to a secondary intervention, no strict fast-track concept was applied. All laparoscopic and open procedures were performed by a gastrointestinal surgeon with comparable special experience in colorectal surgery. The laparoscopic procedures were all done without ‘hand-assistance’. The operation procedure started with the patient positioned in the lithotomy position. The optical trocar was placed 1 cm superior to the umbilicus in the midline. After placement of the optical trocar, further trocars were positioned according to the extent and localization of adhesions. Whenever possible, trocar placement was performed as shown in Fig. 1. This positioning allowed for the mobilization of the remaining colon up to the mid-transverse colon, including the splenic flexure if necessary. Complete adhesiolysis and complete mobilization of the proximal colonic segment for the formation of a colorectal anastomosis followed after temporary reversal of the enterostoma and provisional closure of the abdominal wall. This procedure allows maximum visibility during intraabdominal laparoscopic dissection. In our procedure, the stapler head is not primarily used in the reversal of the enterostoma, but instead is placed in the proximal colon segment after adhesiolysis and mobilization of the descending colon are complete. The abdominal wall is then re-opened in the area of the former enterostoma for the conduction of this procedure. Primary placement of the stapler head before stoma reversal and intra-abdominal mobilization does not compromise arterial perfusion. However, the compression of small veins can cause outlet
World J Surg
obstruction and lead to incomplete intestinal wall ischemia in the region that feeds the anastomosis. The re-exposition of the colonic segment after re-opening of the fascia additionally allows ruling out perfusion disturbances secondary to vascular compression during mobilization or due to anatomic abnormalities. The latter can be managed by a short re-resection of the colon. A complete exposition of the rectal stump with the reopening of the serosa and complete visualization of the stapler or hand suture is not part of our procedure. Exposing and re-mobilizing the rectal stump without reresection may result in vascular perfusion dysfunction, and ultimately anastomotic leak from dissection-related closure of the vasa recti, which usually run parallel to the rectum wall. Therefore, the rectal stump is mobilized as little as possible, generally not for more than 2 cm, to preserve its blood supply. An end-to-end anastomosis is performed using a circular stapler and the integrity of the anastomosis is checked using an air leak test under saline. Closure of trocar wounds are done with resorbable suture of the fascia for 10-mm trocars (T1, T3 Fig. 1). For 5-mm trocars (T2, T4) only skin suturing is done. Demographic and patient history data; indication for initial Hartmann procedure; intraoperative findings during the Hartmann reversal; operative and postoperative course, especially with respect to operating time, pain medication, return to normal diet, and length of hospital stay, were analyzed to evaluate morbidity and mortality and compared between the laparoscopic and the open group. The classification of adhesions by four degrees of severity is based on histological and morphological criteria classified by Zu¨hlke et al. [9]. Stage one is thin strands and adhesions, stage two broadened strands and the beginning of vascularization, stage three is further broadened strands and plane adhesions, and stage four is plane adhesions conglomerated to organs. The median follow-up was 8 months (range 1–20). All patients underwent rectoscopy, colonoscopy, and sphincter manometry as well as sensory discrimination and continence testing preoperatively. In cases of malignancy, staging was carried out with the appropriate imaging. Due to the retrospective analysis, it was difficult to evaluate the surgeon’s decision to choose the laparoscopic or open approach. Generally, there are absolute contraindications such as severe cardiopulmonary disease and severe chronic obstructive pulmonary disease referring to pneumoperitoneum. Several other relative contraindications, like previous laparotomy, peritonitis, and age of patient are open to opinion, and decisions are dependent on the surgeon’s experience. Statistical analysis was performed using the Statistical Package for Social Sciences for Windows (SPSS 18.0, IBM, Armonk, NY, USA). The Pearson v2 test was used to compare categorical variables, and the Mann–Whitney
U test was used to compare continuous variables as we could not assume normal distribution because of the relatively small numbers. Statistical significance was set at p \ 0.05.
Results During the observational period (January 1999–December 2011), 70 patients underwent a Hartmann reversal procedure after partial colectomy with a terminal colostomy of the colon descendens or sigmoid colon. In 24 patients, the procedure was carried out via laparoscopic access. A conversion to laparotomy was necessary in three of 24 patients (12.5 %). The reasons for conversion were profound enteric adhesions with the anterior abdominal wall and the inability to insert the optical trocar under direct sight in one patient, and in the remaining two patients, the reasons for conversion were the inability to identify the rectal stump due to a conglomerate of loops of small intestine within the lesser pelvis. The data for these three patients were included in the laparoscopic group (LG) and evaluated with an ‘intention to treat’ analysis. The percentage of laparoscopic procedures increased as experience with the technique increased during the observational period. A total of 46 patients underwent a primary laparotomy. The data were analyzed as the open group (OG). Relevant demographic and patient history data are displayed in Table 1. There was a noticeable difference in median age between the groups [LG 46 (27–84); OG 60 (18–84)], which was statistically significant (p = 0.019). There were no significant differences between the LG (N = 24) and the OG (N = 46) in terms of gender distribution, body mass index (BMI), American Society of Anesthesiologists (ASA) classification, and previous abdominal surgery. As shown in Table 1, perforated sigmoid diverticulitis was the most common pathology to be treated with a Hartmann procedure in both groups [LG 15 (62.5 %), OG 18 (39.1 %)]. Another disease with a high prevalence was colorectal cancer at the rectosigmoid junction with bowel obstruction [LG 1 (4.2 %), OG 11 (23.9 %)]. The prevalence differed significantly (p = 0.037) between the two groups. Other reasons included ischemic events [LG 3 (12.5 %), OG 4 (8.7 %)], iatrogenic perforations [LG 1 (4.2 %), OG 3 (6.5 %)], perforations due to M. Crohn LG 3 [(12.5 %), OG 3 (6.5 %)], anastomotic leaks [LG 0, OG 1 (2.2 %)], perforation due to radiation [LG0, OG 1 (2.2 %)], and others [LG 1 (4.2 %), OG 5 (10.9 %)]. The overall mean timespan from Hartmann procedure to Hartmann reversal was 194 days (91–294). In the LG, the mean time was only 167 days (91–241), whereas it was
123
World J Surg Table 1 Demographic and patient history data All patients (N = 70)
Laparoscopic group (N = 24)
Open group (N = 46)
p value
Age, years [median (range)]
56 (18–84)
46 (27–84)
60 (18–84)
0.019
Gender (F : M)
34 : 36
10 : 14
24 : 22
NS
BMI, kg/m2 (range)
26.3 (15.9–35.7)
26.5 (19.5–34.9)
25.9 (15.9–35.7)
NS NS
ASA classification [N (%)] I
3 (4.3)
2 (8.3)
1 (2.2)
II
34 (48.6)
13 (54.2)
21 (45.7)
NS
III
33 (47.1)
9 (37.5)
24 (52.2)
NS
IV
0
0
0
NS
Previous abdominal surgeries [N (%)] Necrotizing pancreatitis
1 (1.4)
0
1 (2.2)
NS
Anal fistula Appendectomy
3 (4.3) 6 (8.5)
1 (4.1) 2 (8.2)
2 (4.3) 4 (8.7)
NS NS
Other
9 (12.8)
3 (12.5)
5 (10.9)
NS
Primary operation [N (%)] Sigmoid diverticulitis
33 (47.1)
15 (62.5)
18 (39.1)
NS
Malignant disease
12 (17.1)
1 (4.2)
11 (23.9)
0.037
Ischemic event
7 (10.0)
3 (12.5)
4 (8.7)
NS
Iatrogenic perforation
4 (5.7)
1 (4.2)
3 (6.5)
NS
M. Crohn perforation
6 (8.6)
3 (12.5)
3 (6.5)
NS
Anastomosis insufficiency
1 (1.4)
0
1 (2.2)
NS
Radiation stenosis
1 (1.4)
0
1 (2.2)
NS
Other
6 (8.6)
1 (4.2)
5 (10.9)
NS
Time to reversal (mean; days)
194 (91–294)
167 (91–241)
208 (118–294)
NS
ASA American Society of Anesthesiologists, BMI body mass index, NS not significant
