Predicting acute renal failure in Bothrops snakebite patients in a tertiary reference center, Western Brazilian Amazon

Acute Kidney Injury (AKI) is the main systemic complication and cause of death in viperid envenomation. Although there are hypotheses for the development of AKI, the mechanisms involved are still not established. The aim of this study was to evaluate the clinical-laboratorial-epidemiological factors associated with AKI in victims of Bothrops sp envenomation. This is an observational study carried out at the Fundação de Medicina Tropical Dr. Heitor Vieira Dourado. AKI was defined according to the guidelines of the Acute Kidney Injury Network (AKIN). Among the 186 patients evaluated, AKI was observed in 24 (12.9%) after 48 hours of admission. Stage I was present in 17 (70.8%) patients, II in 3 (12.5%) and III in 4 (16.7%). Epidemiological characterization showed predominance of men, occurrence in rural areas, aged between 16–60 years, feet as the most affected anatomical region, and time to medical assistance less than 3 hours. Hypertension and diabetes were the comorbidities identified. Most of the accidents were classified as moderate, and clinical manifestations included severe pain, mild edema, local bleeding and headache. Laboratory results showed blood uncoagulability, hypofibrinogenemia, leukocytosis, increase of creatine kinase, and high lactate dehydrogenase levels. Multivariate analysis showed an association with high LDH levels [AOR = 1.01 (95% CI = 1.01–1.01, p<0.002)], local bleeding [AOR = 0.13 (95%CI = 0.027–0.59, p<0.009)], and the presence of comorbidities [AOR = 60.96 (95%CI = 9.69–383.30; p<0.000)]. Herein, laboratory markers such as high LDH levels along with local bleeding and comorbidities may aid in the diagnosis of AKI.

Introduction which can exemplify AKI. Besides, existing publications on the region only describe the frequency of the event, with no clinical studies especially on AKI [5,26,50]. Although studies point to hypotheses for the development of AKI, its mechanisms have not yet been fully established. Thus, the description of the renal injury pathogenesis based on clinical characteristics presented by patients could contribute to the clarification of the development of acute renal injury induced by envenomation. The aim of this study was to evaluate the clinical-laboratorial-epidemiological factors associated with AKI in victims of envenomation by Bothrops snakes.

Study design
This is an observational study carried out from August 2014 to August 2016 at the Fundação de Medicina Tropical Dr. Heitor Vieira Dourado (FMT-HVD), a reference hospital for snakebite treatment in Manaus, capital city of Amazonas, Brazil, and the only institution in this city offering antivenom therapy. This study was approved by the Ethics Review Board of FMT-HVD (approval n˚602.907-0/2013). All participants signed a consent form after the explanation of the study aims. Under 18 years of age, the consent was signed by parents or guardians. The individual in this manuscript has given written informed consent (as outlined in PLOS consent form) to publish these case details.

Sample size calculation
Sample calculation was based on two studies conducted in the Amazon region, with an expected frequency of AKI after envenomation of 12% [5,51], at an 80% power and 5% of significance level. The minimum of 174 individuals was necessary adjusting by 15% of lost to follow-up.

Case selection
All patients included in the study were hospitalized due to the snakebite. Eligible patients were those diagnosed with clinical signs of envenomation by Bothrops genus. Pregnant women or individuals that underwent previous antivenom therapy in another hospital unit were not included in this study. According to the Brazilian Ministry of Health Guideline [12], envenomation by Bothrops species is classified as follows: I) mild envenomation, characterized by changes in clotting time or not, pain and edema involving one to two segments of the affected limb; II) moderate envenomation, with changes in clotting time or not, pain, evident edema involving three or more segments of the affected limb and local or systemic bleeding without hemodynamic repercussion; III) severe envenomation, that may present changes in clotting time, severe pain, severe hardened edema in the limb, severe hemorrhagic conditions with hemodynamic repercussion, compartment syndrome, and AKI. In addition to the signs of envenomation, patients should have at least two serum creatinine assessments, where the first creatinine test was performed prior to the antivenom therapy.

Case definition of AKI
At least two measurements of creatinine levels were obtained from each patient during the study period. Acute kidney injury was defined according to the Acute Kidney Injury Network (AKIN) Guideline [52], as follows: Stage I, defined by an increase of >0.3 mg/dL or up to 199% of baseline creatinine levels; Stage II, defined by an increase of 200-299%; and Stage III, an increase in more than 300% in baseline creatinine or serum creatinine levels higher than 4.0 mg/dL with an abrupt elevation of at least 0.5 mg/dL. All patients diagnosed with AKI were referred to the institution's nephrologist. Those classified as stage III AKI underwent renal replacement therapy.

