分枝桿菌感染： 第一部份：膿瘍分枝桿菌肺部疾病之抗生素處方及臨床效果 第二部份：Rifabutin之療劑監測

Mycobacterial Infections： Part I：Antibiotic Regimens and Clinical Outcomes of Mycobacterium abscessus complex Lung Diseases Part II：Therapeutic Drug Monitoring of Rifabutin

龎艾昕(Ai-Hsin Pang)

指導教授 ： 林淑文

國立臺灣大學/醫學院/臨床藥學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202104202

摘要

分枝桿菌可大致分為結核分枝桿菌和非結核分枝桿菌兩類，分別會造成結核病和非結核分枝桿菌感染。近幾年由於公共衛生政策對結核病的重視和執行，結核病的發生率和盛行率正逐漸下降。然而相反地，非結核分枝桿菌的感染，特別是肺部疾病，卻有逐漸上升的趨勢。其中，膿瘍分枝桿菌是最難治療的非結核分枝桿菌，因其對於多種抗生素本身即有抗藥性。因此，許多研究者正在努力尋找新的藥物，甚至是將舊有的結核病藥物重新再測試是否能對抗該菌。Rifabutin即為其中一種。它在體外試驗、動物實驗上都顯示對膿瘍分枝桿菌有效果。但由於使用rifabutin容易產生副作用，因此學者們也建議對rifabutin進行療劑監測。
在本論文中，第一部分將會呈現和討論目前治療膿瘍分枝桿菌肺部疾病的抗生素處方和臨床效果。接著在第二部分則會討論執行rifabutin療劑監測的結果。
第一部分：膿瘍分枝桿菌肺部疾病之抗生素處方及臨床效果
背景
膿瘍分枝桿菌為一群生長快速的非結核分枝桿菌，可能感染免疫低下的病人或者是免疫相對正常但有特定危險因子的病人。最常見的感染部位為肺部。膿瘍分枝桿菌肺部感染的治療需要在初始治療（initial phase）時使用至少三種抗生素，其中包含至少一種靜脈注射藥物，而接著在持續治療（continuation phase）時使用至少兩種口服（含吸入劑型）抗生素。由於膿瘍分枝桿菌本身對多種抗生素有抗藥性，在長期治療下也可能誘導出抗藥性基因，因此其治療效果大部分不盡理想。
雖然目前已有治療指引提出治療膿瘍分枝桿菌肺部感染時應使用靜脈注射藥物及多種抗生素併用，然而臺灣醫師在治療膿瘍分枝桿菌肺部感染時是否遵循治療指引建議仍未知，其相對應的治療效果也應同時評估。
研究目的
本研究目的為了解臺灣醫師在治療膿瘍分枝桿菌感染時的抗生素處方，並找到處方和治療成功率及副作用發生率的相關性。
研究方法
本研究為一回溯性觀察性研究，研究地點為臺灣的六間醫院，並納入在2006年到2020年期間被診斷且接受膿瘍分枝桿菌肺部感染治療的病人，若治療期間小於14天，則會被排除。病人的基本特性、微生物學及影像學資料、抗生素處方則由各間醫院的醫師和研究助理從病歷中擷取。我們使用了微調的微生物治癒定義：若病人有三套陰性檢體，或是在治療期間僅有一套陰性檢體，或是無法取得檢體但臨床症狀改善，則都定義為微生物學上的治癒。副作用也同時被記錄下來。
我們使用卡方檢定、費雪檢定和Mann-Whitney U test來比較兩組間的差異，並用單變項及多變項迴歸分析評估達到微調後微生物治癒的預測因子。同時使用ROC curve評估最終的模型，並利用Youden index來找到最終模型中各因子的最佳切點。
研究結果
本研究最終納入89位病人進行分析。其中29位（32.6%）為男性，診斷時的年齡中位數為61歲。34位病人（38.2%）在療程中有使用靜脈注射抗生素，而其中啟用靜脈注射抗生素的中位數為1天（1-49天），且24位（70.6%）在治療開始後28天內使用靜脈注射抗生素。接近全部（96.6%）的病人皆有使用巨環類抗生素（macrolides）。整體而言，平均治療期間為465.26 （± 498.5）天。共有42位（47.2%）達到微調後的微生物治癒。而32位（36%）病人在治療期間發生副作用。
