再感染對台灣地區結核病發展的重要性

The importance of reinfection in the development of tuberculosis in Taiwan

王振源(Jann-Yuan Wang)

指導教授 ： 楊泮池

共同指導教授 ： 薛博仁(Po-Ren Hsueh)

國立臺灣大學/醫學院/臨床醫學研究所/博士(2007年)

https://doi.org/10.6342/NTU.2007.00673

摘要

結核病（tuberculosis）是一個古老的傳染病，但仍是目前世界上死亡人數最高的一種傳染疾病。台灣地區，近年來結核病的發生率仍居高不下，民國九十三年和民國九十四年，連續兩年發生率皆突破每十萬人口 70 人。傳染病防治上的明顯失敗，意味著必須重新檢討結核病的傳播途徑和發病原因，重新規劃有效的防疫措施，以杜絕結核病的發生。以往認為結核病發病的原因是因為原本蟄伏在體內的結核菌再活化而導致疾病，並非再感染到新的結核菌株。但近年來的研究發現，在結核病患者當中，大部分是屬於群聚感染，而且，與一般的傳染病一樣，某些特殊的環境、或特定的帶原者不斷地散播，造成結核病的疫情，可能是結核病之所以無法被撲滅的主因。
結核病的發病機轉，究竟是因為內在的潛伏性結核菌再度活化，或者是外來的結核菌再度感染，是一個十分重要的議題，因為這將會嚴重影響公共衛生政策的制定。撲滅結核病最好的方法，當然是兩方面都可以顧及，但因為公共衛生以及醫療資源是有限的，實際上兩者無法兼顧。所以如果是再活化這一個機轉比較重要，那麼就應該要將更多的資源投注在發病個案的掌握及治療；相反的，如果是再感染這一個機轉比較重要，那麼除了發病個案的管理治療以外，就必須要將一部份的資源用在新個案的尋找上面。這樣，才能夠用有限的資源，達到最大的成效。
直到西元 2005 年的六月，一共有 23 篇文章探討再感染和再活化這個議題。將這些研究結果歸類後可以發現，在發生率小於每年每十萬人口中五十人的地區，再感染比率平均為 18.3%；但如果是在結核病發生率介於五十人到三百人的地區，再感染比率介於 20% 到 36% 之間；而在發生率超過三百人的地區，再感染的比率會高達七成以上。這暗示著結核病的發生率愈高，再感染比率也就愈高。因此我們假設，在台灣這個結核病的盛行地區，再感染這一個途徑，在病因學上扮演著相當重要的角色。我們將分四個部份進行研究設計與分析。
在研究的第一個部份，針對整體一般結核病個案和散播性結核病患者的菌株進行基因型鑑定，分析群聚感染所佔的比率，推測近期傳染的重要性。在1999 年到 2005 年隨機挑選的 250 株菌株中，有 205 株成功得做完基因型鑑定。總共有 66 個基因型，其中 156（76.1%）個患者屬於群聚感染。此外，由 1995 年到 2004 年這十年當中，64 位散播性結核病患者的菌株中，鑑定結果顯示總共有 22 種基因型，47（73%）個患者屬於群聚感染。
在研究的第二個部份，我們將針對不同的族群，利用結核菌特有抗原與血液中 T 細胞共同培養後，偵測 γ 型干擾素的產量。藉由分析不同族群中結核菌的感染率，希望可以證實的確有接觸後感染或再感染的情況發生。總共有 13 個醫護人員、43 位結核病患的同住家屬、以及 49 位培養確定為結核病之患者接受免疫檢測，結果有 1（8%）位醫護人員、24（56%）位家屬、和43（88%）位結核病患有陽性反應。
在研究的第三個部份，針對臨床分離的結核菌株、抗酸性塗片陽性的痰液檢體，偵測是否同時存在有北京家族或非北京家族的結核菌株，以了解多重感染的比率。由 1999 年到 2004 年這六年的 100 株菌株中，有 17（17%）株在解凍後無法培養出結核菌。剩下的 83 株當中，14（17%）株為多重感染。另外，在 2007 年 10 月 1 日到 15 日，13 個結核菌聚合酶連鎖反應陽性的抗酸性塗片陽性痰液檢體中，2（15%）個是多重感染。
