Diagnostic Value of Contrast-Enhanced Fluid-Attenuated Inversion Recovery MR Imaging of Intracranial Metastases

Methods

Sixty-nine patients (33 women, 36 men; mean age, 51 years; age range, 5–77 years) with a known primary malignancy and clinically suspected cranial metastases underwent MR imaging at 1.0 T. The distribution of the primary neoplasms was as follows: lung carcinoma in 31 patients, breast carcinoma in 15, both lung and breast carcinomas in one, stomach cancer in five, lymphoma in four, urinary bladder cancer in three, renal cell carcinoma in two, glioblastoma multiforme in two, colon cancer in one, pancreatic cancer in one, ovarian cancer in one, nasopharyngeal carcinoma in one, malignant schwannoma in one, and malignant melanoma in one. Institutional board approval and informed consent was obtained from each patient before their examination.

Following the acquisition of sagittal scout T1W images, axial T2W FLAIR and spin-echo T1W images were obtained before and after the intravenous administration of gadopentetate dimeglumine (GD) 0.1 mmol/kg over 1 minute. T1W (TR/TE/NEX = 380–460/8–17/3) and FLAIR (TR/TE_eff/TI/NEX = 8002/149–192/2000/1) images were alternately obtained as the first contrast-enhanced images. The acquisition time was 2 minutes 50 seconds to 6 minutes 50 seconds for T1W sequences, and 2 minutes 40 seconds to 3 minutes 10 seconds for FLAIR sequences. All images were acquired with a section thickness of 5 mm, an intersection gap of 2 mm, a field of view of 18 × 18–22 × 22 cm, and a matrix of 256 × 192.

The images were transferred to a workstation and reviewed by three experienced radiologists at one sitting. Each radiologist independently evaluated precontrast and postcontrast FLAIR and T1W imaging side by side at the same time. In cases of disagreement, the final judgement was reached by consensus of the three reviewers. Parenchymal and leptomeningeal-cisternal metastases were assessed for lesion conspicuity and degree of contrast enhancement. In addition, the number of parenchymal metastases and the extent of the leptomeningeal-cisternal metastases were assessed.

Results

MR examinations revealed parenchymal metastases in 33 of 69 patients. Ten patients had solitary metastasis, while two patients had two parenchymal metastases, two patients had three lesions, and 19 patients had more than three. Four patients had associated cranial-nerve (CN) metastases, while five patients had associated leptomeningeal-cisternal metastases. Six patients had only leptomeningeal-cisternal metastases. Table 1 shows the distribution of cranial metastases.

In the evaluation of parenchymal metastases, postcontrast FLAIR images demonstrated more parenchymal metastases in five patients than did T1W images, whereas they showed an equal number in 20 patients and fewer lesions in eight (Fig 1). When we compared the conspicuity of parenchymal metastases, contrast-enhanced FLAIR imaging was superior to contrast-enhanced T1W imaging in five of 33 patients, equal in one, and inferior in 27 (Fig 1). Similarly, the degree of contrast enhancement was superior with the postcontrast FLAIR sequence in five patients, equal in five, and inferior in 23 (Fig 1). In our evaluation of the number, conspicuity, and enhancement of parenchymal metastases, the postcontrast FLAIR sequence was superior to the postcontrast T1W sequence in the same five patients. In all of these patients, FLAIR imaging was obtained as the second contrast-enhanced sequence. Table 2 summarizes results from the evaluation of parenchymal metastases.

• Download figure
• Open in new tab
• Download powerpoint

Fig 1.

Images obtained in a 57-year-old-man with small-cell lung cancer reveal multiple, enhancing, solid and cystic parenchymal metastases. The number, conspicuity, and contrast enhancement are superior on the postcontrast FLAIR image compared with the postcontrast T1W image. In this patient, FLAIR was performed as the second postcontrast sequence.

A, Contrast-enhanced T1W image.

B, Nonenhanced FLAIR image.

C, Contrast-enhanced FLAIR image.

Leptomeningeal-cisternal metastases were found in 11 patients. Five of these patients also had positive CSF cytologic results. In eight patients, postcontrast FLAIR imaging surpassed postcontrast T1W imaging in terms of lesion conspicuity, extension, and degree of enhancement (Fig 2). FLAIR images were acquired as the second postcontrast sequence in five of these eight patients. Table 3 summarizes results from the evaluation of leptomeningeal-cisternal metastases.

• Download figure
• Open in new tab
• Download powerpoint

Fig 2.

Images obtained in a 45-year-old-woman with breast cancer. The conspicuity, extension, and contrast enhancement of the lesions are superior on the postcontrast FLAIR image (C) than on the postcontrast T1W image (A). The leptomeningeal enhancement of frontal lobes (white arrow) is not conspicuous in A as compared with C. For this patient, T1W imaging was performed as the second postcontrast sequence.

A, Contrast-enhanced T1W image shows abnormal enhancement in the basal cisterns (arrow) and leptomeninges consistent with metastases.

B, Nonenhanced FLAIR image.

C, Contrast-enhanced FLAIR image also shows the abnormal enhancement of the leptomeninges (white arrow) and basal cisterns (black arrow).

