The Usefulness of Diffusion Weighted and Contrast Enhanced Magnetic Resonance Imaging in Characterization of Inconclusive Ovarian Mass

Introduction: Ovarian cancer is the commonest reason for death in females due to gynecologic malignancy around the world. In contrast to other gynecologic tumours, the definitive diagnosis is accomplished days after of the index surgery by the histopathology. Preoperative assessment based on conventional MRI is not accurate. Information is expanding increasing about the ability of new MRI modalities to assess ovarian mass. Aim of the study: To assess the ability of dynamic contrast-enhanced MRI (DCE–MRI), and Diffu-sion-weighted image (DWI) to describe uncertain ovarian masses. Patients and Methods: This is a retrospective study. Patients were referred from ra-dio-diagnosis department and gynecology department of Qena faculty of medicine hospitals, South Valley University. Patients had uncertain adnexal masses at ultrasound. Magnetic resonance examination was doneutilizing 1.5 Tesla machine. The protocol included T1WI, T2WI, T1WI following contrast, and DWI. Results: We included 44 patients with different forms of complex cystic and solid ovarian masses. The final pathology of the ovarian masses was 18 benign, 4 borderline, and 22 malignant. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for DWI were 100%, 94.4%, 96.3%, 100%, and 97.7% respectively. The performance of DWI was higher than the conventional MRI and DCE-MRI. Conclusion: DCE-MRI and DWI have accepted ability to recognize malignant ovarian mass.


Introduction
Ovarian cancer is one of the commonest cancers in the world. It is the commonest cause of death in women due to gynecologic malignancy around the world. It has a 5 year survival of 40% [1]. In contrast to other gynecologic cancers, the definitive diagnosis is achieved days after of the index surgery. The diagnosis is by pathology. Preoperative diagnosis based on clinical, laboratory, radiological, and even frozen section is not free of errors [2]. Preoperative biopsy is not offered except when the planed treatment is neoadjuvant chemotherapy or in palliative settings [3].
DWI is another modality in MRI. It depends on the different microstructure, cellular density, and microvasculature of malignant tissue. In ordinary body structure, water molecules move freely. Water in malignant tissue has restricted ability to diffuse. The apparent diffusion coefficient (ADC) value is a quantitative measure of diffusion. The computer unit in the MRI machine changes these quantities into a user-friendly qualitative image. This is the DWI. Tissues with limited diffusion are white on DWI and hypointense on the ADC map [5].
Earlier reports on the ability of DWI to recognize malignant ovarian tumour have found that DWI is not useful [6] [7]. Later reports found that DWI is useful in discrimination between benign and malignant ovarian mass [8] [9] [10]. A more recent study found a sensitivity of 84%, and a specificity of 89% [11].
Information is expanding on the diagnostic ability of DCE-MRI and DWI. In this study, we expect to evaluate the ability of DCE-MRI, and DWI to describe uncertain ovarian masses.
Aim of the study: to assess the ability of dynamic contrast-enhanced MRI (DCE-MRI), and Diffusion-weighted image (DWI) to describe uncertain ovarian masses.

Patients and Methods
Design: a comparative study.

Methods
We did transabdominal ultrasound and transvaginal ultrasound for all cases. We used GE Logic P6 ultrasound machine (GE Healthcare Medical system; USA).
We pursue the International Ovarian Tumor Analysis (IOTA) rules to characterize ovarian mass [12]. MR assessment was done at the magnetic resonance unit, Qena faculty of medicine Hospital, South Valley University. We used 1.5 Tesla amachine with body coil as a transmitter and a receiver of radio frequency signals (Achieva, Philips Medical System; Netherland). The MR assessment included T1WI, T2WI, post-contrast fat-suppressed T1WI, and DWI. DWI was done at b0, b500, b1000. Descriptive analysis was done. Data from the MR assessment included the mean size of the cyst or mass, the ADC value, and the morphologic criteria suggesting malignancy. We had executed an individual analysis for conventional MRI, DCE-MRI and DWI concerning their diagnostic performance in the characterization of ovarian masses/cysts. Masses are sent for histopathology after operations.

