Anzeige
Erschienen in:
01.07.2019 | Head and Neck
Assessment of the cancerization risk for oral potentially malignant disorders by clinical risk model combined with autofluorescence and brush biopsy with DNA-image cytometry
Erschienen in: European Archives of Oto-Rhino-Laryngology | Ausgabe 9/2019
Einloggen, um Zugang zu erhaltenAbstract
Purpose
To explore the feasibility of assessing the cancerization risk of oral potentially malignant disorders (OPMD) through a clinical risk model combined with autofluorescence and brush biopsy with DNA-image cytometry.
Methods
We collected the baseline clinical data of 269 patients; then, performed autofluorescence, brush biopsy with DNA-image cytometry and histopathological examination. Then, we obtained the significant factors by univariate logistic analysis, constructed the clinical risk model by multiple logistic regression and selected the optimal cutoff value according to the maximum Youden index. Finally, we calculated the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the clinical risk score ≥ cutoff value, autofluorescence and brush biopsy with DNA-image cytometry, and plotted the receiver-operating characteristic (ROC) curves and decision curve analysis (DCA).
Results
The clinical risk model is represented by the formula: 1 × gender + 1.6 × age group + 1 × lesion site + 1.4 × local stimulus + 1.5 × drink. The area under the curve (AUC) was 0.83, and the optimal cutoff score was 3. The AUC indicated that the clinical risk score ≥ 3 (0.74) and autofluorescence (0.77) had a certain diagnostic values, while brush biopsy with DNA-image cytometry (0.92) displayed a good value. Besides, the DCA showed that all three tests had clinical significance.
Conclusions
The cancerization risk of patients can be assessed by the clinical risk model combined with sequence application of autofluorescence and brush biopsy with DNA-image cytometry, to decide whether histopathological examination or other intervention measures should be selected.