Colistin use in pediatric intensive care unit for severe nosocomial infections: experience of an university hospital

Colistin, an antibiotic with a bactericidal effect against gram-negative bacteria, was introduced for use in the 1950s. However, an abundance of side effects and the availability of safer alternatives have limited its use [1, 2]. Multi-drug resistance (MDR) is defined as the resistance of microorganisms to at least three different groups of antibiotics with intrinsic activity against gram-negative bacteria [3]. Nosocomial infections caused by MDR microorganisms have been increasing with frequency and today are a common problem in all areas of the world [4–6]. The problem of resistance to existing drugs, and the slowing rate at which new drugs are being invented in this pharmacologic area, have caused reconsideration of colistin for use in this context [4, 7–9]. Studies on the efficacy and safety of colistin therapy in children are rare [4, 8, 10–12]. This study investigated the efficacy and safety of colistin therapy in children with severe nosocomial infections in pediatric intensive care unit (PICU).

The study was performed at a PICU of 200-bed university children hospital. Forty-one nosocomial infection episodes that were treated with intravenous colistin and occurred in thirty-five children between March 2008 and March 2013 were analyzed retrospectively. Neonatal patients and patients who received colistin treatment for less than 4 hours were excluded from the study. Information recorded included clinical and demographic characteristics, medical history, presence of medical devices such as central catheters, urinary catheters, and endotracheal or tracheostomy tubes, concomitant drug use especially can effect neurological or renal systems, clinical and laboratory findings, type and antimicrobial susceptibility of the isolated microorganism, type of infection that required colistin treatment (bloodstream infection, ventilator-associated pneumonia, etc.), dose, administration route, duration and side effects of colistin, results of antimicrobial therapy, and prognosis. The levels of blood urea nitrogen (BUN), creatinine, C-reactive protein (CRP) and the counts of leukocytes and platelets were recorded (at the beginning and during treatment). Clinical and laboratory data were obtained from patients’ cards and from the hospital computer database system. Pediatric risk of mortality (PRISM) and pediatric logic organ dysfunction (PELOD) scores were used to determine the severity of illness and organ dysfunction respectively. The medical records of each patient were investigated also for signs of neurotoxicity (neuromuscular blockade, seizures, change in level of consciousness, etc.) during treatment.

The diagnosis of infection was made by clinical and laboratory findings and through positive culture. The diagnosis of nosocomial infection and specific infections such as catheter-related infection, ventilator-associated pneumonia (VAP) was made on the basis of the criteria of the Centers for Disease Control and Prevention [13]; sepsis was defined according to the definition of the International Pediatric Sepsis Consensus Conference [14].

Some of the important terms used in this paper were defined as follows: Multi-drug resistance was defined as the resistance of microorganisms to at least three different groups of antibiotics with intrinsic activity against gram-negative bacteria (e.g., beta-lactams, aminoglycosides, carbapenems, and quinolones) [3]. Empirical treatment was defined as the beginning of colistin treatment based on surveillance data, or on inadequate response of the infection to other antimicrobials [10]. Response to treatment was defined as improved clinical and laboratory findings of index infection, and eradication of the pathogen microorganism isolated by initial culture. If the patient was alive at the end of treatment, prognosis was evaluated as a survey. Patients who died during treatment were also assessed in order to find out if the death had been caused by infection. Nephrotoxicity was defined in patients with normal baseline renal functions as serum creatinine levels higher than >1.1 mg/dl, or by a >50% reduction in creatinine clearance or by the need for renal replacement treatment (RRT) at any time during treatment.

Descriptive statistics were used in the data analysis. The study was approved by the Ondokuz Mayis Universitiy Human Studies Ethics committee. (No:2011/2062)

In 5 patients, intravenous colistin therapy was started empirically because of fever that was unresponsive to broad-spectrum antimicrobial treatment and progressive sepsis. In other patients, indication of colistin treatment was culture-proven infection. The most frequently isolated microorganism, the most common isolation site and type of infections were A. baumannii, tracheal aspirate, and VAP respectively.

Only three of thirty-four episode used concomitant nephrotoxic agent with colistin developed renal impairment. All three of these patients received at least one nephrotoxic agent such as aminoglycosides, amphotericin B or a glycopeptides together with colistin. In patient 8, peritoneal dialysis was started after 13 days of colistin treatment because of oliguric renal failure and remained for 14 days. Gentamicin treatment was discontinued and doses of colistin were adjusted according to the creatinin clearance in this patient. Level of creatinine was 5.6 mg/dl at the end of 22 days of colistin treatment and returned to normal value 18 days after the end of treatment. Acute renal failure developed after eight days of treatment in patient 12. Colistin treatment was discontinued and bloodstream infection caused by P. aeruginosa was treated successfully with meropenem in this patient despite in-vitro resistance. In patient 15 who had ataxia telengiectasia, non-oliguric renal failure developed on the fourth day of colistin treatment due to severe sepsis and septic shock. Colistin was continued in this patient because of severe sepsis and immune deficiency. The patient died on the eighth day of treatment before RRT was performed. Possible cause of renal failure in this patient was refractory shock and multiorgan dysfunction syndrome. We did not find neurological side-effects (change of consciousness, seizures, visual disturbances, or neuromuscular blockade) in any patients. However, most of the patients were comatose or were receiving sedative-analgesic agent.

In twenty-eight (68.3%) episodes, positive clinical response was achieved. Twelve patients died while receiving colistin therapy. Six of these patients died because of infection-related reasons, and others died of primary underlying disease. Renal impairment was developed in one patient who died due to infection-related reasons; however, RRT was not performed due to sufficient urine output. Positive clinical and microbiological results were obtained in other patients (Table 3).

