嗜麥芽寡養單胞菌菌血症回溯性研究:預後因子和抗生素治療結果

Abstract

背景： 近年來嗜麥芽寡養單胞菌菌血症（Stenotrophomonas maltophilia bacteremia）在世界各地發生的頻率逐漸增加，抗藥性也逐漸上升，再加上其所造成的高死亡率，使得尋找更好的抗生素治療更顯得格外重要。 目的： 研究臺大醫院造成嗜麥芽寡養單胞菌菌血症的危險因子，死亡率和預後因子，並分析不同的抗生素療法，包括單一或合併抗生素對治療結果的影響，以期能找出降低感染及感染後有效治療的方法。 研究設計、地點和對象： 本研究於國立台灣大學醫學院附設醫院，一家位於台灣北部的教學醫院，以病歷回顧進行單醫學中心回溯性研究，蒐集在2009年7月1日到2010年12月31日期間發生S. maltophilia 菌血症成年病人之資料進行分析，但若同一病人發生許多次S. maltophilia菌血症，只納入第一次事件。 研究方法： 從紙本或電子病歷蒐集病人基本資料、潛在疾病/合併症、菌血症前30天的各種感染和抗生素使用，以及各種可能造成感染的危險因子，包括菌血症前各種管路 （catheters）、侵入性處置或手術、免疫低下、長期住院等，並且紀錄S. maltophilia血液培養、敏感性試驗、菌血症後30天內的感染，此外還記錄S. maltophilia菌血症發作時的臨床症狀及併發症、相關檢驗數據、抗生素治療，發作後D7、D14、D30、出院時的治療結果。主要觀察終點為D14的死亡率。 統計方法包括Fisher’s exact test、χ2 test、T-test、Mann-Whitney U test，並且以單變項和多變項迴歸分析相關危險因子及D14、D30的預後因子。以Kaplan-Meier method繪製存活曲線，並用Log-rank test比較。 研究結果： 有102個感染S. maltophilia 菌血症之成年病人最後被納入分析中，其中52人屬於單一菌血症，其餘的50人是多重菌種感染，兩組平均年齡都約為60歲，男女比為61:41，院內感染的病人佔了大多數（87.3%），主要發生在普通病房（52.9%） 和加護病房（35.3%），感染源大多不明（58.8%），有26.5%的病人是由呼吸道轉移而來的菌血症，再來才是來自導管（14.7%）。Charlson’s comorbidity score中位數為4，合併症/潛在疾病中以血液疾病（80.4%）佔最大宗，其次是惡性腫瘤（71.6%），再來是心血管疾病（55.9%）和其他疾病。感染S. maltophilia菌血症的可能危險因子分析中有56.9%的人住院超過14天，有54.9%屬於免疫低下，在過去30天有92.2%的人使用過抗生素，有33.3%接受過手術，至於在發作前3天內有89.2%使用侵入性導管，使用呼吸器的有24.5%，氣切的有8.8%，使用全靜脈營養的則有19.6%。菌血症發作時病人Pitt bacteremia score中位數為2分，APACHE II score中位數為21分，發生併發症的有39.6%，以敗血性休克最多。 抗生素敏感性試驗結果發現aminoglycosides、carbapenems和monobactam幾乎都是呈現100%抗藥性，敏感性較好的抗生素主要包括co-trimoxazole、ceftazidime、ticarcillin/clavulanate、fluoroquinolones、minocycline。至於經驗性抗生素的適當性與使用適當抗生素的不同延遲時間都並未發現和病人的D14、D30死亡率有相關，也並未找出較好的單一或合併抗生素可以有較低的死亡率。 死亡率方面，感染S. maltophilia菌血症的全部病人在D14有35%的死亡率，在D30則高達47%，影響D14死亡的獨立危險因子是敗血性休克（OR=34.67, 95% CI= 6.79-176.91, p<0.0001），而與D30死亡相關的獨立危險因子是腎臟疾病（OR=13.71, 95% CI=1.31-143.44, p=0.0288）。 結論： 本研究中感染S. maltophilia菌血症的全部病人在D14有35%的死亡率，而在D30則有47%的死亡率。經驗性抗生素的適當性與使用適當抗生素的不同延遲時間都並未發現和D14、D30死亡率有顯著相關，本研究中也並未發現使用後有顯著較低的D14和D30死亡率的確切性抗生素。與D14和D30死亡相關的獨立危險因子分別是敗血性休克和腎臟疾病。

