Nonoperative Treatment Versus Appendectomy for Acute Nonperforated Appendicitis in Children

Objective: The aim of this study was to evaluate the safety and feasibility of nonoperative treatment of acute nonperforated appendicitis in children during 5 years of follow-up. Methods: A 4-year follow-up of a previous randomized controlled pilot trial, including 50 children with acute nonperforated appendicitis, was performed. The patients were initially randomized to nonoperative treatment with antibiotics or appendectomy with 1-year follow-up previously reported. Data were extracted from the computerized notes and telephone interviews. The primary outcome was treatment failure, defined as need for a secondary intervention under general anesthesia, related to the previous diagnosis of acute nonperforated appendicitis. Results: The children were followed up for at least 5 years [median 5.3 (range 5.0–5.6)] after inclusion. There were no failures in the appendectomy group (0/26) and 11 failures in the nonoperative group (11/24). Nine failures had occurred during the first year after inclusion, 2 of whom had histologically confirmed appendicitis. There were 2 further patients with recurrent acute appendicitis 1 to 5 years after inclusion. Both these patients had uncomplicated laparoscopic appendectomies for histologically confirmed acute appendicitis. There were no losses to follow-up. Conclusions: At 5 years of follow-up 46% of children treated with antibiotics for acute nonperforated appendicitis had undergone an appendectomy, although acute appendicitis was only histologically confirmed in 4/24 (17%). Treatment with antibiotics seems to be safe in the intermediate-term; none of the children previously treated nonoperatively re-presented with complicated appendicitis.

A cute appendicitis is one of the most common surgical emergencies worldwide. In Western countries, 8%-10% of the population will have acute appendicitis at some point in their life 1 and 1%-8% of children presenting with acute abdominal pain to the emergency room are diagnosed with acute appendicitis. 2 Despite the increasing knowledge regarding nonoperative treatment of appendicitis, an early surgical intervention is still the standard treatment.
A recent meta-analysis of nonoperative versus surgical treatment of acute uncomplicated appendicitis in adults suggested that antibiotic treatment without surgery is an effective alternative in this group of patients. 3 A recently published 5-year follow-up of the randomized controlled trial (RCT) comparing nonoperative treatment with surgery for acute uncomplicated appendicitis in adults supports this strategy. 4 A systematic review and meta-analysis of the same topic in children with acute uncomplicated appendicitis showed conflicting data, although, in the conclusion, the surgical treatment was favored. 5 However, this meta-analysis included only 1 single randomized controlled study, which was our pilot trial, whereas the rest of the studies were small, uncontrolled studies based on prospective patient-choice cohorts or retrospective physician-choice cohort. 5 The period of follow-up in the included studies varied from 1 to 4 years.
We have previously reported the feasibility of nonoperative treatment in children in a prospective randomized controlled pilot trial with a short-term follow-up of 1 year. 6 But, for nonoperative treatment to be a relevant alternative to surgical treatment, the longterm outcome needs to be further evaluated and be acceptable.
The aim of this study was to present 5-year follow-up data of the previous randomized controlled pilot trial. The study was designed to assess the hypothesis that nonoperative treatment of acute nonperforated appendicitis in children is safe and feasible in the intermediate-term follow-up.

Trial Design
This was a single-center 5-year follow-up study of a previous RCT comparing nonoperative treatment with antibiotics and appendectomy for acute nonperforated appendicitis in children.

Participants
In the RCT, children between 5 and 15 years of age with clinical diagnosis of acute nonperforated appendicitis, who before the trial would have undergone surgery, were invited to participate. A prior history of nonoperative treatment of acute appendicitis was an exclusion criterion for the RCT. The clinical diagnosis of acute nonperforated appendicitis was based on clinical examination, laboratory tests, and radiological imaging.

Study Settings
between February and October 2012. This is a tertiary pediatric surgical center covering the greater Stockholm area with a population of approximately 2.5 million inhabitants in 2012. At the time of the trial, all patients younger than 15 years of age with acute appendicitis in the Stockholm area were treated at this hospital.

