Mitarai Satoshi, Kobayashi Ikuo, Abe Chiyoji, Wada Masako, Suzuki Katsuhiro, Takashima Tetsuya, Kawabe Yoshiko, Machida Kazuko, Tano Masao, Takigawa Shuichi, Kamada Arisu, Shigetoh Eriko, Fujii Shunji, Mori Kenichi, Suyama Hisashi, Yano Shuichi, Kawashiro Takeo, Ogata Hideo
Bacteriology Division, Mycobacterium Reference Centre, Research Institute of Tuberculosis, JATA, 3-1-24 Matsuyama, Kiyose-shi, Tokyo 204-8533, Japan.
Kekkaku. 2007 May;82(5):449-54.
To evaluate the accuracy of drug susceptibility testing to isoniazid with BACTEC MGIT 960 (MGIT AST) comparing with the standard proportion method using Ogawa medium.
A total of 1109 M. tuberculosis strains, which were selected from the collection of RYOKEN drug resistance survey in 2002, were selected and subjected to the susceptibility testing to isoniazid using MGIT AST and 1% Ogawa standard methods. The results from MGIT AST were compared with the judicial diagnosis by Ogawa. The sensitivity to detect drug resistance, the specificity for susceptible strain, the efficiency of overall agreement, and kappa coefficient were calculated to evaluate the performance. The treatment process, outcome and prognosis were analysed for the patients on whom the tests showed discrepant results.
Compared with the judicial results, the sensitivity, specificity, efficiency, and kappa coefficient of MGIT AST were 100%, 97.1%, 97.3%, and 0.798, respectively. The strains, which showed discrepant results between MGIT AST and Ogawa, were all susceptible by Ogawa and resistant by MGIT AST. A total of 11 out of 30 discrepant cases were followed clinically and no relapse cases were identified, irrespective of the modification of the treatment regimen. As for the proportion of primary INH drug resistance in the present study, it was 5.3% with MGIT AST but was 2.7% with Ogawa, and the difference was statistically significant (p = 0.005).
The discrepancies on the results of drug susceptibility testing of M. tuberculosis strains to isoniazid between MGIT AST and 1% Ogawa proportion method have been reported. In the present study, the sensitivity, specificity, and overall efficiency of MGIT AST on the prevalent strains in Japan were all beyond 95%, and considered sufficient as the anti-tuberculosis drug susceptibility testing (AST), though 2.7% of discrepancy was observed. Even for the discrepant cases, there was no difference in the treatment outcome and prognosis. Thus, MGIT AST was confirmed as a reliable AST method comparable to Ogawa standard. However, MGIT AST might increase the proportion of INH resistance if it was used as a major AST method, compared with Ogawa.
采用BACTEC MGIT 960（MGIT药敏试验）评估结核分枝杆菌对异烟肼药敏试验的准确性，并与使用小川培养基的标准比例法进行比较。
从2002年的RYOKEN耐药性调查收集的菌株中选取1109株结核分枝杆菌，采用MGIT药敏试验和1%小川标准方法进行异烟肼药敏试验。将MGIT药敏试验的结果与小川法的判定结果进行比较。计算检测耐药性的敏感性、对敏感菌株的特异性、总体一致性效率和kappa系数以评估性能。对试验结果出现差异的患者的治疗过程、结局和预后进行分析。
与判定结果相比，MGIT药敏试验的敏感性、特异性、效率和kappa系数分别为100%、97.1%、97.3%和0.798。MGIT药敏试验和小川法结果出现差异的菌株，小川法判定均为敏感，MGIT药敏试验判定为耐药。30例差异病例中共有11例进行了临床随访，无论治疗方案是否调整，均未发现复发病例。本研究中，MGIT药敏试验检测的原发性异烟肼耐药比例为5.3%，而小川法检测的为2.7%，差异具有统计学意义（p = 0.005）。
已有报道MGIT药敏试验和1%小川比例法对结核分枝杆菌菌株对异烟肼的药敏试验结果存在差异。在本研究中，MGIT药敏试验对日本流行菌株的敏感性、特异性和总体效率均超过95%，尽管观察到2.7%的差异，但仍被认为足以作为抗结核药敏试验（AST）。即使对于差异病例，治疗结局和预后也没有差异。因此，MGIT药敏试验被确认为一种与小川标准相当的可靠AST方法。然而，与小川法相比，如果将MGIT药敏试验用作主要的AST方法，可能会增加异烟肼耐药的比例。
