Purpose

To evaluate the values of kinetic parameters derived from dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) in distinguishing uterine endometrioid adenocarcinoma (EAC) from adenocarcinoma of cervix (AdC) since they correspond to different diagnosis and prognosis.

Introduction

It is critical to identify the primary sites of tumor before treatments for patients with uterine adenocarcinoma due to their different treatments and prognoses between uterine endometrioid adenocarcinoma (EAC) and adenocarcinoma of cervix (AdC)¹. Whereas there are still some uncertain cases where inconclusive results of cervical biopsy or curettage specimens happen due to the overlaps of histologic appearance and limited value of immunohistochemical markers between EAC and AdC². Previous studies had controversial result in differentiation using MR morphological information^3,4. Clinical observation and research results showed different perfusion characteristics between EAC and AdC. Therefore, in this study, we attempt to use parameters derived from DCE-MRI to investigate the perfusion difference between adenocarcinoma of endometrium and cervix.

Material

Sixty newly diagnosed patients with distinctive pathology underwent DCE-MRI before treatments. Two observers independently calculated the tumor diameters and the parameters of DCE-MRI (including quantitative parameters: K^trans, K_ep and V_e; semi-quantitative parameters: TTP (time to peak), Maxconc(initial area under the gadolinium curve), AUC (Initial area under the gadolinium curve) and Maxslop, using both population and individual based arterial input function (AIF). Inter-observer consistency was evaluated using intra-class correlation (ICC) and Bland-Altman plots. Comparisons between EAC and AdC were performed via Independent sample t-test or Mann–Whitney U test according to the normal distributions of data. Receiver operating characteristic (ROC) curve and area under curve (AUC) were used to analyze the diagnostic efficiency of significant parameters and those combinations.

Results

Inter-observer reproducibility (ICC = 0.938-0.987) was excellent with relatively higher ICC and smaller SD on Bland-Altman plot, especially when calculated via population AIF(Fig.1). The tumor diameters were not correlated with tumor types. All the parameters, except K_ep derived from population AIF and TTP derived from both sets of AIFs, were lower in EAC than in AdC. AUC is 0.895 on V_e and 0.888 on Maxconc when calculated with population AIF, and is 0.915 when calculated from the combination of V_e and Maxconc. The diagnostic efficiency to differentiate EAC with AdC achieved an accuracy level of 85.00% with Maxconc alone, and a specificity level of 90.91% with the combination of V_e and Maxconc(Fig.2, Fig.3).

Discussion and Conclusion

The results of our study displayed significantly higher K^trans, Maxconc, AUC and Maxslop in both sets of AIF in AdC when compared with those in EAC, indicating that AdC had higher perfusion than EAC, which were consistent with previous studies . The measurements by population AIF showed fairly good agreement with the smaller SD and 95% limits of agreement on Bland–Altman plots and the standard deviations of mean values of quantitative parameters were relatively larger when using individual AIF, therefore, population AIF was recommended with better repeatability. Our results suggested that the two adenocarcinomas may have similar TTP. The original small number of patients with inconclusive adenocarcinoma by cervical biopsy, posed challenges to collect patients, especially with unified sequence. Therefore, we have to add some patients with distinctive EACs confirmed by endometrial curettage, which maybe lead to insignificance in morphological evaluation, and needed more case in further study. In conclusion, DCE-MRI could be used as a reliable supplementary MRI method for quantitatively evaluating the perfusion difference between the EAC and AdC, and to assist the diagnosis determination of uterine adenocarcinoma with uncertain biopsy pathology.

Acknowledgements

No acknowledgement found.

References

[1] Pecorelli S, Pasinetti B, Angioli R, et al. Systemic therapy for gynecological neoplasms: ovary, cervix, and endometrium. Cancer Chemother Biol Response Modif.2005, 22:515-544.
[2] Khoury T, Tan D, Wang J, et al. Inclusion of MUC1 (Ma695) in a panel of immunohistochemical markers is useful for distinguishing between endocervical and endometrial mucinous adenocarcinoma. BMC Clin Pathol.2006,6:1.
[3] He H, Bhosale P, Wei W, et al. MRI is highly specific in determining primary cervical versus endometrial cancer when biopsy results are inconclusive. Clin Radiol.2013, 68(11):1107-1113.
[4] Ramirez PT, Frumovitz M, Milam MR, et al. Limited utility of magnetic resonance imaging in determining the primary site of disease in patients with inconclusive endometrial biopsy. Int J Gynecol Cancer.2010, 20(8):1344-1349.
[5] Bourgioti C, Chatoupis K, Panourgias E, et al. Endometrial vs. cervical cancer: development and pilot testing of a magnetic resonance imaging (MRI) scoring system for predicting tumor origin of uterine carcinomas of indeterminate histology. Abdom Imaging.2015, 40(7):2529-2540.