208 days (118–294) in the OG. The difference was not statistically significant. An overview of the peri- and postoperative course of the patients and the intraoperative findings is shown in Table 2. The median operating time was 210 min (range 75–245) in the LG, which was significantly longer (p \ 0.0001) than in the OG [median 166 (range 66–230)]. Adhesions were generally more severe in the OG. In stage III of the Zu¨hlke adhesion score [9], differences between the groups [LG 1 (4.2 %); OG 14 (30.4 %)] were statistically significant (p = 0.011). The length of intensive care unit (ICU) stay was also significantly (p = 0.003) shorter in the LG [LG 0 days (0–2); OG 1 day (0–12)]. Significant advantages associated with a laparoscopic procedure were observed for the return to normal bowel function (measured as a clinical parameter by bowel sounds; LG 3 days [2–4]; OG 4 days [2–9]; p \ 0.0001), the return to normal diet (LG: 3 days [2–6]; OG 4 days [2–10]; p = 0.001), and in the postoperative continuous use of analgesics (non-steroidal anti-inflammatory drug [NSAID] or morphine preparations) in days (LG 7 days [6–10]; OG 12 days [6–30]; p \ 0.0001). The length of the overall hospital stay (LOS) was also significantly shorter in the LG, with a median of 9 days (7–14) compared with 15 days (8–163) in
123
the OG (p \ 0.0001). The LOS of 163 days in one patient in the OG was the result of a prolonged postoperative recovery period related to a high number of comorbidities. Excluding this case from analysis, statistical significance still remains. Table 3 shows the analysis of morbidity and mortality. There was no statistically significant difference between the groups with respect to minor complications [LG 3 (12.5 %) vs. OG 12 (26.1 %)]. Two events of pneumonia and one event of a paralytic ileus were observed in the LG whereas four patients in the OG developed pneumonia and seven patients experienced a paralytic ileus in this group. Paralytic ileus was defined as a postoperative intestinal atony that was managed with conservative treatment and no need for a surgical intervention. In one patient, a pre-existing absolute arrhythmia and atrial fibrillation resulted in cardiac decompensation and had to be monitored in the ICU. As shown in Table 3, major complications were statistically significantly less observed in the LG [LG 4 (16.6 %) vs. OG 27 (58.7 %); p = 0.001]. Unplanned return to theatre during index admission was necessary in seven patients (15.2 %) in the OG. The operative interventions were necessary because of wound dehiscence (N = 4; 8.7 %), anastomotic leak (N = 2; 4.3 %) and a postoperative
World J Surg Table 2 Peri- and postoperative outcome
Operation time (median; min)
Table 3 Postoperative complications
All patients (N = 70)
Laparoscopic group (N = 24)
Open group (N = 46)
179 (66–365)
210 (75–245)
166 \0.0001 (66–365)
All patients (N = 70)
p value
Adhesions (Zuhlke HV, 1990) [N (%)] I
9 (12.9)
5 (20.8)
4 (8.7)
NS
II
44 (62.9)
18 (75.0)
26 (56.5)
NS
III
15 (21.4)
1 (4.2)
14 (30.4)
0.011
IV
2 (2.9)
0
2 (4.3)
NS
Intensive care unit, days
1 (0–12)
0 (0–2)
1 (0–12)
0.003
Return to normal diet, days
4 (2–10)
3 (2–6)
4 (2–10)
0.001
Return to normal bowel function, days
4 (2–9)
Postoperative analgesia, days
10 (6–30)
7 (6–10)
12 (6–30)
\0.0001
Length of hospital stay, days
13 (7–163)
9 (7–14)
15 (8–163)
\0.0001
Laparoscopic group (N = 24)
Open group (N = 46)
p value
Minor complications Total
15 (21.4)
3 (12.5)
12 (26.1)
NS
Pleural effusion Cardiac events
0 1 (1.4)
0 0
0 1 (2.2)
NS NS
Pneumonia
6 (8.6)
2 (8.3)
4 (8.7)
NS
Paralytic ileus
8 (11.4)
1 (4.1)
7 (15.2)
NS
Urinary dysfunction
0
0
0
NS
Sensory discrimination dysfunction
0
0
0
NS
Major complications 3 (2–4)
4 (2–9)
\0.0001
Data are presented as median (range) unless otherwise indicated NS not significant
hematoma (N = 1; 2.2 %). The anastomotic leaks were treated with direct suture of the small defects as severe peritonitis was absent in both patients. A protective loopileostomy was done in both patients, which was reversed 3–5 months later. In both the LG and the OG, all wound healing disturbances could be treated conservatively. After discharge from the hospital, 13 patients (28.2 %) in the OG underwent a re-operation. The main reason for re-operation was the occurrence of incisional hernias (N = 10; 21.7 %), followed by small bowel obstruction due to adhesions (N = 3; 6.5 %). The small bowel obstruction events occurred at 8, 9, and 15 months after discharge from hospital. In the LG, two patients (8.3 %) underwent re-operation. One patient had a small bowel obstruction (11 months after discharge) due to intestinal adhesions and the other had a trocar hernia at ‘T3’ position (10 mm) (Fig. 1). Two patients in the OG with malignant disease (carcinomas of the rectosigmoid junction [T3N1M0G2, T3N2M0G2], both receiving adjuvant chemotherapy with 5-FU) developed peritoneal carcinomatosis and pulmonary and hepatic metastases. These patients died 18 and
Total
31 (44.3)
4 (16.6)
27 (58.7)
0.001
Early (within hosp.)
16 (22.8)
2 (8.3)
14 (30.4)
0.037
Wound healing disturbance
9 (12.8)
2 (8.3)
7 (15.2)
NS
Wound dehiscence
4 (5.7)
0
4 (8.7)
NS
Anastomosis insufficiency
2 (2.8)
0
2 (4.3)
NS
Postoperative Hematoma
1 (1.4)
0
1 (2.2)
NS
Late (after hosp.) Small bowel obstruction
15 (21.4) 4 (5.7)
2 (8.3) 1 (4.2)
13 (28.2) 3 (6.5)
0.054 NS
Scar-/trocar hernia
11 (15.7)
1 (4.2)
10 (21.7)
NS
Lethal course
0
0
0
NS
Data are presented as N (%) unless otherwise indicated NS not significant
19 months after the Hartmann reversal. No perioperative mortality was observed in either group.
Discussion A Hartmann procedure in a patient with normal function of the anal sphincter results in a major reduction in patient quality of life because of the formation of a colostomy [10]. This may result in progressive social isolation [6]. The Hartmann reversal procedure is thus an integral part of the treatment. Early planning of the reversal within the initial surgical workup can have positive effects on the entire course of the treatment [11]. The length of the interval before Hartmann reversal is determined by the underlying disease itself and the
123
World J Surg
physical status of the patient [12]. A review of the literature revealed that most surgeons recognize an interval of 15–17 weeks after the initial Hartmann procedure as ideal for the most common indication, perforated sigmoid diverticulitis. At that time, tissue remodelling and regeneration will have stabilized [13, 14]. In the patients included in our analysis, the earliest re-operation was performed 13 weeks postoperatively. Overall, the reversal was carried out an average of 28 weeks (6–47 months) after the Hartmann procedure. In the LG, a reversal was performed on average after 24 weeks, in the OG on average in week 30. The difference is not significant but shows a greater timespan in the OG. Reasons for this effect might be that patients in the OG were significantly older than in the LG. Additionally, the percentage of malignant disease as indication for the primary operation is significantly higher in the OG. Due to the non-significant difference, an impact on complications cannot be presumed in this study. In current literature, different results have been reported. Whereas Lin et al. [15] found no associations between time to reversal and complications, Pearce et al. [16] reported that risks of postoperative complications in reversals were significantly greater in patients who underwent the procedure within 6 months after primary operation versus those whose reversal was delayed longer than 6 months. Conversely, further studies have reported that a longer time to reversal ([9 months) was associated with greater risks of postoperative complications [17–19]. In our institution, the indication for reversal was only made after post-hospital rehabilitation treatment was complete and after the patients reached a required preoperative fitness level. In patients with malignant disease, a lack of scientific studies regarding the best timeframe for a Hartmann reversal procedure was observed. In our opinion, re-anastomosis in patients with advanced malignancy (i.e. local or distant metastases, stage T4) should be avoided due to the reduced overall prognosis. However, tumor patients should not be excluded from Hartmann reversal procedures in general, especially if they have an early stage malignancy. Although the patient may experience progressive disease, the reversal procedure increases their individual quality of life [6]. Generally, tumor progression manifests itself primarily with hepatic and pulmonary metastases and only rarely with a local recurrence. We also observed tumor progression in our cohort in two of our patients. Neither patient experienced a local tumor recurrence. Our current standard-operating procedure is to perform the reversal procedure no earlier than 6 months after the primary operation and after early tumor progression has been ruled out. Many authors argue against pre-operative bowel preparation. On the contrary, we consider this procedure essential for a successful laparoscopic reversal. The bowel