Interventions
After providing written informed consent, the children were randomly allocated to either appendectomy or nonoperative treatment with antibiotics. Patients randomized to surgery were subjected to a standard appendectomy. This included fluid resuscitation and preoperative antibiotic prophylaxis with 20 mg/kg of metronidazole. All patients underwent a laparoscopic operation, even if the surgical modality was not specified in the study protocol. The further postsurgical treatment with antibiotics was depending on the intraoperative findings. The patients with a phlegmonous appendicitis did not received any further treatment. Those with gangrenous appendicitis received 24 hours of intravenous (iv) trimethoprim/sulfametoxazol and metronidazole. In the cases of perforated appendicitis, they received at least 3 days of iv trimethoprim/sulfametoxazol and metronidazole, depending on clinical improvement.
The children randomized to nonoperative treatment were given iv meropenem (10 mg/kg Â 3 per 24 hours) and metronidazole (20 mg/kg Â 1 per 24 hours) for at least 2 days. As soon as the children were tolerating oral intake, they were given oral ciprofloxacin (20 mg/kg Â 2 per 24 hours) and metronidazole (20 mg/kg Â 1 per 24 hours) for a total of 10 days treatment.
For this 5-year follow-up, we performed a review of the computerized notes to assess for failures, completed with a telephone contact with all patients and/or parents. This was done to find any interventions happening outside the reach of our computerized notes system and also to complete a short questionnaire.

OUTCOMES
The primary endpoint was treatment failure, defined as a need for secondary intervention under general anesthesia, related to the previous diagnosis of acute nonperforated appendicitis. This endpoint measure was designed to be applicable to both treatment groups despite a diverging panorama of complications. Appendectomy (for recurrent appendicitis or surgeon/patient/parent decision), surgery for or drainage of a deep abscess, surgery for malignancy of the appendix or caecum or surgery for mechanical bowel obstruction were the suggested reasons for treatment failure.
The secondary endpoints were hospital readmission, length of hospital stay at readmission, accident and emergency (A&E) admission related to failure or complication of initial treatment, and total cost. Acute appendicitis was confirmed by histopathology according to Carr. 7

Sample Size
The initial RCT was designed as a pilot trial to assess the feasibility of a multi-center RCT. Therefore, a power calculation was not performed. Fifty patients were included in the study based on the historic admission figures, an estimated recruitment rate of 30% and the ambition to conclude the enrolment within a 6-month period. For this follow-up, no patients were added to or excluded from the original cohort.

Randomization
Allocation to groups (1:1 ratio) was made via weighted minimization at the time of enrolment in the study. 6 The following criteria were used: age (5-10 years or 11-15 years), sex (male or female), and duration of symptoms (<48 or !48 hours). All factors were weighted equally. A computer-based randomization program (Simin v 6.0; Institute of Children Health, London, United Kingdom) was used for randomization. After acceptance of enrolment, the attending pediatric surgeon used the computer-based randomization program to put in the relevant minimization criteria and received the random treatment allocation at the same time.

Blinding
For the RCT, it was not possible to blind patients, parents or surgeons due to the nature of the treatment modalities. Blinding was not performed during the 5-year follow-up.

Statistical Methods
Data are presented as the proportion of patients or median (range). Data were compared using the Mann-Whitney U-test, Fisher exact test, or with a Kaplan-Meier survival analysis, using IBM SPSS Statistics version 22.
This trial is reported in accordance with the consolodated standards of reporting trials-guidelines. 8

Ethical Approval
The study was approved by the Regional Ethics Review Board in Stockholm, Sweden, ref 2011/1234-31/4.

RESULTS
A total of 225 children with a clinical diagnosis of acute appendicitis were screened for eligibility during the study period, and 168 met the eligibility criteria. The total number of patients excluded due to various reasons is shown in Fig. 1.
We randomized 51 patients. One patient withdrew consent after randomization but before initiation of treatment, leaving 26 randomized to surgery and 24 to nonoperative treatment. At inclusion, all 50 patients had a radiological evaluation. 46 patients had only an abdominal ultrasound, 4 patients had both an abdominal ultrasound and computed tomography (CT) scan. All patients were diagnosed with acute appendicitis; 12 patients had appendicitis with appendicolith. There was 1 patient, who had both an ultrasound and CT scan, randomized to nonoperative treatment, who underwent an appendectomy because of ongoing abdominal pain, where the initial diagnosis was acute appendicitis, but after appendectomy the senior radiologist changed the diagnosis to mesenteric lymphadenitis. The 1-year follow up was finished in October 2013. The results after 1 year of follow-up have previously been reported. 6 The 5-year follow-up of all the 50 participants enrolled in the study was performed between September 26 and November 15, 2017. Each patient had been followed up for at least 5 years [5.3 years (5.0-5.6)] after enrolment. During the follow-up, there were 9 patients who had a radiological examination in the nonoperative group, 7 patients had only an ultrasound, 1 patient had both an ultrasound and CT-scan; and 1 patient had only CT-scan. The results were: 6 patients were diagnosed with acute appendicitis, among them 1 patient was diagnosed with suspicion of perforation; 1 patient had normal appendix; and 1 patient had an appendix with a diameter exceeding 6 mm, but there were no signs of inflammation.
Baseline demographic and admission characteristics were similar between patients randomized and those who declined to participate. However, symptom duration was shorter in the randomized children compared to children not invited to participate (Table 1). There were no differences between the patients randomized to nonoperative treatment compared to those randomized to surgery ( Table 2).

