A 42-year-old man presented with a two-month history of recurrent seizures, abnormal behaviors including wandering and urinary incontinence and progressive cognitive decline. He was admitted with status epilepticus with a Glasgow Coma Scale score of E1V1M1, prompting initiation of antiepileptic therapy. Despite treatment, he remained comatose and developed right upper limb paresis. Laboratory tests showed normal thyroid function including free triiodothyronine, free thyroxine and thyroid-stimulating hormone, but anti-thyroglobulin (228 IU/mL) and anti-thyroid peroxidase (489 IU/mL) antibodies were markedly elevated. Ant-NH2-terminal of α-enolase (NAE) antibodies was negative. CSF analysis showed normal cell count, protein and glucose levels, with negative cytology and bacterial cultures. In both serum and CSF, tissue-based assays with immunostaining using rat brain frozen sections showed a negative neuropil pattern. Live neuron assays using primary cultured hippocampal neurons from E18 rats were additionally performed, but no reactivity was observed. Cell-based assays for CSF antibodies against intracellular antigens (amphiphysin, AGNA-1, ANNA-1, ANNA-2, ANNA-3, CRMP5, GAD65, GFAP, neurochondrin, NIF, PCA-1, PCA-2, PCA-Tr, PDE10A, Septin-7 and TRIM46) and cell surface antigens (AMPA receptor, CASPR2, DPPX, GABA-A/B receptor, mGluR1, IgLON5, LGI1 and NMDA receptor) were all negative. Brain MRI revealed a hyperintensity in the left-dominant bilateral occipital cortices on diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) (Figure 1 A), without other parenchymal abnormalities including cerebellum (Figure 1 B). N-isopropyl-p-[¹²³I] iodoamphetamine single-photon emission CT (¹²³I-IMP SPECT) revealed increased regional cerebral blood flow (rCBF) in the left-dominant bilateral cerebral cortices and decreased rCBF in the medial frontal lobes and cerebellum (Figure 2 A and B). He underwent two courses of high-dose intravenous methylprednisolone therapy (1000 mg/day) on days 6–8 and 19–21, leading to a complete recovery of consciousness. Based on the clinical and laboratory findings, a diagnosis of Hashimoto’s encephalopathy was established. ^{Reference Mattozzi, Sabater and Escudero1} Post-recovery examination revealed dysarthria, limb and truncal ataxia and saccadic eye movements, consistent with cerebellar ataxia. Follow-up MRI showed resolution of the occipital cortical hyperintensity on DWI and FLAIR. Subsequent SPECT demonstrated improvement in cerebellar hypoperfusion, though partial rCBF abnormalities in other regions persisted (Figure 2 C and D).

Figure 1.

Brain MRI on day 6 demonstrated hyperintensity in the left-dominant bilateral occipital cortices on diffusion-weighted imaging and FLAIR sequences (yellow arrows, (A), without other parenchymal abnormalities. FLAIR images showed no abnormal findings in the cerebellum (B). FLAIR = fluid-attenuated inversion recovery; MRI = magnetic resonance imaging.

Figure 2.

Initial ¹²³I-IMP SPECT (A) and the 3D SSP hyperperfusion and hypoperfusion maps (B) on day 14 revealed increased rCBF in the left-dominant bilateral cerebral cortices (yellow arrowhead) and decreased rCBF in the medial frontal lobes and cerebellum (red arrowheads). Follow-up ¹²³I-IMP SPECT (C) and the 3D SSP maps (D) on day 35 demonstrated improvement in cerebellar hypoperfusion, although partial perfusion abnormalities persisted in other regions. Color scale indicates the degree of perfusion change: red, severe; yellow, moderate; green/blue, mild. Regions with Z-scores>2 represent statistically significant alterations in blood flow. ¹²³I-IMP = N-isopropyl-p-[¹²³I] iodoamphetamine; rCBF = regional cerebral blood flow; SPECT = single-photon emission computed tomography (CT); SSP = stereotactic surface projection.

There are no universally established diagnostic criteria for Hashimoto’s encephalopathy, and various diagnostic frameworks have been proposed. ^{Reference Mattozzi, Sabater and Escudero1–Reference Shaw, Walls, Newman, Cleland and Cartlidge4} Recently, the exclusion of alternative etiologies, particularly through serum and CSF antineuronal antibody testing, has become increasingly emphasized, as applied in this case. ^{Reference Mattozzi, Sabater and Escudero1,Reference Graus, Titulaer and Balu2} Moreover, Hashimoto’s encephalopathy has been classified into several clinical subtypes; this case corresponds to the possible autoimmune encephalitis subtype. ^{Reference Mattozzi, Sabater and Escudero1} Regarding SPECT findings, most reports describe cerebral hypoperfusion, ^{Reference Muramatsu, Ikawa and Yoneda5} with only one case showing hyperperfusion. ^{Reference Okitsu, Sugaya, Yoneda and Takahashi6} To our knowledge, this is the first report of a mixed perfusion pattern. Such patterns have been observed in autoimmune encephalitis, ^{Reference Fukumoto, Miyamoto, Fujita, Murakami, Matsui and Izumi7–Reference Ohta, Seki, Dalmau and Shinohara9} suggesting a potential link to the possible autoimmune encephalitis subtype within Hashimoto’s encephalopathy. In autoimmune encephalitis, SPECT hyperperfusion likely reflects active inflammation, whereas hypoperfusion indicates neuronal dysfunction. ^{Reference Fukumoto, Miyamoto, Fujita, Murakami, Matsui and Izumi7,Reference Matsubayashi, Furuki and Obayashi8} Pathological studies of Hashimoto’s encephalopathy have revealed lymphocytic infiltration in the meningeal and parenchymal microvasculature. ^{Reference Nolte, Unbehaun, Sieker, Kloss and Paulus10} These findings may account for the patient’s persistent impaired consciousness, seizures and right-sided paresis as a result of cortical hyperperfusion associated with active inflammation, whereas the cerebellar ataxia may be attributed to cerebellar hypoperfusion reflecting underlying neuronal dysfunction. The left-dominant cortical hyperintensity on MRI corresponded to a hyperperfusion pattern on SPECT, which may indicate inflammatory changes. This case highlights a novel SPECT perfusion pattern and the potential utility of SPECT in Hashimoto’s encephalopathy.

This case has several minor diagnostic limitations. A cell-based assay for anti-myelin oligodendrocyte glycoprotein (MOG) antibodies was not performed. However, tissue-based assays using both serum and CSF did not demonstrate an oligodendrocyte staining pattern suggestive of the presence of anti-MOG antibodies. In addition, anti-NAE antibodies were negative in this case. Although anti-NAE antibodies are highly specific for Hashimoto’s encephalopathy, their sensitivity is known to be low. ^{Reference Mattozzi, Sabater and Escudero1} Importantly, autoimmune encephalitis was carefully excluded through comprehensive autoantibody screening using a combination of tissue-based assays and live neuron assays in both serum and CSF, as well as multiple cell-based assays in CSF. Taken together, these limitations do not preclude the diagnosis of Hashimoto’s encephalopathy in the present case.