High Impact of Preferential Flow on ^99mTc-MAA and ⁹⁰Y-Loaded Microsphere Uptake Correlation

TO THE EDITOR: It was with great interest that we read the article entitled “Predictive Value of ^99mTc-MAA SPECT for ⁹⁰Y-Labeled Resin Microsphere Distribution in Radioembolization of Primary and Secondary Hepatic Tumors” by Ilhan et al. in the November 2015 issue (1).

The goal of their study was to evaluate the capacity of ^99mTc-labeled macroaggregated albumin (MAA) uptake to predict ⁹⁰Y-labeled resin microsphere uptake on posttherapeutic ⁹⁰Y bremsstrahlung SPECT imaging. The authors found a significant yet low correlation between ^99mTc-MAA and ⁹⁰Y-microsphere uptake in the different tumors tested. When ^99mTc-MAA uptake was found to be high, high ⁹⁰Y-microsphere uptake was almost always observed (i.e., 97% of cases), whereas low ^99mTc-MAA uptake correlated with high ⁹⁰Y-microsphere in 67% of cases.

These results are in clear contradiction to those of several other studies that have demonstrated the accuracy of ^99mTc-MAA–based dosimetry in the prediction of response and survival, suggesting a good correlation between ^99mTc-MAA and ⁹⁰Y-microsphere uptake (2–5).

This study therefore requires further discussion and clarification. For many years now, there has been debate surrounding the question of whether ^99mTc-MAA is a good surrogate for ⁹⁰Y-microsphere distribution, as ^99mTc-MAA particles are not of exactly the same size and density as ⁹⁰Y-microspheres.

The first point clearly recognized when considering ^99mTc-MAA as a surrogate for ⁹⁰Y-microsphere distribution is the fact that ^99mTc-MAA and ⁹⁰Y-microsphere injection should be performed precisely at the same site, meaning the same artery and at the same distance from the arterial bifurcation, with application of the same angulation of the microcatheter in the arterial lumen. In this study, no details were provided on either the exact microcatheter positioning or the accuracy of the repositioning. Where there were discrepancies between ^99mTc-MAA and ⁹⁰Y-microsphere uptake, this point should have been further assessed. A second key point to carefully monitor is the vasoactive status of the arterial tree at the time of ^99mTc-MAA or ⁹⁰Y-microsphere injection. Less is known on this subject than on microcatheter positioning, yet there have been cases of huge discrepancies between ^99mTc-MAA and ⁹⁰Y-microsphere uptake related not to inaccurate catheter repositioning but to vasospasm observed on only 1 of the 2 angiographies performed, namely the diagnostic angiography (6). However, this issue has not been addressed in the Ilhan et al. study.

Presently, the debate centers not only on whether ^99mTc-MAA alone is a good surrogate for microsphere distribution but also on whether the treatment simulation (including preferential flow at diagnostic angiography and ^99mTc-MAA scintigraphy) is a good surrogate for ⁹⁰Y-microsphere distribution. This is of major interest, as it means that during the diagnostic angiography, special care should be taken to use ^99mTc-MAA as a ⁹⁰Y-microsphere surrogate.

Figure 3 of this interesting paper illustrates the probable uncertainties of real microcatheter positioning or different arterial vasoactive statuses. On the ⁹⁰Y-microsphere scan, after injection of the tracer into the right hepatic artery we can see uptake only in the right lobe, with clear uptake in the tumor. On the ^99mTc-MAA scan, however, no uptake at all is observed in the tumor and uptake is only faint in the right lobe, whereas strong uptake is seen in the left lobe. There are only two possible explanations for this result: either the microcatheter position has moved during ^99mTc-MAA injection from the right hepatic artery into the left hepatic artery because of instability, or vascular flow in the right hepatic artery during ^99mTc-MAA injection was dramatically impaired (such as through a huge vasospasm or an arterial dissection) and ^99mTc-MAA flowed back toward the left hepatic artery. In this example, the discrepancy is more than likely accounted for by an abnormality occurring in the diagnostic angiography rather than by the difference between ^99mTc-MAA particles and ⁹⁰Y-microspheres.

Arterial vasospasm can be provoked by any intraarterial procedure and is especially common in prolonged procedures, cases involving anatomic difficulties, or interventions such as coil embolization. When radioembolization is involved, arterial spasm is considered more likely in diagnostic angiography, which consists of full arterial mapping and coil embolization, as necessary.

When using resin microspheres, as in this study, it is generally recommended that systematic coiling of digestive arterial branches be performed in order to avoid gastrointestinal irradiation due to resin microsphere backflow during injection. In the Ilhan et al. study (1), systematic coiling and arterial spasm should have been documented and analyzed, at least when discrepancies between ^99mTc-MAA and ⁹⁰Y-microsphere uptake were observed. Almost all the discrepancies observed in this study involved an association between low ^99mTc-MAA uptake and high ⁹⁰Y-microsphere uptake, strongly favoring the occurrence of vasospasm during ^99mTc-MAA injection. Indeed, for lesions with low grade 4 ^99mTc-MAA uptake, 21% had grade 1 ⁹⁰Y-microsphere uptake and 45.6% had grade 2 ⁹⁰Y-microsphere uptake.

Lastly, in their discussion, the authors presented the hypothesis that discrepancies between their results and those previously published could be explained by their use of glass microspheres (1). The type of microsphere used may, indeed, affect the correlation between ^99mTc-MAA and ⁹⁰Y-microsphere uptake, as glass microspheres are less embolic than resin ones. Nevertheless, it is our experience that the gentler diagnostic angiography approach we use could, once again, account for this discrepancy. As glass microspheres carry a lower risk of backflow during injection and do not require systematic coiling, coiling is used in less than 10% of procedures in our institution. Elsewhere, Lau et al. (2) published on hepatocellular carcinoma and Lam et al. (5) on metastatic disease (taking specific care with the catheter repositioning), and both reported a good dose–response relationship based on ^99mTc-MAA dosimetry using resin microspheres, indicating that the kind of microsphere used may not be the key parameter in reproducibility between ^99mTc-MAA and ⁹⁰Y-microsphere uptake.

To conclude, one must keep in mind that many confounding factors may affect the correlation between ^99mTc-MAA uptake and ⁹⁰Y-microsphere uptake. The use of ^99mTc-MAA as an accurate ⁹⁰Y-microsphere surrogate has to be anticipated. To avoid spasm, the gentlest approach possible should be taken during diagnostic angiography, and effort should be taken to ensure the exactness of catheter repositioning.

• © 2016 by the Society of Nuclear Medicine and Molecular Imaging, Inc.