Role Of Magnetic Resonance Imaging With Diffusion-Weighted Images In Assessment Of Patients with Pre-eclampsia In Pre And Postpartum Periods

INTRODUCTION

Preeclampsia (PE) is a common complication that affects around 5% of pregnant women. It is typically present after 20 weeks of pregnancy with microvascular hemolysis, decreased platelet level, increased blood pressure, proteinuria, and impaired renal function.^1,2

The exact PE etiology is still not well-known. Nevertheless, it was proposed that the disease is due to the ischemic placenta that arises from an abnormal remodeling of the spiral arteries between the maternal, and placental arteries.^3,4

 It is considered as one of the principal causes for maternal, fetal, and neonatal mortality with or without headache, dizziness, and nausea.⁵

 Severe PE, which develops before 34 weeks of gestation, is characterized clinically as early-onset type.⁶

In 10% of women with PE, there are mental adverse effects, while in women with early-stage disease, this risk increases to 15%.⁷

If left untreated, pregnant women who have PE will be seriously affected by eclampsia, stroke, liver rupture, lung edema, or insufficient renal failure.⁸ Moreover, the risk of bronchopulmonary dysplasia and cerebral paralysis in children who come to mothers with PE is higher, caused by early delivery.^{9, 10}

Neurological complications can also differ because of the differing pathogenesis between PE two types; however, previous studies have focused on neurological complications of patients suffering from PE, but there is no research into comparing the neurological complications of early-onset patients against patients suffering from late-onset PE.¹¹

Neuroimaging's diagnostic value for neurological diseases is self-evident; however, neuroimaging was criticized in pregnant patients because of the potential risk to fetus from computed tomography (CT).¹² However, the modality has become widely applied in pregnant patients and has increased the detection rate greatly with continuous improvement in magnetic resonance imaging (MRI).¹³

However, the standard of selecting the imaging methods for various neurological manifestations is still not consistent. In addition, the clinical factors that influence imaging outcomes are still controversial. Recently, diffusion-weighted MRI (DW-MRI) introduced into clinical practice as a non-invasive method that can detect cerebral infarction hours before changes become apparent.¹⁴

Moreover, it can differentiate between cytotoxic edema form vasogenic edema. However, its predictive value in predicting the neurological complications of PE is not established yet.¹⁵ Therefore, in this study, we aim to evaluate the predictive value of DWI in predicting the neurological manifestation of PE.

PATIENT AND MATERIALS

The present manuscript was approved by local ethics committee and prepared in concordance with the recommendations of Strengthening the Reporting of Observational studies in Epidemiology (STROBE).¹⁶

Study Design and Patients:

A prospective study was conducted on 44 patients presented from the Obstetrics department with peri-partum pre-eclampsia/eclampsia manifestations as confirmed by clinical examination and laboratory findings- and they were referred to the radiology department during the period from January 2019 to December  2019. Follow up was done within 10-14 days for patients with restriction of diffusion in DWI. We excluded patients with poor diagnostic quality images due to patient non- cooperation; patients with absolute contra-indications for MRI examination such as patients with metallic surgical clips, joint prothesis, metallic bone plates, or patients who have cardiac pacemakers or surgical clips in or around the eyeballs; and/or patients with history of claustrophobia.

MRI Examination and Data Collection

Eligible patients underwent detailed history taking regarding the onset and course of pre-eclampsia and of associated symptoms as well as signs. All patients were investigated for laboratory parameters such as renal function tests, platelet count, liver function tests, coagulation profile and fundoscopy. Then, all patients were examined on a 1.5T GE signa Excite machine (GE Healthcare, Waukisha, WI). The range of days between the first presentation and having the MRI study was 2-3days.

The examination included a customized brain MRI study composed of  FLAIR, T1 weighted, T2 weighted images and axial DWI. Follow up MRI followed the same sequences. Gadolinium enhanced T1 weighted image and magnetic resonance venography (MRV) were done for patient with venous infarction for confirmation of the diagnosis (Table 1).

Table 1: MRI Sequences Parameters

Study’s outcomes:

The primary outcome in the present study was the MRI and DWI features of patients presented with neurological complications of pre-eclampsia.

Statistical Analysis:

SPSS version 19 (SPSS Inc., Cary, NC) was used for the statistical analysis. Descriptive analysis was used, first, including mean and standard deviation for quantitative data, e.g. age, systolic and diastolic blood pressure. Frequency tables were used for qualitative data, e.g. number of presenting symptoms and signs. Independent sample T-test was used to compare the difference of means in quantitative data, i.e. age, systolic and diastolic blood pressure. Chi square test was used for qualitative data. Spearman correlation test was used to test the correlation between the occurrence of fits and blood pressure.

RESULTS

The mean age of the included patients was 26.9 ±2.9 years old. The admission systolic blood pressure and diastolic blood pressure were 181.8 ±15.2 and 97.7 ± 8.6mmHg, respectively. Twenty-nine patients (65.9%) had neurological manifestations of pre-eclampsia at time of admission. Fits was the most common presenting symptom. It was recorded in 86.7 % of Patient. Altered conscious level was noted in 36.4% of patients. Visual disturbance (including blurring of vision) accounted for 40.9% of patients. Comparing the clinical data between patients with and patients without neurological manifestation, it was found that there was a significant difference between both groups regarding the mean systolic and diastolic blood pressures as well as the proteinuria p <0.05; (Table 2).

