Incidence and risk factors of abdominal compartment syndrome in pediatric oncology patients: a prospective cohort study

Abdominal compartment syndrome (ACS) has been the subject of increasing research over the past decade owing to its effects on morbidity and mortality in critically ill patients. This study aimed to determine the incidence and risk factors of ACS in patients in an onco-hematological pediatric intensive care unit in a middle-income country and to analyze patient outcomes. This prospective cohort study was conducted between May 2015 and October 2017. Altogether, 253 patients were admitted to the PICU, and 54 fulfilled the inclusion criteria for intra-abdominal pressure (IAP) measurements. IAP was measured using the intra-bladder indirect technique with a closed system (AbViser AutoValve®, Wolfle Tory Medical Inc., USA) in patients with clinical indications for indwelling bladder catheterization. Definitions from the World Society for ACS were used. The data were entered into a database and analyzed. The median age was 5.79 years, and the median pediatric risk of mortality score was 7.1. The incidence of ACS was 27.7%. Fluid resuscitation was a significant risk factor for ACS in the univariate analysis. The mortality rates in the ACS and non-ACS groups were 46.6% and 17.9%, respectively (P < 0.05). This is the first study of ACS in critically ill children with cancer. Conclusion: The incidence and mortality rates were high, justifying IAP measurement in children with ACS risk factors.


Introduction
Over the last decade, research on abdominal compartment syndrome (ACS) has increased, as it affects several physiological systems and influences the mortality and survival of critically ill patients [1][2][3]. Although the relationship between intra-abdominal pressure (IAP) and respiratory function was first described in 1863, The Abdominal Compartment Society (WSACS) did not publish the first consensus with specific definitions for children until 2013. The main difference between the adult and child definitions was the IAP cutoff point. [2].
Oncology patients may have several risk factors for ACS; however, no study has yet investigated this specific group. Therefore, the main objective of this study was to determine the incidence of ACS in children admitted to onco-hematological pediatric intensive care units; the second objective was to analyze the risk factors and compare the outcomes.

Ethics statements
This study was performed with the principles of the Declaration of Helsinki and approved with free and informed consent by the Ethics Committee for Analysis of Research Projects of the Clinical Direction of Hospital das Clínicas of Universidade de São Paulo. The protocol was preregistered and supported by the São Paulo Research Foundation (FAPESP) (process 2011/50647-5). Communicated

Measurement of intra-abdominal pressure
The pediatric patients underwent indwelling bladder catheterization using a Foley bladder probe. The IAP measurement was performed according to Cheatham and Safcsak's method [4], and an anti-reflux valve (AbViser AutoValve ® , Wolfle Tory Medical Inc., USA) was interposed between the vesical probe and the collecting system. The patient was placed in a supine position at 0º, and the system was placed at the height of the iliac crest and connected to a digital monitor using a transducer. Saline solution (1 ml/ kg) was infused intravesically. The patient's first expiration measurement after the infusion was considered the IAP and was recorded in the annotation plan. A minimum volume of 3 mL of saline for instillation and a maximum of 20 mL were used [5,8]. The IAP measurements were classified according to WSACS rules [2]. The IAP was measured at the time of bladder catheter placement and every 6 hours for the first 10 days. After this period, if the probe was not removed, one measurement per day was recommended (repeated after 6 hours if the IAP was > 10 mmHg) or according to the recommendations of the PICU staff. A urine culture was recommended before and 24 hours after the initial IAP measurement to track previous urinary infections. Urine culture was requested for children with continued bladder catheter placement at regular intervals of three days or as indicated by the PICU staff.

Definitions
Intra-abdominal hypertension (IAH) was defined as a sustained measurement of IAP > 10 mmHg (at least two measurements within a maximum interval of 6 hours) [2]. ACS is defined as a sustained measurement of IAP > 10 mmHg and any new organ dysfunction that could be explained by IAH [2]. Organ dysfunction was assessed using the Pediatric Logistic Organ Dysfunction II score [9]. Sepsis is defined as the presence of a systemic inflammatory response syndrome associated with a suspected or proven infection [10].

