Another Spurious Description of Hypoerfixation of HMPAO

Shintani and colleagues (1) describe four cases of cerebral infarction in which single-photon emission computed tomography (SPECT) imaging of cerebral perfusion was performed using hexamethylpropyleneamine oxime (HMPAO). The authors conclude that “hyperfixation (of HMPAO) can occur in all stages of infarction, and does not appear to depend on recanalisation.” An examination of the evidence presented by the authors to support this statement is merited, particularly in light of recent calls for increased use of this important imaging technique for stroke (2).

SPECT scans were acquired at the following times after symptom onset: case 1—4 hours; case 2—2, 7, and 20 days; case 3—17 days; and case 4—5 and 16 days. Therefore, in only one case, and for only one SPECT scan (case 1), do the findings relate to what now is considered to be the acute stage of stroke when thrombolytic and neuroprotective therapy are to be most effective. The authors admit that this particular patient was the subject of a previous single case report (3), but what is more disconcerting is their failure to refer to the criticisms published in this journal of their use of the term “spurous hyperfixation” in that article (4, 5).

We believe it is unnecessary to repeat all of the points we made in our previous correspondence (4). Nonetheless, we reiterate that the term “hyperfixation,” when applied to the uptake of HMPAO, refers to a proved phonomenon (6, 7) that occurs in the late subacute stages of stroke (10 to 28 days after onset). Furthermore, the uptake of HMPAO is higher than would be expected from the true level of CBF. This overestimation may be present even when the uptake is below that of the adjacent or contralateral normal tissue, and, on average, is equal to about 20% of the expected value (7). In the earlier stages of stroke (up to 10 days) Sperling and Lassen (7) quite categorically stated that no hyperfixation was present even when HMPAO uptake was increased.

The existence of hyperfixation can be substantiated only when HMPAO SPECT is performed in conjunction with a “gold standard” method of measuring blood flow, such as ¹³³Xenon SPECT or C¹⁵O₂ PET. Shintani and colleagues did not perform such measurements, and cannot claim that hyperfixation exists in any of their cases.

What the authors did find was high uptake of HMPAO in areas of the brain, which eventually were associated with tissue loss on structural images. The authors readily jump to the conclusion, as they did in their previous publication (3), that this represents paradoxical hyperfixation. Of course, this finding has two possible explanations: 1) it is casued by a high CBF (hyperperfusion), or 2) it occurs in an area of low CBF, but where the trapping of HMPAO is enhanced (hyperfixation).

Reperfusion hyperemia has been described in many classic studies (8–10) in both the acute and subacute stages of stroke. Whether this hyperfusion is beneficial (nutritional) or nonnutritional, and thus whether the region involved progresses to infarction, depends on the magnitude of the initial ischemia and the timing of the reperfusion. In either situation, it represents a true increase in cerebral perfusion demonstrable by PET with C¹⁵O₂ (11) and SPECT with ¹³³Xenon (12). It is present in a region consistently much larger than the area of infarction seen in later structural imaging, although within this region there may be smaller areas of poorly perfused tissue that cannot be detected by poor resolution imaging systems, such as the one used by the present authors.

The mere existence of high HMPAO uptake in any stage of stroke does not necessarily imply hyperfixation of this tracer, yet that appears to be the premise of this article. The only explanation for the conclusion drawn by the authors is that they have made the assumption that because the tissue has progressed to infarction, then cerebral blood flow must have been low during the SPECT scans. Published literature on hyperperfusion attests to the naiveté of such an assumption, even in subacute stroke.

It is disappointing that the authors have been allowed to repeat such an unscientific approach to an important subject. Hyperfixation of HMPAO is a phenomenon that should be considered as one of the possible causes of increased uptake of HMPAO in subacute stroke. It has never been shown to present in acute stroke, when the use of HMPAO SPECT is likely to be most effective. At both time periods, HMPAO SPECT continues to be a reliable indicator of reperfusion, even when hyperfixation is present (7).

Further investigation of hyperfixation should be encouraged at centers equipped to perform such studies.