Discussion
The hypothesis of the APPAC trial is that the majority of patients with uncomplicated AA can be cured with wide-spectrum antibiotics avoiding a large number of unnecessary appendectomies and this hypothesis is supported by previous randomized studies. Acute appendicitis is one of the most common urgent conditions seen in general surgery practice. Although the exact mechanisms leading to this condition are still obscure, it is likely that luminal obstruction by external (lymphoid hyperplasia) or internal (sticked fecal material, appendicolith) compression plays a key pathogenetic role. The luminal obstruction leads to increased mucus production, bacterial overgrowth, and stasis, which increase appendiceal wall tension. Consequently, blood and lymph flow is diminished, and necrosis and perforation follow. As these events occur over time, it is conceivable that early surgical intervention prevents progression of disease. However, epidemiologic studies on incidence of nonperforated and perforated AA suggest that nonperforated and perforated AA may have different pathogenetic mechanisms strongly supporting our study hypothesis in re-evaluating the dictum that surgical removal of the appendix is always necessary for AA.
The best design for a therapeutic trial is a randomized placebo-controlled, double-blind study, but with the interventions used in the APPAC trial the concealment would not be possible and therefore a randomized open design was chosen. As concealment is lacking in all randomized trials comparing appendectomy with antibiotic therapy, the main focus should be on the safety of antibiotic treatment and the reduction in surgically-related morbidity and cost savings by using antibiotic therapy. Our power analysis and study hypothesis are based on the self-evident fact that efficacy of surgical treatment will be clinically superior to antibiotic therapy for uncomplicated AA – no appendix, no appendicitis – and therefore the primary end-point is treatment efficacy in both study arms. The primary endpoint of 30-day post-intervention peritonitis in the study of Vons et al. is not clearly defined and, in addition, the definition varies between treatment arms. In the study by Hansson et al. nearly half of the patients randomized to antibiotic group crossed over to the appendectomy group prior to receiving any drug and were classified as antibiotic treatment failures. Regarding these study designs, particular attention should be made to identify a clear and concise definition of efficacy to be used for both the conservative and surgical treatments, standardizing the different treatment procedures as much as possible even though there is an intrinsic difficulty in defining a common outcome for both treatment arms.
Before enrolling patients into a randomized trial, the diagnosis of AA needs to be confirmed by CT, but this inclusion criterion has been used so far in only one study. In contrast to this study by Vons et al., we have determined the presence of intraluminal appendicolith as an important exclusion criterion, as it has earlier been reported to predict negative outcome of non-operative management and to predict complicated AA. Indeed, if Vons et al. had excluded the patients with an appendicolith from their analysis, no significant difference in the incidence of post-intervention peritonitis between the treatment groups would have been noticed in their study.
The antibiotic therapy has been suboptimal in many previous randomized studies, as for example in the study by Vons et al. amoxicillin-clavulanic acid was used even though this combination has been associated with considerable Escherichia coli non-susceptibility. Furthermore, the use of this combination may play a role in both the initial antibiotic treatment failures and the recurrence of AA considering that this antibiotic treatment is not recommended to be used in the non-operative treatment of AA. The most common organism in AA is Escherichia coli, and the next most common is Enterococcus and other Streptococcus species. Pseudomonas, Klebsiella, and Bacteroides species are less commonly isolated. Accordingly, the selection of antibiotics should cover both aerobic and anaerobic bacteria. In the present study ertapenem was chosen for the antibiotic therapy, because it is is a broad-spectrum antibiotic with a single-dose daily administration and the efficacy of ertapenem monotherapy in serious intra-abdominal infections has been demonstrated.
The results of our interim analysis (n = 161) corresponded both with the hypothesis of our study and the sample size calculation. Vons et al. reported a recurrence rate of 26% in the antibiotic group. However, 68% of the patients in their study did not require appendectomy supporting our study hypothesis, that the majority of patients (> 70%) with uncomplicated AA can be treated successfully with antibiotics and unnecessary appendectomies can be avoided resulting in reduced morbidity and mortality of surgical treatment of AA, enormous cost savings and allocation of surgical resources to other emergency operations. Since so far only a small number of RCTs (< 1000 patients) with somewhat impaired methodological quality are available, more well-designed RCTs are urgently needed to both conclusively define the role of antibiotic therapy in the management of uncomplicated AA and to assess the predictive markers for successful non-operative treatment of uncomplicated AA.