Clinical and laboratorial parameters
Data collection included sociodemographic features such as sex, age (in years), anatomical region of bite, area of occurrence (rural or urban), work-related bite, time to medical assistance, clinical and laboratory characteristics (at the moment of the antivenom therapy-D0, first day-D1, second day-D2, third day-D3 and seventh day-D7 of hospitalization), including vital signs, severity classification (mild, moderate or severe), local and systemic manifestations, clinical outcome (discharge or death) and presence of comorbiditites declared by the patient.

Statistical analysis
A database was built using Epi Info 3.5.1™ (Center for Disease Control and Prevention, Atlanta, USA) forms fed by two independent investigators. Data analysis was carried out using the statistical package STATA v13 (Stata Corp, College Station, USA). Explanatory variables were grouped into hierarchical blocks [44,54]. The proximal block was composed of laboratory markers at admission, intermediate block by clinical findings at admission, and distal block by demographic and epidemiological variables. Univariate regression analysis was performed for each block individually. Variables with significance level of p<0.2 were included in the multivariate analysis per block. All variables with significance level of p<0.05 in the multivariate analysis per block were further included in the overall model (all blocks together). Crude Odds Ratio (OR) and Adjusted Odds Ratio (AOR) with their respective confidence intervals (CIs) were calculated for each hierarchical level and for the global model. The precision of the final model was assessed through Hosmer-Lemeshow goodness-of-fit test [30,45]. The significance level was considered for p<0.05.

Patients' characterization
A total of 345 patients were assessed for eligibility in the study period. From the total, 187 met the inclusion criteria, with one patient lost to follow-up due to hospital evasion (Fig 1). Enzyme immunoassay confirmed the diagnosis of Bothrops sp. envenomation in all patients with complete follow-up.
Patients were mostly male (82.3%), from rural areas (87.1%) and aged between 16-60 years (82.8%). The main anatomical region bitten was foot (66.1%). Most bites were not related to work (59.7%). Time to medical assistance was mostly less than 3 hours (57.5%), the means this time was 4.5 hours and most accidents were moderate (49.0%). Mild and moderate cases were similarly distributed in patients with AKI. A tourniquet on the affected limb was made in 24.7%. No sociodemographic association was found with AKI (Table 1). No patient declared the usage of nonsteroidal anti-inflammatory drugs upon admission. Early adverse reactions to antivenom were observed in 28 patients (16.5%). No late adverse reactions were detected.
Laboratory results revealed that patients with AKI had a slight thrombocytopenia and leukocytosis, increase in creatine kinase, as well as increase in lactate dehydrogenase. Clotting time was uncoagulable in 97.3%. The mean concentration of circulating venom was 44.42 ng/ mL (Table 3). Among these parameters, high LDH levels (p<0.001) were associated with the development of AKI. The first blood evaluation occurred in the same time that the patient arrived in the hospital for assistance before administration of the antivenom.
Clinical outcomes and risk factors for AKI. Of the 186 patients in the study, 24 (12.9%) developed AKI after 48 hours of admission, and 162 (87.1%) did not develop such clinical condition. Stage I was observed in 17 (70.8%) patients, stage II in 3 (12.5%), and stage III in 4 (16.7%) patients (Fig 1). All patients who had a diagnosis of AKI were referred to the institution's nephrologist. Those classified as stage III underwent dialysis. No patients had Chronic  (Table 4).

