[Analysis of correlation between magnetic resonance imaging features and prognosis of invasive pituitary adenoma]

Objective: To explore the application of conventional nuclear magnetic resonance imaging, diffusion-weighted imaging and dynamic contrast-enhanced imaging in invasive pituitary adenoma, and to study the correlation between its imaging features and prognosis. Methods: The imaging manifestations (including T1WI, T2WI, ADC and dynamic contrast-enhanced MR), clinical features and pathological data of 47 cases of invasive pituitary adenomas (IPA) in Tianjin Huanhu Hospital from June 2017 to June 2018 were retrospectively analyzed, and compared with 16 cases of non-invasive pituitary adenoma (NIPA). The differences in tumor/white matter signal intensity ratio on T1WI and T2WI images, MR signs, ADC values and time signal-curve type, peak enhancement time and average enhancement rate between the two groups were observed. Postoperative MR enhanced scan and related hormone results were reviewed to evaluate the prognosis of all patients. Logistic stepwise regression analysis was used to analyze the relationship between IPA imaging characteristics and prognosis. Results: The location, shape, size, Knosp classification and ADC value of tumors were significantly different between IPA and NIPA groups (χ(2) values were 41.344, 21.818, 28.596 respectively, t values were 5.869, 11.720 respectively, all P<0.05). The majority of T1WI, T2WI and enhanced heterogeneous signals occurred in IPA group (P<0.05). IPA showed two-stage enhancement characteristics of rapid enhancement and fast subsidence phase. There were significant differences in peak enhancement time between intrasellar tumors and invasive cavernous sinus tumors. There were no significant differences in signal intensity of T1 and diffusion limitation and two-stage enhancement characteristics between the two groups (all P>0.05). There was a significant correlation between the ratio of tumor body/white matter signal on T2WI, ADC value and the constitution of invasive pituitary tumors (all P<0.05). In IPA, 9 cases(19%) were totally resected, 25 cases(53%) subtotally resected and 13 cases(28%) partially resected. Postoperative symptoms were improved in 29 cases(62%), alleviated in 6 cases(13%), unhealed in 3 cases(6%), assisted drug therapy in 7 cases and radiotherapy in 15 cases. All cases were followed up for 7 to 16 months, with a median time of 12.5 months. Eight cases recurred, with a recurrence rate of 17%. There were significant differences in the total/subtotal resection rate, the improvement rate of symptoms and the need for adjuvant treatment among IPA with different texture (all P<0.05).Knosp grade, tumor size and the ratio of tumor body to white matter on T2WI were associated with IPA recurrence (all P<0.05). Conclusion: It is helpful to diagnose IPA by analyzing the distribution, shape, size, ADC value and signal difference of tumors in sella turcica region through MR and DWI signs. Combined with dynamic contrast enhancement features, the invasion of IPA and its surrounding tissues can be early judged, and the texture of the tumors can be assessed by imaging signs, so as to better evaluate the therapeutic effect and prognosis of IPA.