我們發現早期使用靜脈注射抗生素有較高的勝算達到微調後的微生物治癒（aOR 8.577, 95%CI: 2.309-31.859），影像學分數和治療期間的勝算比則較低（aOR分別為0.827, 95%CI: 0.697-0.981和0.996, 95%CI: 0.996-0.999）。而啟用靜脈注射藥物的最佳切點為25天。
結論
早期使用靜脈注射抗生素可能較容易達到好的治療效果，而其切點為25天。病人的初始影像學分數較低及早期使用靜脈注射抗生素或許可以預測治癒的機率。未來若有更大型的前瞻性研究，將可以更確立治癒率和病人特性、微生物學及影像學證據、處方等之間的關聯性。
第二部分：Rifabutin之療劑監測
背景
Rifabutin為一rifamycin類抗生素，目前可以用在治療結核病及非結核分枝桿菌感染上。雖然rifabutin與rifampin相比，為一較弱的CYP3A4誘導劑，但仍然會和其他CYP3A4的受質、抑制劑和誘導劑產生交互作用，進而影響rifabutin的血中濃度。
過高的rifabutin血中濃度常和嗜中性白血球低下、白血球低下、血小板低下等血液方面的不良反應，甚至和葡萄膜炎相關。因此在過去關於結核病病人的研究中，建議將目標峰濃度訂在0.45-0.9 μg/mL。然而，rifabutin的療劑監測大多應用在HIV病人上，在其他族群的病人仍較少，故仍需更多的研究。
研究目的
本研究目的為監測臨床上病人服用rifabutin後的血中最高濃度，並評估濃度、副作用和臨床療效的相關性。
研究方法
本研究為一前瞻性的觀察性研究，研究地點為台大醫院，包括門診和住院病人，研究期間為2021年2月1日至2021年7月31日。20歲以上的病人在服用完rifabutin後七天可納入研究，並抽取服藥後3小時和6小時的血液檢體，使用HPLC進行分析。病人的基本資料和檢驗數據擷取自病歷，而副作用和臨床療效則會在整個療程中持續監測到療程結束後七天或直到2021年7月31日。藥物不良反應會藉由Naranjo scale來評估和rifabutin的相關性。兩組間的差異會藉由費雪檢定、卡方檢定和Mann-Whitney U test來比較，使用rifabutin前後的檢驗數值變化則會用Wilcoxon signed-rank test來評估。並利用簡單和複線性迴歸來尋找可能影響rifabutin C3的因子，最後使用羅吉斯迴歸來尋找副作用發生的預測因子。而rifabutin濃度的最佳切點則是使用自助抽樣法重複抽樣2000次來估計。
研究結果
在研究期間，共有13位病人加入研究，總共進行了21次的療劑監測。在這13位病人中，總共有13個C3、5個C6和2個C0濃度。所有濃度皆小於0.5 μg/mL，其中有7個濃度低於定量極限。在其他可定量的11個濃度中，平均C3和C6濃度分別為0.39 ± 0.07 μg/mL和0.28 ± 0.03 μg/mL。Rifabutin的中位數使用天數為65天（49-133天）。
複線性迴歸結果顯示體重、AST和C3有正相關。使用rifabutin前後的白血球中位數從5.9 k/μL（IQR：4.8-10.4）下降到4.6 k/μL （IQR：3.7-7.0）(p <0.05)。而利用自助抽樣法（bootstrap method）重複抽取的濃度來尋找和血液不良反應相關的C3濃度切點時，白血球下降的C3為0.395 µg/mL (95%CI: 0.382-0.429)。然而在單變項及多變項羅吉斯迴歸中，C3和其他因子都未發現和白血球低下、嗜中性白血球低下、血小板低下或其他副作用有相關性。
結論
我們開發並確效了rifabutin的血中濃度分析方法。此研究收納的病人群rifabutin濃度皆偏低。然而，此研究受限於樣本數小，且侷限於一間醫學中心。未來持續納入新使用rifabutin的病人，對於尋找影響濃度的因子及副作用發生的預測因子將更有幫助。
總結第一部分及第二部分
總結來說，由於非結核分枝桿菌感染盛行率越來越高，更複雜且個人化的治療在所難免。早期使用靜脈注射抗生素來治療膿瘍分枝桿菌肺部疾病可能會有較好的治療效果，而臨床上是否需要進行rifabutin的療劑監測仍需要更多的研究來證實。

