For diagnosing lung cancerous and precancerous lesions, autofluorescence bronchoscopy (AFB) presented considerably high sensitivity but low specificity, which may be improved when AFB was combined with white light brochoscopy (AFB + WLB). The aim of this study was to compare the diagnostic performance between AFB and AFB + WLB directly.
Pubmed, Scopus, Embase, Web of Science, ProQuest, the Cochrane Library and Ovid were searched from inception to Mar 20, 2015, for elegible studies containing sufficient data from both of tehcniques to construct 2 Ã 2 table with confirmation by histopathology. Pooled sensitivity, specificity, diagnostic odds ratio (DOR) and the area under the receiver-operating characteristic curve (AUC) were estimated by random-effect model. Quality assessment and heterogeneity were assessed.
7 comparative studies involving with 904 patients and 2740 biopsy specimens were included. The summary sensitivity, specificity, DOR and AUC of AFB were 88% (95%CI 65%-97%), 63% (49-75), 12 (3-54) and 77% (73-81) respectively, and those of AFB + WLB were 90% (77-96), 54% (39-68), 11 (4-34) and 78% (74-81). In heterogeneity assessment, study quality (according to QUADAS-2) was indicated as the effect on AFB + WLB data, corresponding P value was 0.01 and I2 was 78% (51-100). However, lower specificity of AFB + WLB (vs AFB) was presented and no significant difference (P < 0.05) for comparison existed among studies with high or moderate&low quality. Same situation was shown in all results of summary and exploratory subgroup analysis.
|Histopathology||Tech||Sen (%)||95%CI (%)||P||Spe (%)||95%CI (%)||P|
|INV- > SEV||A||86||50-100||0.239||89||80-97||0.629|
|A + W||91||62-100||79||58-99|
|INV- > MOD||A||95||86-100||0.151||63||45-81||0.178|
|A + W||96||87-100||60||40-79|
|CIS- > MOD||A||74||34-100||0.109||52||30-75||0.140|
|A + W||77||43-100||49||23-74|
|A + W||91||63-100||23||7-39|
INV = Invasive carcinoma CIS = Carcinoma in situ SEV = Severe dysplasia MOD = Moderate dysplasia ASD = Angiogenic Squamous Dysplasia.
Insufficient evidence indicated the diagnostic performance of AFB + WLB superior to AFB alone for lung cancerous and precancerous lesions.
Clinical trial identification
It is a systematic review and meta-analysis, and all data come from published articles, therefore, it is non-essential to apply for protocol number of clinical trial.
All authors have declared no conflicts of interest.