Renewed Interest in Acute Appendicitis: Are Antibiotics Non-Inferior to Surgery or Possibly Clinically Superior? What Is Long-Term Follow-Up and Natural Evolution of Appendicitis Treated Conservatively with “Antibiotics First”?

To the Editor:

Recently there has been increased, renewed interest in the management of acute appendicitis [1], specifically regarding the debate on the efficacy of conservative treatment with antibiotics versus surgery and the possible short- and long-term outcomes of appendicitis treated with the “antibiotics first” strategy.

We read with interest the article by Flum [2], which outlines effectively the issues related to the treatment of appendicitis with either a straightforward appendectomy or by non-operative management (NOM). We would like to highlight a few areas that were described in the article as unclear or unavailable data. The statement “data from longer follow-up periods after initial successful NOM are unavailable” should be noted as incorrect as is the statement “unclear whether the likelihood of appendectomy continued to increase or stabilized over time.” The results of the Non Operative Treatment for Acute Appendicitis (NOTA) Study [3] were recently published. This long-term (two-year) study focused on the natural history of a prospective observational cohort of 159 patients who were treated conservatively with antibiotics for right iliac fossa (RIF) pain and suspected appendicitis. Keep in mind that when managing these patients non-operatively, we must emphasize that the diagnosis of appendicitis remains only a diagnosis of probability and therefore a clinical suspicion. In our patient cohort study, appendectomy after initial successful treatment with antibiotics (159 − 19 = 140 patients) occurred in 5.7% of those initially successfully NOM-treated patients (8/140) after a completed two-year follow-up period.

Short-term (less than seven days) treatment failures were 11.9% and the incidence of finding pathologically confirmed appendicitis in the patients who underwent appendectomy was 17 of 19 (89%). At 15 d, no early recurrence was recorded. At six months, 17 additional recurrence episodes (10.7%) occurred. At one year, only three additional recurrences later than six months from the index episode occurred (recurrence rate at one year: 12.6%; 20/159). Two-year follow-up showed overall recurrence rate of 13.8% (22/159).

Overall, 14 of 22 patients who experienced recurrences were again treated non-operatively with an additional cycle of amoxicillin-clavulanate with successful outcome, whereas eight patients who developed a recurrence required surgery (six had acute appendicitis and two had pelvic inflammatory disease).

We also read with interest the results of the Finnish APPAC Trial [4], which evaluated the efficacy of treating uncomplicated appendicitis with antibiotics. We do, however, have some concerns regarding a few possible methodological issues for this trial.

The primary end point of the study was to demonstrate non-inferiority of antibiotics versus appendectomy in the treatment of uncomplicated acute appendicitis; the pre-specified non-inferiority margin was 24%, which the current study actually failed to reach. Although this margin of 24% was not specified in the initial ClinicalTrials.gov record [5], the following statement can be found in the Methods section: “We tested the hypothesis that antibiotic treatment was non-inferior to appendectomy. Based on prior studies [6], we assumed that there would be a 24% difference in treatment efficacy between the surgical and antibiotic group.” How can the authors have chosen this margin as the referenced study was published in April 2012? The APPAC Trial Protocol was registered on the Clinical Trials database in November 2009 and the accrual closed in June 2012. Does it mean that the 24% non-inferiority margin was decided around or after the accrual closing date? The non-inferiority margin of 24% should have been estimated at the beginning of the study to allow a reliable calculation of the sample size, as stated in the Statistical Analysis section. How can these data, published in 2012, have been used to calculate sample size in 2009? Furthermore, in the referenced meta-analysis from Mason et al. [6], the reported difference in the treatment efficacy was approximately 32% (40.2% failure rate for antibiotics versus 8.5% for appendectomy).

Another concern is the termination of the enrollment, which was stopped prematurely after recruiting 530 patients. The investigators “believed” that a power of 0.86, instead of the originally calculated 0.90, was adequate. Could the authors report on what data or calculations they decided to terminate enrollment? Another issue is in the Results section, Table 3. The length of stay is shown to be shorter in the surgical group. Is a median of three days with 25th and 75th percentiles being 2–3 versus 3–3, respectively, be statistically significant? What was the mean of these numbers? Even if statistically significant, is this difference clinically significant? Regarding patient selection, what selection criteria was used in the emergency department to request computed tomography (CT) scans in patients with suspected appendicitis? Was it only the presence of RIF pain or were clinical scores used? Did all patients presenting to the emergency department with RIF pain undergo CT and be considered candidates for the study, regardless of their clinical findings, duration of pain, associated symptoms, physical examinations, and laboratory tests? In addition to the risks of radiation exposure for every patient presenting with RIF pain—even more so if young—we must emphasize that several patients may present CT findings suggesting appendicitis but not have an actual acute appendicitis, Conversely, patients with a clinically clear acute appendicitis may not have a visible appendix on CT scan. Additionally, an appendix diameter of more than 6 mm on the CT scan can be found in a large proportion of a control population [7,8].

Finally, we would like to comment on the relevant clinical meaning of the results of the current trial, which actually failed to prove non-inferiority of antibiotics to surgical treatment. Nonetheless, can antibiotics be considered truly inferior to surgery when including a greater incidence of complications in the surgical group? The overall complication rate was 20.5% versus 2.8%, with no patient developing intra-abdominal abscess, including those who underwent delayed appendectomy, versus a relevant incidence of surgical site infection (SSI), incisional hernias, pain, and adhesive obstruction symptoms. Time lost from work, visual analog scale, use of pain medications and concerns of poor cosmesis are not of secondary importance. Finally, the costs are not shown: It would be interesting the analysis of the costs of intravenous antibiotic regimen versus surgery (open and/or laparoscopic) and more interestingly compare with costs of previous studies using oral antibiotics [3]. Oral antibiotic course with amoxicillin/clavulanate for treating suspected appendicitis might be equally effective and as inexpensive as $16.50 per patient, with the overall costs of NOM and antibiotics being $352.00 per patient. With oral antibiotics, length of stay can also be as short as 0.4 days [3].

A Consensus Conference of international experts of the World Society of Emergency Surgery was held in Jerusalem, Israel, in July 2015 [9]. Several questions regarding diagnosis and management of acute appendicitis were discussed and recommendations proposed for approval, to be included in the World Society of Emergency Surgery guidelines on acute appendicitis.