Diffusion Weighted Magnetic Resonance imaging in Differentiation between Different Vertebral Lesions Using Apparent Diffusion Coefficient (ADC) mapping as a Quantitative Assessment Tool

INTRODUCTION

An MRI (Magnetic Resonance Imaging) scan is the best way to diagnose, stage, and follow up on Musculoskeletal diseases. When compared to traditional MR scans, Diffusion-weighted imaging (DWI) is a more recent development .¹ DWI provides vital information on cellular water flow in terms of both qualitative and quantitative metrics. It is the tiny self-diffusion of water molecules in tissues that controls DWI signal intensity. Organelles, macromolecules, and cell membranes limit the mobility of internal water molecules, yet these molecules can move more freely because of their increased flexibility. Water mobility at the molecular level will be restricted in locations where cellularity is high (as a result of which there are more intracellular ingredients and cell membranes). ²

Diffusion-weighted MRI is being effectively employed in the Central Nervous System (CNS),

particularly in detection of acute stroke as well as in the differentiation of distinct components of brain cancers. Outside the brain, diffusion measures of cancers have also been recorded.³

DWI is a quick MRI procedure that allows for the evaluation of many kinds of vertebral and discal lesions in the spine without contrast medium.⁴

Malignant lesions show primarily hyperintense in diffusion-weighted imaging of the spine. The presence of a hypointense signal in malignant vertebral lesions might be explained by irradiation lesions prior to imaging or sclerotic metastases with extremely low water content.⁵

In cases with degenerative alterations, DWI did not reveal significant signal intensity at the region of endplate anomalies.⁶

Diffusion weighted imaging may also be employed to monitor bone marrow neoplasms after treatment. When bone marrow illness is effectively treated, tumor cell  mortality causes greater water diffusivity, which is shown as higher ADC levels.⁷

Aims of the work: The goal of this research is to see how useful both qualitative and quantitative DWI assessments are in differentiating between various spinal tumors, as well as to find a cut-off ADC values that distinguishes benign from malignant tumors.

PATIENTS AND METHODS

Between February 2018 and October 2021, individuals described with or suspected of having spinal lesion of any cause were studied at the Radio-Diagnosis Department, Faculty of Medicine, Al Azhar University (Sayed Galal hospital), National Cancer Institute, and other radiological institutions. For this ethics committee-approved research, informed consent was acquired.

There were 50 patients in all (27 men and 23 females) ranging in age from 2 to 88 years old (average 49).

Inclusion criteria:

Patients diagnosed with vertebral lesion or vertebral bone marrow abnormality. Patients suspected to have vertebral lesion. Exclusion criteria including patients with pacemaker ,patients with claustrophobia.

The following was done to the patients, Out-patients were given a brief medical history; In-patients, on the other hand, were given more thorough clinical and laboratory data ,Radiological investigations using MRI examination (All patients were subjected to MRI examination). Some patients received IV contrast "gadolinium D.T.P.A" , Also plain X-ray, CT, or bone scan were done for some of the patients.

Pathological examination results were avail-able in some cases form which biopsy was taken during our study.

Technique for MR imaging was carried out using a standard 1.5 Tesla unit (Intera and Achieva, Philips) , sense spine coil was used  , Spine MRI sequences on a regular basis (Sagittal T1WI, T2WI & STIR , Axial T1WI , T2WI ).

After a manual iv gadolinium-DTPA injection of 0.1mmol/kg (in certain cases), contrast enhanced pictures were obtained.

DWI of spine free breathing and inversion recovery were used to conduct DWI utilizing single-shot spin-echo echo-planar sequences. A total of three diffusion-sensitive b-gradients of 0, 50, and 800s/mm2 were employed. The imaging process took six and a half minutes in total.

Image analysis using workstation was used to load all of the images (Philips Medical Systems). An examination of the bone marrow was performed, as well as imaging interpretation .

Qualitative analysis on T1-weighted images, the non-degenerated intervertebral disc signal intensity, sub-cutaneous tissue, and muscular tissue was visually compared to that of the bone marrow signal. We employed previously published diagnostic criteria for spinal bone marrow assessment for MRI interpretation.

