Discussion
In this retrospective study we found that pediatric patients with an active oncologic diagnosis who underwent appendectomy did not have higher odds of suffering post-operative complications including superficial or deep SSI’s, superficial or deep wound dehiscence, deep organ space infections or suture line dehiscence as compared to patients without cancer. The literature on post-operative morbidity in cancer patients, particularly in pediatrics, is sparse. Indeed, most have focused specifically on patients with neutropenia, rather than cancer, although there is an obvious and nearly 1:1 overlap in these groups among available studies on appendicitis.
Some studies have suggested an elevated risk of surgical complications in pediatric neutropenic patients and by extrapolation led to inference that appendicitis should be treated non-operatively if possible1,2. Pudela et al. reported a case series of 3 neutropenic patients, all with underlying oncologic diagnoses and each treated differently – one with antibiotics alone, one with interval appendectomy after up-front antibiotic therapy, and one with urgent operation – concluding that therapy should be tailored, but implying that medical management may be the prudent course1. Wiegering et al.’s study of 5 neutropenic oncology patients from one institution reported symptom resolution after antibiotic therapy for appendicitis in all patients, with 6-month to 3-year-followup2. One patient underwent prophylactic appendectomy prior to bone marrow transplantation. The acceptance of these studies is buoyed, in part, by the ample literature demonstrating that non-operative management of appendicitis with antibiotic therapy alone can – in the short term – have successful outcomes that approach those of surgical appendectomy3,4.
Other studies evaluating neutropenic pediatric patients with appendicitis have come to opposite conclusions, demonstrating relative safety of surgery in the face of neutropenia, and thus have implied that surgical treatment is preferable to medical treatment of appendicitis5,6,7. Mortellaro et al. included 11 neutropenic patients from 2 institutions, all with an oncologic diagnosis5. All underwent appendectomy and had hospital lengths of stay comparable to the groups’ previously published outcomes data on appendectomy in non-neutropenic patients. Although the paper did not report on any complications in the group of 11, the authors concluded that appendectomy in neutropenic patients was“tolerated well with a low risk of surgical complications.”Scarpa et al.’s show heterogeneous treatment strategies among 30 neutropenic children with appendicitis (90% of whom had a hematological malignancy) from 8 French institutions6. Among the 30, 6 underwent immediate operation, 17 underwent delayed operation after initial antibiotic therapy, while the remaining 7 were treated with antibiotics alone. Wound complications were reported in 2 patients but it was unclear to which surgical group they belonged.
All of the available studies were conducted as single institution or small multi-center data collections, and individually included no more than 30 patients. They also were very inconsistent and vague in reporting surgical complications.
The literature on overall surgical intervention in neutropenic patients, pediatric or adult, is quite sparse. Jolissaint et al. reported on nearly 250 neutropenic adult patients from a single institution undergoing abdominal surgery, finding that ANC < 500 and emergency surgery carried a relatively higher risk of morbidity and mortality compared to less severe neutrophilia and/or elective surgery8. Badgwell et al. concluded in a single-institution study of 60 neutropenic adult patients undergoing urgent abdominal surgery that delay until severe neutropenia can be improved – usually by administration of colony-stimulating factors – may have beneficial effects9. Grant et al., examining 4389 adult patients undergoing chemotherapy for cancer in the NSQIP database, showed that leukopenia itself was not an independent risk factor for perioperative morbidity or mortality10. The authors did not, however, compare these results to the non-cancer population.
Further muddying the waters is a 2015 report by Gulack et al. using the NSQIP adult database to examine the association of leukopenia and perioperative complications in patients undergoing exclusively emergency surgery11. Among the 20443 included patients undergoing emergency laparotomy, leukopenia was independently associated with increased odds of both post-operative morbidity and mortality.
Our study was strengthened by higher overall numbers than other pediatric reports (95 patients in total with a cancer diagnosis), a comparison cohort group of patients without cancer that allowed for risk stratification and multivariable logistic regression to identify covariates independently significant for increase odds of developing a wound complication, as well as the unique feature of the NSQIP database’s reporting structure that actively seeks out evidence of 30-day complications by a trained analyst. In addition to looking for 30-day complications, NSQIP reports 30-day hospital readmissions, which can also be an important metric to examine. Although quite a high proportion of the 95 patients with a cancer diagnosis were readmitted within 30 days of operation, only a scant 5% of readmissions were related to the surgery itself (compared to the 85% surgery-related rate among readmissions in the non-cancer cohort); this again suggests that operating on appendicitis in children with a synchronous oncologic diagnosis may be safe and prudent.
Our study has several limitations that are worth noting. There are those relevant to any database study, including its retrospective nature, risk of coding and reporting error, and missing data points. The last was addressed by a sensitivity analysis that did not change the overall outcomes based on missing WBC and ASA Class data points. In addition, NSQIP-P reports only total WBC counts but does not include neutrophil counts, so we were unable to include neutropenia specifically as a risk factor. The flag for cancer diagnosis includes both “active cancer diagnosis” and “undergoing chemotherapy” so we were unable to categorize specifically treatment vs overall diagnosis as possible risk factors. Specific type of cancer diagnosis was also not uniformly available, further limiting our ability to stratify our findings.
In conclusion, pediatric patients undergoing treatment for cancer do not have increased odds of suffering post-operative wound complications following appendectomy compared to the general population. These findings refute beliefs that surgical risks are prohibitively high in patients with appendicitis who have a synchronous cancer diagnosis and support surgical appendectomy as the appropriate first-line treatment. Further studies may help define any subgroups that are particularly at risk for surgical complications – such as patients with markedly low ANCs, those with specific types of cancer – although the feasibility of these studies is questionable given the small patients numbers involved.