Results
We investigated 50 patients with 50 adnexal lesions. Six masses were omitted, three of them due to loss of follow up. Unfortunately, we excluded one mass because of lack of pathology (patient died before operation). Two masses are excluded, as they were broad-ligamentary fibroid. The remaining 44 patients with 44 complex ovarian masses were included in our study. Benign masses included six serous cystadenoma, five mucinous cysadenoma, three mature cystic teratoma, two ovarian fibroma, and fibrothecoma, and one tubo-ovarian abscess. There were four Borderline tumors (two serous and two mucinous). There were 22 invasive malignant masses (seven Serous cyst-adenocarcinoma, five Mucinous cyst-adenocarcinoma, three Metastatic krukenburg, three Immature teratoma, two fibrosarcoma, and two clear cell carcinoma).
The analysis of the size of the lesions is appeared in Table 1. The malignant and borderline ovarian lesions were bigger than the benign lesions.  and 12 masses showed mixed signals (Figures 2-5).  Table 2). For the sake of statistical evaluation, we incorporated the borderline tumors into the classification of malignant tumors. The histopathological reported 26 malignant and 18 benign lesions. The performance of the preoperative imaging is outlined in Table 3. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for DWI were 100%, 94.4%, 96.3%, 100%, and 97.7% respectively. The performance of DWI was higher than the conventional MRI and DCE-MRI. Borderline and malignant ovarian tumors showed restricted diffusion. Benign tumors had facilitated diffusion (Figures 2-5).

Discussion
Ovarian carcinoma has the most elevated death rate among tumours of the female genital tract. Differentiation between benign and malignant ovarian tumor is vital to offer the appropriate plan for each case [3] [13].
Conventional MRI assesses morphologic criteria of the lesion, such as wall thickening, intra luminal papilla, mural nodules, thick septae, and signal intensity on T1WI and T2WI. None of these criteria can consistently segregate benign from malignant lesions [4]. Development of novel MRI modalities like DCE MRI and DWI improves the diagnostic performance of MRI [4].
We had executed an individual analysis for conventional MRI, DCE-MRI and DWI concerning their diagnostic performance in the characterization of ovarian masses/cysts. We found that conventional MRI had 88.5% sensitivity and 72.2% specificity. This looks well with a meta-analysis of the value of MRI in characterization of ovarian mass/cyst in women with non-conclusive ultrasound evaluation. They found that the sensitivity and specificity was 76% and 97%, respectively [14].
We found that DCE-MRI had 92.3% sensitivity and 88.8% specificity. This compares favourably to conventional MRI in our study. So, adding DCE to the MRI increased the accuracy of examination. Systematic review showed that DCE-MRI has 81% sensitivity and 98% specificity [14]. However, a more recent study showed 83% sensitivity and 75% specificity [15]. Malignant masses showed  more intense enhancement than benign lesions. Difference was clearer in the early phase of the contrast study than the late phase [16] [17].
Our analysis revealed that DWI has 100% sensitivity, 94.4% specificity, 96.3% PPV, 100% NPV, and 97.7% accuracy. The performance of DWI was higher than conventional MRI and DCE-MRI. We found that all malignant lesions and one  case of dermoid cyst demonstrated a high signal on DWI. This may be ascribed to keratinized substance in dermoid cyst. These results are consistent with the conclusions in the previous researches. They showed that most of the malignant ovarian masses and some of the dermoid cysts had high intensity on DWI. Most of the benign lesions had low signal intensity on DWI [6] [18] [19].

Conclusion
DCE-MRI and DWI have accepted ability to distinguish between benign and malignant ovarian mass.

Recommendations
We recommend adding DCE-MRI and DWI to the conventional MRI in the assessment of ovarian mass. This will aid in accurate diagnosis. Further researches are needed to build up cumulative evidence of the value of novel signals of MRI.

Authors' Approval
Authors read and approve the manuscript. All authors have fulfilled the requirements for authorship. Each author believes that the manuscript represents honest work. This manuscript is not considered for publishing anywhere else.

Conflicts of Interest
We have no conflict of interest to disclose. We received no funding.