Effectiveness and nephrotoxicity of colistin monotherapy vs. colistin–meropenem combination therapy for multidrug-resistant Gram-negative bacterial infections


Corresponding author and reprint requests: M. E. Falagas, Alfa Institute of Biomedical Sciences (AIBS), 9 Neapoleos Street, 151 23 Marousi, Greece
E-mail: m.falagas@aibs.gr


A retrospective cohort study evaluated the effectiveness and nephrotoxicity of intravenous colistin monotherapy vs. colistin–meropenem combination therapy for patients with multidrug-resistant Gram-negative bacterial infections. Fourteen patients received intravenous colistin monotherapy and 57 received colistin–meropenem. No significant differences were found concerning clinical response of the infection (12/14 (85.7%) vs. 39/57 (68.4%), p 0.32) and development of nephrotoxicity (0/14 (0%) vs. 4/57 (7%), p 0.58). A favourable association was revealed between survival and treatment with colistin monotherapy compared to colistin–meropenem (0/14 (0%) vs. 21/57 (36.8%) deaths, p 0.007), even after adjusting for the variables for which significant differences were found.

Most recent clinical studies with polymyxins have reported data concerning the clinical outcome for patients treated with combinations of polymyxins and various other antimicrobial agents [1–5]. However, data concerning the comparative effectiveness and toxicity of colistin monotherapy vs. colistin–β-lactam combination therapy in patients other than those with cystic fibrosis are lacking.

A retrospective cohort study was conducted at the tertiary-care Henry Dunant hospital in Athens, Greece. The study was approved by the Institutional Review Board of the hospital. Patients with infections caused by multidrug-resistant Gram-negative bacilli who were hospitalised during the period October 2000 to May 2005 and were managed with intravenous colistimethate sodium were identified from the pharmacy electronic databases and included in the study. Two physicians (PIR and SKK) reviewed the patients' records and classified the outcomes. Patients were excluded from further analysis if they had received intravenous colistin therapy for <72 h. This group of patients was compared with patients treated concomitantly with colistin and meropenem for at least 50% of the duration of the treatment, or alternatively for ≥7 days. For patients who received colistin monotherapy or colistin–meropenem combination therapy for more than one episode of infection, only the first episode was included in the analysis.

Definitions of episodes of infections and outcomes, including cure of the infection and nephrotoxicity, as well as the microbiological methods used, have been described previously [3]. Specifically, ‘cure’ was defined as a resolution of presenting signs and symptoms of infection by the end of colistin treatment; ‘improvement’ was defined as a partial resolution of presenting signs and symptoms; and ‘deterioration’ was defined as persistence or worsening of presenting signs and symptoms during colistin administration. ‘Normal renal function’ was defined as a serum creatinine level ≤1.3 mg/dL, with ‘deterioration’ defined as an increase of >50% of the baseline creatinine level to a value >1.3 mg/dL, or as a decline in renal function requiring renal replacement therapy. ‘Baseline creatinine’ was defined as the creatinine level on the initial day of intravenous colistin administration. Intermediate sensitivity was considered as resistance.

The distribution of variables in the two groups was analysed using the chi-square test or Fischer's exact test for categorical variables, and Student's t-test or the Mann–Whitney test for normally and non-normally distributed continuous variables, respectively. Variables associated with mortality in the bivariable analysis (p <0.05) were included in a multiple logistic regression model.

Of 257 patients who received intravenous colistin during the study period, 71 qualified for inclusion in the present analysis. Of these, 14 received colistin monotherapy and 57 received colistin–meropenem combination therapy. The mean daily dose of colistin administered to the monotherapy group was 4.6 ± 2.3 MIU, with a mean duration of administration of 14.2 ± 7.3 days. In the colistin–meropenem group, the mean daily dose of colistin was 5.5 ± 2.2 MIU, with a mean duration of administration of 17.8 ± 11.4 days, and meropenem was administered at a mean daily dose of 4.8 ± 1.6 g for a mean of 15.7 ± 9.6 days.

Table 1 compares the baseline (non-outcome) variables in the two groups. No statistically significant differences were found with respect to demographical and clinical characteristics, except that the urinary tract was the site of infection (p 0.009), and Pseudomonas aeruginosa (p 0.05) was the responsible pathogen more commonly among the colistin monotherapy group than among the colistin–meropenem group.

Table 1.   Comparison of the severity of illness and co-morbidity of patients treated with either colistin monotherapy or colistin–meropenem
n = 14a
n = 57a
  • a

    n refers to the first episode of infection for which patients received colistin monotherapy or the combination of colistin and meropenem.

  • b

    The denominator refers to the number of pathogens isolated from the 57 patients. From one patient, two microorganisms were isolated from a single specimen; thus 58 was the total number of responsible pathogens.