Discussion
AKI is the main complication observed in snakebite envenomation [38]. It is estimated that between 1.6 to 38.5% of the victims develop AKI [36]. This wide variation in frequency is due to unclear establishment of the diagnostic criteria. In this study, AKI was observed in 12.9% of cases, which is similar to previous studies in the Brazilian Amazon by Feitosa et al. and Souza et al. [5,50]. The development of AKI is an unestablished event [45] and, given that, there are proposals for some triggering events of hemodynamic, nephrotoxic and immunological nature [38]. As for circulatory abnormalities, hemorrhagic conditions, hypo or hypertension, vascular changes, intravascular hemolysis and disseminated intravascular coagulation (DIC) are described in the literature [29,43,45].
Vascular disorders occur through hemorrhagic and coagulant action induced by venom components of Bothrops snakes that promote endothelial injury by hemorrhagins [13] and consumption of coagulation factors [55]. In South America, this disorder has been caused by snakes of the Viperidae family [31,32]. In this study, changes were caused specifically by Bothrops snakes, also described in other papers on the Amazon region [5,6,49,56]. In human envenomation, it is has been shown that experimental models with venoms from B. jararaca, B. moojeni and B. alternatus, induced changes in tubular, glomerular and interstitial structures [39,41,57] indicating the action of venom toxins on the renal parenchyma occurs within a few hours [32]. However, the involvement of immune response in the AKI development still requires more experimental data. Its harmful effect may be caused by deposition of immunocomplexes, IgM and C3 molecules in mesangial areas characterizing glomerulonephritis.
Changes in the renal parenchyma reflect in the clinical presentation, defined and classified according to the International Society of Nephrology's guideline of clinical practice (KDIGO) [52]. Diagnosis is based on serum creatinine and/or urinary volume and classified into three stages. In this study, stage I was presented by most patients (70.8%), while stage II occurred in 12.5% and stage III, which is indicative of renal replacement therapy (RRT), was present in 16.7% of patients. Due to hospital routine, it was not possible to obtain urinary volume measurement of all patients, which is a limitation of the study. The recommended treatment in these conditions is complementary, and it is necessary to adopt specific initial measures for envenomation and later for complications, supportive measures for hemodynamic disorders and specific actions for AKI [36].
Early and adequate administration of antivenom is crucial. Coagulation factors return to baseline levels between 6 to 24 hours after antivenom therapy [12,21,58]. This fact minimizes venom-induced actions, of circulatory and immunological order, that are related to renal disorder. The management of AKI ranges from the use of antidiuretics to increase urinary flow [59] to renal replacement therapy (RRT) such as peritoneal dialysis [26] and hemodialysis. In India, about 13.5% of the cases developed AKI, of which 48.3% needed dialysis [29]. It is important to note that the most common snakes that cause accidents in India are Daboia russellii and Echis carinatus that present venoms with hemotoxic and myotoxic actions [30,44].
In this study, there was no significant association between AKI and epidemiological aspects. These were similar to previous studies where it was seen a greater predominance of male individuals, living in rural areas, in productive age group, and having the feet as the main    anatomical region reached [7,8,26]. This fact may be explained by the greater exposure to leisure and occupational activities [7].
Regarding clinical aspects, local inflammatory events such as pain and edema were the most commonly identified [5,21,26], venom toxins from Bothrops snakes present proteolytic action through inflammatory mediators that act by activating nociceptors and promoting changes in lymphatic vessels [16].
The majority of patients presented blood uncoagulability (97.3%). The pathophysiological mechanism of hemostatic disorders relies on the consumption of coagulation factors and procoagulant activation, as well as thrombin-like enzymes [60]. According to the guidelines of the Brazilian Ministry of Health, clotting time is not a diagnostic criterion for the definition of envenomation [12]. Studies indicate 23 to 43.1% of patients who are victims of accidents in the Amazon region present blood uncoagulability, varying from case to case [5,7,26]. In addition, thrombocytopenia is a result of platelet aggregation inhibition by serine proteases [60]. Another parameter related to hemostatic changes induced by the venom was the presence of hemorrhagic conditions, also showed by Athappan et al. [29].
In this study, most patients who developed AKI had comorbidities namely hypertension and diabetes. It is known that such comorbidities lead to increased risk for the development of this condition [51], as a result of long-term damage to target organs such as heart and kidneys. Also, the difference in serum lactate dehydrogenase levels among patients with AKI was significant. This is a marker of cell injury as shown by Castro et al. in experimental models of B. jararaca-induced tubular injury [39].
In the multivariate analysis, changes in lactate dehydrogenase levels followed by local bleeding and presence of comorbidities, were independently associated with the onset of AKI. This data was shown by Aye et al. where it was observed that the administration of antivenom more than 2 hours from bite, leukocytosis, DIVC, rhabdomyolysis, hyponatremia and microscopic hematuria were factors associated with the clinical picture of AKI [61]. In a retrospective study, independent factors associated with AKI at admission were the presence of hemorrhagic changes and long hospital stay [32]. Thrombocytopenia and liver enzyme release were shown to be statistically significant in those events according to Naqvi et al. [62]. In India, a prospective observational study found that patients who took longer to medical assistance and presented hypotension, albuminuria, anemia, changes in bleeding time, prothrombin time and total bilirubin serum levels, developed AKI [63].
This study had limitations. First, it may underestimate the real prevalence of AKI since only snakebites that took no longer than 24 hours to hospital admission were included in this study. Besides, it was not possible to perform urine volume measurement and calculate creatinine clearance during the study. Imaging tests would also be a good tool to complement diagnosis although not used in this study. For logistical reasons, it was not possible to follow-up participants at the end of the study either personally or by telephone, to identify the chronicity of the case.

Conclusion
Acute renal failure was an important systemic complication observed in patients envenomed by Bothrops snakes. This study may aid in understanding some of the toxin-induced changes in the human body. The associated risk factors for AKI were high lactate dehydrogenase levels followed by local bleeding and the presence of comorbidities, majorly hypertension, suggesting that these laboratory and clinical aspects may aid in the diagnosis, management, clinical evolution and possibly reduce the frequency of these events in those victims. Further prospective studies are needed in order to elucidate the pathogenesis of AKI and to identify the different factors involved.