Published ahead of print on June 30, 2011
doi: 10.3174/ajnr.A2622

American Journal of Neuroradiology 32:E142, August 2011
© 2011 American Society of Neuroradiology

A. M. McKinney^a, B. Sarikaya^a, J. Spanbauer^a and B. D. Lohman^a
aDepartment of Radiology

E. Uhlmann^b
bDepartment of Neurology 
Hennepin County and University of Minnesota Medical Centers 
Minneapolis, Minnesota

We read with great interest the description by U-King-Im et al¹ in the February issue of the American Journal of Neuroradiology (AJNR) of 4 patients with “acute hyperammonemic encephalopathy” on diffusion-weighted imaging (DWI). In the September 2010 issue of AJNR, we used a similar term of “acute hepatic encephalopathy” (most of the patients had hyperammonemia), and the terms could perhaps be considered interchangeable (notably, both would result in the acronym AHE).² We thank them for describing their findings, which are similar to the cases we described that, in our opinion, lie at the severe end of the spectrum of AHE. Of particular note is that 2 of their 4 patients died. In our study, 3 of 5 patients died; they had a similar distribution on DWI, which we termed “diffuse cortical involvement.”

U-King-Im et al¹ limited this description to the cingulate and insular gyri, but review of their available images demonstrates the abnormalities to be more extensive than those 2 regions. We do not point this out for the purpose of criticism; rather, we agree that these findings should alert the radiologist to the possibility of AHE. Thus, the combination of these 2 studies would indeed suggest (though preliminarily) that “diffuse cortical involvement” or alternatively “cingulate and insular involvement” may portend a poor outcome, though AHE is potentially reversible with therapy (such as lactulose).

They also similarly noted thalamic brain stem involvement in 1 patient, which we noted in most of our 20 patients. Hence, it would be of interest to us whether subtle involvement of the thalami or dorsal brain stem existed on further review of fluid-attenuated inversion recovery (FLAIR) or DWI in these 4 patients or if they may have other patients who may have been eventually excluded from their study or not included due to stringent criteria, for example due to confounding diagnoses with preliminarily negative findings on MR imaging (we have noted that such situations uncommonly occur). Such confirmation of milder cases limited to the thalami, for example, would help solidify our findings that AHE occurs along a spectrum, with multifocal diffuse cortical findings being at the severe end.

We proposed the terminology “acute hepatic encephalopathy” rather than “acute hyperammonemic encephalopathy” for several reasons: 1) Although some correlation likely exists, the degree of correlation between serum ammonia levels and AHE severity (the “ammonia hypothesis”) is still controversial. 2) Ammonia levels may be mildly elevated in patients with chronic cirrhosis without symptoms of AHE. 3) Severely encephalopathic patients may have normal ammonia levels. 4) Serum ammonia has been shown to be a poor predictor as a single test for the presence of AHE.^3,4 More recent evidence suggests rather that there is a synergistic effect between ammonia and various other inflammatory cytokines that results in excess glutamine within astrocytes, leading to osmotic swelling of the astrocytes and the subsequent brain edema as well as other neurocytotoxic effects.⁵

Ultimately, the diagnosis of AHE still remains a clinical one, where serum ammonia and other ancillary tests are supplementary. Indeed, our study did not find a significant correlation between plasma ammonia levels and the initial clinical severity of AHE, though we found that the initial clinical severity, the plasma ammonia level, and the extent of FLAIR and DWI abnormalities on MR imaging all significantly correlated with patient outcome.² Thus, while ammonia is inextricably linked to the pathogenesis of AHE, it may be a bit misleading to use the term “hyperammonemicencephalopathy,” because ammonia is but 1 (albeit quite important) precursor to the development of encephalopathy. Both our study and that of U-King-Im et al¹ would suggest that prospective monitoring of serum ammonia levels and MR imaging findings, along with other clinical and laboratory tests, could further delineate the pathogenesis of this potentially reversible disorder.

References

1. U-King-Im JM, Yu E, Bartlett E, et al. Acute hyperammonemic encephalopathy in adults: imaging findings. AJNR Am J Neuroradiol 2011;32:639–42[Abstract/Free Full Text]
2. McKinney AM, Lohman BD, Sarikaya B, et al. Acute hepatic encephalopathy: diffusion-weighted and fluid-attenuated inversion recovery findings, and correlation with plasma ammonia level and clinical outcome. AJNR Am J Neuroradiol 2010;31:1471–79[Abstract/Free Full Text]
3. Arora S, Martin CL, Herbert M. Myth: interpretation of a single ammonia level in patients with chronic liver disease can confirm or rule out hepatic encephalopathy. CJEM 2006;8:433–35[Medline]
4. Nicolao F, Efrati C, Masini A, et al. Role of determination of partial pressure of ammonia in cirrhotic patients with and without hepatic encephalopathy. J Hepatol 2003;38:441–46[CrossRef][Medline]
5. Shawcross D, Wright G, Olde Damink S, et al. Role of ammonia and inflammation in minimal hepatic encephalopathy. Metab Brain Dis 2007;22:125–38[CrossRef][Medline]