關鍵字

膿瘍分枝桿菌肺部感染 ； 非結核分枝桿菌 ； 靜脈注射抗生素 ； 早期使用 ； 臨床效果 ； Rifabutin ； 結核菌 ； 療劑監測 ； 藥物不良反應

並列摘要

Mycobacteria can generally be divided into two major family of bacteria, Mycobacterium tuberculosis and nontuberculous mycobacteria (NTM), which cause tuberculosis (TB) and NTM infections, respectively. As the prevalence and incidence of TB are decreasing due to public health awareness and control, those of NTM infections, especially NTM lung diseases, are increasing gradually. Among various kinds of NTM, Mycobacteria abscessus complex is the most difficult to treat because of its natural resistance against many antibiotics, and therefore, researchers are striving to develop new drugs or repurpose old TB drugs to treat the infections. Rifabutin is one of the repurposed drugs that was found active against M. abscessus complex in in vitro studies and animal models. However, it is also associated with a lot of adverse effects, so many researchers have recommended therapeutic drug monitoring of rifabutin.
In the first part of my thesis, antibiotic regimens and clinical outcomes of M. abscessus complex lung diseases in Taiwan will be presented and discussed. Then, in the second part, results of therapeutic drug monitoring of rifabutin will be discussed.
Part I：Antibiotic Regimens and Clinical Outcomes of Mycobacterium abscessus complex Lung Diseases
Background
Mycobacterium abscessus complex (MABC) is a group of rapidly growing mycobacteria that are known to cause infections in both immunocompromised and relatively immunocompetent hosts with certain risk factors. The lung is the most frequent site of infection. Treatments of Mycobacterium abscessus complex lung diseases (MABC-LD) require a combination of at least 3 drugs including at least 1 intravenous agent for a minimum of 1 month as the initial phase of therapy and then a combination of at least 2 oral/inhaled agents as the continuation phase of therapy. The clinical outcomes of MABC-LD are usually poor because MABC possesses different kinds of intrinsic resistance genes against many antibiotics and can acquire resistance during treatment.
Although guidelines have recommendations regarding the use of IV and multiple antibiotics, Taiwanese physicians’ adherence to treatment guidelines of MABC-LD is not known and the clinical outcomes of treatment of MABC-LD should be assessed as well.
Objectives
The objective of this research was to identify the current use of antibiotics in Taiwanese physicians when treating MABC-LD and to determine the association between therapies and clinical outcomes including treatment success and development of adverse effects.
Methods
A retrospective observational study was conducted in 6 hospitals in Taiwan. Patients diagnosed with and received treatment of MABC-LD during 2006 and 2020 were included, whereas patients whose treatment duration was shorter than 14 days were excluded. Patient characteristics, microbiologic and radiologic data, and treatment regimens were abstracted from medical records by physicians and research assistants in respective hospitals. Modified microbiological cure was used as a treatment outcome comprising 3 consecutive negative culture results, 1 or 2 negative culture results if further culture results could not be obtained, and no culture results, but the patient was considered clinical cure were defined as cure. Adverse effects were recorded as well.
Fisher’s exact tests, Chi-square tests, and Mann-Whitney U tests were used to compare the difference between the 2 groups. Univariable and multivariable logistic regression was used to assess the contributing factors to modified microbiologic cure. ROC curve analysis was performed to evaluate the performance of the final model and the Youden index was employed to define the optimal cutoff value for each factor selected into the final model.
Results
A total of 89 patients were included in our final analysis. Among them, 29 (32.6%) were male, and the median age at diagnosis was 61 years old. Thirty-four of the enrolled patients (38.2%) had received intravenous (IV) antibiotics in their treatment course. Of all patients receiving IV antibiotics, the median time to IV initiation was 1 day (IQR: 1-49), and 24 (70.6%) of them had regimen containing IV agents within 4 weeks. Almost all patients (96.6%) received macrolides during their treatment. The median treatment course duration was 280 days (IQR: 198-608). There were 42 patients (47.2%) who achieved a modified microbiological cure. Thirty-two patients (36%) developed adverse effects.
In our study, we have found that early antibiotics use was associated with modified microbiological cure (aOR 8.577, 95% CI: 2.309-31.859), radiological score and treatment course were negatively associated with modified microbiological cure (aOR 0.827, 95% CI: 0.697-0.981 for a radiological score; aOR 0.998, 95% CI: 0.996-0.999 for treatment course). The optimal cutoff value for initiation of IV antibiotics was 25 days.
Conclusion
Early initiation of IV antibiotics was associated with better treatment outcomes and the optimal cutoff point was determined to be 25 days. Low radiological score at presentation, and early initiation of IV antibiotics may predict the probability of cure. In the future, larger and prospective studies are needed to define a clearer association between patient characteristics, regimens, and treatment outcomes.
Part II：Therapeutic Drug Monitoring of Rifabutin
Background
Rifabutin is a rifamycin antibiotic used in the treatment of TB and NTM infections. Although rifabutin is a less potent inducer of CYP3A4, it still interacts with other CYP3A4 inducers and inhibitors and thus, influencing plasma concentration of rifabutin.
High rifabutin concentration has been associated with adverse drug reactions such as leukopenia, neutropenia, thrombocytopenia and more seriously, uveitis. Previous studies suggested the targeted peak concentration be kept at 0.45-0.9 μg/mL in patients with tuberculosis. However, few studies have been published regarding the use of therapeutic drug monitoring of rifabutin.
Objectives
Our study aimed to monitor the plasma concentration of rifabutin and to evaluate the relationship between its concentration, adverse effects and clinical outcomes.
Methods
A prospective observational study was performed at National Taiwan University Hospital from February 1st, 2021 to July 31st, 2021. Adult patients taking rifabutin for over 7 days in the study period were enrolled. Blood samples collected 0, 3 and 6 hours (C0, C3 and C6) after oral rifabutin were analyzed using a validated high-performance liquid chromatography (HPLC) method. Patients’ data were collected from electronic medical records. Adverse events and clinical outcomes were assessed throughout the treatment course until the end of treatment or July 31st, 2021. All ADRs were evaluated by Naranjo scale.
Fisher’s exact test, Chi-square tests, and Mann-Whitney U tests were used to compare the differences between the two groups, Wilcoxon signed-rank test was used to evaluate the value change before and after the initiation of rifabutin. Linear regression was used to identify contributing factors that influence rifabutin C3. Logistic regression was used to identify possible factors that led to the development of adverse effects. Bootstrap method was used estimate the optimal cutoff value for rifabutin C3.
Results
During study period, there were 13 patients enrolled and 21 TDMs performed. Among 13 patients with mycobacterial infections, there were 13 C3 concentrations, 5 C6 concentrations, and 2 C0 concentrations drawn. All concentrations were below 0.5 μg/mL and 7 of them even below quantification limits. The mean C3 and C6 were 0.39 ± 0.07 and 0.28 ± 0.03 μg/mL in 11 quantifiable concentrations, respectively. The median duration of rifabutin use as of the day of TDM was 65 (IQR: 49-133) days.
Multiple linear regression showed that body weight and AST might have positive effect on C3. The median WBC changed from 5.9 (4.8-10.4) to 4.6 (3.7-7.0) k/μL (p <0.05) before and after the initiation of rifabutin, and four patients (30.8%) developed leukopenia. The optimal cutoff value for leukopenia estimated by sampling 2000 times by bootstrap method was 0.395 µg/mL (95%CI: 0.382-0.429). However, neither C3 nor other factors were associated with leukopenia, neutropenia, thrombocytopenia or any other adverse effects in univariable and multivariable logistic regressions.
Conclusion
We have developed and validated a liquid chromatography method for determining plasma concentration of rifabutin. Plasma concentrations of rifabutin were generally low in our patient population. The study was limited by small sample size in a single medical center. Continuous enrollment of patients newly starting on rifabutin is needed to identify contributing factors associated with C3 and adverse events.
Summary of Part I and Part II
In conclusion, since NTM infections are getting more and more prevalent, more complex and individualized therapies are required. Early initiation of IV antibiotics for MABC-LD may be beneficial and whether incorporation of rifabutin TDM in clinical practice helps or not needs further studies.

