Comparison and evaluation of vertebral metastasis with MRI imaging and radionuclide bone scan

 

 

Abstract               Among the primary malignant tumors, most of the tumors spread into the spinal column, and few of the important primary neoplasm’s causing spinal metastasis are carcinoma prostate, breast, lung, thyroid and renal neoplasms. Carcinoma prostate and breast malignancy are the most common malignant neoplasm causing vertebral metastasis, constituting 47 to 85 % of among skeletal metastasis. Dissemination of malignant cells into the spinal column may occur through the venous system, occasionally through arterial system or through the batsons plexuses. Spinal metastasis account for 39 % of skeletal metastasis. The aim of this study to compare the efficacy of MR Imaging and Radionuclide bone scan in detection of vertebral metastasis. Radionuclide bone scan can detect the osteoblastic activity and in this way this modality is more sensitive in detection of cortical metastasis as compared with intramedullary metastasis. MRI is very sensitive in detection of intramedullary metastasis of vertebrae.

Key Words: Vertebral metastasis, MR Imaging, Radionuclide bone scan, Carcinoma breast, Carcinoma prostate

 

 

INTRODUCTION

Deposition of malignant cells into the bone cause destruction of bone and reactive new bone formation. Hydroxyapatite crystals are deposited in the osteoid matrix in new bone formation and in which osteoid matrix are formed by osteoblast. Radionuclide bony scan detect the activity of new bone formation. Bone lesions causing active new bone formation which also include fracture, infection and other processes.² Early bone metastasis start in the marrow, so small pure intramedullary deposit are not detected by bone scan due to insufficient osteoblastic activity ³. MR imaging is sensitive in detecting the intramedullary deposits and MRI has limited use due to its cost, longer duration and time. Bone marrow is filled with higher percentage of fat components with hemopoietic component. It is readily imaged by T1 W images. Tumor tissue appear hypo intense on T1 W images and hyper intense on STIR sequences. MRI Imaging can be a alternative to detect the early intramedullary vertebral metastasis and early initiation of treatment.

 

MATERIAL AND METHODS

Patients of all age groups and with proven malignancy and clinically suspected cases of metastasis were included in this study. All the patients who were studied have undergone a detailed clinical evaluation, a skeletal survey and other relevant investigations prior to Magnetic Resonance Imaging and bone scan. Pregnancy, patient with claustrophobia, benign bone islands, hemangiomas, patient with pacemaker were excluded from this study. Studies of non cooperative patient causing suboptimal images were also excluded. MR imaging and bone scan were done within 15 days of each others. Total of 120 patients were studied. These studies were conducted in the department of Radiology in collaboration with Nuclear medicine department. All parameters like size, site of the lesions were studied and compiled and analyzed.

Conventional whole-body scinti graphy were performed 3 hours after intravenous administration of 20 mCi (740 M Bq) or Tc-99m MDP on Siemens ECAM (Hi Definition) dual head gamma camera to be used for the whole body dynamic scan and requisite static images. No specific patient preparation was required. Jewelry, belts, change, or external prostheses and braces to avoid artifacts were removed before taking the scan. Three hour delayed images were obtained over the entire body in the anterior and posterior positions in all patients.MR Imaging was done on a Siemens 1.5 Tesla super conductive magnet. Axial, coronal and sagittal images of spine done. T1W, T2 W, STIR sequences were acquired.

 

Table 1:

TR = Time of repetition TE= Time of echo TI = Time of Inversion FOV=Field of view

Bone scan regions were read positive using the accepted subjective criteria such as the increased intensity of uptake as compared with adjacent structure, number, location and pattern of distribution. Only the regions examined by both Magnetic Resonance Imaging and scintigraphic studies were analyzed.MR images were read positive if a focus of low signal intensity on T1-weighted images and high signal intensity on STIR or T2-weighted images and enhancement in post contrast study. Lesions were further grouped according to the site, size (greatest dimension of lesion) and location like cortical or intramedullary and sub cortical.

 

RESULTS

Total of 120 cases of vertebral metastases were finally included and the findings were analyzed and compiled.

Of the 120 patients studied, the youngest patient was aged 17 months and the eldest was 75 years. Majority of the patients (30.0%) fell into the age group of 51-60 years. The second commonest age group was 41-50 years (33%)

 

Table 2: Age distribution of the study population

The study included 64 males and 56 Females, forming 53% and 47% respectively.