123
preparation not only results in augmentation of the essential free space necessary for laparoscopy, but also limits bacterial contamination after iatrogenic bowel wall lesions, as the lavage reduces bacterial concentrations [20, 21]. The choice of operative approach depends on the individual expertise of the operating surgeon as well as the technology available [22]. Contraindications for a laparoscopic reversal, such as previous open surgeries and status post peritonitis, which were established in the early days of colorectal laparoscopic surgery, are no longer absolute [23]. Current studies and our own results have demonstrated the feasibility of laparoscopic reversal after a Hartmann procedure [7, 24]. The primary entry sites for the optical trocar differ greatly in the existing literature [8, 25, 26]. Placement of the first trocar in the area of the repositioned enterostoma, the Hasson technique, and our technique, which includes placement of the first trocar supraumbilically in the midline under direct sight after primary enterostoma repositioning, are all common and possible approaches. In our patient cohort, only one patient required primary conversion to an open procedure due to severe enteric adhesions to the parietal peritoneum. The conversion rate varies greatly among the existing studies and is reported to be between 9 and 22 % [7, 8, 13, 27]. Our conversion rate of 12.5 % was thus absolutely comparable. Most authors indicated the inability to identify the stump as the most common cause for a conversion [25, 26, 28]. In contrast, the three causes for conversion in the present study were the result of inter-enteric and enteric adhesions with the parietal peritoneum and conglomerate formation within the lesser pelvis. However, conversion should be performed when an adequate exposition is impossible, e.g. due to severe adhesions. Another indication for a conversion is apparent or assumed injuries to the serosa of the small intestine that cannot be managed without exposing the patient to inadequate risk. Contrary to the results of some authors, we observed a prolonged operation time associated with the laparoscopic procedure in comparison with the open Hartmann reversals [29]. However, our findings are in agreement with most other studies on laparoscopic versus open colorectal procedures, in which comparable and often prolonged operation times for the laparoscopic approach were observed [8]. The main reasons for longer operation times are laparoscopic dissection of adhesions and the identification of the rectal stump, which are described as challenging within the literature [30]. The prolonged operation time we observed in our cohort had no measurable negative influence on the postoperative outcome, especially with respect to morbidity and mortality (e.g. pneumonia). Compared with other planned open colorectal surgical procedures, the open Hartmann reversal procedure is
World J Surg
associated with an elevated morbidity of 10–41 %, which our own results confirm [29]. The major complications we observed in our cohort in the immediate postoperative course, such as anastomosis leak (N = 2, 2.8 %), wound dehiscence (N = 4, 5.7 %), and impairment of subcutaneous wound healing (N = 9, 12.8 %), also reflect the most common major complications described by other authors. The results of preceding studies varied in terms of the incidence of wound infections (8–20 %) and anastomotic leaks (up to 12 %) [8]. The high number of minor complications (26.1 %) in the OG, which was also described by Mazeh et al. [13], is further evidence for the severity of the entry trauma associated with this procedure. The present study is biased by a selection of younger patients for the laparoscopic approach, particularly in the first years after the implementation of this procedure. A potential limitation of this bias is that there were no significant differences concerning the ASA classification of the patients in the two groups. With increasing experience in the laparoscopic approach, further results from other study groups have shown that older patients with considerable comorbidities also benefit from less invasive methods [31]. The accepted advantages of laparoscopy for colorectal operations, including reduced entry trauma, more accurate tissue dissection at 129 magnification and the associated reduction in immune response, as well as the reduction in intraoperative morphine levels, appear to lead to a reduction in postoperative morbidity for the laparoscopic Hartmann reversal procedure [22, 32]. This assumption is supported by the reduction in minor and major complications by 15 and 43 %, respectively, as observed in our patient cohort. The most severe procedure-associated complication, anastomotic leak, was only found in the OG in our cohort (N = 2, 4.3 %) [33]. The preservation of the integrity of the abdominal compartment, especially with respect to physiological function of the anterior abdominal wall, as well as the minimal operation trauma, was a reason for the postoperative reduction in analgesic drugs. We also observed significant advantages in the earlier return of normal bowel function, a shorter recovery time, and a shorter LOS associated with the LG. The minimally invasive approach was also associated with a lower prevalence (p = 0.054) of late complications such as incisional hernias (OG 21.7 % vs. LG : 4.2 %) and adhesion ileus (OG 6.5 % vs. LG 4.2 %), both of which required surgical intervention. There are no published data available on the long-term course of patients who underwent laparoscopic Hartmann reversal. Nevertheless, comparative studies on laparoscopic versus open sigmoid resection have also shown a profound tendency of less incisional hernias and thus indirectly support this observation in our study [34]. The laparoscopic Hartmann reversal procedure offers more accurate tissue dissection as well as an additional diagnostic tool without increasing entry trauma. Previously unknown
pathology, such as, for example, metastases, fistulas, development of progressive diverticular disease or hernias, may be detected or ruled out [30]. It is also possible to treat some of the pathologies immediately during the same procedure or take biopsies for further analysis and the planning of additional therapies. Finally, the laparoscopic approach has proven to be comparable to the open operation regarding economical and technical aspects and especially with regard to patient safety [35, 36]. Consequently, the reversal of Hartmann’s procedure can usually be commenced laparoscopically providing there is no contraindication or expectation of extreme difficulties with adhesions and mobilization.
Conclusion The laparoscopic Hartmann reversal procedure is safe and effective. Compared with the open approach, it should be seen as equivalent regarding morbidity and mortality. The prolonged operation time is not associated with a negative influence on the postoperative course of the patients in this cohort. The reduction of entry and operation trauma is beneficial for patients with less incisional hernias, less consumption of analgesics, an earlier return of normal bowel function, and a shorter LOS. Previous studies were able to show that economic disadvantages resulting from the prolonged operation time and instrument costs are more than equalized by the advantages associated with laparoscopy, as mentioned previously. In the absence of contraindications, the operation can generally be commenced laparoscopically. Nevertheless, the technical requirements and the laparoscopic expertise necessary for this approach are seriously limiting the widespread use of this and many other laparoscopic procedures, despite its proven advantages. Disclosure statement Drs. M. Zimmermann, M. Hoffmann, P. Hildebrand, K.-F. Meyer, T. Jungbluth, U.-J. Roblick, H.-.P. Bruch, and E. Schloericke have no conflicts of interest or financial ties to disclose.
