Acute kidney injury in children with cancer admitted in an intensive care unit

Acute kidney injury (AKI) is a frequent complication found in children with oncological disease, being associated with chronic kidney disease and increase of death risk.1 Many risk factors are related with AKI in patients with cancer, such as tumor lysis syndrome, vascular obstruction, and malignant cell infiltration in the urogenital system. The use of nephrotoxic drugs, such as antimicrobials, chemotherapy and iodinated contrast also predispose to the development of the problem.2 Besides that, the use of more aggressive therapy for the cancer treatment can result in prolonged immunosuppressive conditions, predisposing to the development of infections, septic shock and, consequently, kidney dysfunction due to organic hypoperfusion.3

Park et al. (2019)4 in their retrospective study with pediatric patients with cancer, observed the presence of AKI in 52.6% of the children in the first year of diagnosis.4 The progressive worsening of the kidney function may require admittance in an intensive care unit (ICU), in consequence of the disease evolution or due to the complications resulting from the treatment. The coexistence of AKI with other organic dysfunctions, such as heart failure or liver failure makes the clinical handling of the child more complex, and it can evolve to anuria, volume overload, and severe metabolic acidosis.5 Asperen et al. (2019)6 identified that AKI with the need of continuous renal replacement therapy (RRT) increases 3.2 times the risk of death when compared to the patients that do not need this treatment.6

Despite the importance of the theme in the context of the pediatric patient with oncologic disease, there are few studies in the medical literature evaluating the interaction of AKI with multiple organ dysfunction syndrome. Besides that, few authors used standardized criteria of AKI, such as the Kidney Disease Improving Global Outcomes (KDIGO) classification system.7 The main objective of this paper is to evaluate the association between kidney failure and death in children with oncological disease admitted in an intensive care unit.

It is a retrospective cohort study involving children with cancer in a pediatric intensive care unit in a quaternary hospital that is a referral in pediatrics, with the capacity of 22 general beds between November of 2016 and January of 2019. The data were extracted from the electronic medical records and the study was approved by the Research Ethics Committee in the institution.

The inclusion criteria were being a carrier of oncologic disease, which were either solid or hematologic, length of stay in the ICU ≥ 24 hours and being between 1 month and 18 years old. The exclusion criteria were chronic nephropathy with estimated creatinine clearance <15ml/min/1,73m³, history of nephrectomy or urogenital disease.

All the patients of the sample were scanned as to the presence of AKI during the admittance in the ICU for a maximum period of 15 days. They were classified by the presence or absence of acute kidney injury. The main outcome was death in the ICU. Other analyzed variables were age, gender, pediatric risk of mortality 2 (PRISM) score, reason for admittance, kind of oncologic disease, history of autologous or allogeneic bone marrow transplant (BMT), use of recent chemotherapy, period between cancer diagnosis and admittance, multiple organ dysfunction syndrome, basal creatinine, stages of AKI, nephrotoxic drug (NTD), use of loop diuretics, volume overload, renal replacement therapy, and length of stay in the ICU.

The population of pediatric ICU admitted patients during the studied period was 810 children, excluding 726 children, resulting in a sample of 84 patients. In a decreasing order of frequency, 18 (21.4%) patients had acute lymphoid leukemia, 14 (16.7%) acute myeloid leukemia, 10 (11.9%) medulloblastoma, 9 (10.7%) glioma, 7 (8.3%) lymphoma, 5 (6%) neuroblastoma, 4 (4.7%) ependymoma, 4 (4.7%) rhabdomyosarcoma, 3 (3.6%) Ewing's sarcoma, 2 (2.4%) craniopharyngioma, 2 (2.4%) choroid plexus tumor, 1 (1.2%) adrenal carcinoma, 1 (1.2%) hepatocarcinoma, 1 (1.2%) osteosarcoma, 1 (1.2%) pleomorphic sarcoma, 1 (1.2%) teratoid rhabdoid tumor and 1 (1.2%) tongue tumor.

The sample showed a discrete male predominance (51.1%), with age medians and death expectation of 6.5 years and 1.9%, respectively. The main reason for the ICU admission was for monitoring, followed by hemodynamic shock, respiratory failure, and reduced level of awareness. Among the patients that were admitted for monitoring, 5 (12.2%) showed a risk of tumor lysis syndrome and 36 (87.8%) were in a post-surgical state. Fifty-three percent of the children had solid tumors, 39% had recently received the cancer diagnosis, 32.1% went through chemotherapy less than 4 weeks before ICU admission, 29.8% had febrile neutropenia, and 11.9% had a history of bone marrow transplant. The median of the ICU length of admittance was 5 days and 23 (27.3%) children died (Table 1).

Among the research subjects, 43 (51.2%) developed AKI, in which most of them were in the first 48 hours of admittance. Of these subjects, 12 (27.9%) were in stage 1, 17 (39.5%) stage 2 and 14 (32.6%) stage 3.

The AKI diagnosis occurred through the urine flow rate in only 7.1% of the sample, being that most of it was made through the serum creatinine level. Besides that, 38% of it developed volume overload, 53.6% used furosemide and 8.3% needed renal replacement therapy. The most used classes of nephrotoxic drugs by the pediatric intensivist were antibiotics (39.3%), antivirals (9.5%) and nonsteroidal anti-inflammatory (4.8%). The variables that had a statistically significant association with the stages of acute kidney injury were bone marrow transplant (p=0.035), admission by hemodynamic shock (p=0.019), PRISM score (p= 0.036), volume overflow (p<0.001) and SDMO (p<0.001). More advanced stages of AKI were associated with RRT (p<0.001), longer ICU admission length (p=0.006) and death (p=0.003) (Table 2).

