Ketamine, a known antagonist of N-methyl-D-aspartic (NMDA) glutamate receptors, had been used as an anesthetic, particularly for pediatric or for cardiac patients. Ketamine employed in prescribed medical conditions had its advantages as it did not increase intracranial pressure during neurosurgery (Schmittner et al., 2007) and had no postoperative neurological damage when used in cardiopulmonary bypass patients (Smith et al., 2006). Medically, ketamine has also been proposed for anticonvulsive control (Dickenson and Ghandehari, 2007) and for controlling injury after stroke via it action on the glycine, zinc, and magnesium components of the glutamate binding sites (Collins et al., 1989); thus, protecting neuronal loss after stroke (Meldrum et al., 1987) and preventing of the spread of depolarization after injury (Hertle et al., 2012). This drug is, however, not without flaw. Being one of the noncompetitive NMDA receptor antagonists like phencyclidine and MK801, it would produce schizophrenia like psychosis in human (Dickerson and Sharp, 2006). Unfortunately, ketamine has now become an abusive drug in many parts of the world and chronic and prolonged usage led to damages of many organs in experimental animals (Yeung et al., 2009; Chan et al., 2011; Tan et al., 2011a; Wai et al., 2012; Wong et al., 2012). The damage on the nervous system included neuronal loss, synaptic changes, changes in functional magnetic resonance imaging (fMRI) activities, and the formation of mutated tau protein in neurons as described in models of rodents and monkeys (Yeung et al., 2010a; Sun et al., 2011; Yu et al., 2012).

In the higher primate, chronic treatment of ketamine induced changes of apoptotic markers in the prefrontal cortex and abnormal behavior in movement, walking, jumping, and climbing (Sun et al., 2012). fMRI in monkeys also revealed hyperactivity in entorhinal cortex, striatum regions, but hypoactivity in midbrain and visual cortex (Yu et al., 2012). In the mice, prefrontal hippocampal damages (Yeung et al., 2010a; Wai et al., 2013), pain altercations, and schizophrenic like behavior (Becker et al., 2003, 2006) had been documented and GABA receptor and changes of dopaminergic neurons were recorded in mouse model (Tan et al., 2011b, 2012). In the human brain, so far, few studies appeared in the literature. Our group, for example, revealed fMRI hypoactivities in the cerebellum of addicts (Chan et al., 2012). Narendran et al. (2005) reported that ketamine addicts exhibited selective up-regulation of dopamine D1 receptor via biochemistry. No study was ever put forward summarizing human central nervous system (CNS) lesions as yet. This paper described for the first time via magnetic resonance imaging (MRI) changes in the addicts of 0.5–12 years of ketamine addiction, and demarcated the possible brain regions susceptible to ketamine damages.

The results of lesions observed in all the 21 ketamine addicts were depicted in Table 2. Those who had two or less regions in the brain with lesions were classified as light damage. Those that had three to four regions in the brain with lesions were classified as moderate damage, and those with five or more regions with lesions were classified as severe damage. The MRI lesions initially were observed as hyperintense spots (holes or patches) of degeneration in the superficial white matter of the cortex which appeared as early as 1 year after ketamine addiction (Figure 1A), while each lesions spread to the internal capsule by 3 years of addiction (Figure 1B). Slightly after, patches of hyperintense degeneration spots appeared in the basal forebrain (Figure 2A), cerebellum, and pons (Figure 2B), and diencephalon at 4 years of addiction (Figure 2C). Likewise, diffusion blockage was illustrated by FLAIR image in the parahippocampal gyrus and insula, also by 4 years of addiction (Figure 3A), while atrophy of the parahippocampal gyrus was observed a bit later by 5 years of addiction (Figure 3B). Atrophy of the other parts of cortex was first noted after 4 years of addiction, usually with atrophy on only a small region of the cortex (Figure 4) and extended to two or three regions (usually frontal, parietal, and occipital) of the cortex by 7 years of addiction (Figure 5). Hyperintense lesions were also observed in the corpus striatum by 6 years (Figure 6). In this patient cohort, one patient had a combination of drugs and was taking ketamine together with amphetamine and ecstasy. He demonstrated early atrophy of cortex after taking the three drugs together in 0.5 years, in which the basal prefrontal gyrus rectus already exhibited significant atrophy (Figure 7A) when compared with control (Figure7B). Similarly, cortical atrophy also occurred early in another patient who had used a high dose of ketamine, in this case 3 g per day for 3 years (Figure 8). After 7 years of addiction, in all other patients, lesions then appeared in the midbrain (Figure 9). From 10 to 12 years of addiction, all lesion sites were as those described above.

Medical histories indicated that addicts over 4 year of ketamine addiction displayed memory deficits and anxiety depression while those from the 5 to 7 years developed definitive ataxia and by 7 years and above, dyskinesia became obvious.