Data collection
Data collection was conducted during three periods (May to December 2015, April to June 2016, and January to October 2017) using a questionnaire administered by two trained researchers. Variables were entered into an Excel database (Microsoft Corp., Albuquerque, NM, USA).
The seven risk factors that were analyzed were as follows: 1) abdominal surgery or any procedure that required an abdominal wall incision; 2) abdominal infection or radiological or laboratory evidence of abdominal cavity infection; 3) acidemia, arterial pH < 7.20; 4) coagulopathy, platelet count < 55.000/mm 3 or international normalized ratio > 1.5, or activated partial thromboplastin time > 2 times the reference value; 5) fluid resuscitation or infusion of > 40 mL/kg of crystalloid or colloid weight; 6) invasive or noninvasive mechanical ventilation; and 7) presence on palpation or confirmation by existing examination of an abdominal mass.
Five outcomes were investigated as follows: 1) length of stay in the PICU, 2) days of mechanical ventilation, 3) use or absence of vasopressors or inotropic drugs, 4) use or absence of fluid removal therapy, and 5) PICU mortality.

Statistical analysis
Convenience sampling was also conducted. Incidence was calculated by dividing the number of new ACS cases by the total number of samples. Continuous data with a normal distribution were reported as the means and standard deviations and compared between the two groups (those with and without ACS) using the t-test. Continuous variables without a normal distribution were expressed as median, maximum, and minimum values and compared between the groups using the Mann-Whitney U test. Categorical variables were reported as frequencies and were compared using the likelihood ratio test. To identify the independent risk factors for ACS, we included variables that showed statistically significant differences in the univariate analysis in a logistic regression model, and the results were expressed as odds ratios and 95% confidence intervals [6]. Statistical analyses were performed using the Statistical Package for Social Sciences (Version 25.0; IBM Corp., Armonk, NY, USA). Statistical significance was set at P < 0.05.

Patient demographics
During the three study periods, 253 patients were admitted to the PICU with a median age of 7.79 years (Tables 1 and 2). Patients were primarily male, eutrophic, and had malignant onco-hematological diseases (85%), with leukemia in 34.8%, and 52 (20.6%) had undergone bone marrow transplantations. Sepsis/septic shock was the main cause of PICU admission, and the mortality rate in the general population was 14.6%. The average length of PICU stay was 12.32 days (range, 1-95 days; median, 6 days). The median pediatric risk of mortality (PRISM) IV score was 5 (1.8%) and ranged from 0 to 24 (0.1-61.2%). The mean Pediatric Index of Mortality (PIM) 3 score was 0.059 (range 0.001-0.742). Patient body mass index ranged from 10.26 42.26 kg/m 2 (median: 17.07 kg/m 2 ).
Sixty-four (25.2%) of the patients admitted had previously been admitted with an indwelling bladder catheter, and eight of them had the catheter changed to the study material ® (AbViser AutoValve ® , Wolfle Tory Medical Inc., USA) between the second and sixth days after admission. Fortyeight of the remaining 189 patients had a bladder catheter placed after admission, and the rate of indwelling bladder catheterization was 41% (112 patients). Two patients were excluded because of hematuria. Ultimately, 54 patients fulfilled the inclusion criteria for IAP measurement (Fig. 1).
Patients included in the sample were primarily male with a median age of 3.49 months (12 days to 17.92 months). Eighty-seven patients had malignant onco-hematologic diseases, and 40% had leukemia, myeloproliferative disease, or myelodysplasia. Ten patients underwent bone marrow transplants. The main causes of PICU admission were sepsis, septic shock, discomfort or respiratory failure, and the risk of tumor lysis. The mortality rate of the patients was 29.6%. The median PRISM IV score was 7 (1.8%), with minimum to maximum values ranging from 0 to 18 (0.1-31.5%). The median PIM 3 score was 0.059 (minimum to maximum values: 0.001-0.403). The patients' body mass index ranged from 11.06 to 31.11 kg/m 2 (median: 16.44 kg/m 2 ), and the median length of stay in the PICU for these patients was 20.5 days (minimum to maximum values: 5 to 89). No statistically significant differences were found comparing the demographic characteristics of the groups with and without ACS with the sample, as shown in Tables 1 and 2.
Thirty-one patients (57.4%) from the sample had IAH, with 11 presenting IAH on admission and 20 developing the condition during the first week of their PICU stay. Eleven patients died (35.4%), seven of whom developed ACS before death. Patients with an abdominal mass developed IAH (32.3%) more frequently than those without an abdominal mass (8.7%) (P = 0.039).

