Detection of contrast in arthrogram protocols

[AndrewBworth]

Hello All,

I have recently had some insight into a problem I saw approximately 4 months ago: A patient was ordered for a shoulder arthrogram. For those who don't know, an MRI arthrogram is a procedure where a mixture of x-ray and MRI contrasts and I believe saline is directly injected into a joint. The x-ray contrast allows confirmation by fluoroscopy of the injection, and the MRI contrast allows a more direct assessment of injuries like labral tears. This test usually precedes a surgery.
Anyway, the patient was injected and confirmed by fluoroscopy. When the patient was scanned however, no contrast could be seen in the joint. The tech opened up the coverage of the scan to see if the contrast was in the wrong place - and in doing so accidentally set the TR beyond T1 weighting (slightly). The dye was never found, and the patient had to be repeated another week (an angry patient). It was decided that since the scan did not have the required weighting, it was inconclusive as to whether or not the injection had missed the mark.
The next time I did an arthrogram I paid close attention - the contrast is bright on both T1 and T2 weighted scans, so I realized that we had probably arrived at the wrong conclusion. But in fact it was more obvious than that. With the injection confirmed by fluoroscopy, we know that the injection didn't miss. However, only trace amounts of MRI contrast are used to prepare the mixture. The reason the injection wasn't seen is almost certainly because the gadolinium contrast was forgotten.
Supposedly, "missing" the injection is a problem that occurs in hospitals more often than anyone would like to admit. If this ever happens to you, run both a T1 scan, and a fat-sat T2 - I suspect you will find the "missing" injection.