並列關鍵字

Mycobacterium abscessus complex lung diseases ； nontuberculous mycobacteria ； intravenous antibiotics ； early initiation ； clinical outcomes ； Rifabutin ； tuberculosis ； therapeutic drug monitoring ； adverse effects

參考文獻

Falkinham JO, 3rd. Environmental Sources of Nontuberculous Mycobacteria. Clin Chest Med 2015;36(1):35-41. DOI: 10.1016/j.ccm.2014.10.003.

Wolinsky E. Mycobacterial Diseases Other Than Tuberculosis. Clinical Infectious Diseases 1992;15(1):1-12. DOI: 10.1093/clinids/15.1.1.

Johansen MD, Herrmann J-L, Kremer L. Non-Tuberculous Mycobacteria and the Rise of Mycobacterium abscessus. Nature Reviews Microbiology 2020;18(7):392-407. DOI: 10.1038/s41579-020-0331-1.

Honda JR, Knight V, Chan ED. Pathogenesis and Risk Factors for Nontuberculous Mycobacterial Lung Disease. Clin Chest Med 2015;36(1):1-11. DOI: 10.1016/j.ccm.2014.10.001.

Kwak N, Dalcolmo MP, Daley CL, et al. Mycobacterium abscessus Pulmonary Disease: Individual Patient Data Meta-Analysis. European Respiratory Journal 2019:1801991. DOI: 10.1183/13993003.01991-2018.