Abdominal compartment syndrome
The incidence of ACS was 27.7%. Patients with ACS had a higher risk of mortality at admission, as demonstrated by the high median PRISM IV score (8 vs 6) and higher mortality (46.6% vs 19.9%, P = 0.031).

Risk factors
We examined the risk factors for ACS (Table 3), and nearly all were found more frequently in the ACS group than in the non-ACS group, except for the use of mechanical ventilation. Fluid resuscitation was a significant risk factor for ACS (P = 0.02; (OR] = 4.4; 95% CI: 1.18-16.37).

Outcomes
Mortality was 2.6 times higher in the ACS group than in the non-ACS group (46.6% vs 17.9%, P = 0.031). Patients with ACS required more fluid removal therapy and inotropic drugs than those in the non-ACS group (60% vs 20.5%, P = 0.031 and 46.7% vs 41%, P = 0.707, respectively), and they had longer durations of mechanical ventilation and PICU stay (Table 4).

Side effects
Five patients developed urinary fungal infections, three were treated with antifungal agents, and two were in end-of-life care. None of the patients had serious outcomes secondary to these infections (Supplemental 1).

Discussion
This was the first prospective cohort study conducted in a population of critically ill children with onco-hematological diseases following the WSACS 2013 Guideline concerning pediatric-specific recommendations. The first articles on the incidence of IAH/ACS in pediatric patients were published in 1993. The incidence ranged from 0.6% to 34% depending on the classification criteria used (IAP measurements of 10, 12, or 15 mmHg), and the study population differed between the studies, with some patients being treated clinically and others surgically [7,[11][12][13][14][15]. In 2006, there was consensus among experts during the Second World Congress on ACS to unify the definitions and pathophysiology of IAH and ACS, and updates were published in 2013 with specific data for pediatric patients [2].
Studies with new classification criteria specific to the pediatric population reported an IAH incidence of 12.6% and a prevalence of 43.9% [16]. Thabet et al. conducted a study in a tertiary and multidisciplinary PICU and reported IAH and ACS incidences of 12,6, and 4%, respectively, which were much lower than those reported in our study (57.4% and 17.7%) [13].
To the best of our knowledge, there have been no published data regarding the incidence of IAH or ACS in children with onco-hematological diseases. The higher incidence of ACS and IAH shown in our study could be explained by the specific features of patients with oncohematological diseases related to direct or indirect chemotherapy or other immunosuppressive therapy effects. This group of patients also had a greater presence of risk factors such as abdominal surgery, abdominal infection, acidemia, coagulopathy, fluid resuscitation, mechanical ventilation, and abdominal masses. Previous studies on IAH or ACS in onco-hematology patients were only case reports or isolated cases in multidisciplinary samples. In a descriptive, observational, and retrospective study of 23 children who underwent decompressive laparotomy, they found a case of ACS in a child with Wilms' tumor [17]. Another descriptive, observational, and prospective study of 14 patients with ACS described a case of a 7-year-old boy who survived until his PICU stay and had abdominal Burkitt's lymphoma [7].
Additionally, Egyed et al. reported a case of an adult with abdominal Burkitt's lymphoma who developed ACS at the time of diagnosis and survived after decompressive laparotomy and chemotherapy [18]. Lode et al. reported a case of a 10-month-old girl who developed abdominal compression with renal failure, severe bleeding, and tumor lysis syndrome. The diagnosis of abdominal compartment syndrome (ACS) was established by intragastric pressure (max. 17 cmH2O). Based on the deteriorating clinical condition, the sustained rise of the intraabdominal pressure (IAP) together with the newly developed organ dysfunction, it was decided to perform an enterostomy to decompress the abdominal cavity. Surgical decompression by enterostomy, local and systemic bleeding control with platelets and coagulation factors, and anti-infective and TLS therapy were effective in stabilizing the patient's condition. This allowed initiation of multimodal antineoplastic treatment according to protocol NB 2004 [19].