Reply

Published ahead of print on June 30, 2011
doi: 10.3174/ajnr.A2650

American Journal of Neuroradiology 32:E143, August 2011
© 2011 American Society of Neuroradiology

J. M. U-King-Im^a, E. Yu^a, E. Bartlett^a, R. Soobrah^a and W. Kucharczyk^a
aDepartment of Diagnostic Imaging 
University Health Network 
University of Toronto 
Toronto, Ontario, Canada

We thank Dr McKinney and colleagues for their interest in our article. The authors generally agree with our findings and raise several interesting issues. We are grateful to the Editor-in-Chief for the opportunity to respond to some of the main issues raised.

Dr McKinney writes that we described only lesions in the cingulate and insular cortex. This is not the case. Throughout our article, we stated that all our cases had additional findings such as more variable additional cortical involvement or basal ganglia changes.¹ The point we were rather hoping to put across wasthat while such changes were not specific (eg, diffuse cortical involvement could be seen in a range of conditions including diffuse hypoxic injury), symmetric involvement of the cingulate and insular cortex was seen in all 4 cases and could therefore represent a more specific finding for hyperammonemic encephalopathy. Such symmetric findings are not commonly seen in other conditions, as far as we are aware.

Dr McKinney and colleagues also argue that the terms “acutehyperammonemic encephalopathy” and “acute hepatic encephalopathy” could be interchangeable. They then suggest that their proposed terminology would be more accurate, given that in acute hepatic failure, hyperammonemia is only 1 of the precursors of encephalopathy.² Respectfully, we have to disagree. The use of this terminology would have to assume that all cases of hyperammonemic encephalopathy are caused by hepatic failure; this is also not the case. In our small series, 3 patients had hepatic failure, but 1 patientwas post–orthoptic lung transplant without evidence of hepatic failure. Moreover, in the pediatric literature, there are many described cases of hyperammonemic encephalopathy, without any link to liver failure.^3,4 Many of these cases, (eg, due to urea cycle disorders) had radiologic findings similar to ours with cingulate and insular cortex involvement. ⁴ We argue that this provides compelling evidence that the changes we describe are related to hyperammonemic encephalopathy rather than hepatic encephalopathy per se. We believe that these findings can be seen in hyperammonemic encephalopathy due to various causes, such as hepatic failure, urea cycle disorders, post–lung transplantation, and so forth.

It is unclear from our data whether cortical involvement lies at the severe end of the spectrum, as Dr Mc Kinney suggestsbecause 2 of our patients survived despite having cortical involvement. This survival suggests that these changes could potentially be reversible. Data from other series hopefully could shed more light on prognosis. Only 1 of our patients had brain stem and thalamic involvement, and we are unable to comment further on the basis of our small series.

In summary, we suggest that the pattern of symmetric cingulate and insular cortical involvement, with additional and more variable cortical involvement, should alert the radiologist to the possibility of hyperammonemic encephalopathy. The cause of the hyperammonemia is nonspecific and could include hepatic failure, lung transplantation, urea cycle disorders, and so forth. Despite the differences in viewpoints, both Dr Mc Kinney’s and our articles should increase the radiologic recognition of this potentially reversible condition. ^1,2

References

1. U-King-Im JM, Yu E, Bartlett E, et al. Acute hyperammonemic encephalopathy in adults: imaging findings. AJNR Am J Neuroradiol 2011;32:413–18. Epub 2010 Nov 18[Abstract/Free Full Text]
2. McKinney AM, Lohman BD, Sarikaya B, et al. Acute hepatic encephalopathy: diffusion-weighted and fluid-attenuated inversion recovery findings, and correlation with plasma ammonia level and clinical outcome. AJNR Am J Neuroradiol 2010;31:1471–79[Abstract/Free Full Text]
3. Blaser S, Feigenbaum A. A neuroimaging approach to inborn errors of metabolism. Neuroimaging Clin N Am2004;14:307–329, ix[CrossRef][Medline]
4. Bindu PS, Sinha S, Taly AB, et al. Cranial MRI in acute hyperammonemic encephalopathy. Pediatr Neurol2009;41:139–42[CrossRef][Medline]