Combined measurement of miRNA-183, HE4, and CA-125increases diagnostic efficiency for ovarian cancer

Jing Liang, Xi Yang, Lei Liu, Lei Qiao, Peng Peng, Jianlong Zhou, R Siegel, D Naishadham, A Jemal, O Foley, J Rauh-Hain, M Del Carmen, M Hall, C Gourley, I Mcneish, J Ledermann, M Gore, G Jayson, J Bian, B Li, X Kou, T Liu, L Ming, M Tsujiura, D Ichikawa, S Komatsu, A Shiozaki, H Takeshita, T Kosuga, Y Zhao, G Guo, W Li, E Zhu, X Luo, H Chen, L Zhang, J Du, H Wang, B Wang, B Marzec-Kotarska, M Cybulski, J Kotarski, A Ronowicz, R Tarkowski, G Polak, K Niu, W Shen, Y Zhang, Y Lu, L Wang, M Zhu, A Ren, H Wu, W Han, R Tan, R Tu, Y Wang, R Guo, W Sheng, L Yin, H Zheng, D Yang, F Song, Y Wen, J Xu, C Wu, X Che, D Yu, T Zhang, W Zhu, J He, D Chen, B Zhang, L Xu, H Ma, X Chen, Y Ba, L Ma, X Cai, Y Yin, K Wang, J Lu, G Getz, E Miska, E Alvarez-Saavedra, J Lamb, D Peck, R Eitan, M Kushnir, G Lithwick-Yanai, M David, M Hoshen, M GlezermanEuropean 2020 · DOI

Abstract

This study aimed to determine the role of miR-205, miR-182, and miR-183 expression in the serum of ovarian cancer patients for early diagnosis. The expression of these microRNAs, alongside traditional markers CA-125 and HE4, was analyzed in 101 ovarian cancer (OC) patients, 50 patients with benign ovarian diseases (BOD), and 50 healthy volunteers, with results validated in an independent cohort of 201 subjects. Binary logistic regression identified a diagnostic model combining miR-183, HE4, and CA-125 as the most effective, achieving an AUC of 0.951. This triple-marker approach significantly improves diagnostic sensitivity and specificity compared to single-marker tests.

The Challenge of Early Detection

Ovarian cancer is frequently diagnosed at advanced stages, where the five-year survival rate is less than 15%. Conversely, early detection can increase this rate to 90%. Current diagnostic standards, such as cytologic examinations and serum biomarkers like CA-125, often lack the necessary balance of sensitivity and specificity. CA-125 is frequently elevated in benign conditions, leading to false positives, while newer markers like HE4, though more specific, may miss early cases. This research explores whether circulating microRNAs (miRNAs) can serve as stable, sensitive biomarkers to bridge this diagnostic gap.

Methods and Marker Selection

The researchers evaluated three miRNAs (miR-205, miR-182, and miR-183) known to be upregulated in ovarian cancer, alongside protein markers HE4 and CA-125. Serum samples were collected from OC patients, BOD patients, and healthy controls. miRNAs were quantified using quantitative Real-Time PCR (qRT-PCR), while protein markers were detected via chemiluminescence microparticle immunoassay.

Serum levels of miR-205, miR-183, miR-182, HE4, and CA-125 across the ovarian cancer (OC), benign ovarian disease (BOD), and control groups. Significant differences are observed in the OC group compared to others.

Binary logistic regression was employed to develop a predictive model, and Receiver Operating Characteristic (ROC) curves were used to evaluate diagnostic efficiency. The study utilized a training set (n=201) to build the model and a validation set (n=201) to confirm its robustness on independent data.

Key Results: The Power of Combination

The study confirmed that while individual markers like miR-182 and CA-125 are highly sensitive, they suffer from low specificity (~0.27). In contrast, miR-183 emerged as a strong candidate, correlating with clinical stage and metastasis.

The most effective diagnostic model was established using the following equation: $$Y = \text{Logit}(P) = -5.457 + 5.365 \times \text{miR183} + 0.019 \times \text{HE4} + 0.004 \times \text{CA125}$$

ROC curve analysis comparing the diagnostic performance of individual markers and the combined triple-marker model. The combined model (F, G) shows superior Area Under the Curve (AUC) values.