Patients with neurological manifestations, revealed normal MRI findings in 7 patients (24.1%) and abnormal features in 22 patients (75.9%).Patients with neurological manifestations who revealed abnormal MRI were classified according to the radiological findings in our study into patients with typical posterior reversible encephalopathy syndrome (PRES) accounting for 27.5%, patients with atypical PRES accounting for 41.3%, one patient with venous infarction accounting for 3.4% and also one patient with subarachnoid hemorrhage accounting for 3.4%. All patients with atypical PRES had restriction of diffusion figure (1).

In addition, we found that the parietal lobe was the most frequently affected lobe, followed by the occipital lobe. The parieto-occipital regions were the most commonly affected areas and represented the typical site of PE table (3).

Patient`s characteristics were presented in (Table-1). There was no statistically significant difference between both groups regarding age, sex, BMI, comorbidities and urologic surgical history, (p-value >0.05) The stone characteristics are shown in table(2).

In this study, complete staghorn stones represented (24%) in OSS group and (20%) in PNL group while partial staghorn stones represented (67%) in OSS group and (80%) in PNL group with no significant difference.

In OSS group, the used technique was extended pelolithotomy in 20 cases, pyelolithotomy plus one or two nephrotomies in 4 cases and anatrophic nephrolithotomy in 1 case. In PNL group, the access was achieved through one punctured calyx in 22 cases and through two calyces in 3 cases.

There is statistically significant difference between both groups regarding to the operative time with shorter duration for OSS group (131.48 min +10.2) vs (174.63 min +14.9) (p-value <0.001).

 Mean creatinine rise in OSS group was (0.19 ± 0.1 mg/dl) and in PNL group was (0.14 ± 0.07 mg/dl) which was statistically significant (p-value = 0.04). Regarding mean operative Hb loss, it was (1.5 ± 0.85 mg/dl) in OSS group versus (1.65 ± 0.90 mg/dl) in PNL group with no significant difference table (7).

There is no statistically significant difference between study groups as regards the different complications, either intraoperative or postoperative and also according to modified Clavien postoperative grades of complications table (8).

Stone-clearance in OSS group was (92%) and in PCNL group was (84%) with no significant difference.

Hospital stay and recovery time were significantly shorter in PNL group than in OSS group.

Comparing the gray and white matter involvement with the diffusion characteristics, it was noted that most cortical involvement revealed restricted diffusion of cytotoxic edema, all white matter involvements revealed enhanced diffusion of vasogenic edema, while mixed grey and white matter involvement revealed mixed diffusion properties (vasogenic and cytotoxic) with the cytotoxic edema located mainly in the grey matter (p<0.0005) table (4).

The ADC values recorded from the regions of cytotoxic edema ranged from 0.12 to 7.55 x10-11, with an average of 3.95 x 10-11.The ADC value recorded from the single venous infarction lesion was 6.98 x10-12. All patients with DWI evidence of cytotoxic edema underwent follow up MRI studies after 2 weeks. The lesions were fully reversible except in one patient. The residual infarct in this patient was much smaller than the initial area of restricted diffusivity.  The ADC value of the irreversible lesion was 6.2 x10-11, while the mean ADC value of reversible lesions was 3.95 x 10-11. There was no significant difference between both numbers (p = 0.627).

Figure 1 : (a & b) Axial FLAIR images showing high signal intensity area in posterior parietal regions at both sides involving deep white matter and extend to involve cortex as well as hyperintense signal area involving  basal ganglia at both sides. (c) Axial T2WI showing hyperintense signal in posterior parietal region. (d) Axial DWI showing hyperintense signal in posterior parietal region. (E) ADC map which shows reduced ADC value.  (F) Axial FLAR Follow up is normal. These features are consistent with PRES with both typical (bilateral posterior parietal regions) and atypical sites (basal ganglia). Restriction of diffusion denoting cytotoxic edema with ischemic area. With complete resolution in follow up studies.

Figure 2: (a) Axial FLAIR shows abnormal high signal intensity area seen involving both parietal regions. (b) Axial T21WI shows abnormal high signal intensity areas in parietal region. (c) Axial DWI shows abnormal high signal intensity area seen involving the parietal region confirmed on the (d) ADC map which shows reduced ADC value.  Interpretation: abnormal signal intensity areas are consistent with PRES which shows typical regional distribution (in parietal region). On DWI not all abnormal signal intensity areas seen in FLAIR images show restriction in diffusion denoting the presence of two types of edema (vasogenic and cytotoxic).

Table 2: The presenting symptoms in patients with pre-eclampsia (number and percent):

Table 3: The distribution of signal abnormalities among the anatomical regions of the brain in relation to the number of the patients with PRES whether typical or atypical:

NB: Bilateral affection of the same anatomic region was considered as a single region involvement.

Table 4: The distribution of grey and white matter involvement among patients with PRES (n = 20)