儿童肺结核支气管肺泡灌洗液Xpert MTB/RIF检测的诊断准确性研究

目的探讨支气管肺泡灌洗液的Xpert MTB/RIF检测对儿童肺结核的诊断价值。方法有疑似活动性结核病的临床症状和体征，行纤维支气管镜检查并留取了支气管肺泡灌洗液的患儿为研究对象。以支气管肺泡灌洗液作为病原学检测标本，分别以病原学诊断活动性结核病和临床诊断活动性结核病为金标准，以Xpert MTB/RIF为待测标准，考察Xpert MTB/RIF对病原学和临床诊断结核的诊断价值。基于样本中结核分枝杆菌（MTB）的Ct值反映所检测样本中MTB的载量；通过对利福平耐药位点的检测，对MTB菌株进行药物敏感性检测。结果符合本文纳入标准的疑似结核病患儿351例，男198例，女153例，年龄（5.9±3.9）岁。肺结核患儿125例，其中病原学诊断结核43例（34.4%），临床诊断结核82例（65.6%）；肺结核合并支气管结核51例（40.8%），单纯肺结核74例（59.2%）；非结核呼吸道感染性疾病226例，肺炎支原体肺炎187例（82.7%），细菌性肺炎39例（17.3%）。Xpert MTB/RIF在病原学诊断和临床诊断结核的敏感度分别为79%(95%CI：63%~89%)和51%(95%CI：40%~62%)，差异有统计学意义（χ2=9.18，P=0.002）；肺结核合并支气管结核的敏感度为80%（66%~90%),单纯肺结核的敏感度47%(35%~59%)；病原学诊断和临床诊断结核、合并支气管结核和单纯肺结核特异度均为100%(95% CI：97.9%~100%)。Xpert MTB/RIF检测病原学诊断结核敏感度高于临床诊断结核，肺结核合并支气管结核敏感度高于单纯肺结核，差异有统计学意义（χ2=13.88，P<0.001）；支气管肺泡灌洗液的MTB核酸检出载量病原学诊断结核高于临床诊断结核，差异有统计学意义（χ2=7.37，P=0.025）。76例Xpert MTB/RIF检测阳性患儿中，利福平耐药2例（2.6%）。结论支气管肺泡灌洗液的Xpert MTB/RIF检测在儿童肺结核诊断中具有较高价值，可在缺乏细菌学诊断证据的临床诊断结核病儿童中发现MTB，具有较高的敏感度，有助于提高儿童结核病病原学检出率。

Evaluation of Xpert MTB/RIF in the diagnosis of pulmonary tuberculosis in children using bronchoalveolar lavage fluid

ObjectiveEvaluate the value of Xpert MTB/RIF in the diagnosis of pulmonary tuberculosis (TB) in children using bronchoalveolar lavage fluid (BALF). MethodsChildren were enrolled if (i) they had suspected symptoms of TB; (ii) they had the indications to undergo fiberoptic bronchoscopy; and (iii) informed consent had been provided to undergo fiberoptic bronchoscopy. With the golden standard of bacteriologically and clinical evidence, the diagnostic value of Xpert MTB/RIF was evaluated in children with bacteriologically confirmed and clinically diagnosed TB using BALF. The MTB load of the samples were reported as semiquantitative readouts based on the Ct. Rifampicin (RIF). The sensitivity testing were detected through the RIF-resistance-determining region of rpoB gene. ResultsFinally, 351 children suspected with pulmonary TB ［(5.9±3.9) years old, 198 boys and 153 girls ］ were included in the study. 125 children were finally diagnosed as having pulmonary TB, including 43 (34.4%) cases with bacteriologically confirmed TB and 82 (65.6%) cases classified as clinically diagnosed TB. Among them, 51（40.8%）children were accompanied by tracheobronchial tuberculosis (TBTB). 226 children were finally diagnosed with respiratory tract infection, including 187(82.7%) cases with Mycoplasma pnuemoniae pneumonia and 39(17.3%) cases with bacterial pneumonia. The sensitivity of Xpert MTB/RIF were 79%(95%CI: 64-89), 51%(95%CI: 40-62), 80%（66%~90%) and 47%(36%~59%) respectively in children with bacteriologically confirmed TB, clinically diagnosed TB, pulmonary TB accompanied by TBTB and pulmonary TB. The specificity of Xpert MTB/RIF were 100% (95% CI:97.9%~100%) in the above four subgroups. The sensitivity of Xpert MTB/RIF in bacteriologically confirmed TB children was significantly higher than that of the test in children with clinically diagnosed TB (χ2=9.18，P=0.002). The sensitivity was significantly higher in children accompanied by TBTB, when compared with those who were not accompanied by TBTB (χ2=13.88，P<0.001). The load of MTB nucleic acid detected in bronchoalveolar lavage was higher in bacteriologically confirmed cases than that in clinically diagnosed cases (χ2=7.37，P=0.025). Among the 76 children with positive Xpert MTB/RIF results, two cases (2.6%) were detected to be infected with rifampicin resistant MTB. ConclusionXpert MTB/RIF using bronchoalveolar lavage fluid had good sensitivity and added clinical value by assisting the rapid and accurate diagnosis of pulmonary TB in children with negative bacteriological results.