In the training set, this combined model achieved an AUC of 0.951, with a sensitivity of 97% and specificity of 85%. These results were successfully replicated in the validation set (AUC 0.920, sensitivity 93.9%, specificity 81.6%). For comparison, miR-183 alone had an AUC of 0.77, and HE4 alone was 0.847, demonstrating the clear advantage of the multi-marker approach.

Conclusion

The combination of miR-183, HE4, and CA-125 provides a highly accurate diagnostic "fingerprint" for ovarian cancer. By leveraging the sensitivity of CA-125, the specificity of HE4, and the refining power of miR-183, this model offers a promising tool for early diagnosis and clinical evaluation. Future research should focus on larger multi-center trials and specifically target Stage I and II patients to further validate its utility as an early warning system.

Summary updated

Abstract

Objective: This study aimed to determine the role of miR-205, miR-182, and miR-183 expression in the serum of ovarian cancer patients in the early diagnosis of ovarian cancer. Materials and Methods: The expression of miR-205, miR-182, miR-183, CA-125, and HE4 was detected in the sera of 101 patients with ovarian cancer, 50 patients with benign ovarian diseases, and 50 healthy volunteers. The results were validated in 98 patients with ovarian cancer, 50 patients with benign ovarian diseases, and 53 healthy volunteers. The expression of miR-205, miR-182, miR-183, CA-125, and HE4 was subjected to ROC analysis and binary logistic regression analysis. Results: The sensitivity of miR-182 and CA-125 was highest (0.901% and 0.832, respectively), but the specificity was low (both 0.27) in the early diagnosis of ovarian cancer. HE4 had the highest specificity in the early diagnosis of ovarian cancer. The sensitivity, specificity, and AUC of HE4 were 0.842, 0.81, and 0.847, respectively. Binary logistic regression analysis showed that three variables were suitable for the diagnostic model: Y=Logit(P)=-5.457+5.365*miR183+0.019*HE4+0.004*CA125. Based on the diagnostic model, ROC analysis showed that the sensitivity, specificity, and AUC were 0.97, 0.85, and 0.951, respectively. Statistical validation showed that the sensitivity, specificity and AUC were 0.941, 0.86, and 0.951, respectively. Conclusion: miR-183 has high specificity and sensitivity in the diagnosis of ovarian cancer. Measurement of miR-183 combined with HE4 and CA-125 is of value for the early diagnosis and evaluation of ovarian cancer.
BibTeX
@inproceedings{liang2020combined,
  author = {Jing Liang and Xi Yang and Lei Liu and Lei Qiao and Peng Peng and Jianlong Zhou and R Siegel and D Naishadham and A Jemal and O Foley and J Rauh-Hain and M Del Carmen and M Hall and C Gourley and I Mcneish and J Ledermann and M Gore and G Jayson and J Bian and B Li and X Kou and T Liu and L Ming and M Tsujiura and D Ichikawa and S Komatsu and A Shiozaki and H Takeshita and T Kosuga and B Li and Y Zhao and G Guo and W Li and E Zhu and X Luo and H Chen and L Zhang and L Zhang and J Du and H Wang and B Wang and B Marzec-Kotarska and M Cybulski and J Kotarski and A Ronowicz and R Tarkowski and G Polak and K Niu and W Shen and Y Zhang and Y Zhao and Y Lu and L Wang and M Zhu and A Ren and H Wu and W Han and R Tan and R Tu and Y Wang and R Guo and R Guo and W Sheng and L Yin and H Zheng and L Zhang and Y Zhao and D Yang and F Song and Y Wen and J Xu and C Wu and X Che and L Wang and D Yu and T Zhang and W Zhu and J He and D Chen and B Zhang and L Xu and H Ma and X Chen and Y Ba and L Ma and X Cai and Y Yin and K Wang and J Lu and G Getz and E Miska and E Alvarez-Saavedra and J Lamb and D Peck and R Eitan and M Kushnir and G Lithwick-Yanai and M David and M Hoshen and M Glezerman},
  title = {{Combined measurement of miRNA-183, HE4, and CA-125increases diagnostic efficiency for ovarian cancer}},
  booktitle = {European},
  year = {2020},
  doi = {10.31083/j.ejgo.2020.01.4788},
}
Figure