Outcomes
In the appendectomy group, there were no complications after the initial treatment (0/26). All patients underwent laparoscopic appendectomy without perioperative complications. Histological examination confirmed acute appendicitis in all cases. In total in the surgical group, there were 20 patients with histopathological diagnosis of phlegmonous appendicitis, 4 patients had gangrenous appendicitis, and 2 patients had perforated gangrenous appendicitis.  One patient reported cosmetic dissatisfaction with the infraumbilical scar at 1-year follow-up, but at 5-year follow-up the patient was satisfied. There were no complications registered at 5-year followup. No patients were readmitted related to the previous appendectomy during the 5-year follow-up period.
In the nonoperative group, there were 11 failures which resulted in appendectomy at the 5-year follow-up (11/24). Nine of the 11 failures occurred during the first year after the inclusion, in 2 of these patients acute appendicitis was verified histologically, as previously reported. Another 2 patients had undergone appendectomy between 1 and 5 years of follow up. In both these cases of late failure, acute appendicitis was confirmed by histopathology. All patients with failure in the nonoperative group underwent uncomplicated appendectomies, 10 laparoscopic and 1 open procedure. The indication for appendectomy, histopathological appearance of the appendix and resolution of symptoms of these patients are shown in Table 3.
The rate of appendectomy over time in the nonoperative group is presented in Fig. 2, both for all appendectomies and for appendectomy with histopathologically confirmed appendicitis. In the nonoperative group, there were 11 appendectomies. Overall, there was no surgery or drainage of deep abscess, no surgery for malignancy of appendix or caecum, or surgery for mechanical bowel obstruction.   During 5-year follow up, there were no visits to the A&E department registered in the surgical group. In total, there were 12 patients in the nonoperative group who visited the emergency department at least once. One patient visited the emergency department 3 times before the admission for appendectomy. The nonoperative treatment group can be divided into 2 subgroups. The first group are the patients who underwent appendectomy. In this group 1 patient had complaints of abdominal pain both before study inclusion and after the appendectomy. In the subgroup of the patients who still had not undergone appendectomy at the 5-year follow up, there were 5 patients who had abdominal pain problems. Of these patients, 1 did not visit the emergency department, 1 had a phone contact with our hospital, and 3 patients visited the emergency department 2 times. At the end of the follow-up period they had no problem with abdominal pain related to the previous diagnosis of acute appendicitis. The

DISCUSSION
This is the first study assessing 5-year outcome of nonoperative treatment of acute nonperforated appendicitis in children. Most recent studies of nonoperative treatment of acute appendicitis have focused on the success of treatment or recurrence of symptoms necessitating appendectomy within the first year. Of the 24 children randomized to nonoperative treatment, 13/24 (54%) have not had an appendectomy within 5 years, and only 4/24 (17%) have had an appendectomy for histologically confirmed recurrent acute appendicitis. Of the 15 children not having an appendectomy within the first year, 2 (13%) had an appendectomy within the following 4 years. During 5-year follow-up interviews, all parents expressed satisfaction with their choice to participate in the study, also in cases where their children underwent appendectomy during the study period.
Importantly, none of the children with recurrence presented with perforated appendicitis, and all had an uncomplicated course, suggesting that initial nonoperative treatment is safe in the mediumterm, although the number of patients is limited.
Two patients randomized to surgery (2/26; 8%) had a finding of perforated appendicitis. They did not have any clinical or radiological signs of perforation during initial evaluation of eligibility for the study. It is not possible to identify when perforation occurred and we do not know this was due to radiologically missed perforated appendicitis, or to perforation occurring between recruitment and surgery.

Limitations
There are a few limitations of this study. First, the study was designed as a pilot RCT with relatively small sample size not powered to detect differences in treatment efficacy. The purpose was to determine whether treatment with antibiotics is feasible and safe. However, the outcome data can be used to design and perform future prospective multicentre studies. The second limitation is that we did not have a protocol how to deal with patients in the nonoperative treated group presenting with recurrent abdominal symptoms. As this was the first randomized controlled study of nonoperative treatment of acute nonperforated appendicitis in children and we did not want to risk any missed neoplasm, we were liberal with appendectomy. In most cases, the indication for surgery was mild abdominal pain and a decision made by an attending surgeon and parents. There is a possibility that parents of the children in the nonoperative treatment group were concerned about abdominal pain as a potential symptom of recurrent appendicitis and, therefore, wanted their child to undergo an appendectomy. However, the surgical treatment was successful and without any perioperative complications.