In our study of 120 patients with vertebral metastases, the number of cases in each primary carcinoma is given in the underlying table ( Table no : 3)

Table 3: Primary Carcinomas

The commonest was carcinoma breast with 38 cases (32%). The second commonest was carcinoma prostrate with 32 cases (27%) and the third commonest was lung carcinoma with 9 cases (15%).

4. Age Wise Distribution of Primary Carcinomas

Table 4 shows the age wise distribution of the primary carcinomas. Commonest age group for primary carcinomas was 51-60 years with 42 cases (35%) and the second commonest was 41-50 years with 40 cases (33%). Carcinoma breast was commonest in the age group between 41-50 years with 14 cases (37%). Carcinoma prostrate was commonest in the age group between 51-60 years with 14 cases (44 %). Carcinoma lung was commonest in the age group between 41-50 years with 6 cases (33%).

Table 4: Age wise distribution of primary carcinomas

 

5. Analysis of results of Bone Scan Readings as per site.

In all 120 patients, each region was scored as involved or not involved by metastases depending on whether the bone scan readings were positive or negative. The commonest region positive for metastatic involvement, in bone scan, was lower lumbar region with 66 (55%) positive regions and the second commonest was upper lumbar region with 60 positive regions (50%).

6. Analysis of MRI findings as Per Region

Result of all 120 patients were compiled and analyzed. The readings are given in Table 5. The commonest region positive for metastatic involvement, in MRI was lower lumbar region with 92 (77%) positive regions. The lower thoracic and upper lumbar regions were the second commonest with 86 (72%) positive regions each.

 

Table 5: MRI readings as per region

 

7. Comparison of Regions Read Positive and Negative on Bone Scan And MRI

Table 6 summarizes the comparison of regions read positive or negative on bone scans and MRI.

Table 6: Comparison of regions read positive and negative

BS-Bone scan, MRI- Magnetic Resonance Imaging

Of the 720 regions, 266 regions (37%) were read positive on both bone scan and MRI and 298 regions (41%) were read negative on both. 16 regions (2.0%) were read positive on bone scan and negative on MRI and 140 (20%) regions were read positive on MRI and negative on bone scan.

Total number of vertebrae involved by metastases

In the 120 patients evaluated in our study, a total of 3480 vertebrae were studied. Table 12 shows the number of vertebrae positive and negative on bone scan and MRI respectively.

 

Table 7: Number of vertebrae involved by metastases

2732 (79%) vertebrae were read as negative on both modalities. 748 (22%) vertebrae were read as positive on MRI or bone scans. 728 (3.0%) vertebrae were positive on MRI and 568 (16.0%) vertebrae were positive on bone scan. 548 (16%) vertebrae were read as positive on both bone scan and MRI. 180 (5%) vertebrae were read as positive on MRI and negative on bone scan. 20 (0.6%) vertebrae were read as positive on bone scan and negative on MRI. The size of each lesion in 364 vertebrae, that were read as positive on MRI, was measured at its greatest dimension on MRI. The lesions were categorized as small (<2 cm) or large (>2 cm) and the results are tabulated in Table 8.

 

Table 8: Correlation of positive bone scans with lesion size on MRI

X² chi square value = 301.69 Degrees of freedom = 15 P value = 0.0000

548 vertebrae (75) were read as positive on bone scan and 180 vertebrae (25%) negative. 164 small lesions (22.5%) and 564 large lesions (77.5%) were identified in the 728 vertebrae. Out of the 164 small lesions, only 4 lesions (2%) were read as positive and 160 lesions (98%) were read as negative. Out of the 564 large lesions 544 lesions (96) were read as positive and 20 lesions (4%) were read as negative.

Correlation of positive bone scan with lesion location on MRI

The lesions in all 728 vertebrae that were read as positive on MRI, were classified as intramedullary, sub cortical and transcortical. This was based on the relationship between the lesion and cortical bone. The results are tabulated in Table 14. 548 vertebrae (75) were read as positive on bone scan and 180 vertebrae (26) negative. 140 intramedullary (19%), 152 sub cortical (21%) and 436 transcortical, (60.5) lesions were identified in the 428 vertebrae. Out of the 140 -intramedullary lesions, none of the lesions (0%) were positive on bone scan. Out of the152 sub cortical lesions, 116 lesions (76%) lesions were positive on bone scan and 36 lesions (24) were negative. Out of the 436 -transcortical lesions, 432 lesions (99%) were positive on bone scan and 4 lesions (0.9%) were read as negative.