Four patients had CN metastases. In one patient, FLAIR images revealed involvement of the prechiasmal portion of the right optic nerve and extension to the optic chiasm and right optic tract. These findings were depicted as abnormal hyperintensity (Fig 3A) and marked contrast enhancement (Fig 3B) on precontrast and postcontrast FLAIR images, respectively. Only optic chiasmatic involvement was detected on the postcontrast T1W image, with no abnormal enhancement of the right optic nerve or right optic tract (Fig 3C). In another patient with involvement of the right optic tract, the metastasis was detected on precontrast and postcontrast FLAIR images (as abnormal hyperintensity) but not on the postcontrast T1W image. In a third patient with involvement of both the optic chiasm and left optic tract, the lesions were again detected on only precontrast and postcontrast FLAIR images, appearing as abnormal hyperintensity. In the patient with bilateral metastases of CNs VII and VIII, postcontrast T1W and FLAIR images were equal in the evaluation of lesion conspicuity, extension, and degree of contrast enhancement. Table 4 summarizes the results from evaluation of the CN metastases.

• Download figure
• Open in new tab
• Download powerpoint

Fig 3.

Images obtained in a 54-year-old-man with small-cell lung cancer. In this patient, FLAIR imaging was the second postcontrast sequence.

A and B, Nonenhanced (A) and contrast-enhanced (B) FLAIR images show involvement of the prechiasmal portion of the right optic nerve (white arrow) and extension to the optic chiasm (black arrow) and right optic tract (arrowhead). These findings appear as abnormal hyperintensity in A and as abnormal contrast enhancement in B.

C, Contrast-enhanced T1W image. Although involvement of optic chiasm is shown as abnormal contrast enhancement (arrow), involvement of the right optic nerve and right optic tract could not be detected.

Discussion

GD is frequently used as an intravenous MR imaging contrast agent to improve lesion detection and characterization. Although gadolinium shortens both the T1 and T2 relaxation times of tissues in which it accumulates, T1 shortening is the prominent effect at common doses. For intra-axial brain lesions, the blood-brain barrier must be disrupted for GD to enter the extracellular space. For extra-axial lesions, enhancement is observed in lesions with relatively high vascularity (9).

Contrast-enhanced T1W imaging is still the criterion standard in the detection and characterization of intracranial lesions. Although FLAIR images are commonly thought of as T2W images with dark CSF (because of the mild T1 effect induced by the long inversion time), T1 shortening produced by GD can be seen as hyperintensity on FLAIR images. Thus, lesions that show enhancement on postcontrast T1W images also show enhancement on postcontrast FLAIR images (1, 19, 20). When the FLAIR sequence is performed only after the injection of contrast material, the observed hyperintensity may be due to either T2 lengthening or T1 shortening; this feature may limit the usefulness of a postcontrast FLAIR sequence alone. To overcome this difficulty, the FLAIR sequence should be performed both before and after the administration of contrast material (9).

In parenchymal metastases with long T2 relaxation times, GD enhancement of the hyperintense lesion may be difficult to see on postcontrast FLAIR images. At postcontrast T1W imaging, it is easier to detect a hyperintense lesion surrounded by a hypointense edematous area. Hence, T1W imaging is superior to contrast-enhanced FLAIR imaging for detecting parenchymal tumors (9). In our study, the conspicuity of parenchymal metastases was superior on postcontrast T1W images compared with postcontrast FLAIR images in 27 of 33 patients.

Wessbecher et al (21) studied precontrast and postcontrast long-TR (spin-echo and T2W) images with known enhancing intracranial lesions. They compared these images with postcontrast T1W images and found enhancing lesions on postcontrast long-TR images that were not depicted on postcontrast T1W images. This observation was attributed to delayed enhancement because the long-TR images were obtained after the T1W images were. The concern that delayed enhancement might effect our results led us to perform postcontrast FLAIR and T1W sequences alternately after the administration of GD. Postcontrast FLAIR images were superior to postcontrast T1W images in depicting the number, conspicuity, and enhancement of parenchymal metastases in five patients. Because FLAIR imaging was performed as the second contrast-enhanced sequence in all of these patients, the findings were attributed to the delayed enhancement.

Contrast-enhanced FLAIR images are highly effective in the detection of sulcal or meningeal infection, inflammation, and metastases (1, 9, 15, 16). Because of the suppression of CSF signal intensity, there is marked delineation of lesions, especially meningeal lesions, that abut the border of the CSF. In addition, the fact that slow-flowing blood is not usually hyperintense on postcontrast FLAIR images but frequently hyperintense on postcontrast T1W images partly accounts for the clearer distinction of enhancing meninges and enhancing cortical veins with FLAIR imaging (9). Our results support this finding. In eight of 11 patients, the conspicuity, extension, and enhancement degree of leptomeningeal-cisternal metastases were superior on contrast-enhanced FLAIR images compared with postcontrast T1W images. FLAIR was the second postcontrast sequence in five of these eight patients. Because contrast-enhanced T1W images were acquired after contrast-enhanced FLAIR images in three patients, the findings were not attributed to delayed enhancement.

In the limited number of CN metastases, contrast-enhanced FLAIR images were superior to contrast-enhanced T1W images in terms of the conspicuity, extension, and enhancement of neural metastases. Larger series with CN metastases are needed to compare these two postcontrast sequences.

Conclusion

A postcontrast FLAIR sequence is a valuable adjunct to postcontrast T1W imaging, and the routine use of FLAIR imaging before and after the administration of contrast material is particularly useful for delineating leptomeningeal-cisternal and CN metastases.