Generalizability
The study includes a representative spectrum of pediatric patients with acute nonperforated appendicitis. All patients who, before the study, would have undergone a surgery and had no clinical or radiological suspicion of perforated appendicitis, could have participated in the study. We did not use any clinical scoring system as we feel that the available scoring systems are most useful in situations where abdominal ultrasound and CT are not routinely available. In our centre, we have a negative appendectomy rate of 3.8%. 9 However, there were 37 (22%) patients who had never been asked to participate in the study for unknown reasons. One possible motive explaining this could be surgical bias, in cases where children had longer duration (more than 48 hours) of symptoms, the urge of the surgeon to do an appendectomy. We did not exclude patients with appendicolith on radiological imagining.

Interpretation
We demonstrated that nonoperative treatment of acute nonperforated appendicitis can be reliably and effectively delivered in children.
To date there have been a few studies with various study designs comparing nonoperative treatment of acute nonperforated appendicitis in children but no fully powered RCTs. 2,3,5 Minneci and co-workers published a prospective patient-choice cohort study with patient and parents' choice between appendectomy and nonoperative management. In the study, they included 102 patients (age 7-17) with acute uncomplicated appendicitis, where 65 patients chose appendectomy, and 37 chose nonoperative management. The success rate of the nonoperative management was 89% at 30 days follow-up, and 75% at 1-year follow-up. There were 9 patients who underwent an appendectomy during 1-year follow-up. Even though they tried to minimize bias by evaluating the eligible patients by 4 trained surgeons, there still is the selection and preference bias which cannot be ignored. Due to the lack of randomization, there is a likelihood of bias towards the less severely ill patients in preference toward nonoperative treatment. One may assume that the groups differed in severity of illness and/or pain (not published in baseline clinical characteristics). The fewer disability days in the nonoperative group might also suggest that they were not particularly unwell. Only a true RCT will minimize the probability of these differences influencing trial results.
As this is the first RCT in children, we are not able to compare our findings directly with other study results in children. Many studies, for diseases occurring both in adults and children, are usually initially performed in adults, followed by studies in children. Salminen and co-workers have recently published a 5-year follow-up of antibiotics therapy for acute uncomplicated appendicitis in the ''APPAC (the APPendicitis ACuta)''-trial in adults. There was 39% (100 of the 256 patients) recurrence after 5-year follow-up. Most of these patients (70/100, 70%) had their recurrent appendicitis within 1 year from the initial presentation. Fifteen patients underwent surgery during the initial hospitalization. 76 out of 85 patients had an uncomplicated appendicitis confirmed histopathologically. There were 7 patients without acute appendicitis on histopathology. Complicated appendicitis was found in 2 patients who underwent appendectomy in second to fifth year. In this study, the researchers excluded patients with appendicolith because of the concern of not definitive relationship between appendicolith and recurrence of Acute Appendicitis (AA).
In our findings, there were only 4 children with acute appendicitis on histopathology in the nonoperative group. Most of the histopathological findings were an appendix without inflammation, but with fibrotic changes. It is difficult to say if these changes occurred postinflammatory, or it could have been fibrous obliteration from the beginning, or chronic fibrosis, or neurogenic appendicopathy. 10 The clinical symptoms of these patients are not distinguishable from patients with AA. 10 We have no way of knowing how many patients in the nonoperative group had no acute appendicitis as we did not operate on them. We are aware that there can be some patients who did not have an acute appendicitis at the time of randomization. We only know that the 2 groups of patients were very similar at the time of randomisation. It is difficult to know the exact treatment effect of antibiotics. There is the possibility of spontaneous resolution of symptoms. However, these are 2 entities that need to be addressed in future. It is not possible to separate these 2 effects in our study.
In addition, there is a concern about missing a neoplasm in the appendix, or leaving a chronical infection which could lead to increased incidence of bowel cancer in these patients. Enblad and co-workers have recently published a population-based study from the Swedish National Inpatient Register, showing that patients with nonsurgical treatment of appendicitis have an increased short and long-term incidence of bowel cancer. 11 They state that patients with nonoperative treatment of acute appendicitis had an elevated risk of right-sided colon cancer up to 5 years after appendicitis. However, there was only 1% of patients with bowel cancer in the age group 10 to 19. There is no more specific description of which type of cancer, or how advanced the disease they had was in this age group, no information about how they were diagnosed or treated. We do not know the causation between nonoperative treatment of acute appendicitis and cancer. Underlying malignancy can present for the first time with the symptoms of acute appendicitis, and it is important to be aware of it. However, we cannot be certain how they got diagnosed with acute appendicitis, and how they were treated, if not surgically. We do not currently support routine use of nonoperative treatment with antibiotics outside the scope of RCTs or other prospective studies.