Table 9: Correlation of positive bone scans with lesion location on MRI

Figure 1:

Figure 2:

Figure 3

DISCUSSION

In our study, 120 patients with a proven malignancy and clinical and laboratory suspicion of vertebral metastases were included between the ages of 17 months to 75 years. The mean age of the patients was 43 years and 64 males (53) and 56 females (47%) formed part of our study. The commonest age group of patients with vertebral metastases was 51-60 years. This was in consonance with the finding of various authors as mentioned in Table 10.

 

Table 10: Commonest age group of patient with vertebral metastases

The commonest primary tumor in our study was carcinoma breast with 38 cases (32%). Various authors have also reported similar frequency of primary tumors, as mentioned in Table 11.

 

Table 11: Frequency of commonest primary tumor

The second commonest tumor in our study was carcinoma prostrate with 32 cases (27%). Carcinoma lung was the third commonest with 18 cases (15%). Carcinoma breast was commonest in the age group between 41-50 years with 14 cases (37%). Carcinoma prostrate was commonest in the age group between 51-60 years with 14 cases (44%). Carcinoma lung was commonest in the age group between 41-50 years with 6 cases (33%).Out of the total 720 regions, bone scan was positive in 282 regions (39%) and negative in 438 regions (61%). The commonest region positive for metastatic involvement, in bone scan, was lower lumbar region with 66 (55) positive regions and the second commonest was upper lumbar region with 60 positive regions (505). This was comparable to the previous studies as shown in Table 12.

 

Table 12: Commonest region positive for metastases in bone scan

Out of the total 720 regions, MRI was positive in 406 regions (56%) and negative in 314 regions (44%)

 

Table 13: Commonest region positive for metastases in MRI

 

The commonest region positive for metastatic involvement, in MRI, was lower lumbar region with 92 (77%) positive regions, this was comparable to the previous studies as shown in Table 13. The lower thoracic and upper lumbar regions were the second commonest with 86 (72%) positive regions each, whereas in other studies upper lumbar region was the only second commonest region. Of the 720 regions 266 regions (37%) were read positive on both bone scan and MRI and 298 regions (41%) were read negative on both. 16 regions (2%) were read positive on bone scan and negative on MRI and 140 (20%) regions were read positive on MRI and negative on bone scan.A total of 3480 vertebrae were evaluated in our study of 120 patients. 2732 (79%) vertebrae were read as negative on both modalities. 1748 (22 %) vertebrae were read as positive on MRI or bone scans. 728 (21%) vertebrae were positive on MRI and 568 (16%) vertebrae were positive on bone scan. 548 (16 %) vertebrae were read as positive on both bone scan and MRI. 180 (5%) vertebrae were read as positive on MRI and negative on bone scan. 20 (0.6%) vertebrae were read as positive on bone scan and negative on MRI. MR imaging is a sensitive method for detecting the intra medullary vertberal metastases⁽⁹⁾. However, it is not cost-effective in examining bones with small marrow cavities, because it cannot examine the entire skeletal system in single study. The extra-vertebral metastases (detected by bone scan) were therefore not evaluated by MRI in our study. The sensitivity of MRI in respect of bone scan, in detection of vertebral metastases, was 97% and the specificity was 94%. This was comparable to the previous study of Eustace S et al [8], who also reported a sensitivity of 96.5%

Correlation of positive bone scan with lesion size on MRI

The size of each lesion in 728 vertebrae, that were read as positive on MRI, was measured at its greatest dimension on MRI and the lesions were categorized as small (<2 cm) or large (>2 cm). 164 small lesions (23%) and 564 large lesions (78%) were identified in the 728 vertebrae. Out of the 164 small lesions, only 4 lesions (2 %) were read as positive and 160 lesions (98%) were read as negative. Out of the 564 large lesions, 554 (97%) were read as positive and 20 lesions (3 %) were read as negative. Percentage of small sized lesions diagnosed by bone scan was significantly lower than the percentage of large size lesions diagnosed. Pearson’s chi-square test (x² Chi square value = 301.69) confirmed a statistically highly significant correlation between the lesion size and bone scintigraphci results (p value = 0.0000). this was in consonance with previous study of Taoka T et al ¹.