在研究的第四個部份，由臺大醫院培養確定為結核病的個案中，挑出完治後復發的個案，收集前後兩次發病的結核菌株行基因型比對，結果顯示，在 1999 年到 2004 年這六年中一共有 2071 位培養確定為結核病的患者，共有 61 位完治後復發（2.9%），其中有 49 位兩次發病的結核菌株均有保存下來。基因型比對後發現 25（51%）位患者是因為再感染所引起的復發。
第一部份的研究結果與之前針對台灣各地結核菌株的分子流行病學研究結果相同，超過七成屬於群聚感染，這表示台灣地區的結核病患中，近期傳染而造成發病的比率很高，也就意味著再感染這一個致病機轉可能是相當重要的。
第二部份的研究結果顯示，結核病患者的家屬在密切接觸病患之後，受到結核菌再次感染的比率很高，這一點，也間接證實了近期暴露之後所造成的再感染，可能比以往被感染的比率還高。
由於北京家族的結核菌株佔台灣整個結核菌族群的一半左右，因此，若不考慮菌株之間的不同，多重感染的情況中，應該有一半會是混合著一株北京家族結核菌株和一株非北京家族結核菌株的組合。由第三部份的研究結果可以知道，臺大醫院結核病患者當中，多重感染的比率大約在 15% 至 17% 左右。根據這樣的發現，我們可以推論，台灣地區可能有超過三成的結核病患者是多重感染。而多重感染，廣義來說，其實也算是一種再感染，因此，再感染確實在台灣地區結核病的發病機轉上，佔有很重要的角色。
我們第四部份的研究顯示，在整整六年的期間，所有培養確定為結核病的患者，有 2.9% 的個案復發，其中超過一半是因為再感染所引起。這樣的結果表示治療過後的結核病患，每年結核病的發生率是 0.25% 左右，比一般民眾得到結核病機率大了四倍左右。這個發現表示曾經得過結核病的患者，更容易再次得到結核病，再感染的機會也比一般人高。
根據最新的結核病治療標準指引，對於一般病患，標準的治療是合併四種藥物；但如果患者之前已經有過結核病或曾經接受治療，一開始的治療會建議併用五種藥物。如果復發個案可以提早知道是再感染而發病，那麼和一般結核病患者一樣，僅只使用四種藥物，而不需額外再加一種。但如果知道是因為再活化而造成復發，一開始使用五種藥物治療就有其必要性。我們的結果顯示，在各種臨床的徵象上，包括年齡、性別、系統性疾病、臨床症狀及其時間、以及胸部 X 光的表現，再感染的患者與再活化的患者並沒有顯著的不同，這也就代表著，臨床醫師根本無法根據患者的臨床表現猜測究竟是再活化還是再感染。因此，在結核病發生率較高的地區，既然再感染所佔的比率可能大過一半以上，利用分子生物學的方法快速得知結核菌株的抗藥性，以決定治療的處方，將是十分重要的課題。
既然再感染在台灣地區以及其他結核病盛行地區很重要，那麼主動、積極、和即時地篩檢，找出病患將是十分重要的防疫措施。目前世界衛生組織以及各國結核病醫學會並不推薦大規模的結核病篩檢以降低結核病的發生率，但對於某些結核病的高危險群，像是住院病患、門診長期追蹤病患、囚犯、無家可歸的人、以及藥物濫用者等等，規則篩檢究竟有沒有實際上的幫助，目前仍然有爭議。其實，在這幾個特殊的環境中，由於人與人之間的接觸相當頻繁，加上宿主的抵抗力也較差，因此，也正是再感染比率很高的地方。而我們已經知道，結核病發生率高的國家，再感染的比率也隨著增高，因此，大規模篩檢的可行性和防疫上的成本效益，事實上還是需要再次確認，但究竟要挑出哪些高危險族群進行篩檢、用什麼方法做篩檢、以及多久篩檢一次，仍需要更多的研究才可以確定。
總而言之，我們的研究，分四個部份，利用結核菌株的群聚、多重感染分析、和結核病暴露者的免疫反應、以及復發患者的菌株基因型比對等方法，證實在台灣地區，的確有再感染的現象存在，而且比再活化來得更重要，未來台灣地區公共衛生政策的擬定上，除了加強落實短程直接觀察治療法以外，規則的篩檢以積極發掘結核病個案也是不容忽視的方案。期待我們的努力與大家的投入，可以讓未來的台灣，不再是一個結核病的盛行地區。