References 1. Hartmann H (1923) Note sur un procede nouveau d’extirpation des cancers de la partie du colon. Bull Mem Soc Chir Paris 49:1474–1477 2. Desai DC, Brennan EJJ, Reilly JF, Smink RDJ (1998) The utility of the Hartmann procedure. Am J Surg 175(2):152–154 3. Seah DW, Ibrahim S, Tay KH (2005) Hartmann procedure: is it still relevant today? ANZ J Surg 75(6):436–440 4. Farthmann EH, Ruckauer KD, Haring RU (2000) Evidence-based surgery: diverticulitis–a surgical disease? Langenbecks Arch Surg 385(2):143–151
123
World J Surg 5. Varma S (2009) Issues in irrigation for people with a permanent colostomy: a review. Br J Nurs 18(4):15–18 6. Fucini C, Gattai R, Urena C, Bandettini L, Elbetti C (2008) Quality of life among five-year survivors after treatment for very low rectal cancer with or without a permanent abdominal stoma. Ann Surg Oncol 15(4):1099–1106 7. Petersen M, Kockerling F, Lippert H, Scheidbach H (2009) Laparoscopically assisted reversal of Hartmann procedure. Surg Laparosc Endosc Percutan Tech 19(1):48–51 8. Khaikin M, Zmora O, Rosin D, Bar-Zakai B, Goldes Y, Shabtai M, Ayalon A, Munz Y (2006) Laparoscopically assisted reversal of Hartmann’s procedure. Surg Endosc 20(12):1883–1886 9. Zu¨hlke HV, Lorenz EM, Straub EM, Savvas V (1990) Pathophysiology and classification of adhesions. Langenbecks Arch Chir Suppl II Verh Dtsch Ges Chir 1009–1016 (German) 10. Vermeulen J, Gosselink MP, Busschbach JJV, Lange JF (2010) Avoiding or reversing Hartmann’s procedure provides improved quality of life after perforated diverticulitis. J Gastrointest Surg 14(4):651–657 11. Taylor C, Morgan L (2010) Quality of life following reversal of temporary stoma after rectal cancer treatment. Eur J Oncol Nurs 15(1):59–66 12. Albarran SA, Simoens C, Van De Winkel N, da Costa PM, Thill V (2009) Restoration of digestive continuity after Hartmann’s procedure: ASA score is a predictive factor for risk of postoperative complications. Acta Chir Belg 109(6):714–719 13. Mazeh H, Greenstein AJ, Swedish K, Nguyen SQ, Lipskar A, Weber KJ, Chin EH, Divino CM (2009) Laparoscopic and open reversal of Hartmann’s procedure—a comparative retrospective analysis. Surg Endosc 23(3):496–502 14. Keck JO, Collopy BT, Ryan PJ, Fink R, Mackay JR, Woods RJ (1994) Reversal of Hartmann’s procedure: effect of timing and technique on ease and safety. Dis Colon Rectum 37(3):243–248 15. Lin FL, Boutros M, Da Silva GM, Weiss EG, Lu XR, Wexner SD (2013) Hartmann reversal: obesity adversely impacts outcome. Dis Colon Rectum 56(1):83–90 16. Pearce NW, Scott SD, Karran SJ (1992) Timing and method of reversal of Hartmann’s procedure. Br J Surg 79:839–841 17. Roque-Castellano C, Marchena-Gomez J, Hemmersbach-Miller M et al (2007) Analysis of the factors related to the decision of restoring intestinal continuity after Hartmann’s procedure. Int J Colorectal Dis 22:1091–1096 18. Fleming FJ, Gillen P (2009) Reversal of Hartmann’s procedure following acute diverticulitis: is timing everything? Int J Colorectal Dis 24:1219–1225 19. Schmelzer TM, Mostafa G, Norton HJ et al (2007) Reversal of Hartmann’s procedure: a high-risk operation? Surgery 142:598–606 20. Muzii L, Bellati F, Zullo MA, Manci N, Angioli R, Panici PB (2006) Mechanical bowel preparation before gynecologic laparoscopy: a randomized, single-blind, controlled trial. Fertil Steril 85(3):689–693 21. Zmora O, Lebedyev A, Hoffman A, Khaikin M, Munz Y, Shabtai M, Ayalon A, Rosin D (2006) Laparoscopic colectomy without mechanical bowel preparation. Int J Colorectal Dis 21(7):683–687
123
22. Svenningsen PO, Bulut O, Jess P (2010) Laparoscopic reversal of Hartmann’s procedure. Dan Med Bull 57(6):A4149 23. Riansuwan W, Hull TL, Millan MM, Hammel JP (2009) Nonreversal of Hartmann’s procedure for diverticulitis: derivation of a scoring system to predict nonreversal. Dis Colon Rectum 52(8):1400–1408 24. Slawik S, Dixon AR (2008) Laparoscopic reversal of Hartmann’s rectosigmoidectomy. Colorectal Dis 10(1):81–83 25. Sosa JL, Sleeman D, Puente I, McKenney MG, Hartmann R (1994) Laparoscopic-assisted colostomy closure after Hartmann’s procedure. Dis Colon Rectum 37(2):149–152 26. Vacher C, Zaghloul R, Borie F, Laporte S, Callafe R, Skawinski P, Leynau G, Domergue J (2002) Retablissement de la continuite digestive apres intervention de Hartmann par voie laparoscopique. Etude retrospective de la Societe francaise de chirurgie endoscopique (SFCE). Ann Chir 127(3):189–192 27. Rosen MJ, Cobb WS, Kercher KW, Heniford BT (2006) Laparoscopic versus open colostomy reversal: a comparative analysis. J Gastrointest Surg 10(6):895–900 28. Costantino GN, Mukalian GG (1994) Laparoscopic reversal of Hartmann procedure. J Laparoendosc Surg 4(6):429–433 29. Di Carlo I, Toro A, Pannofino O, Pulvirenti EPE (2010) Laparoscopic versus open restoration of intestinal continuity after Hartmann procedure. Hepatogastroenterology 57(98):232–235 30. Sneider EB, Cahan MA, Litwin DEM (2010) Laparoscopic repair of acute surgical diseases in the 21st century. Minerva Chir 65(3):275–296 31. van de Wall BJ, Draaisma WA, Schouten ES, Broeders IA, Consten EC (2010) Conventional and laparoscopic reversal of the Hartmann procedure: a review of literature. J Gastrointest Surg 14(4):743–752 32. Huynh H, Trottier DC, Soto CM, Moloo H, Poulin EC, Mamazza J, Boushey RP (2011) Laparoscopic colostomy reversal after a Hartmann procedure: a prospective series, literature review and an argument against laparotomy as the primary approach. Can J Surg 54(2):133–137 33. Boccola MA, Lin J, Rozen WM, Ho Y (2010) Reducing anastomotic leakage in oncologic colorectal surgery: an evidencebased review. Anticancer Res 30(2):601–607 34. Siddiqui MR, Sajid MS, Khatri K, Cheek E, Baig MK (2010) Elective open versus laparoscopic sigmoid colectomy for diverticular disease: a meta-analysis with the Sigma trial. World J Surg 34(12):2883–2901. doi:10.1007/s00268-010-0762-3 35. Alkhamesi NA, Martin J, Schlachta CM (2011) Cost-efficiency of laparoscopic versus open colon surgery in a tertiary care center. Surg Endosc 25(11):3597–3604 36. Fleshman J, Sargent DJ, Green E, Anvari M, Stryker SJ, Beart RW Jr, Hellinger M, Flanagan R Jr, Peters W, Nelson H, for The Clinical Outcomes of Surgical Therapy Study Group (2007) Laparoscopic colectomy for cancer is not inferior to open surgery based on 5-year data from the COST Study Group trial. Ann Surg 246(4):655–662
Laparoscopic Versus Open Reversal of a Hartmann Procedure: A Single-Center Study Markus Zimmermann • Martin Hoffmann • Tilman Laubert • Karl-Frederik Meyer • Thomas Jungbluth • Uwe-Johannes Roblick Hans-Peter Bruch • Erik Schlo¨ricke
•
Ó Socie´te´ Internationale de Chirurgie 2014
Abstract Purpose Re-anastomosis after a Hartmann procedure is associated with a higher morbidity and mortality than other elective colorectal operations. The goal of this comparative study was to evaluate whether laparoscopic reversal is a justified operative approach, although the initial operation is most often an emergency laparotomy. Methods A retrospective analysis was conducted on data collected on all 70 patients who underwent laparoscopic and open reversal of a Hartmann procedure at the Department of Surgery, University of Schleswig–Holstein, Campus Lu¨beck, between January 1999 and December 2011. Together with general demographic data, the analysis included the indication for the initial Hartmann procedure, time to reversal, intraoperative findings, the choice of operative method, operating time, postoperative pain control, return of normal bowel function, length of hospital stay, and peri- and postoperative morbidity and mortality.
M. Zimmermann (&) M. Hoffmann T. Laubert H.-P. Bruch Department of Surgery, University of Schleswig–Holstein, Campus Lu¨beck, Ratzeburger Allee 160, 23538 Lu¨beck, Germany e-mail: [email protected] K.-F. Meyer U.-J. Roblick AGAPLESION DIAKONIEKLINIKUM HAMBURG, Hohe Weide 17, 20259 Hamburg, Germany
Results In most patients, the Hartmann procedure was performed after a perforated sigmoid diverticulitis. We were not able to find any statistically significant differences with respect to gender, body mass index (BMI) and American Society of Anesthesiologists classification between the laparoscopic group (LG) (N = 24 patients) and the open group (OG) (N = 46). In the LG, patients were significantly younger (p = 0.019). The median operating time was 210 min (75–245) in the LG, which was significantly longer than in the OG (166 min; 66–230). The statistical analysis of the duration of postoperative analgesic therapy (LG 7 days [6–10]; OG 12 days [6–30] ), return to normal diet (LG 3 days [2–6]; OG 4 days [2–10] ), return of normal bowel function (LG 3 days [2–4]; OG 4 days [2–9] ) and length of hospital stay (LOS) (LG 10 days [8–13]; OG 15 days [8–163]) detected significant differences in advantage for the LG. Unplanned return to theatre during index admission was only necessary in the OG (N = 7, 15.2 %). With a median follow-up of 8 months (range 1–20), we observed a comparable number of minor complications in both groups but a significantly higher number of major complications in the OG (N = 27, 58.7 %) (p = 0.001). Conversion occurred in three cases (12.5 %). There was no mortality in either of the two groups. Conclusions This study was able to demonstrate the feasibility of the laparoscopic approach. In terms of postoperative results it should be seen as equivalent to the open procedure. However, the laparoscopic approach requires profound surgical expertise. The indication should be made after a careful risk/benefit analysis for each individual patient.