Demographical data
 Age (years)56.2 ± 19.861.5 ± 18.90.43
 Gender (male)8/14 (57.1%)34/57 (59.6%)0.86
 APACHE II score14.3 ± 7.415.4 ± 6.60.56
 Malignancy3/14 (21.4%)10/57 (17.5%)0.71
 Heart dysfunction8/14 (57.1%)30/57 (52.6%)0.76
 Lung dysfunction2/14 (14.3%)18/57 (31.6%)0.20
 Diabetes mellitus3/14 (21.4%)15/57 (26.3%)0.71
 Urogenital disorder4/14 (28.6%)14/57 (24.6%)0.76
 Chronic renal failure0/14 (0%)6/57 (10.5%)0.59
 Hepatic disease0/14 (0%)6/57 (10.5%)0.59
 Haematological disorder2/14 (14.3%)3/57 (5.3%)0.25
 Neurological disorder8/14 (57.1%)18/57 (31.6%)0.08
Previous hospitalisation10/14 (71.4%)37/57 (64.9%)0.64
Previous antibiotic use10/14 (71.4%)33/57 (57.9%)0.35
Previous surgery8/14 (57.1%)34/57 (59.6%)0.86
Duration of hospitalisation
   before the first day of
   colistin treatment (days)
23.3 ± 37.116.5 ± 12.00.69
Admission to the ICU11/14 (78.6%)52/57 (91.2%)0.18
Duration of ICU stay (days)10.5 ± 11.517.2 ± 18.60.69
Mechanical ventilation support7/14 (50%)36/57 (63.2%)0.37
Special treatments10/14 (71.4%)47/57 (82.5%)0.45
 Anti-tumour treatment1/14 (7.1%)4/57 (7%)1
 Steroid treatment5/14 (35.7%)22/57 (38.6%)0.84
 Blood transfusion9/14 (64.3%)40/57 (70.2%)0.75
 Haemodialysis2/14 (14.3%)9/57 (15.8%)1
Urinary catheter14/14 (100%)46/57 (80.7%)0.11
Gastrostomy/colostomy1/14 (7.1%)1/57 (1.8%)0.36
Type of infection
 Pneumonia6/14 (42.9%)23/57 (40.4%)0.86
 Urinary tract infection4/14 (28.6%)3/57 (5.3%)0.009
 Abdominal2/14 (14.3%)5/57 (8.8%)0.62
 Spondylodiscitis1/14 (7.1%)1/57 (1.8%)0.36
 Surgical site infection/skin
 and soft-tissue infection
0/14 (0%)3/57 (5.3%)1
 Bacteremia1/14 (7.14%)16/57 (28.1%)0.16
 Catheter-related0/14 (0%)5/57 (8.8%)0.58
 Empirical0/14 (0%)1/57 (1.8%)1
Pathogen isolatedb
 Acinetobacter baumannii5/17 (29.4%)32/58 (55.2%)0.06
 Pseudomonas aeruginosa9/17 (52.9%)16/58 (27.6%)0.05
 Klebsiella pneumoniae2/17 (11.8%)7/58 (12.1%)1
 Stenotrophomonas maltophilia1/17 (5.9%)0/58 (0.0%)0.23
 Enterobacter cloacae0/17 (0%)1/58 (1.7%)1
 Escherichia coli0/17 (0%)1/58 (1.7%)1
 None0/17 (0%)1/58 (1.7%)1
Antimicrobial susceptibility
 Multidrug-resistant7/17 (41.2%)32/58 (55.2%)0.31
 Susceptible only to polymyxin10/17 (58.8%)26/58 (44.8%)0.41

No statistically significant difference regarding clinical response (cure and improvement) of the infection and occurrence of nephrotoxicity was found between the two groups (12/14 (85.7%) vs. 39/57 (68.4%), p 0.32, and 0/14 (0%) vs. 4/57 (7%), p 0.58, respectively). In-hospital mortality was less common among the colistin monotherapy group than among the colistin–meropenem group (0/14 (0%) vs. 21/57 (36.8%), p 0.007). The favourable association between colistin monotherapy and survival remained even after adjusting for these two variables, for which statistically significant differences were found in the bivariable analyses.

In an era of expanding antimicrobial resistance, the use of colistin for the treatment of infections caused by Gram-negative bacteria has re-emerged. The safety and effectiveness of colistin have been questioned in the past, and its use was restricted mainly to patients with cystic fibrosis [6,7]. The present retrospective comparative study sought to evaluate the safety and effectiveness of colistin monotherapy and colistin–meropenem combinations in the treatment of infections caused by multidrug-resistant Gram-negative bacteria. The main finding of the study was that colistin monotherapy was not inferior to colistin–meropenem therapy with respect to cure of the infection and nephrotoxicity. Indeed, better survival was noted in patients treated with colistin monotherapy, and this association remained even after adjusting for the variables for which statistically significant differences between the compared groups were found (i.e., site of infection and pathogen).

Four patients in the colistin–meropenem group experienced nephrotoxicity and three died. Development of acute renal failure, especially in critically-ill patients, is associated with increased mortality [8,9], and deterioration of renal function of any pathogenesis during intravenous colistin treatment, in addition to an age of >50 years, was found to be associated independently with mortality in a study of 43 patients with Gram-negative bacterial infections [4]. This may partially explain the difference in rates of mortality observed in the present study, although the small sample size meant that a significant association was not detected. In addition, there was a predominance of urinary tract infections among the colistin monotherapy group, compared with more severe infections among the colistin–meropenem group.

It should be noted that combinations of colistin with various antimicrobial agents have been reported to have in-vitro synergic activity against multidrug-resistant Gram-negative bacteria [6,10]. Synergic activity of colistin with ceftazidime against P. aeruginosa has also been reported in an in-vitro study, as well as with rifampicin against P. aeruginosa and Acinetobacter baumannii strains [11,12].

The retrospective nature and the relatively small number of patients treated with colistin monotherapy were the main shortcomings of the present analysis. In addition, there were too few patients to examine the comparative effectiveness and toxicity of colistin in combination with antibiotics other than meropenem, and microbiological synergy studies of colistin with meropenem were not performed.

In conclusion, the effectiveness of colistin monotherapy did not appear to be inferior to that of colistin–meropenem combination therapy for patients with multidrug-resistant bacterial infections. These limited and preliminary data provide support for further studies of colistin monotherapy in this patient population.