This study revealed the lesions in many regions of the brain of ketamine addicts. These lesions appeared as minute patches in the first year and became larger sites of atrophy by 4 years of addiction. The majority of the addicts was on dosage of 1 g per day and used ketamine daily for several years. In this work, since the volunteers were mostly below 30 years old and only 2 individuals above 30 years old, a comparison of the effect of age upon addiction was not conclusive in this stage, even though we had seen no worse in the aged group (above 30 years old) when compared with the slightly younger old. A study of age response would definitely be conducted in future. However, it is well-known that ketamine addicts were usually young as represented in this cohort. The brain regions affected were prefrontal, parietal, occipital, limbic, brainstem, and corpus striatum. The lesions affected both the gray and white matter, i.e., neurons and nerve fibers in the human; these were similar to those reported earlier by us in the mice and the monkey (Yu et al., 2012). This MRI study also collated with the work by Morgan and Curran suggesting a loss of memory via psychological examination in chronic ketamine abuses (Morgan and Curran, 2006). In animals, prefrontal cortex apoptosis, mutated tau aggregation, brainstem chemical changes, and cerebellar apoptosis had been reported (Mak et al., 2010; Yeung et al., 2010b; Sun et al., 2011, 2012; Tan et al., 2011b, 2012; Yu et al., 2012; Wai et al., 2013). In fact, in the mice model, it had been documented both neurons and fibers (white matter) were both targets like in this report consisting of human subjects (Mak et al., 2010; Yeung et al., 2010b). Along with structural changes, fMRI and functional studies confirmed functional and cognitive derangements (Morgan and Curran, 2006; Sun et al., 2011, 2012; Chan et al., 2012; Yu et al., 2012). This human MRI brain imaging on the ketamine addicts thus consolidated that the areas of lesion in mice, monkey, and human were essentially similar. We now have clear and unequivocal proof of damages in the CNS upon chronic use of ketamine in human. As for the combination of drugs, we had only one individual on three types of drug—amphetamine, ecstasy and ketamine. The individual was only on such drugs for half a year with low dosage of ketamine, but his lesions were more vigorous than the addicts who had been on the drug for 3 years. This preliminary observation suggested severe detrimental effects on brain upon the combination of abusive drugs in the addicts.

Hyperdensity spots in computerized axial tomography (CAT) scan usually pointed to demyelination or metabolic changes (Brismar and Ozand, 1994; Matsushima et al., 1997). For MRI, while hyperintensive spot in T1 imaging pointed to toxicity or metabolic lesion, T2 hyperintensive could refer to demyelination and hardening of arteries (Hyttinen et al., 2009). Other studies on T1, T2, and FLAIR MRI indicated even more possibilities for hyperintensive spotty or non-spotty lesions (Cakirer et al., 2003). For instance, T1 hyperintensity could be related to infarct, infection, axonal damage, hemorrhage or vascular change and neoplasm (Cakirer et al., 2003), T2 hyperintensite image might indicate cytotoxicity, vasogenic edema whilst metabolic toxicity, encephalopathy, vasogenic edema would restrict diffusion and contributed to FLAIR hyperintensity (Cakirer et al., 2003). In our patients, these lesions observed were probably related to cytotoxicity, axonal damage, and vasogenic edema which we indicated as “degeneration” while atrophy of cortex was the last sequel of the events.

There were some studies on the neuroimaging of addicts using different drugs. Reneman et al. (2006), for example, using proton magnetic resonance spectroscopy (1H-MRS) suggested the cortical and subcortical reduction of serotonin transporter in ecstasy patients. Also using magnetic resonance spectroscopy, de Win et al. (2008a) reported thalamic damage by ecstasy. Apart from the thalamus, other subcortical structures in the globus pallidus and putamen were also affected by esctacy (de Win et al., 2008b). Using simple MRI and diffusion tensor imaging (DTI) techniques, other areas of the brain were reported to be damaged by ecstasy, e.g., the limbic cortex (Thompson et al., 2004), the medial temporal cortex (Thompson et al., 2004), callosum (Salo et al., 2009), limbic system (Sowell et al., 2010), and hippocampus (Den Hollander et al., 2012). Front-cortical and striatal damages were featured in the cocaine treated rat by MRI (Gozzi et al., 2011), callosum damage by DTI in human cocaine addicts (Moeller et al., 2007) and damage in frontal white matter in prenatally cocaine exposed human addicts (Smith et al., 2001). In the heroin addicts, for example, decreased gray matter was observed in the frontal, cingulate, and occipital cortices (Wang et al., 2012). In the cocaine addicts, cue-induced cocaine crowing involved the left dorsolateral frontal cortex and the anterior cingulate (Bolla et al., 1998; Maas et al., 1998; Wexler et al., 2001; Goldstein et al., 2004). These areas were also found to be involved in methamphetamine craving (Yin et al., 2012). Toda revisited these damaged areas in a review and proposed that these areas had glutamatergic projections to the nucleus accumbens (Toda, 2012). In this study, the ketamine damaged regions in the human subjects were put forward for the first time and resolved that they included diffusely many regions: frontal, parietal, occipital cortices, parahippocampal gyrus, striatum (including caudate), cerebellum, and brainstem. However, no obvious involvement was observed in the cingulate gyrus. Some of these areas were the same as cocaine and ecstasy patient, e.g., frontal, striatum, and limbic. However, ketamine presented a rather diffuse effect in many other of areas of the brain which might differ from cocaine, heroin, or methamphetamine.

This piece of work further suggests that addicts even on a single drug alone like ketamine might lead to atrophy of the brain after a few years of addiction. As addicts usually take ketamine at different time and ad lib, it was very difficult to compare dosages. In this work, there were hints that increasing dosage or combination of ketamine with other abusive drug would hasten the damages. It is important therefore to compare the years of duration and with a certain level of dosage which at least led some usable data. Further work should emphasize on the comparison of the interaction of these abuse drugs.