T. Jungbluth Klinikum Wolfsburg, Sauerbruchstr. 7, 38440 Wolfsburg, Germany
Introduction
E. Schlo¨ricke Westku¨stenklinikum Heide, Esmarchstraße 50, D-25746 Heide, Germany
Although Hartmann’s procedure [1] was originally introduced for the management of a malignant left-sided tumor, today the
123
World J Surg
most common reason for it to be performed is complicated or perforated diverticular disease [2]. The increasing experience of colorectal surgeons, together with major progress in interventional endoscopic techniques, such as stent implantations and surveillance programs, are causes for the decline in this indication over the past few years [3]. The current most important consideration for the Hartmann procedure is to avoid a postoperative anastomotic leak in order to reduce morbidity and mortality. In particular, emergency operations involving peritonitis and sepsis for perforated sigmoid diverticulitis are an indication for this procedure [4]. In such patients, a primary anastomosis is associated with a high risk of anastomotic leakage due to disturbances of the vascular supply following systemic reactions to peritonitis and vasoconstrictor drugs that are administered to the patients [2]. Despite technical achievements in the management of enterostomas and the advantages of regular bowel irrigation [5], the formation of an artificial anus leads to a considerable reduction in the individual quality of life [6]. Thus, it is understandable that most patients opt for a stoma reversal as soon as possible. The reversal procedure is generally associated with a high morbidity and mortality [7]. The elevated complication rate is a result of the changed anatomy, the presence of interenteric adhesions, the destruction of anatomic layers and compromised vascular perfusion after the resection of the superior rectal artery or the inferior mesenteric artery [8]. The underlying comparative study between the laparoscopic and the open approach for the reversal of the Hartmann procedure is intended to clarify the extent to which morbidity and mortality is influenced by the choice of the operative approach. Additionally, possible difficulties in tissue dissection and possible solutions associated with a laparoscopic reversal procedure after a primary open Hartmann procedure are discussed.
Patients and methods Prospective data of all 70 patients who underwent a reversal of a Hartmann procedure in the Department of Surgery (University Hospital of Schleswig–Holstein, Campus Lu¨beck, Germany) during a period of 12 years (January 1999–December 2011) were collected in an electronic database and analyzed retrospectively. All patients with a terminal sigmoideostoma or descendostoma after partial colectomy were included in the study. On the preoperative day, bowel preparation was carried out in all patients, and preoperative fasting was started 12 h before intervention. Postoperative, low-calorie parenteral nutrition was administered and patients were allowed water in sips 6 h after intervention. On the first postoperative day, oral nutrition was started with tea, soup, and yogurt. Further diet restoration
123
T1
T2
T4: recommended extra trocar to aid mobilization of the splenic flexure
Colostoma T3
Fig. 1 Recommended trocar placement
was carried out according to signs of normal bowel function (bowel sounds, flatulence, and stool). Due to a secondary intervention, no strict fast-track concept was applied. All laparoscopic and open procedures were performed by a gastrointestinal surgeon with comparable special experience in colorectal surgery. The laparoscopic procedures were all done without ‘hand-assistance’. The operation procedure started with the patient positioned in the lithotomy position. The optical trocar was placed 1 cm superior to the umbilicus in the midline. After placement of the optical trocar, further trocars were positioned according to the extent and localization of adhesions. Whenever possible, trocar placement was performed as shown in Fig. 1. This positioning allowed for the mobilization of the remaining colon up to the mid-transverse colon, including the splenic flexure if necessary. Complete adhesiolysis and complete mobilization of the proximal colonic segment for the formation of a colorectal anastomosis followed after temporary reversal of the enterostoma and provisional closure of the abdominal wall. This procedure allows maximum visibility during intraabdominal laparoscopic dissection. In our procedure, the stapler head is not primarily used in the reversal of the enterostoma, but instead is placed in the proximal colon segment after adhesiolysis and mobilization of the descending colon are complete. The abdominal wall is then re-opened in the area of the former enterostoma for the conduction of this procedure. Primary placement of the stapler head before stoma reversal and intra-abdominal mobilization does not compromise arterial perfusion. However, the compression of small veins can cause outlet
World J Surg
obstruction and lead to incomplete intestinal wall ischemia in the region that feeds the anastomosis. The re-exposition of the colonic segment after re-opening of the fascia additionally allows ruling out perfusion disturbances secondary to vascular compression during mobilization or due to anatomic abnormalities. The latter can be managed by a short re-resection of the colon. A complete exposition of the rectal stump with the reopening of the serosa and complete visualization of the stapler or hand suture is not part of our procedure. Exposing and re-mobilizing the rectal stump without reresection may result in vascular perfusion dysfunction, and ultimately anastomotic leak from dissection-related closure of the vasa recti, which usually run parallel to the rectum wall. Therefore, the rectal stump is mobilized as little as possible, generally not for more than 2 cm, to preserve its blood supply. An end-to-end anastomosis is performed using a circular stapler and the integrity of the anastomosis is checked using an air leak test under saline. Closure of trocar wounds are done with resorbable suture of the fascia for 10-mm trocars (T1, T3 Fig. 1). For 5-mm trocars (T2, T4) only skin suturing is done. Demographic and patient history data; indication for initial Hartmann procedure; intraoperative findings during the Hartmann reversal; operative and postoperative course, especially with respect to operating time, pain medication, return to normal diet, and length of hospital stay, were analyzed to evaluate morbidity and mortality and compared between the laparoscopic and the open group. The classification of adhesions by four degrees of severity is based on histological and morphological criteria classified by Zu¨hlke et al. [9]. Stage one is thin strands and adhesions, stage two broadened strands and the beginning of vascularization, stage three is further broadened strands and plane adhesions, and stage four is plane adhesions conglomerated to organs. The median follow-up was 8 months (range 1–20). All patients underwent rectoscopy, colonoscopy, and sphincter manometry as well as sensory discrimination and continence testing preoperatively. In cases of malignancy, staging was carried out with the appropriate imaging. Due to the retrospective analysis, it was difficult to evaluate the surgeon’s decision to choose the laparoscopic or open approach. Generally, there are absolute contraindications such as severe cardiopulmonary disease and severe chronic obstructive pulmonary disease referring to pneumoperitoneum. Several other relative contraindications, like previous laparotomy, peritonitis, and age of patient are open to opinion, and decisions are dependent on the surgeon’s experience. Statistical analysis was performed using the Statistical Package for Social Sciences for Windows (SPSS 18.0, IBM, Armonk, NY, USA). The Pearson v2 test was used to compare categorical variables, and the Mann–Whitney
U test was used to compare continuous variables as we could not assume normal distribution because of the relatively small numbers. Statistical significance was set at p \ 0.05.