The survival analysis identified the highest death risk among the patients with acute kidney injury stages 2 or 3 when compared to children without AKI or stage 1 (log-rank test, p=0.045) (Figure 1). The variables that showed a statistically significant association with death were hematologic neoplasm (p=0.009), PRISM score ≥5% (p= 0.008), febrile neutropenia (p=0.026), SDMO (p<0.001), and renal replacement therapy (p=0.006) (Table 3).

After logistic regression, the only independent predictors of mortality were cardiovascular dysfunction (p=0.017) and neurological dysfunction (p=0.023).

It has been observed that all patients with up to 1 organic dysfunction were dismissed from the ICU. On the other hand, 33.3% of the children with 2 or 3 organic dysfunctions and 70.8% of patients with 4 or more organic dysfunctions evolved to death (p<0.001).

Our study identified that 51.2% of the sample developed some degree of acute kidney injury, being that more advanced stages of this complication showed a higher association with death. The incidence of renal dysfunction in severe pediatric patients may vary from 5% to 82% depending on the characteristics of the subjects studied.11 12 The following must be taken into consideration: the underlying disease, the severity of the condition, the comorbidities, and the AKI definition that was used. Thus, children with cancer must be under constant care regarding this problem, as a result of the exposure to many risk factors.5

On the other hand, we noticed that the renal dysfunction does not compromise the body homeostasis of the critical patient in isolation. It is generally associated with other organic dysfunctions that, as a whole, result in a higher risk of death. That becomes clear when we observe that the only independent predictors of mortality in the study were cardiovascular and neurological dysfunctions, which were also found by other authors.13 Dursun et al. (2009)14 and Akhtar et al. (2011)15 identified that sepsis, use of vasoactive drugs, and mechanical ventilation are variables that are most associated with death in this population.14 15

The cardiovascular dysfunction is characterized by the body's inability to supply the tissues with blood in a satisfactory way. This results in anaerobic metabolism, lactic acidosis and multiple organ dysfunction syndrome.10 16 Prolonged and refractive hemodynamic shock increase the number of organic dysfunctions, which results in a higher risk of death. Haase et al. (2011)17 identified a mortality rate of 9%, 36%, 44%, and 76% of the patients with 1, 2, 3, and 4 organic dysfunctions respectively in their study with 226 children in severe stages of cancer. Fiser et al. (2005)18 and Ali et al. (2016)19 also observed a higher rate of organic dysfunctions among children with septic shock that evolved to death (p=0.001). Our study showed similar results, and thus the identification and early treatment of the cardiovascular dysfunction is fundamental for the increase of the patients' survival especially those with a case of febrile neutropenia.3 20 In an associated way, the indication of pediatric ICU admittance must not be postponed, since an elevated PRISM score is also associated with a worsened clinical evolution.13 14

Patients with a history of BMT may show additional risk factors for the development of AKI, such as graft-versus-host disease, thrombotic microangiopathy, immunosuppressive nephrotoxicity, and sinusoidal obstruction syndrome.21 22 Our study identified that 80% of these patients showed some level of AKI. Koh et al. (2018)23 reported that the incidence of renal dysfunction in patients with BMT may vary from 21 to 84%. Kizilbash et al. (2016)24 noticed that AKI occurs early in patients with hematopoietic stem cell transplantation, regardless of whether the source is autologous or allogeneic.

Once the treatment of children with cancer involves the administration of large amounts of volume through the use of antimicrobials, hyperhydration, sedatives, chemotherapy, and transfusions, these patients show a higher risk of volume overload, especially in the presence of AKI with reduced urine flow rate.25 26 Patients with positive cumulative fluid balance show longer mechanical ventilation time and greater need for RRT.27 Raymakers-Janssen et al. (2019)28 observed in their multicentric study with pediatric cancer patients that fluid overload (p=0.003) and a need for inotropic support at the beginning of the continuous renal replacement therapy (p=0.004) are independent predictors of death. Cortina et al. (2019)29 verified that 77.8% of children with oncologic-hematologic disease that need renal replacement therapy in the ICU died (p< 0.001). We found similar results, since 71.4% of the patients that started this procedure evolved to death (p=0.006).

In view of these findings, avoiding advanced stages of AKI and its possible complications seem to be the most plausible strategy. The systematic monitoring of the renal function through the serum creatinine level and the urine flow rate must be the rule, so as to make an early diagnosis.7 Arterial hypotension must not be tolerated, with indication of aggressive measures for the changing of the condition.30 Besides that, adjusting the dose of medications according to the estimated glomerular filtration rate, maintaining a state of normovolemia, and avoiding the use of nephrotoxic drugs when possible is indicated.31

As positive points, our study was performed in a general ICU, reference in pediatrics, with a diverse sample of neoplasms and a research protocol with specific definitions. This allows us to generalize the data to other units. A holistic analysis of the critically ill patients, made it possible to minimize the risk of confusion bias in the association between AKI and death, when taking the multiple organ dysfunction syndrome into consideration. However, some limitations are shown. Firstly, the study is a retrospective study with a sample size that may be considered small when compared to other papers in the literature. Secondly, the iodinated contrast was not included in the list of nephrotoxic medications due to the absence of specification of types and their respective osmolarities. Thirdly, some risk factors such as graft-versus-host, thrombotic microangiopathy, and sinusoidal obstruction syndrome were not evaluated in this study. Finally, it was not possible to assess the dry weight of all the patients in the study, being opted in the use of the weight in the ICU admittance.

Children with oncological disease showed many risk factors of acute kidney injury, and this complication is associated with the increase of mortality rate, especially when associated with multiple organ dysfunction syndrome.