Quantitative analysis measurements in Areas of Interest allowed the radiologist to estimate bone marrow signal strengths (ROI). Hyperintense lesions (with a "b 800" b-value), which correlate to aberrant signal intensity fluctuations on both T1 and T2 MRI, were manually delineated inside the ROI. If a patient had a diffuse vertebral marrow lesion, the ROI was drawn as big as feasible in order to prevent vertebral end plate degeneration and block access to the basivertebral vein plexus. Furthermore, by inserting ROI at the middle of the vertebral bodies, the ADC value of normal/normal seeming vertebral bodies was assessed. The diameter of the zones of interest ranged from 5 to 15mm. At least three ROIs were administered to each subject. The ADC values were automatically created employing software produced by the MR scanner provider, and the ADC quantitative feature was displayed in the following ranges: 10–3mm2/s (Philips Medical Systems). The average ADC value was computed and reported for each patient's 3 areas of interest. The "standard method" for classifying the spinal marrow tumors was the ultimate diagnosis, which was determined based on biopsy findings or clinical and radiologic follow-up outcomes.

RESULTS

A sum of 50 patients (27 men and 23 females) with ages ranging from 2 to 88 (average 49) were assessed radiologically for spinal lesion of any cause.

Classification of the vertebral bone marrow abnormalities:

The research comprised 26 patients with benign and 24 patients with malignant spinal marrow infiltrative tumors (Table 1). Clinical, radiographic, and histological evidence are used to make this diagnosis. Nine individuals had inflammatory/infectious spondylitis or spondylodiscitis, 11 had benign neoplastic marrow lesions, three had diffuse yellow to red marrow reconversion, and three had degenerative marrow alterations among the 26 patients with benign marrow lesions.

Metastatic lesions (n=17), lymphomatous infiltration (n=1), and multiple myeloma (n=1) were among the 24 individuals with malignant spinal marrow lesions that are infiltrative or multifocal.

Group                                                No. of patients

Inflammatory                                              9

Benign neoplastic lesion                          11

Malignant neoplastic lesion                       24

Red marrow                                                3

Degenerative                                               3

Table 1: Classification of the vertebral bone marrow abnormalities.

Inflammatory lesions:

On DWIs, all lesions in the inflammatory group revealed increased signal intensity. The median ADC value of pyogenic spondylodiscitis was high (mean; 1.454 X 10–3mm2/s, SD ±0.16 X 10–3mm2/s) while TB spondylodiscitis were of lower ADC values (mean; 1.191 X 10–3mm2/s, SD ±0.29 X10–3 mm2/s) (Table 2).

Table 2: DWI signal and ADC values in inflammatory lesions.

Benign neoplastic lesions:

A total of four cases with benign neoplastic vertebral lesions were found: Three cases with hemangioma and one case with osteoblastoma.

On DWIs, hemangiomas revealed lower signal intensity while the osteoblastoma revealed higher signal intensity. The median ADC value of the hemangioma was high (median; 1.406 X 1 0–3mm2/s, SD ±0.13 X 10–3mm2/s) while osteoblastoma was of lower ADC value (0.818 X 10–3mm2/s) (Table 3).

Table 3: DWI signal and ADC values in benign neoplastic lesions.

Malignant Neoplastic lesions:

A total of nineteen cases with malignant verte-bral lesions were found: Seventeen patients with metastatic lesions, a case of lymphoma and a case of multiple myeloma. The case of lymphoma and one of the patients with metastasis were presenting for follow-up after receiving therapy.

On DWIs, eighteen malignant lesions revealed high signal intensity while the case of multiple myeloma revealed low signal intensity. The median ADC value of the non-treated malignant tumors was (median; 0.982 X 10–3mm2/s, SD ±0.17 X 10–3 mm2/s) while the two lesions under therapy were of higher ADC values (1.612 & 1.483 X 10–3 mm2/s) (Table 4).

Table 4: DWI signal and ADC values in malignant neoplastic lesions.