ACS and IAH are thought to be caused by a variety of pathophysiological mechanisms. A more modern theory suggests that both often result from a two-step process that begins with reperfusion of an ischemic injury and results in a vicious cycle known as "intestinal stress syndrome" or "intestinal permeability syndrome." The resuscitation phases of different types of shock can cause intestinal reperfusion injuries after ischemia that induce systemic and intestinal inflammatory responses. Pro-inflammatory mediators increase mesenteric and intestinal wall capillary permeability, leading to fluid leakage between the mesentery and intestinal wall, bacterial translocation, and absorption of bacterial endotoxins [20].
The second step is a result of bowel wall edema that increases the IAP, generates compression of the intra-abdominal lymphatic system, and decreases the flow of lymph out of the abdominal cavity [21]. The increase in IAP progressively decreases blood flow to the intestinal mucosa, generating ischemia and further increasing intestinal permeability. Thus, there is an influx of proinflammatory mediators into the systemic circulation, which facilitates the worsening of visceral edema and an increase in IAP following the vicious cycle of acute intestinal stress [22]. Bulky fluid resuscitation aggravates this vicious cycle by generating hemodilution, decreasing the oncotic pressure of the intestinal mucosa, and favoring an increase in hydrostatic pressure, further increasing IAP [20]. The increase in pressure in the abdominal cavity is transmitted to the interstitial space and microvascularization, generating a decrease in abdominal perfusion pressure and, consequently, in blood flow to the intracavitary organs, resulting in ischemia, congestion, and edema of the intra-abdominal organs. Visceral edema contributes to an increase in IAP, contributing to the vicious cycle described above and leading to progressive organ dysfunction. When the IAP exceeds the abdominal perfusion pressure, blood flow to the organs is interrupted, resulting in cell death. Therefore, ischemia and necrosis are the causes and consequences of ACS [23].
The effects of IAP on the respiratory system have also been investigated. The diaphragm elevation and consequently a reduction in lung volume was observed, with a significant increase in the degree of atelectasis in the posterior lung regions [24].
In the cardiovascular system, the effects of an increased IAP are multifactorial and occur via three processes: 1) vascular intra-abdominal compression with venous stasis in the inferior vena cava, generating a reduction in preload; 2) elevation of the diaphragm leading to cardiac compression and an increase in intrathoracic pressure, which consequently generates a decrease in cardiac contractility, aggravation of the decrease in preload, and a reduction in biventricular diastolic volume; and 3) intra-abdominal organic compression generating activation of the renin-angiotensin-aldosterone system, thereby promoting an increase in the afterload. These three effects generate a decrease in coronary perfusion secondary to a reduction in cardiac output, increasing myocardial damage and the need for vasoactive drugs [3].
The most significant effect of IAP on the renal system is related to blood flow. There was a decrease in the glomerular filtration rate and consequent activation of the renin-angiotensinaldosterone system. The attempt to correct cardiac output with fluids did not result in the reversal of renal dysfunction, but the reversal of acute ACS did [25].
Moreover, Citerio et al. demonstrated that the increase in IAP was also transmitted to the intracranial cavity [26].
The present study has some limitations. It had a singlecenter design, and we could not perform sample calculations. In addition, we studied only 54 patients, which may have limited our ability to detect any associations. Finally, this study was observational with selection bias and confounding factors. Future studies should include a larger sample size and investigate whether early intervention can prevent mortality in this group of patients.
The high incidence, serious consequences caused by IAH/ ACS, and high mortality rate substantiate the importance of monitoring IAP in children with onco-hematological diseases who are admitted to the PICU and have risk factors for IAH/ACS.