Nosocomial infections caused by multidrug-resistant gram-negative microorganisms are an important cause of morbidity and mortality in intensive care units [15, 16]. We previously reported the results of a multicenter study investigating the efficacy and safety of colistin treatment in PICU [12]. In this study we aimed to present our experience about the use of colistin in severe nosocomial infections in our PICU during a longer period of time and with newly emerged patients.

A few studies have investigated the efficacy and safety of colistin therapy in children thus far, and the majority of existing studies have been performed in specific patient groups such as newborns and patients with cystic fibrosis or burn injuries [4, 11, 17]. No special patient group was singled out in our study, differentiating it from previous studies. The results of present study have suggests that colistin may be a viable option in the treatment of serious pediatric nosocomial infections caused by MDR gram-negative bacteria.

The optimum pediatric dosing for colistin is not clear due to lack of the pharmacokinetic and pharmacodynamic studies in this age group. However, an intravenous dose of 2.5-5 mg/kg/day has been recommended [2, 4, 11, 18]. In our study we found the average dose of 4.9 mg/kg/day divided into 2 or 3 equal doses effective and safe.

The major side effects of treatment with colistin are neurotoxicity and nephrotoxicity [4, 8, 11, 12, 19]. Recent pediatric studies have reported a nephrotoxicity rate of 0–14.3% [4, 8, 10–12, 18]. This rate is lower than that given in previously reported results in adults (5.8-26.8%) [20–22]. The main mechanism of renal injury is acute tubular necrosis. Renal injury usually occurs in the early stages of treatment and characterized by a decrease in creatinine clearance and increase in serum urea and creatinine levels [2, 17]. In our study renal toxicity occurred at the median 8th day of treatment (relatively late). Less commonly, hematuria, proteinuria, and cylindruria may also be seen [17]. Similar to our study sepsis, hypotension and the use of other nephrotoxic drugs due to nosocomial infections may potentiate the nephrotoxicty of colistin [4, 10, 12, 17]. Many studies have reported that the use of other nephrotoxic drugs such as aminoglycosides together with colistin facilitates the deterioration of renal function [4, 10, 12, 17, 23–26]. In our study colistin was used with other nephrotoxic agents such as aminoglycoside, vancomicyn or amphotericin B in twenty-nine of forty-one episodes of nosocomial infection and only three patients developed nephrotoxicity. Similarly, Jajoo and colleagues [17] reported that nephrotoxicity occured in only one of six newborns after the administration of netilmicin. Interestingly, Li and colleagues [25] reported that colistin is less nephrotoxic than amikacin or tobramycin. Both adult and pediatric studies have reported that the nephrotoxicity of colistin is reversible and renal function returns to normal with discontinuation of treatment [2, 8, 19]. In our study treatment with gentamicin was halted in one patient because of nephrotoxicity, but colistin therapy was continued due to life-threatening sepsis. With peritoneal dialysis, renal function returned to normal levels 18 days after the end of colistin treatment in this patient. In other patient, colistin treatment was changed with meropenem and renal function returned to normal levels within days. In our unit, none of the patiens used colistin developed nefhrotoxicity in the last two years. None of the patient used colistin in the last two years developed nephrotoxicity. We attribute this to the increased sensitivity of our team regarding renal toxicity, to closer monitoring of fluid balance and renal functions in colistin users, and to the avoidance of the use of concomitant nephrotoxic agents.

There are very few reported neurological side effects of colsitin in pediatric studies [8, 11, 12, 17], and neurological side effects were not observed in our study. The challenge of detecting neurological side effects in patients with underlying neurological disease may be the reason for this. Other possible reasons include the fact that drugs frequently used in the intensive care unit (such as sedative/analgesic drugs, or neuromuscular blocking agents) can mask neurological side effects and neurological side effects may also be overlooked in retrospective studies. Finally, the use of sodium colistimethate (which is less toxic than colistin) in our study may be another reason for the lack of neurological side effects [8, 16].

Favorable clinical response rates have reportedly been higher in pediatric studies than in adult studies (76–86.5% versus 57-73%, respectively) [8–10, 20, 27]. A favorable outcome occurred in 68.3% courses and was consistent with the previous studies. In addition, positive clinical and microbiological response was achieved in 6 out of 7 episodes developed in 4 patients with primary immune deficiency.

The synergistic effect of concomitant use of colistin with rifampicin, carbapenems, piperacillin/tazobactam, or ciprofloxacin has been reported in previous studies [24, 25, 28–32]. Similar to previous studies, we used colistin in various combinations with other antimicrobial agents. In this study, colistin was combined with an antibiotic against on gram-negative bacteria in almost all cases. We believe that combination regimen can contribute to the success rate of treatment.

Prognosis in the case of a bacterial infection caused by MDR is affected by the presence of underlying disease, sepsis, or organ failure. Of the 35 patients in our study, 30 had underlying chronic diseases and 4 had immunodeficiency. A mortality rate of 2-28% has been reported in previous studies [8, 11, 12, 17]. Our results are similar to previous reports. While 6 out of the 12 patients who died during the study period died of infection-related reasons, the others died because of their primary diseases such as multiple trauma, malignancy and congenital heart diseases.

The major limitations of this study include the retrospective design and the insufficient number of patients which is too limited to determine the true incidence of drug-related side effects. Additionally, patients had taken many drugs before and during treatment with colistin. Therefore, it was not possible to associate all the effects and side effects that occurred during the study with colistin alone.

In conclusion, the results of this study have shown that colistin is effective in the treatment of serious nosocomial infections caused by MDR gram-negative bacteria in children. Renal function test must be closely monitored during the treatment period. We suggest avoiding the use of other nephrotoxic drugs with colistin to minimize the nephrotoxicity. However, more prospective controlled trials are needed to objectively evaluate the efficacy and side effects of this treatment.