Appendectomy Versus Observation for Appendicitis in Neutropenic Children With Cancer

This retrospective study was completed by the Pediatric Surgical Oncology Research Collaborative, a multi-institutional group of North American pediatric surgeons devoted to the surgical care of children with cancer. Institutional review board approval and a waiver of informed consent was obtained at each participating institution, coordinated by a central institutional review board at Cincinnati Children’s Hospital Medical Center. Data for this study were collected from 16 institutions.

Inclusion criteria for the study included patients aged 0 to 18 years with severe neutropenia (absolute neutrophil count [ANC] <500/μL) due to cancer therapy at the time of appendicitis diagnosis between January 1, 2012, and December 31, 2018. Diagnosis of appendicitis was confirmed by imaging (ultrasound, MRI, or computed tomography [CT] scan), and patients with evidence of neutropenic colitis were excluded. Any mention of colitis on any imaging, including the possibility of or concern for colitis, led to exclusion. Patients with congenital causes of neutropenia were excluded. Standardized data collection was performed by using a Research Electronic Data Capture database created by investigators. The data instrument included patients’ demographics, malignancy type and treatment regimen, presenting signs and symptoms of appendicitis, treatment approach, surgical and infectious complications, and resultant chemotherapy delays or alterations (Appendix 1).

Descriptive statistics were calculated for all variables collected to examine baseline demographic and clinical factors. Categorical variables were analyzed by using χ² contingency tests or Fisher’s exact tests, depending on frequencies, and continuous variables were analyzed by using the Student’s t test for means or Kruskal-Wallis test for medians; P values ≤.05 were considered significant. All analyses were completed by using Stata 15 (Stata Corp, College Station, TX).¹⁴ 

A total of 66 patients from 16 North American institutions were included in the analysis. Baseline demographics for patients who underwent upfront appendectomy and initial nonoperative management are compared in Table 1. The median age at the time of appendicitis diagnosis was 11 years (range 1–17 years, interquartile range [IQR] = 5–14), and 50% (n = 33) of the patients were male. The majority (53%) of patients were Non-Hispanic white. Sixty-two percent of the patients had leukemia, 12% had a brain tumor, and 6% had lymphoma. Twenty-four percent (n = 16) of the patients had a history of stem cell transplant. There were no differences in sex, race and/or ethnicity, malignancy type, or age between the initial management groups.

The upfront appendectomy and initial nonoperative management groups did not differ at presentation in terms of median ANC, white blood cell count, platelet count, maximum temperature on the day of diagnosis, evidence of complicated appendicitis, symptom duration, or diagnostic imaging modality (Table 2).

Initial management strategy included upfront appendectomy within 24 hours in 41% (n = 27) of patients, as compared to initial nonoperative management in 56% (n = 37) (Fig. 1). The rate of upfront appendectomy ranged from 0% to 100% at the different institutions. Among all patients, granulocyte colony-stimulating factor was given in 44% (n = 29) to stimulate count recovery. Among patients treated with initial nonoperative management, 46% (n = 17) underwent subsequent delayed appendectomy during the same hospitalization. Delayed appendectomy was performed because of count recovery in 6 patients and failure of nonoperative management in 11. The median number of days from diagnosis of NA to delayed appendectomy was 5 days. For the other 20 patients who had initial nonoperative management and no delayed appendectomy during that hospitalization, the ultimate outcome included unplanned return for appendectomy or further care of the appendix in 3, routine interval appendectomy during another hospitalization in 2, and definitive nonoperative management without appendectomy in 15. All appendectomies were performed laparoscopically. Operative findings were reported as acute appendicitis in 21 patients, gangrenous or suppurative in 3, and perforated in 6.

In Table 3, we display the complications, including bowel obstruction, bowel resection, abscess, and Clostridioides difficile infection, which occurred in 7% of patients who had upfront appendectomy compared with 27% who had initial nonoperative management (P = .05). Within the initial nonoperative group, the subset of patients who had delayed appendectomy during the same admission had an even higher complication rate (59%) compared with the upfront appendectomy group (P < .01) (Table 3). There were no cases of postoperative fistulae or wound dehiscence and no deaths due to appendicitis or complications in either cohort (Table 3). Average LOS was significantly longer in the initial nonoperative management group compared with the upfront surgery group (12 vs 29 days; P < .01) (Table 4).

Oncology treatment was delayed >48 hours in 36% (n = 23) of patients, including 2 to 7 days in 14% (n = 9), 7 to 14 days in 9% (n = 7), 14 to 21 days in 5% (n = 3), and >21 days in 6% (n = 4). One patient additionally required alteration in their chemotherapy regimen. Therapy alteration or delay occurred in 22% (n = 6) of patients with upfront appendectomy and 46% (n = 17) of those with initial nonoperative management (P = .05) (Table 4). Delays or alteration in chemotherapy regimens occurred at similar rates in patients with perforated versus nonperforated appendicitis (P = .51).

