by D. Brandt, M.D., P. Gervaz, M.D., Y. Durmishi, M.D., A. Platon, M.D., Ph. Morel, M.D., P. A. Poletti, M.D.
CT<scan–guided percutaneous abscess drainage of Hinchey Stage II diverticulitis is considered the best initial approach to treat conservatively the abscess and to subsequently perform an elective sigmoidectomy. However, drainage is not always technically feasible, may expose the patient to additional morbidity, and has not been critically evaluated in this indication. This study was undertaken to compare the results of percutaneous drainage vs. antibiotic therapy alone in patients with Hinchey II diverticulitis.
This was a case-control study of all patients who presented in our institution with Hinchey Stage II diverticulitis between 1993 and 2005. Thirty-four patients underwent abscess drainage under CT-scan guidance (Group 1), and 32 patients were treated with antibiotic therapy alone (Group 2), in most cases because CT-scan-guided abscess drainage was considered technically unfeasible by the interventional radiology team. Initial conservative treatment was considered a failure when: 1) emergency surgery had to be performed, 2) signs of worsening sepsis developed, and 3) abscess recurred within four weeks of drainage.
The median size of abscess was 6 (range, 3–18) cm in Group 1 and 4 (range, 3–10) cm in Group 2 (P = 0.002). Median duration of drainage was 8 (range, 1–18) days. Conservative treatment failed in 11 patients (33 percent) of Group 1, and in 6 patients (19 percent) of Group 2 (P = 0.26). Ten patients (29 percent) in Group 1 and five patients (16 percent) in Group 2 underwent emergency surgery (P = 0.24); there were four postoperative deaths (26.6 percent) in this subgroup. Twelve patients (35 percent) in Group 1 and 16 patients (50 percent) in Group 2 subsequently underwent an elective sigmoid resection (P = 0.31). In this subgroup of patients, there was neither anastomotic leakage nor postoperative death.
Emergency surgery for Hinchey Stage II diverticulitis carries a high mortality rate and should be avoided. To achieve this, antibiotic therapy alone seems to be a safe alternative, whenever percutaneous drainage is technically difficult or hazardous. Actually, our data did not demonstrate any benefit of CT scan-guided percutaneous abscess drainage, suggesting that the role of interventional radiology techniques in this indication deserves further critical evaluation.
According to many expert committees, percutaneous abscess drainage (PAD) associated with intravenous antibiotics is considered the standard therapeutic approach for patients with Hinchey Stage II diverticulitis.
The rationale behind this strategy is twofold:
- initial conservative treatment of sepsis, and
- subsequent elective sigmoidectomy with primary colorectal anastomosis.
This approach has been widely adopted and has benefited from the rapid development of radiologic interventional procedures during the last two decades, as well as by the recognition of the role of CT scan as the initial imaging modality for suspected diverticulitis. Besides the fact that this therapeutic approach has entered surgical practice without any Type 1 evidence, there are several limitations to percutaneous drainage of diverticular abscesses. First, feasibility is not always possible because of the presence of small-bowel loops in contiguity with the fluid collection. Second, drainage, when feasible, is not always successful; reported failure rates for diverticular abscesses range from 15 to 30 percent, and when PAD fails, the mortality rate may be as high as 75 percent.
Third, when feasible and initially successful, abscess recurrence and/or fistula may occur and compromise the performance of an elective surgical resection. In addition, the literature on the subject is somewhat confusing; most series report mixed cases of abdominal fluid collections or consider inappropriate end points to evaluate the technique. Thus, the question is to determine whether a conservative treatment with systemic antibiotic therapy alone could achieve similar or even better rates of elective sigmoid resection than CT-scan–guided drainage in patients with Hinchey II diverticulitis.
Because of the low incidence of this clinical presentation, it is likely that the answer to this question will require a multicenter trial. In the meantime, this study was undertaken to retrospectively compare the results of both options in our institution.
Patients and methods
All patients with a diagnosis of Hinchey II diverticulitis who were admitted in our institution between January 1993 and June 2005 were identified by searching the interventional radiology database. The study protocol was submitted to the local ethical committee, and permission was obtained from the patients for our review of their medical records. Each patient_s medical record was reviewed separately by a surgeon (DB, PG, YD) to determine the demographics, size and number of abscesses, date and duration of drainage, and the outcome of PAD or conservative treatment. Initial treatment was considered a failure when:
- the general condition of the patient did not improve and signs of worsening sepsis developed,
- abscess and/or fistula recurred within four weeks of drainage, and
- emergency surgical resection with or without a colostomy had to be performed.