The ADC obtained values from the ADC maps generated automatically. (Table 7) vary substantially (p-value <0.001, highly substantial) between malignant (median, 0.982 X 10–3mm2/s; SD ±0.17 X 10–3mm2/s) and the following benign marrow entities; inflammatory pyogenic marrow tumors (median, 1.454±0.322 X 10–3mm2/s; SD ±0.16 X 10–3mm2/s) and the degenerative marrow changes (median, 0.569 X 10–3mm2/s; SD ±0.06 X 10–3 mm2/s). ADC values of malignant lesions also differed substantially (p-value 0.002, substantial) from the hemangiomas (mean, 1.406±0.13 X 10–3 mm2/s; SD ±0.16 X 10–3mm2/s).

The ADC values of malignant tumors and TB spondylodiscitis were found to be similar(median, 1.191 X 10–3mm2/s; SD ±0.29 X 10–3 mm2/s), yet with a statistically substantially variance (p-value 0.034, significant) in their ADC values. An overlap between the ADC values of malignant tumors and the red marrow (median, 0.613 X 10–3 mm2/s; SD ±0.08 X 10–3mm2/s) was also detected, yet with a statistically significant difference (p-value 0.002, significant) in their ADC values.

The ADC value of osteoblastoma (0.818 X 10–3mm2/s) was within the range of ADC values of the malignant bone marrow lesions.

The ADC values of post therapy neoplastic lesions "metastasis" 1.612 X 10–3mm2/s and "lymphoma" 1.483 X 10–3mm2/s were greater than those of pretherapy malignant lesions (median, 0.982 X 10–3mm2/s; SD ±0.17 X 10–3mm2/s).

Receiver Operating Characteristics (ROC) curve was utilized to define the best cut off value of ADC (10–3mm2/sec) among benign and malignant lesions which was < 1.237, With a sensitivity of 100 % and a specificity of 61.9 %, the predictive accuracy of 66.7 is 66.7 percent, the negative prediction accuracy is 100 percent, and the diagnostic accuracy is 72 percent.

( A )                              ( B )                              ( C )                                 ( D )                                   ( E )

Fig. 1: 45 year old female known to have breast cancer presenting with back pain. MRI of the lumbosacral spine revealed multiple metastatic deposits: ( A ) T1, ( B ) T2, ( C ) STIR ( D ) DWI and ( E ) ADC sagittal images revealing diffuse alteration of the bone marrow signal eliciting low signal on both T1 & T2 WIs with multiple p–a3tchy 2areas of bright signal intensitieson STIR images and patchy areas of restricted diffusion. ADC value: 1.061 X 10 mm /s.

Fig. 2: 64 year old male known to have multiple myeloma presenting with back pain. MRI of the dorsolumbar spine: ( A ) T1,( B ) T2, ( C ) STIR ( D ) DWI and ( E ) ADC sagittal images of a case of multiple myeloma revealed diffuse alteration of the bone marrow signal of the lumbar vertebrae eliciting decreased signal on both T1 & T2 WIs and rather diffuse increased signal intensity on STIR images with no areas of restricted diffusion noted. ADC value: 0.975 X 10 –3 mm2/s

                                                                                (F)

Fig. 3: 39 year old female with neck pain and right upper limb brachialgia. MRI of the cervical spine: ( A ) T1,  ( B ) T2,( C ) STIR, ( D ) DWI , ( E ) ADC sagittal images & ( F ) axial T2* image of a case of osteoblastoma of D1 right transverse process revealed a well defined lytic expansile lesion of right transverse process of D1. The lesion elicits isointense signal on both T1 & T2 WIs and bright signal on STIR images with

a corr–e3spon2ding area of restricted diffusion. ADC value: 0.818 X 10 mm /s

Fig. 4: 60 year old male "prisoner" presenting with back pain. MRI of the lumbosacral spine: ( A ) T1, ( B ) T2, ( C ) STIR ( D ) DWI and ( E ) ADC sagittal images of a case of pyogenic spondylodiscitis. Sagittal T1-weighted image shows diffuse low signal intensity of L3 and L4 vertebrae. The sagittal T2-weighted picture and sagittal STIR demonstrate a strong association.L3 and L4 vertebrae signal intensity, as well as the intervening L3/4 disc, with L3 and L4 vertebrae and intervening disc diffusion restriction. ADC value: 1.369 X 10 –3 mm2/s.