Background The frequency of Stenotrophomonas maltophilia bacteremia gradually increased in recent years in the world. Besides, the increasing resistance rate of Stenotrophomonas maltophilia to many antibiotics and the considerable mortality all contribute to the need of better antibiotic treatments. Objectives The objectives of this study are to investigate the risk factors, mortality, prognostic factors of Stenotrophomonas maltophilia bacteremia in National Taiwan University Hospital, and to analyze the treatment outcomes with different antibiotic regimens, including comparison of the treatment outcome with monotherapy versus combination therapy. With those data, we hope we can find a way to reduce the occurrence of S.maltophilia bacteremia and find the best treatment regimens. Study design and study population This is a single center retrospective study performed through review of medical charts of patients diagnosed and treated at National Taiwan University Hospital (NTUH), a teaching hospital in northern Taiwan. All adult patients who developed S. maltophilia bacteremia between July 1st , 2009 and December 31st , 2010 were included. If patients had multiple episodes of S. maltophilia during the study period, only the first episode was included in this study. Methods Data were collected from paper or electronic medical charts, including patient profiles, underlying diseases/comorbidities, any infection and antibiotic used within 30 days prior to S. maltophilia bacteremia, and risk factors of S. maltophilia bacteremia, such as catheters, invasive procedures, surgery, immunocompromised status, prolonged hospitalization, etc. Besides, blood culture, sensitivity test, infections within 30 days after S. maltophilia bacteremia onset, clinical presentation, complication, laboratory data, antibiotic treatment, and treatment outcomes on D7, D14, D30, Day of discharge were recorded. The primary endpoint is the D14 mortality. Statistical methods included Fisher’s exact test, χ2 test, T-test, Mann-Whitney U test. Risk factors and prognostic factors attributing to D14, D30 mortality were analyzed by univariate and multivariate logistic regression. Survival curves were drawn by Kaplan-Meier method and compared by Log-rank test. Results One hundred and two patients with S. maltophilia bacteremia were included in the analysis. Among these, 52 were monomicrobial bacteremia and the other 50 were polymicrobial. The mean age was 60.1 years old, and the ratio of male to female was 61:41. Most patients were nosocomial infection (87.3%), and 52.9% of bacteremia onset occurred in general wards, 35.3% in intensive care units. Source of infection was unknown in 58.8% of the patients, and 26.5% was from respiratory tracts and 14.7% was from catheters. The median Charlson’s comorbidity score was 4, and the most common underlying diseases/comorbidity was hematological diseases (80.4%), and then malignancy (71.6%), cardiovascular diseases (55.9%). In the analysis of risk factors associated with S. maltophilia bacteremia, 56.9% patients were prolonged hospitalization (>14 days), 54.9% immunocompromised, 92.2% using antibiotics within 30 days before onset, 33.3% receiving surgery, 89.2% with invasive catheters within 3 days before onset, 24.5% with ventilators, 8.8% tracheostomy, 19.6% with total parenteral nutrition. When bacteremia onset, the median Pitt bacteremia score was 2, and the median APACHE II score was 21. Complications occurred in 39.6% of the patients. The most common complication was septic shock. S. maltophilia was almost 100% resistant to aminoglycosides, carbapenems and monobactam in the susceptibility tests. Better sensitivity appeared with co-trimoxazole, ceftazidime, ticarcillin/clavulanate, fluoroquinolones and minocycline. As to the appropriateness of empirical antibiotic and different delaying times to appropriate antibiotics, there was no significant difference for D14 and D30 mortality. Besides, there was no significantly better monotherapy or combination therapy found to be associated with lower mortality. As to the mortality analysis, the D14 mortality rate of all 102 patients was 35%, and the D30 mortality was 47%. Independent risk factors associated with D14 mortality was septic shock (OR=34.67, 95% CI= 6.79-176.91, p<0.0001) and with D30 mortality was renal diseases (OR=13.71, 95% CI=1.31-143.44, p=0.0288). Conclusions In this study, the D14 and D30 mortality rate in total patients were 35% and 47%, respectively. The appropriateness of empirical antibiotics and different delaying times to appropriate antibiotics were not found to significantly affect D14 and D30 mortality. There was no better definitive antibiotic found to be associated with lower D14 and D30 mortality. The independent risk factors associated with D14 and D30 mortality were septic shock and renal diseases, respectively.