Methadone-induced encephalopathy is a rare event. To date, this phenomenon remains poorly characterized (4,5,21). The brain MRI changes reported in the literature are summarized in Table 4 and include: cerebellum abnormalities (1,4,10-14), bilateral cerebral white matter abnormalities (4,5,11,12,14), signal changes in hippocampus (10), globus pallidus (13), and in a single case report in the head of caudate nuclei (4). In addition, there is a single case report of a 2-year-old infant found to have cerebral white matter, cerebellar, and globus pallidus hypodensities based on computed tomography (CT) scan (22).
In our AOE patients, the most frequent MRI finding was bilateral confluent or patchy cerebral white matter hyperintensity (n= 5). Cerebellar abnormalities were detected in only three cases despite this was the most common observed abnormality in previous studies (1,10,13). A consistent (n= 4) and new finding in these patients was bilateral parieto-occipital cortex T2 and FLAIR hyperintensity. This radiological finding has also been reported in patients with posterior reversible encephalopathy syndrome (PRES; 23). PRES has been reported as a consequence of or in conjunction with a variety of critical illness states including severe hypertension, hemolytic-uremic syndrome, thrombocytopenic thrombotic purpura, and in association with drug toxicities such as cisplatin, cyclophosphamide, interferon (23–25), and opiates such as morphine (26,27). In keeping with the findings in PRES, three of our patients had bilateral parieto-occipital cortex restriction in DWI which was confirmed by ADC sequence. Additionally, restriction was observed in one patient with internal capsule involvement (case 7). Restriction in bilateral cerebral white matter has previously been reported secondary to methadone toxicity (4,11). One study suggested that “deep watershed infarct” resulted in the restriction imaging observed (11). Given our observations and previous published reports, it can be postulated that the changes in AOE due to methadone could result in PRES.
We also had two patients who had internal capsule involvement. This finding is in accordance with previously published reports as a characteristic of heroin toxicity (28). In our both patients, morphine and methadone were detected in urine analysis. Therefore, heroin use cannot be ruled out. Additional confirmatory testing for supplementary heroin metabolites would have been useful in these two individuals. However this was not available in our center. One of them (Case 5) demonstrated lesions in splenium of corpus callosum, a finding never reported before in either heroin or methadone intoxication. This finding may be a transient lesion of splenium and has been associated with various clinical conditions such as seizures, metabolic disturbances, infections, CNS malignancy, and drugs and toxins (antidepressants, antiepileptics, antipsychotics, chemotherapy agents, and pesticides) (15, 28-38). We also had a single patient (case 8) who showed involvement of the globus pallidus and head of caudate nuclei. This finding has been observed in association with methadone toxicity (4,13). Previously, brain imaging changes associated with methadone intoxication were suggested to be as a consequence of hypoxic events secondary to overdose (12). However, hypoxia-associated cerebral adverse effects on imaging seem to be only a result of prolonged hypoxia (39, 40). Majority of our patients did not have a persistent documented hypoxic insult. Brain neuroimaging was performed on admission, and before the worsening of patient’s condition. Secondly, brain and cerebellar damage demonstrated at both diagnosis and follow-up showed a clear-cut prominent involvement of the subcortical white matter. In adulthood, hypoxic-ischemic insults usually result in watershed zone infarcts when mild to moderate, and affect the gray matter in the basal ganglia, thalami, cerebral cortex, cerebellum, and hippocampi when severe. Furthermore, severe insult generally includes a stage of diffuse cerebral edema with loss of differentiation between gray and white matter, a finding that was not noted in the patients reported. Furthermore, acute and early subacute phases of hypoxia-induced encephalopathy primarily affect the basal ganglia, thalamus, and cortex (41). We reported bilateral cerebral white matter and cerebellum abnormalities as the most common brain MRI finding.
To date, only 8 case reports evaluating 11 patients have been published reporting delayed-onset methadone-induced leukoencephalopathy (6,10, 17–20,16), summarized in Table 5. The most frequent imaging findings in case reports of patients with DOL is bilateral cerebral white matter T2 and FLAIR hyperintensity (6,8,9,18,20,16) followed by corpus callosum (9,16) and globus pallidus (8) involvement. This is in keeping with our observation of bilateral cerebral white matter hyperintensity. However, the findings in DOL group are not generalizable, as there were only two cases in this group, who also lacked imaging in their acute phase for comparison with the DOL phase imaging. Furthermore, during examination of DWI and ADC, no restriction was found in either case. Four patients have been described with restriction in DWI scans, although a correlation with ADC was not reported in them (9,17,18,20). It is possible that the restrictions observed in these patients is related to T2 shine through, as this phenomenon has also been observed in our patients.
Almost all published case reports to date are in adult patients, except for a single case of 30-month-old infant. There are no previous publications on DOL due to other reasons (strangulation, CO poisoning, benzodiazepine overdose, etc.) in adults younger than 30 years (7). Since both of our patients were also adults, it is possible that DOL is a phenomenon more common among adult patients. DOL has been previously suggested to be due to hypoxia (6,16). However, given that neither of our patients had history of prolonged unconsciousness or respiratory depression, hypoxia as an etiology can be excluded. The lucid intervals of one to five weeks have been reported in earlier case reports (7), which was reinforced with our cases.