In this North American, multicenter, retrospective cohort study, we found that upfront appendectomy in pediatric patients with NA was associated with a shorter LOS and fewer delays or alterations to their chemotherapy regimen, as compared with those treated with initial nonoperative management, although ANCs were similar between the cohorts. Additionally, patients managed medically who later underwent delayed appendectomy because of either treatment failure or count recovery experienced increased postoperative complications, as compared with patients treated with upfront appendectomy.

The optimal management of NA in children undergoing chemotherapy remains ill defined, and there is significant practice variation. Numerous studies reveal the safety and efficacy of upfront appendectomy.^9,11,15–18  The largest case series of NA in children is a retrospective review from 12 French pediatric oncology units. Of the 30 patients included, 6 underwent early surgery (surgery within 48 hours), 7 patients were medically managed, and 17 were treated with a combined approach consisting of initial antibiotic management and interval appendectomy. Patient outcomes, including death, duration of antibiotics, and LOS, were not significantly different between the groups. Three patients had postoperative wound infections, and a higher percentage of patients in the early surgery group were diagnosed with septic shock (33%) than in the medically managed group (0%).⁷ 

A recent study of 1068 children who had either initial nonoperative management or upfront surgery for treatment of uncomplicated appendicitis found surgical site infection rates of 0.3% in the initial nonoperative group and 1.1% in the upfront surgery group.¹⁹  Operative intervention may be avoided in patients with neutropenia out of concern for postoperative complications, such as delayed wound healing, surgical site infections, postoperative abscess formation, and postoperative sepsis, although there are few (if any) publications in which researchers support this idea in the context of laparoscopic abdominal operations.²⁰  Data on postprocedural complications in children with neutropenia are more robust in children undergoing central-line placement for chemotherapy administration; however, there remains variation in practice and conflicting evidence.^21–23  In a recent review of 542 patients at St Jude Children’s Research Hospital undergoing tunneled venous catheter placement, researchers found no relationship between early postoperative infection and neutropenia.²⁴  In the present analysis of children with NA on chemotherapy, children who underwent delayed appendectomy after count recovery or failure of initial nonoperative management experienced over a threefold increase in postoperative complications, as compared with children who had upfront appendectomy within 24 hours of diagnosis (P < .01). The most common complications were intraabdominal abscess in over a third of children in the delayed appendectomy group and 7% in the upfront appendectomy cohort. There were no postoperative wound infections or mortalities due to appendicitis in either treatment group.

It is unknown whether delays or alterations in chemotherapy secondary to treatment of NA result in differences in oncologic outcomes. The effect of chemotherapy alterations or delays in overall survival and disease-free survival is likely related to cancer type and aggressiveness as well as efficacy of available chemotherapy. In at least 3 studies, researchers have assessed whether delays in the initiation of chemotherapy at the outset of therapy impact short- and long-term outcomes in patients with leukemia, the most common diagnosis in our patient cohort. In an American observational cohort study of 1317 patients with acute myelogenous leukemia (AML), researchers determined that delaying chemotherapy >5 days after diagnosis was associated with inferior rates of complete remission.^12,25  In a single-center, French study of 599 patients, researchers did not identify a relationship between time-to-treatment and complete remission, early death, or long-term survival in patients diagnosed with AML.²⁵  In a 2014 Danish, national cohort study consisting of 1388 patients, researchers found that delays in chemotherapy, particularly >10 days, were associated with an increased risk of death for pediatric patients with AML.¹³  In our study, at least 7 patients had delays beyond this 10-day threshold, although the delay occurred during treatment rather than at the beginning of therapy. Furthermore, children treated with initial nonoperative management of NA had significantly greater delays in chemotherapy, as compared with those undergoing upfront appendectomy. Taken together, delays in definitive treatment of NA results in delays and alterations in chemotherapy and may potentially impact cancer outcomes.

Our analysis is a retrospective study and is, therefore, vulnerable to selection bias. Although the sites included represent diverse practices and patient populations from 2 countries, prospectively assessing management strategy and its effect on patient and oncologic outcomes would be ideal. Also, the centers included in the freestanding, tertiary or quaternary children’s hospitals contributing to this study may not be representative of hospitals in which the majority of patients with appendicitis present. Nonetheless, this study includes data from 16 geographically diverse centers in both the United States and Canada.

Patients with upfront appendectomy had lower complication rates than patients with initial nonoperative management. Second, patients who had upfront appendectomy experienced fewer chemotherapy delays or alterations compared with patients who had initial medical management. Third, patients who had initial medical management had a longer LOS, which is associated with increased costs and decreased quality of life, which are particularly relevant in children with cancer.^26,27  Finally, although concern for postoperative complications, such as delayed wound healing, surgical site infections, postoperative abscess formation, and postoperative sepsis, may cause providers to opt for nonoperative management, serious complications and mortality are rare in both treatment groups, suggesting both strategies have an acceptable safety profile.

In pediatric patients with appendicitis and neutropenia secondary to oncologic treatment, upfront appendectomy resulted in lower complication rates, reduced LOS, and fewer alterations in chemotherapy regimens, as compared to upfront nonoperative management.

AML

acute myelogenous leukemia

ANC

absolute neutrophil count

CT

computed tomography

IQR

interquartile range

LOS

length of stay

NA

neutropenic appendicitis