Results During the observational period (January 1999–December 2011), 70 patients underwent a Hartmann reversal procedure after partial colectomy with a terminal colostomy of the colon descendens or sigmoid colon. In 24 patients, the procedure was carried out via laparoscopic access. A conversion to laparotomy was necessary in three of 24 patients (12.5 %). The reasons for conversion were profound enteric adhesions with the anterior abdominal wall and the inability to insert the optical trocar under direct sight in one patient, and in the remaining two patients, the reasons for conversion were the inability to identify the rectal stump due to a conglomerate of loops of small intestine within the lesser pelvis. The data for these three patients were included in the laparoscopic group (LG) and evaluated with an ‘intention to treat’ analysis. The percentage of laparoscopic procedures increased as experience with the technique increased during the observational period. A total of 46 patients underwent a primary laparotomy. The data were analyzed as the open group (OG). Relevant demographic and patient history data are displayed in Table 1. There was a noticeable difference in median age between the groups [LG 46 (27–84); OG 60 (18–84)], which was statistically significant (p = 0.019). There were no significant differences between the LG (N = 24) and the OG (N = 46) in terms of gender distribution, body mass index (BMI), American Society of Anesthesiologists (ASA) classification, and previous abdominal surgery. As shown in Table 1, perforated sigmoid diverticulitis was the most common pathology to be treated with a Hartmann procedure in both groups [LG 15 (62.5 %), OG 18 (39.1 %)]. Another disease with a high prevalence was colorectal cancer at the rectosigmoid junction with bowel obstruction [LG 1 (4.2 %), OG 11 (23.9 %)]. The prevalence differed significantly (p = 0.037) between the two groups. Other reasons included ischemic events [LG 3 (12.5 %), OG 4 (8.7 %)], iatrogenic perforations [LG 1 (4.2 %), OG 3 (6.5 %)], perforations due to M. Crohn LG 3 [(12.5 %), OG 3 (6.5 %)], anastomotic leaks [LG 0, OG 1 (2.2 %)], perforation due to radiation [LG0, OG 1 (2.2 %)], and others [LG 1 (4.2 %), OG 5 (10.9 %)]. The overall mean timespan from Hartmann procedure to Hartmann reversal was 194 days (91–294). In the LG, the mean time was only 167 days (91–241), whereas it was
123
World J Surg Table 1 Demographic and patient history data All patients (N = 70)
Laparoscopic group (N = 24)
Open group (N = 46)
p value
Age, years [median (range)]
56 (18–84)
46 (27–84)
60 (18–84)
0.019
Gender (F : M)
34 : 36
10 : 14
24 : 22
NS
BMI, kg/m2 (range)
26.3 (15.9–35.7)
26.5 (19.5–34.9)
25.9 (15.9–35.7)
NS NS
ASA classification [N (%)] I
3 (4.3)
2 (8.3)
1 (2.2)
II
34 (48.6)
13 (54.2)
21 (45.7)
NS
III
33 (47.1)
9 (37.5)
24 (52.2)
NS
IV
0
0
0
NS
Previous abdominal surgeries [N (%)] Necrotizing pancreatitis
1 (1.4)
0
1 (2.2)
NS
Anal fistula Appendectomy
3 (4.3) 6 (8.5)
1 (4.1) 2 (8.2)
2 (4.3) 4 (8.7)
NS NS
Other
9 (12.8)
3 (12.5)
5 (10.9)
NS
Primary operation [N (%)] Sigmoid diverticulitis
33 (47.1)
15 (62.5)
18 (39.1)
NS
Malignant disease
12 (17.1)
1 (4.2)
11 (23.9)
0.037
Ischemic event
7 (10.0)
3 (12.5)
4 (8.7)
NS
Iatrogenic perforation
4 (5.7)
1 (4.2)
3 (6.5)
NS
M. Crohn perforation
6 (8.6)
3 (12.5)
3 (6.5)
NS
Anastomosis insufficiency
1 (1.4)
0
1 (2.2)
NS
Radiation stenosis
1 (1.4)
0
1 (2.2)
NS
Other
6 (8.6)
1 (4.2)
5 (10.9)
NS
Time to reversal (mean; days)
194 (91–294)
167 (91–241)
208 (118–294)
NS
ASA American Society of Anesthesiologists, BMI body mass index, NS not significant
208 days (118–294) in the OG. The difference was not statistically significant. An overview of the peri- and postoperative course of the patients and the intraoperative findings is shown in Table 2. The median operating time was 210 min (range 75–245) in the LG, which was significantly longer (p \ 0.0001) than in the OG [median 166 (range 66–230)]. Adhesions were generally more severe in the OG. In stage III of the Zu¨hlke adhesion score [9], differences between the groups [LG 1 (4.2 %); OG 14 (30.4 %)] were statistically significant (p = 0.011). The length of intensive care unit (ICU) stay was also significantly (p = 0.003) shorter in the LG [LG 0 days (0–2); OG 1 day (0–12)]. Significant advantages associated with a laparoscopic procedure were observed for the return to normal bowel function (measured as a clinical parameter by bowel sounds; LG 3 days [2–4]; OG 4 days [2–9]; p \ 0.0001), the return to normal diet (LG: 3 days [2–6]; OG 4 days [2–10]; p = 0.001), and in the postoperative continuous use of analgesics (non-steroidal anti-inflammatory drug [NSAID] or morphine preparations) in days (LG 7 days [6–10]; OG 12 days [6–30]; p \ 0.0001). The length of the overall hospital stay (LOS) was also significantly shorter in the LG, with a median of 9 days (7–14) compared with 15 days (8–163) in
123
the OG (p \ 0.0001). The LOS of 163 days in one patient in the OG was the result of a prolonged postoperative recovery period related to a high number of comorbidities. Excluding this case from analysis, statistical significance still remains. Table 3 shows the analysis of morbidity and mortality. There was no statistically significant difference between the groups with respect to minor complications [LG 3 (12.5 %) vs. OG 12 (26.1 %)]. Two events of pneumonia and one event of a paralytic ileus were observed in the LG whereas four patients in the OG developed pneumonia and seven patients experienced a paralytic ileus in this group. Paralytic ileus was defined as a postoperative intestinal atony that was managed with conservative treatment and no need for a surgical intervention. In one patient, a pre-existing absolute arrhythmia and atrial fibrillation resulted in cardiac decompensation and had to be monitored in the ICU. As shown in Table 3, major complications were statistically significantly less observed in the LG [LG 4 (16.6 %) vs. OG 27 (58.7 %); p = 0.001]. Unplanned return to theatre during index admission was necessary in seven patients (15.2 %) in the OG. The operative interventions were necessary because of wound dehiscence (N = 4; 8.7 %), anastomotic leak (N = 2; 4.3 %) and a postoperative
World J Surg Table 2 Peri- and postoperative outcome
Operation time (median; min)
Table 3 Postoperative complications
All patients (N = 70)
Laparoscopic group (N = 24)
Open group (N = 46)
179 (66–365)
210 (75–245)
166 \0.0001 (66–365)
All patients (N = 70)
p value
Adhesions (Zuhlke HV, 1990) [N (%)] I
9 (12.9)
5 (20.8)
4 (8.7)
NS
II
44 (62.9)
18 (75.0)
26 (56.5)
NS
III
15 (21.4)
1 (4.2)
14 (30.4)
0.011
IV
2 (2.9)
0
2 (4.3)
NS
Intensive care unit, days
1 (0–12)
0 (0–2)
1 (0–12)
0.003
Return to normal diet, days
4 (2–10)
3 (2–6)
4 (2–10)
0.001
Return to normal bowel function, days
4 (2–9)
Postoperative analgesia, days
10 (6–30)
7 (6–10)
12 (6–30)
\0.0001
Length of hospital stay, days
13 (7–163)
9 (7–14)
15 (8–163)
\0.0001
Laparoscopic group (N = 24)
Open group (N = 46)
p value
Minor complications Total
15 (21.4)
3 (12.5)
12 (26.1)
NS
Pleural effusion Cardiac events
0 1 (1.4)
0 0
0 1 (2.2)
NS NS
Pneumonia
6 (8.6)
2 (8.3)
4 (8.7)
NS
Paralytic ileus
8 (11.4)
1 (4.1)
7 (15.2)
NS
Urinary dysfunction
0
0
0
NS
Sensory discrimination dysfunction
0
0
0
NS
Major complications 3 (2–4)
4 (2–9)
\0.0001
Data are presented as median (range) unless otherwise indicated NS not significant
hematoma (N = 1; 2.2 %). The anastomotic leaks were treated with direct suture of the small defects as severe peritonitis was absent in both patients. A protective loopileostomy was done in both patients, which was reversed 3–5 months later. In both the LG and the OG, all wound healing disturbances could be treated conservatively. After discharge from the hospital, 13 patients (28.2 %) in the OG underwent a re-operation. The main reason for re-operation was the occurrence of incisional hernias (N = 10; 21.7 %), followed by small bowel obstruction due to adhesions (N = 3; 6.5 %). The small bowel obstruction events occurred at 8, 9, and 15 months after discharge from hospital. In the LG, two patients (8.3 %) underwent re-operation. One patient had a small bowel obstruction (11 months after discharge) due to intestinal adhesions and the other had a trocar hernia at ‘T3’ position (10 mm) (Fig. 1). Two patients in the OG with malignant disease (carcinomas of the rectosigmoid junction [T3N1M0G2, T3N2M0G2], both receiving adjuvant chemotherapy with 5-FU) developed peritoneal carcinomatosis and pulmonary and hepatic metastases. These patients died 18 and
Total
31 (44.3)
4 (16.6)
27 (58.7)
0.001
Early (within hosp.)
16 (22.8)
2 (8.3)
14 (30.4)
0.037
Wound healing disturbance
9 (12.8)
2 (8.3)
7 (15.2)
NS
Wound dehiscence
4 (5.7)
0
4 (8.7)
NS
Anastomosis insufficiency
2 (2.8)
0
2 (4.3)
NS
Postoperative Hematoma
1 (1.4)
0
1 (2.2)
NS
Late (after hosp.) Small bowel obstruction
15 (21.4) 4 (5.7)
2 (8.3) 1 (4.2)
13 (28.2) 3 (6.5)
0.054 NS
Scar-/trocar hernia
11 (15.7)
1 (4.2)
10 (21.7)
NS
Lethal course
0
0
0
NS
Data are presented as N (%) unless otherwise indicated NS not significant
19 months after the Hartmann reversal. No perioperative mortality was observed in either group.