The medical record also was reviewed to document the occurrence of an eventual surgical resection; the type of surgery (Hartmann procedure vs. sigmoidectomy with primary anastomosis), timing of the procedure (emergent vs. elective), and outcome of surgery were recorded.
Since 1991, CT scan with oral, rectal, and intravenous contrast has been considered in our institution as the standard imaging technique for patients with suspected diverticulitis. When the initial CT scan demonstrated a complicated diverticulitis with an abscess, the indication for drainage and its feasibility were discussed jointly between interventional radiologists and colorectal surgeons. All drainage procedures were performed by experienced interventional radiologists under CT-scan guidance. Depending of abscess location, drainage of collection was achieved through a transgluteal or lateral transabdominal approach. Self-retaining pigtail catheters with distal hydrophilic tip with size ranging from 12 to 14 French were used in all cases. Immediately after catheter placement, a syringe was inserted and the fluid collection was aspirated until the flow ceased; the pus was then sent for bacteriologic analysis. After aspiration, the catheter was left in place to allow gravity drainage and its adequate position was subsequently verified with additional imaging. The interventional radiology team and the surgical staff jointly performed patient follow-up. The outcome of PAD was assessed after the procedure by CT-scan imaging of the lower abdomen to determine abscess resolution/recurrence, as well as by observation of improvement/deterioration in patient_s clinical condition. All patients who underwent a simple fluid evacuation without permanent catheter placement were excluded from the study.
Systemic Antibiotic Therapy
Intravenous antibiotics used were similar in both groups and consisted of a monotherapy (imipenem) or a combination of ceftriaxone/metronidazole or ciprofloxacin/metronidazole. The choice of antibiotic was at the discretion of the emergency unit team. The intravenous antibiotics were usually delivered for ten days. Indications for treating patients with systemic antibiotic therapy alone were the following:
- PAD considered technically difficult/unfeasible by the interventional radiology team because of abscess location or the presence of small-bowel loops in contiguity,
- patient refusal to undergo subsequent drainage, and
- rapid improvement of the clinical condition after admission under systemic antibiotic therapy alone.
Definition and Medical Management of Hinchey II Diverticulitis
The staging of complicated diverticulitis was established according to the original Hinchey classification; in this classification, Stage II is related to a large (>3–4 cm) pus collection, which is at distance (in the pelvis or the abdomen) of the sigmoid colon.
Unlike Stage I, Hinchey Stage II abscesses are not confined within the mesocolon and are considered amenable to percutaneous drainage. All patients had their final diagnosis of complicated diverticulitis confirmed at the time of surgery or by colonoscopy. Five patients initially considered to have complicated diverticulitis underwent eventual surgery, with a final diagnosis of sigmoid adenocarcinoma, and thus were excluded from the study.
Statistical analyses were undertaken by means of the software package, STATGRAPH 3.0\ software for Windows (Statgraph Software, Inc., San Diego, CA). Quantitative data were expressed as medians with ranges. Group comparisons were made by using twosided Fisher_s exact test for categorical variables and two-sided Student_s t-test, or Mann-Whitney U test when indicated, for continuous variables. P e 0.05 was considered statistically significant.
