Severe pneumonia is a serious condition of lower respiratory infection. In addition to common symptom of pneumonia like fever and cough, patients with severe pneumonia also have symptoms of respiratory failure and obvious involvement of other systems [9]. According to different environments obtained, it can be divided into severe community-acquired pneumonia (SCAP) and severe hospital-acquired pneumonia (SHAP). SCAP development rapidly. Even with sufficient anti-infection treatment, its mortality can be as high as 28.8% [10–12]. And some research reported that its mortality can even be 50% [13]. The mortality of SHAP is even higher, as the prevalence of drug-resistant bacteria is becoming increasingly severe [4].
Since 20th century, the increasing bacterial resistance rate has become an important threat to human health, especially CRE, CRAB, and CRPA. These strains have high resistance to common antibiotics and limited options for therapeutic drugs. Polymyxin was recommended by both domestic and international guidelines as an important therapeutic drug for carbapenem-resistant Gram-negative bacterial infections [1, 2]. Polymyxin belongs to the class of peptide antibiotics, clinically including polymyxin B sulfate, colistin sulfate, and colistin methane-sulfonate (CMS). CMS is a precursor drug that is converted into the active ingredient colistin after entering the human body to exert its effects [14]. Both polymyxin B sulfate and colistin sulfate directly enter the bloodstream in an active form and exerted similarly high bactericidal activity when administered topically [15, 16]. While colistin sulfate showed less nephrotoxicity, little has been reported on the bactericidal effect of colistin sulfate in patients with severe pneumonia [16, 17]. After the relaunch of colistin sulfate in May 2019 in China, we organized this study to estimate the anti-bacterial effect and safety of colistin sulfate in severe pneumonia patients.
At present, there is lack of high-quality, large-sample clinical research for severe pneumonia patients, especially the one with MDR bacterial infections. Some studies have shown that HAP is the direct cause of death in critically ill patients, with related mortality ranging from 15.5% − 38.2% [18, 19]. The clinical survey results of 13 large teaching hospitals in China show that the average all-cause mortality of HAP is 22.3%, of which VAP is 34.5% [20]. Another survey in China showed that among 17,358 ICU inpatients in 46 hospitals, the mortality of VAP in mechanically ventilated patients was 21.2% -43.2% [20, 21]. And it had been proved that if the pathogen is MDR or pan-drug resistant (PXD), the attributed mortality can reach as high as 38.8% − 60% [21, 22]. In our retrospective cohort study, we can see that the 28-day mortality after discharge of severe pneumonia patients with suspected MDR bacterial infections included in our cohort study is also high (42/84, 50%), which is consistent with the results of other relative studies.
We divided the patients recruited in our study into two groups based on different colistin sulfate treatment schemes. The basic clinical characteristics of patients in the two groups were comparable, no matter with their age, gender, complication, and detection of pathogens (Table 1). Further study showed clinical outcomes including CT image improvement and hospitalization days, were also comparable between the two groups, except the 28-day mortality after discharge. For patients in TC group, the 28-day mortality is significantly lower than that of the OC group (18/46, 39.1% vs. 24/38, 63.2%, P = 0.048). At present, there is still lack of clinical research on the efficacy of colistin and tigecycline in the treatment of CRE pneumonia, and the results of some existing observational clinical studies are inconsistent [23–24]. A meta-analysis which included 14 studies involving 1163 patients, compared the efficacy and safety of polymyxin E and tigecycline monotherapy or combination therapy for multidrug-resistant/extensively-resistant Gram-negative bacterial infections. The monotherapy with tigecycline significantly reduced the 30-day mortality (OR = 0.35, 95% CI 0.16–0.75, P = 0.007), while the monotherapy with polymyxin E significantly reduced the in-hospital mortality rate (OR = 2.27, 95% CI 1.24–4.16, P = 0.008) [23]. While combined the use of tigecycline and polymyxin E, there is no significant difference in clinical efficacy and mortality. Compared with polymyxin E, tigecycline has a lower incidence of renal injury, whether it is monotherapy or combination therapy. Kimberly et al. retrospectively compared the efficacy of polymyxin E and tigecycline in the treatment of CRAB and/or CRE infections. And the results showed that receiving polymyxin E alone, patients treated with the combination of colistin E and tigecycline have a higher mortality (33/90 37% vs. 0/16 0%, P = 0.002) and a longer hospital stay [29.5 (37.7 ± 30.3) vs 23.3 (23.5 ± 14.0), P = 0.004] than those treated with tigecycline alone [24]. Researchers analyze that it may be due to the higher severity of acute disease in patients receiving polymyxin E treatment, and a significant delay in initiating effective of antibacterial treatment (P < 0.001). Chang et al. conducted a multicenter retrospective study comparing the clinical efficacy of polymyxin B combined with tigecycline, and polymyxin B or tigecycline combined with other drugs in the treatment of HAP caused by CRO [25]. The results showed that the 28-day all-cause mortality of the combination regimen based on polymyxin B was 28.3% (28/99), significantly lower than 39.3% (68/173) of the combination regimen based on tigecycline or 48.9% (45/92) of the polymyxin B combined with tigecycline group (P = 0.014), which is quite consisted with the results in our study. Researchers analyze the reasons and believe that tigecycline has certain limitations in treating pulmonary infections, partly due to the limitations of tigecycline is an antibacterial agent rather than a fungicide. On the other hand, tigecycline has a steady-state distribution volume of about 7–10 L/kg and is widely distributed in human tissues, while the drug concentration in plasma is relatively low [25]. However, it should be noted that Chang et al. did not mention the dosage of tigecycline administered to the included patients in their study [25].
Further we found that sputum culture results were significantly associated with the 28-day mortality after discharge both through the univariate (OR 0.375, 95% CI 0.155–0.910, P = 0.030) and multivariate analysis (OR 0.180, 95%CI 0.040–0.813, P = 0.026). The positive sputum culture results were especially associated with the 28-day mortality after discharge of severe pneumonia patients (univariate analysis: OR 0.233, 95% CI 0.074–0.739, P = 0.013; multivariate analysis: OR 0.073, 95% CI 0.006–0.882, P = 0.040). This suggests the important role of pathogen detection in guiding antibiotic selection, which is consistent with previous research findings [23–25].
Finally, we estimated the safety of colistin sulfate using in patients with severe pneumonia. And found that there were no significant differences before and after the use of colistin sulfate for the white blood count (WBC), neutrophil count, lymphocyte count, platelet count, level of CRP, PCT, ALT, AST, and creatinine, so it will not increase the burden on the liver, kidneys, and blood system. In fact, colistin sulfate exert its effect directly after entering the body, mainly eliminated through non-renal or non-hepatic pathways [26].
To summary, in this study, we confirmed the effectiveness and safety of colistin therapy through a retrospective study of severe pneumonia patients who were considered to have CRE infection in our hospital. Of course, some lacking existed. One thing is that this study is a single center study with a small sample size, which may have significant bias. In addition, prospective studies with large samples and multiple centers are needed for further validation.
Table Legend
Table 1. Clinical characteristics of patients with different colistin schemes.
Table 2. Univariate and multivariate analysis of different treatment regimens and 28-day disease-related mortality after discharge.
Table 3. Comparison of blood biochemical indicators before and after the use of colistin.