Discussion A Hartmann procedure in a patient with normal function of the anal sphincter results in a major reduction in patient quality of life because of the formation of a colostomy [10]. This may result in progressive social isolation [6]. The Hartmann reversal procedure is thus an integral part of the treatment. Early planning of the reversal within the initial surgical workup can have positive effects on the entire course of the treatment [11]. The length of the interval before Hartmann reversal is determined by the underlying disease itself and the
123
World J Surg
physical status of the patient [12]. A review of the literature revealed that most surgeons recognize an interval of 15–17 weeks after the initial Hartmann procedure as ideal for the most common indication, perforated sigmoid diverticulitis. At that time, tissue remodelling and regeneration will have stabilized [13, 14]. In the patients included in our analysis, the earliest re-operation was performed 13 weeks postoperatively. Overall, the reversal was carried out an average of 28 weeks (6–47 months) after the Hartmann procedure. In the LG, a reversal was performed on average after 24 weeks, in the OG on average in week 30. The difference is not significant but shows a greater timespan in the OG. Reasons for this effect might be that patients in the OG were significantly older than in the LG. Additionally, the percentage of malignant disease as indication for the primary operation is significantly higher in the OG. Due to the non-significant difference, an impact on complications cannot be presumed in this study. In current literature, different results have been reported. Whereas Lin et al. [15] found no associations between time to reversal and complications, Pearce et al. [16] reported that risks of postoperative complications in reversals were significantly greater in patients who underwent the procedure within 6 months after primary operation versus those whose reversal was delayed longer than 6 months. Conversely, further studies have reported that a longer time to reversal ([9 months) was associated with greater risks of postoperative complications [17–19]. In our institution, the indication for reversal was only made after post-hospital rehabilitation treatment was complete and after the patients reached a required preoperative fitness level. In patients with malignant disease, a lack of scientific studies regarding the best timeframe for a Hartmann reversal procedure was observed. In our opinion, re-anastomosis in patients with advanced malignancy (i.e. local or distant metastases, stage T4) should be avoided due to the reduced overall prognosis. However, tumor patients should not be excluded from Hartmann reversal procedures in general, especially if they have an early stage malignancy. Although the patient may experience progressive disease, the reversal procedure increases their individual quality of life [6]. Generally, tumor progression manifests itself primarily with hepatic and pulmonary metastases and only rarely with a local recurrence. We also observed tumor progression in our cohort in two of our patients. Neither patient experienced a local tumor recurrence. Our current standard-operating procedure is to perform the reversal procedure no earlier than 6 months after the primary operation and after early tumor progression has been ruled out. Many authors argue against pre-operative bowel preparation. On the contrary, we consider this procedure essential for a successful laparoscopic reversal. The bowel
123
preparation not only results in augmentation of the essential free space necessary for laparoscopy, but also limits bacterial contamination after iatrogenic bowel wall lesions, as the lavage reduces bacterial concentrations [20, 21]. The choice of operative approach depends on the individual expertise of the operating surgeon as well as the technology available [22]. Contraindications for a laparoscopic reversal, such as previous open surgeries and status post peritonitis, which were established in the early days of colorectal laparoscopic surgery, are no longer absolute [23]. Current studies and our own results have demonstrated the feasibility of laparoscopic reversal after a Hartmann procedure [7, 24]. The primary entry sites for the optical trocar differ greatly in the existing literature [8, 25, 26]. Placement of the first trocar in the area of the repositioned enterostoma, the Hasson technique, and our technique, which includes placement of the first trocar supraumbilically in the midline under direct sight after primary enterostoma repositioning, are all common and possible approaches. In our patient cohort, only one patient required primary conversion to an open procedure due to severe enteric adhesions to the parietal peritoneum. The conversion rate varies greatly among the existing studies and is reported to be between 9 and 22 % [7, 8, 13, 27]. Our conversion rate of 12.5 % was thus absolutely comparable. Most authors indicated the inability to identify the stump as the most common cause for a conversion [25, 26, 28]. In contrast, the three causes for conversion in the present study were the result of inter-enteric and enteric adhesions with the parietal peritoneum and conglomerate formation within the lesser pelvis. However, conversion should be performed when an adequate exposition is impossible, e.g. due to severe adhesions. Another indication for a conversion is apparent or assumed injuries to the serosa of the small intestine that cannot be managed without exposing the patient to inadequate risk. Contrary to the results of some authors, we observed a prolonged operation time associated with the laparoscopic procedure in comparison with the open Hartmann reversals [29]. However, our findings are in agreement with most other studies on laparoscopic versus open colorectal procedures, in which comparable and often prolonged operation times for the laparoscopic approach were observed [8]. The main reasons for longer operation times are laparoscopic dissection of adhesions and the identification of the rectal stump, which are described as challenging within the literature [30]. The prolonged operation time we observed in our cohort had no measurable negative influence on the postoperative outcome, especially with respect to morbidity and mortality (e.g. pneumonia). Compared with other planned open colorectal surgical procedures, the open Hartmann reversal procedure is
World J Surg
associated with an elevated morbidity of 10–41 %, which our own results confirm [29]. The major complications we observed in our cohort in the immediate postoperative course, such as anastomosis leak (N = 2, 2.8 %), wound dehiscence (N = 4, 5.7 %), and impairment of subcutaneous wound healing (N = 9, 12.8 %), also reflect the most common major complications described by other authors. The results of preceding studies varied in terms of the incidence of wound infections (8–20 %) and anastomotic leaks (up to 12 %) [8]. The high number of minor complications (26.1 %) in the OG, which was also described by Mazeh et al. [13], is further evidence for the severity of the entry trauma associated with this procedure. The present study is biased by a selection of younger patients for the laparoscopic approach, particularly in the first years after the implementation of this procedure. A potential limitation of this bias is that there were no significant differences concerning the ASA classification of the patients in the two groups. With increasing experience in the laparoscopic approach, further results from other study groups have shown that older patients with considerable comorbidities also benefit from less invasive methods [31]. The accepted advantages of laparoscopy for colorectal operations, including reduced entry trauma, more accurate tissue dissection at 129 magnification and the associated reduction in immune response, as well as the reduction in intraoperative morphine levels, appear to lead to a reduction in postoperative morbidity for the laparoscopic Hartmann reversal procedure [22, 32]. This assumption is supported by the reduction in minor and major complications by 15 and 43 %, respectively, as observed in our patient cohort. The most severe procedure-associated complication, anastomotic leak, was only found in the OG in our cohort (N = 2, 4.3 %) [33]. The preservation of the integrity of the abdominal compartment, especially with respect to physiological function of the anterior abdominal wall, as well as the minimal operation trauma, was a reason for the postoperative reduction in analgesic drugs. We also observed significant advantages in the earlier return of normal bowel function, a shorter recovery time, and a shorter LOS associated with the LG. The minimally invasive approach was also associated with a lower prevalence (p = 0.054) of late complications such as incisional hernias (OG 21.7 % vs. LG : 4.2 %) and adhesion ileus (OG 6.5 % vs. LG 4.2 %), both of which required surgical intervention. There are no published data available on the long-term course of patients who underwent laparoscopic Hartmann reversal. Nevertheless, comparative studies on laparoscopic versus open sigmoid resection have also shown a profound tendency of less incisional hernias and thus indirectly support this observation in our study [34]. The laparoscopic Hartmann reversal procedure offers more accurate tissue dissection as well as an additional diagnostic tool without increasing entry trauma. Previously unknown
pathology, such as, for example, metastases, fistulas, development of progressive diverticular disease or hernias, may be detected or ruled out [30]. It is also possible to treat some of the pathologies immediately during the same procedure or take biopsies for further analysis and the planning of additional therapies. Finally, the laparoscopic approach has proven to be comparable to the open operation regarding economical and technical aspects and especially with regard to patient safety [35, 36]. Consequently, the reversal of Hartmann’s procedure can usually be commenced laparoscopically providing there is no contraindication or expectation of extreme difficulties with adhesions and mobilization.