In the PAD group, 34 patients (17 males; median age, 71 (range, 34–90) years) were considered for analysis (Group 1). Thirty-two patients (19 females; median age, 69 (range 32–95) years) were treated with antibiotic therapy alone (Group 2). The clinical characteristics of patients in both groups are summarized. Of note, the median size of abscess was smaller in Group 2 (median, 6 (range 3–18) cm (Group 1) vs. 4 (range, 3–10) cm (Group 2), MannWhitney U test, P = 0.002). One patient in each group had multiple (n = 2) abscesses. In Group 1, median duration of drainage was 8 (range, 1–18) days. Drainage was considered successful in 23 patients (67 percent). Causes for failure in 11 patients included aggravating sepsis (5 cases), abscess recurrence (5 cases), and fistula formation (1 case). Ten patients who failed PAD underwent an emergency Hartmann procedure with a median delay of 14 (range, 1–65) days between drainage and surgery; in this group three patients (33 percent) died in the immediate postoperative period. Thus, the overall mortality rate in Group 1 was 8.8 percent. Among 23 patients who were successfully drained, 12 patients eventually underwent an elective sigmoid resection with a primary anastomosis, with a median delay of 101 (range, 40–420) days between drainage and surgery. In this group, there was no mortality and no anastomotic leak. Of note, among 11 patients who did not undergo sigmoid resection for various reasons, 2 elderly females eventually presented with a perforated diverticulitis and underwent emergency surgery more than seven years after the initial diagnosis. Clinical outcome of Group 1 patients is summarized in Figure 1. In Group 2, systemic antibiotic therapy was successful in 26 patients (81.3 percent) and 16 (50 percent) of them eventually underwent an elective sigmoidectomy, with a median delay of 88 (range, 48–601) days; again, there was neither anastomotic leak nor postoperative death in this subgroup of patients, who underwent an elective sigmoidectomy. Five of six patients (18.7 percent), who failed the initial therapeutic approach, underwent a Hartmann procedure, and one 88-year-old patient died in the postoperative period; thus, overall mortality in Group 2 was 3.1 percent. Clinical outcome of Group 2 patients is summarized.
This series of 66 patients indicate that emergency surgery for Hinchey Stage II diverticulitis has a >25 percent mortality. Conservative treatment was successful in three out of four patients and was similar in the PAD (67 percent) and the antibiotic therapyalone group (81 percent).
Thirty-five percent of Group 1 patients and 50 percent of Group 2 patients eventually underwent elective sigmoidectomy, with a median delay of 113 (range, 40–600) days between the initial admission and surgery. In this subgroup of patients, there was no anastomotic leak or postoperative death. These data indicate that patients treated with antibiotherapy alone may achieve a clinical outcome similar to patients treated with PAD. The overall mortality rate in our series is 6 percent (4/66 patients), and the mortality rate after emergency surgery is 26.6 percent, which is in accordance with three large series, reporting 5 to 12 percent overall mortality rates and 18 to 23 percent mortality rate after emergency surgery.
In addition, our data demonstrate a 33 percent failure rate of PAD, which also is in accordance with most series from the literature. In summary, data from our series and from others indicate that emergency surgery should be avoided whenever it is possible in Hinchey II diverticulitis and that PAD might fail to achieve this in approximately one-third of these patients. The present paper also focuses on a group of Hinchey II patients who were treated with intravenous antibiotics alone, because abscess location and/ or sentinel small-bowel loops made PAD technically unfeasible. It is noteworthy that the clinical outcome in this group was similar to the group of patients who underwent PAD. Because of the retrospective nature of our study, it is possible that patients who had a more aggressive course of the disease were more likely to undergo PAD because the size of abscess was larger in the PAD group. Nevertheless, it is interesting to note that antibiotics alone offer a decent alternative and were successful in avoiding emergency surgery in > 80 percent of patients who presented in our institution with Hinchey II diverticulitis.
PAD has emerged during the past two decades as the standard treatment for peridiverticular abscesses that are more than 3 to 4 cm in diameter and is becoming increasingly popular, in parallel with the development of interventional radiology techniques. Using a statewide hospital discharge database, Salem et al. demonstrated that the odds of percutaneous abscess drainage for complicated diverticulitis increased 7 percent per year between 1987 and 2001, concomitantly with a 2 percent per year decrease of emergency colectomies. Consequently, it is clinically relevant to determine what is the best therapeutic alternative in Hinchey II patients who, for technical/anatomic reasons, are considered poor candidates to PAD.
Hinchey Stage II diverticulitis is a severe condition, which often affects elderly patients, and has a 5 to 10 percent overall mortality rate. Emergency surgery in this situation, with a mortality rate superior to 25 percent, should be avoided whenever possible. The clinical outcome is favorable in roughly two thirds of patients, and correlates with the response to conservative treatment, whether it is based on PAD or antibiotic therapy alone, both modalities having similar success rates.
The results of this study demonstrate that systemic antibiotic therapy alone is a safe and efficient alternative whenever percutaneous drainage is technically difficult or hazardous.
Multicenter, prospective studies are needed to clarify the role of interventional radiology techniques in this indication.
in Diseases of the Colon and Rectum 2006 Oct; 49(10): 1533-8.