Correlation of positive bone scan with lesion location on MRI

The lesions in all 728 vertebrae, that were read as positive on MRI, were classified as intramedullary, subcortical and transcortical. This was based on the relationship between the lesion and cortical bone. 140 intramedullary (19%). 152 subcortical (21%) and 426 transcortical (60%) lesions were identified in the 728 vertebrae. Out of the 140 -intramedullary lesions, none of the lesions (0%) were positive on bone scan. Out of the 152 subcortical lesions, 116 lesions (76 %) lesions were positive on bone scan and 36 lesions (24 %) were negative. Out of the 436-transcortical lesions, 432 lesions (99%) were positive on bone scan and 4 lesions (0.95) were read as negative. Percentage of intramedullary and subcortical lesions diagnosed by bone scan was significantly lower than the percentage of transcortical lesions diagnosed. Pearson’s chi-square test (X² Chie square value + 301.69) confirmed a statistically highly significant correlation between the lesions size and bone scintigraphic results (p value = 0.0000). this was in consonance with previous study of Taoka T et al.¹ Our study indicate that cortical involvement is the key factor contributing to the difference in detection of vertebral body metastases on MRI and bone scintigraphy and that MRI is a sensitive and specific modality in detection of vertebral metastases.

 

 

CONCLUSION

Total of 120 patients (between the ages of 17 months of 75 years) with a proven malignancy and clinical and laboratory suspicion of vertebral metastases were evaluated on radionuclide bone scan and MR imaging. The commonest age group of patients with vertebral metastases was 51-60 years. The commonest primary tumor was carcinoma breast (38 patients). Carcinoma prostrate was the second commonest (32 patients) and carcinoma lung the third commonest (18 patients).Of the total 720 evaluated regions, bone scan was positive in 282 regions and negative in 438 regions. The commonest region positive for metastatic involvement, in bone scan, was lower lumbar region (66 regions). Of the total 720 evaluated regions, MRI was positive in 406 regions and negative in 314 regions. The commonest region positive for metastatic involvement, in MRI, was lower lumbar region (92 region) . Out of the total 3480 vertebrae that were evaluated in our study, 748 vertebrae were read as positive on MRI or bone scans. Of these 748 vertebrae, 712 (97%) vertebrae were positive on MRI and 568 (76 %) were positive on bone scan. The sensitivity of MRI as against bone scan, in detection of vertebral metastases, was 96.5% and the specificity 94%.The correlation of positive bone scan with lesion size of MRI was found to be statistically highly significant (p value =0.0000). percentage of small sized lesions diagnosed by bone scan was significantly lower than the percentage of large size lesions diagnosed. The correlation of positive bone scan with lesions location on MRI was found to be statistically (p value = 0.0000). percentage of intramedullary and subcortical lesions diagnosed by bone scan was significantly lower than the percentage of transcortical lesions diagnosed. MR imaging was found to have a higher sensitivity and specificity as against radionuclide bone scan in early detection of vertebral metastases.

 

REFERENCES

• Taoka T, Mayr N, Lee HJ, Yuh WTC, Simonson TM. Factors influencing visualization of vertebral metastases on MR Imaging versus boenscintigraphy. AJR. 2001; 176: 1525-1530.
• Dickison F, Liddicoaat A, Dhingsa R. MRI versus radionuclide scintigraphyofor screening in bone metastases. ClinRadiol. 2000; 55 (8):653
• Attenhoefer C. Ghanem M, Moger E. Comparative detectability of bone metastases and impact on therapy of MRI and bone scintigraphy in patients with breast cancer. Eur J Rardiol. 2001; 40: 16-23.
• Avrahami E, Tadmor R, Dally O, Hadar H. Early MR demonstration of spinal metastases in patients with normal radiographs and CT and radionuclide bone scans. J Comput Assist Tomogr. 1989; 13:598-602.
• Frank JA, Ling A, Patronas NJ. Detection of malignant bone tumors: MR imaging vs scintigraphy. AJR. 1990 Nov; 155 (4):1043-48.
• Gosfield E III. AlaviA, Kneeland B. Comparison of radionuclide bone scans and Magnetic Resonance Imaging in detecting spinal metastases. J. Nucl Med. 1993 Dec; 34(12):2191-8.
• Cesani F, Chaljub G, Briscoe E. Assessment of spinal metastases by bone scintigraphy and magnetic resonace imaging. J Nucl Med. 1995; 36 (suppl): 26.
• Eustace S, Tello R, DeCarvalho V. a comparsion of whole body TurboSTR MR imaging and planar Tc-99m diphosphonatescintugraphy in examination of patients with suspected skeletal metastases. AJR. 1997; 169:1655-1661.
• Yuh WT, Quests JP, Lee HJ.Anatomic distribution of metastases in the vertebral body and modes of hematogenous spread. Spine. 1996; 21: 2243-2250.