Conclusion The laparoscopic Hartmann reversal procedure is safe and effective. Compared with the open approach, it should be seen as equivalent regarding morbidity and mortality. The prolonged operation time is not associated with a negative influence on the postoperative course of the patients in this cohort. The reduction of entry and operation trauma is beneficial for patients with less incisional hernias, less consumption of analgesics, an earlier return of normal bowel function, and a shorter LOS. Previous studies were able to show that economic disadvantages resulting from the prolonged operation time and instrument costs are more than equalized by the advantages associated with laparoscopy, as mentioned previously. In the absence of contraindications, the operation can generally be commenced laparoscopically. Nevertheless, the technical requirements and the laparoscopic expertise necessary for this approach are seriously limiting the widespread use of this and many other laparoscopic procedures, despite its proven advantages. Disclosure statement Drs. M. Zimmermann, M. Hoffmann, P. Hildebrand, K.-F. Meyer, T. Jungbluth, U.-J. Roblick, H.-.P. Bruch, and E. Schloericke have no conflicts of interest or financial ties to disclose.
References 1. Hartmann H (1923) Note sur un procede nouveau d’extirpation des cancers de la partie du colon. Bull Mem Soc Chir Paris 49:1474–1477 2. Desai DC, Brennan EJJ, Reilly JF, Smink RDJ (1998) The utility of the Hartmann procedure. Am J Surg 175(2):152–154 3. Seah DW, Ibrahim S, Tay KH (2005) Hartmann procedure: is it still relevant today? ANZ J Surg 75(6):436–440 4. Farthmann EH, Ruckauer KD, Haring RU (2000) Evidence-based surgery: diverticulitis–a surgical disease? Langenbecks Arch Surg 385(2):143–151
123
World J Surg 5. Varma S (2009) Issues in irrigation for people with a permanent colostomy: a review. Br J Nurs 18(4):15–18 6. Fucini C, Gattai R, Urena C, Bandettini L, Elbetti C (2008) Quality of life among five-year survivors after treatment for very low rectal cancer with or without a permanent abdominal stoma. Ann Surg Oncol 15(4):1099–1106 7. Petersen M, Kockerling F, Lippert H, Scheidbach H (2009) Laparoscopically assisted reversal of Hartmann procedure. Surg Laparosc Endosc Percutan Tech 19(1):48–51 8. Khaikin M, Zmora O, Rosin D, Bar-Zakai B, Goldes Y, Shabtai M, Ayalon A, Munz Y (2006) Laparoscopically assisted reversal of Hartmann’s procedure. Surg Endosc 20(12):1883–1886 9. Zu¨hlke HV, Lorenz EM, Straub EM, Savvas V (1990) Pathophysiology and classification of adhesions. Langenbecks Arch Chir Suppl II Verh Dtsch Ges Chir 1009–1016 (German) 10. Vermeulen J, Gosselink MP, Busschbach JJV, Lange JF (2010) Avoiding or reversing Hartmann’s procedure provides improved quality of life after perforated diverticulitis. J Gastrointest Surg 14(4):651–657 11. Taylor C, Morgan L (2010) Quality of life following reversal of temporary stoma after rectal cancer treatment. Eur J Oncol Nurs 15(1):59–66 12. Albarran SA, Simoens C, Van De Winkel N, da Costa PM, Thill V (2009) Restoration of digestive continuity after Hartmann’s procedure: ASA score is a predictive factor for risk of postoperative complications. Acta Chir Belg 109(6):714–719 13. Mazeh H, Greenstein AJ, Swedish K, Nguyen SQ, Lipskar A, Weber KJ, Chin EH, Divino CM (2009) Laparoscopic and open reversal of Hartmann’s procedure—a comparative retrospective analysis. Surg Endosc 23(3):496–502 14. Keck JO, Collopy BT, Ryan PJ, Fink R, Mackay JR, Woods RJ (1994) Reversal of Hartmann’s procedure: effect of timing and technique on ease and safety. Dis Colon Rectum 37(3):243–248 15. Lin FL, Boutros M, Da Silva GM, Weiss EG, Lu XR, Wexner SD (2013) Hartmann reversal: obesity adversely impacts outcome. Dis Colon Rectum 56(1):83–90 16. Pearce NW, Scott SD, Karran SJ (1992) Timing and method of reversal of Hartmann’s procedure. Br J Surg 79:839–841 17. Roque-Castellano C, Marchena-Gomez J, Hemmersbach-Miller M et al (2007) Analysis of the factors related to the decision of restoring intestinal continuity after Hartmann’s procedure. Int J Colorectal Dis 22:1091–1096 18. Fleming FJ, Gillen P (2009) Reversal of Hartmann’s procedure following acute diverticulitis: is timing everything? Int J Colorectal Dis 24:1219–1225 19. Schmelzer TM, Mostafa G, Norton HJ et al (2007) Reversal of Hartmann’s procedure: a high-risk operation? Surgery 142:598–606 20. Muzii L, Bellati F, Zullo MA, Manci N, Angioli R, Panici PB (2006) Mechanical bowel preparation before gynecologic laparoscopy: a randomized, single-blind, controlled trial. Fertil Steril 85(3):689–693 21. Zmora O, Lebedyev A, Hoffman A, Khaikin M, Munz Y, Shabtai M, Ayalon A, Rosin D (2006) Laparoscopic colectomy without mechanical bowel preparation. Int J Colorectal Dis 21(7):683–687
123
22. Svenningsen PO, Bulut O, Jess P (2010) Laparoscopic reversal of Hartmann’s procedure. Dan Med Bull 57(6):A4149 23. Riansuwan W, Hull TL, Millan MM, Hammel JP (2009) Nonreversal of Hartmann’s procedure for diverticulitis: derivation of a scoring system to predict nonreversal. Dis Colon Rectum 52(8):1400–1408 24. Slawik S, Dixon AR (2008) Laparoscopic reversal of Hartmann’s rectosigmoidectomy. Colorectal Dis 10(1):81–83 25. Sosa JL, Sleeman D, Puente I, McKenney MG, Hartmann R (1994) Laparoscopic-assisted colostomy closure after Hartmann’s procedure. Dis Colon Rectum 37(2):149–152 26. Vacher C, Zaghloul R, Borie F, Laporte S, Callafe R, Skawinski P, Leynau G, Domergue J (2002) Retablissement de la continuite digestive apres intervention de Hartmann par voie laparoscopique. Etude retrospective de la Societe francaise de chirurgie endoscopique (SFCE). Ann Chir 127(3):189–192 27. Rosen MJ, Cobb WS, Kercher KW, Heniford BT (2006) Laparoscopic versus open colostomy reversal: a comparative analysis. J Gastrointest Surg 10(6):895–900 28. Costantino GN, Mukalian GG (1994) Laparoscopic reversal of Hartmann procedure. J Laparoendosc Surg 4(6):429–433 29. Di Carlo I, Toro A, Pannofino O, Pulvirenti EPE (2010) Laparoscopic versus open restoration of intestinal continuity after Hartmann procedure. Hepatogastroenterology 57(98):232–235 30. Sneider EB, Cahan MA, Litwin DEM (2010) Laparoscopic repair of acute surgical diseases in the 21st century. Minerva Chir 65(3):275–296 31. van de Wall BJ, Draaisma WA, Schouten ES, Broeders IA, Consten EC (2010) Conventional and laparoscopic reversal of the Hartmann procedure: a review of literature. J Gastrointest Surg 14(4):743–752 32. Huynh H, Trottier DC, Soto CM, Moloo H, Poulin EC, Mamazza J, Boushey RP (2011) Laparoscopic colostomy reversal after a Hartmann procedure: a prospective series, literature review and an argument against laparotomy as the primary approach. Can J Surg 54(2):133–137 33. Boccola MA, Lin J, Rozen WM, Ho Y (2010) Reducing anastomotic leakage in oncologic colorectal surgery: an evidencebased review. Anticancer Res 30(2):601–607 34. Siddiqui MR, Sajid MS, Khatri K, Cheek E, Baig MK (2010) Elective open versus laparoscopic sigmoid colectomy for diverticular disease: a meta-analysis with the Sigma trial. World J Surg 34(12):2883–2901. doi:10.1007/s00268-010-0762-3 35. Alkhamesi NA, Martin J, Schlachta CM (2011) Cost-efficiency of laparoscopic versus open colon surgery in a tertiary care center. Surg Endosc 25(11):3597–3604 36. Fleshman J, Sargent DJ, Green E, Anvari M, Stryker SJ, Beart RW Jr, Hellinger M, Flanagan R Jr, Peters W, Nelson H, for The Clinical Outcomes of Surgical Therapy Study Group (2007) Laparoscopic colectomy for cancer is not inferior to open surgery based on 5-year data from the COST Study Group trial. Ann Surg 246(4):655–662