關鍵字

結核病 ； 再感染 ； 再活化 ； 多重感染 ； 群聚感染 ； 復發 ； 免疫檢測

並列摘要

Tuberculosis (TB) is an infectious disease causing enormous morbidities and mortality for thousands of years. In Taiwan, the incidence of TB was still higher than 70 per 100,000 people in 2004 and 2005. The failure in controlling TB implies that we should re-evaluate the routes of transmission and the mechanisms for the development of TB, and re-design the policies for public health and disease control. It was believed previously that all cases of active TB were resulted from reactivation of dormant bacilli, rather than reinfection of a new TB strain. However, many recent studies revealed that a significant proportion of TB patients belonged to clusters. The results of some studies even suggested that recent transmission could be more important than reactivation in the development of active TB. It is a very important issue whether reactivation or reinfection is responsible for the development of most active TB because it changes the policy in public health. If reinfection is more important, besides case holding, more resources should be distributed on finding source cases (cases finding).
Up to June 2005, a total of 23 studies reporting the relative contributions of reinfection and reactivation in patients with TB relapse were identified. The results of the studies performed in areas with the annual incidence being less than 50 cases per 100,000 people showed that the reinfection proportion was 18.3% in average. When the annual incidence was between 50 and 300 cases per 100,000 people, the reinfection proportion ranged from 20% to 36%. Moreover, in area where the annual incidence was higher than 300 cases per 100,000 people, more than 70% of patients with TB relapse were due to reinfection. These imply that the higher the local incidence of TB, the more important reinfection is. Based on those evidences, we hypothesize that in Taiwan, an endemic area of TB, reinfection should have a major contribution in the development of active TB.
In the first part of our study, we genotyped the clinical isolates of M. tuberculosis from patients with TB or disseminated TB in National Taiwan University Hospital. The proportion of patients with recent transmission was estimated by cluster analysis. From 1999 to 2005, 250 isolates were randomly selected for genotyping. The results were available in 205, belonging to 66 genotypes. Among them, 156 (76.1%) were clusters. In addition, genotyping for 64 isolates from patients with disseminated showed that there was 22 genotypes and 47 (73%) isolates were clusters. Compatible with previous molecular epidemiologic study in Taiwan, our results showed that more than 70% of clinical M. tuberculosis isolates were clustered. These findings and our data suggests that recent transmission has a major contribution in the development of active TB in Taiwan.
In the second part, we included health care workers, culture-confirmed TB patients, and family contacts for ex-vivo immune study, which detecting the production of interferon-γ from T cells after co-culture with M. tuberculosis-specific antigens. A total of 13 health care workers, 49 culture-confirmed TB patients, and 43 family contacts received immune study. The result was positive in 1 (8%), 43 (88%), and 24 (56%), respectively. The results demonstrated that people can be reinfected after contact to patients with active TB. This finding further emphasizes the importance of reinfection in the development of active TB in Taiwan.
In the third part of our study, the proportion of multiple infection in all TB patients was evaluated by detecting the presence of Beijing strain and non-Beijing strain in clinical isolates of M. tuberculosis and acid-fast smear-positive respiratory specimens from TB patients in NTUH. First, 100 clinical isolates from 1999 to 2004 were randomly selected. Among the 83 isolates with results available, 14 (17%) were multiply infected. Furthermore, among 13 TB-PCR-positive and smear-positive respiratory specimens collected during Oct. 1, 2007 and Oct. 15, 2007, 2 (15%) were multiply infected. Because the about half of the clinical M. tuberculosis isolates in Taiwan were Beijing strain, we could roughly estimate that about half of the patients with multiple infection are simultaneously infected by one Beijing strain and one non-Beijing strain. Our results suggested that multiple infection is a common situation in Taiwan, probably occurring in 30% of the TB patients. Compatible with the other results of our studies, this finding demonstrates that reinfection plays an important role in the tuberculosis in Taiwan.
