Is PET or MRI better for detecting tumor recurrence in patients with brain metastases who received radiation therapy? While PET and conventional MRI performed about the same, MRI gets the edge if advanced scanning sequences are used, according to a Korean study in the December issue of the American Journal of Roentgenology.
The findings are an indication of why advanced MRI is gaining popularity for detecting various forms of cancers without the risk of ionizing radiation.
"MRI and PET showed high diagnostic performance for the detection of tumor recurrence after stereotactic radiosurgery in patients with brain metastasis," wrote lead author Dr. Chong Hyun Suh and colleagues from the University of Ulsan College of Medicine in Seoul. However, advanced MRI had a statistically significant edge over PET, they found (AJR, December 2018, Vol. 211:6, pp. 1332-1341).
While stereotactic radiosurgery has become a primary treatment option for patients with brain metastasis, determining treatment response and the potential for tumor recurrence is challenging. Conventional MRI has been the modality of choice for follow-up after radiosurgery, but radiation treatment can cause changes in the irradiated area and edema that can lead to contrast enhancement on scans.
Clinicians have tried to improve MRI follow-up by adding sequences such as dynamic susceptibility-weighted contrast, arterial spin-labeling, diffusion-weighted imaging, and MR spectroscopy to routine scans, while PET proponents have relied on a multitude of radiotracers for detecting tumor recurrence.
To determine which approach might be best, Suh and colleagues performed a meta-analysis of MRI and PET for differentiating between tumor recurrence and radiation necrosis in patients with brain metastasis treated with stereotactic radiosurgery.
The researchers found 20 studies that met their criteria. Eight studies dealt with MRI exclusively, seven studies assessed only PET, and five studies offered head-to-head comparisons. The median interval between stereotactic radiosurgery and MRI or PET ranged from five to 22.7 months.
In the 13 MRI studies and 12 PET studies, pooled sensitivities for MRI and PET were identical at 84%. Pooled specificity for MRI was 88%, compared with 86% for PET. Overall, there was no statistically significant difference in the diagnostic performance of the two modalities (p = 0.80).
In the five direct comparisons, the researchers similarly found no statistically significant difference between the modalities (p = 0.48). MRI and PET had identical pooled sensitivities of 88%, while PET outperformed MRI with a pooled specificity of 86%, compared with 76% for MRI.
The findings changed, however, in a subgroup analysis of nine studies that included advanced MRI techniques, all of which included perfusion MRI. In these studies, MRI achieved a pooled sensitivity of 86% and pooled specificity of 95%. Those increases provided MRI with a statistically significant difference in diagnostic performance over PET (p < 0.01) and conventional MRI (p = 0.01).
