使用Polymyxin合併其他抗生素治療或Polymyxin單一治療Carbapenem-resistant Enterobacteriaceae之結果比較：統合分析與臨床試驗計畫書

Abstract

背景 碳青黴醯類抗生素通常被認為是治療多重抗藥性微生物的最後一道防線，由於碳青黴醯類抗藥性腸道菌的廣泛傳播，以致臨床治療上能使用的藥物有限。在台灣，多黏菌素（polymixin）是針對具碳青黴醯抗藥性腸道菌（carbapenem-resistant Enterobacteriaceae, CRE）感染的首選藥物之一。然而，比較使用多粘菌素合併carbapenem類微生物製劑之合併治療，與單一使用多粘菌素治療 CRE感染的臨床療效之研究，目前仍然有限。因此，本研究以系統性文獻回顧與統合分析方式，探討治療具CRE感染使用多粘菌素合併治療與單一治療之比較。

方法 本研究利用電子資料庫進行主題相關文獻檢索，針對CRE感染，比較使用多粘菌素合併carbapenem類微生物製劑之合併治療，或是單一使用多粘菌素治療之30 天全因死亡率（all-cause mortality）。將符合納入條件的文獻進行資料品質評估，並提取數據整理合併後，使用隨機效應模型估計匯總優勢比（odds ratio, OR）。

結果 根據研究方法，最終選擇了五項文獻共計330名受試者納入統合分析。在30 天之全因死亡率，針對使用多粘菌素的合併治療或是單一使用多粘菌素治療CRE感染，統計上具有顯著差異（pooled OR, 0.53; 95% CI 0.28-1.00, p=0.05）。異質性低（I2=0%）。

結論 本統合分析結果顯示，使用多粘菌素的合併治療或是單一使用多粘菌素治療CRE感染在30天之全因死亡率在統計上具顯著性的差異。然而，本研究結果仍存在許多限制，例如納入的研究族群具有異質性、致病原不同，以及治療策略不同，故依據結果設計了一項具規模性的多中心臨床試驗計畫書。

Background Carbapenems use to be considered as the last-line defense against multidrug-resistant microorganisms. The spread of carbapenem-resistant Enterobacteriaceae (CRE) results in the challenge of limited active drugs. In Taiwan, colistin, which belongs to the polymyxin class, is a prominent medication utilized for treating infections caused by CRE. However, studies to compare to the clinical effectiveness of polymyxin-based combination therapy or polymyxin-based monotherapy on treatment for CRE infection are still rare. Therefore, we conducted this meta-analysis to address the effectiveness compared with polymyxin-based combination therapy for treating CRE infection.

Methods A systematic literature review and meta-analysis were performed by searched electronic publishing to compare the clinical effectiveness of polymyxin-based combination therapy or polymyxin monotherapy on treatment for CRE infection. The primary outcome was 30-day all-cause mortality. Data was retrieved from included studies that met all eligibility criteria and quality assessment. The pooled odds ratio (OR) was estimated using a random-effect model.

Results A total of five studies containing 330 subjects were selected for the meta-analysis. The pooling result shows that combination therapy was significantly better than monotherapy (OR, 0.53; 95% CI 0.28-1.00, p=0.05). Heterogeneity was low (I2=0%).

Conclusion The present study demonstrated that there is significant difference between the combination therapy group and the monotherapy group in 30-day mortality. Our study has several limitations, such us study populations are heterogeneous, the pathogens are different, and the dosing strategies are different. Hence, a well-designed, large-scaled RCT comparing polymyxin-based combination therapy versus polymyxin-based monotherapy is indicated.