In the final part of our study, culture-confirm TB patients who experienced relapse after complete treatment in NTUH were identified. The reinfection proportion was calculated by genotyping the clinical isolates from both episodes of these patients. From 1999 to 2004, a total of 2071 culture-confirmed TB patients were identified. Among them, 61 (2.9%) experienced relapse after complete treatment. The clinical isolates of both episodes were preserved in 49 patients. The results of genotype comparing showed that 25 (51%) were reinfection. The final part of our study revealed that 2.9% of the culture-confirmed TB patients relapsed within the 6-year study period; half of them being reinfection. Therefore, the incidence of TB among cured patients was 0.25%, which was 4-fold higher than in the general population. The finding implies that cured patients have a higher risk on disease due to reinfection.
According to the guideline for the treatment of TB, the standard regimen for general TB is four-combined regimen. However, if resistant TB is suspected, using five-combined regimen until the results of susceptibility test are available is recommended. Therefore, if we can early diagnosis reinfection in patients with TB relapse, four, rather five, drugs are enough for treatment. On the other hand, using five-combined regimen is necessary in patients with reactivation after complete treatment. However, our results suggest that clinical characteristics, including age, sex, underlying co-morbid conditions, symptoms and their duration, and the findings of chest radiographs, fail to differentiate between reinfection and reactivation. Therefore, in endemic areas where the reinfection probably accounts for more than 50% of all TB cases, rapid detection of drug resistance by molecular biologic methods to determine treatment regimen is an important issue.
Since reinfection plays an important role in the development of active TB in Taiwan, aggressive and active surveillance to detect source cases would be important in disease control. At present, mass screening is not recommended by the World Health Organization. However, in certain high-risk groups, such as hospitalized patients, those with underlying co-morbid conditions, prisoners, drug abusers, and the homeless, whether routine screening is cost-effective or not remains undetermined. In addition, since reinfection is common in endemic areas of TB, routine screening for the high-risk patients may be even more feasible in those areas.
In conclusion, by detecting the proportion of clusters and multiple infection in clinical isolates of M. tuberculosis, the presence of immune response in TB contacts, and comparing the genotypes of the clinical isolates in both episodes in patients with TB relapse after complete treatment, we had proven that reinfection not only existed but had a major contribution to the development of active TB in Taiwan. Therefore, the policy of public health and TB control should emphasize both case holding and case finding. With the understanding of the epidemiology of TB, we hope Taiwan will not be an endemic area of TB in the near future.

並列關鍵字

tuberculosis ； reinfection ； reactivation ； multiple infection ； relapse ； immunoassay

參考文獻

1. Medical Research Council. Streptomycin treatment of pulmonary tuberculosis. Brit Med J 1948;2:769-82

2. Medical Research Council. The prevention of streptomycin resistance by combined chemotherapy. Brit Med J 1952

3. Pfuetze KH, Pyle MM, Hinshaw HC and Feldman WH. The first clinical trial of streptomycin in human tuberculosis. Am Rev Tuber 1955;71:752-4

4. Anonymous. Isoniazid in pulmonary tuberculosis. Lancet 1952;2:19-21

6. Seventh report to the Medical Research Council by their Tuberculosis Chemotherapy Trials Committee. Various combinations of isoniazid with streptomycin or with PAS in the treatment of pulmonary tuberculosis. Brit Med J 1955:435-45