key: cord-0006724-x22sl43t authors: Büttner, Andreas; Weis, Serge title: Neuropathological alterations in drug abusers: The involvement of neurons, glial, and vascular systems date: 2006 journal: Forensic Sci Med Pathol DOI: 10.1385/fsmp:2:2:115 sha: 07ebcfc0b4f544a368f0ba9ab0cc61cc5a81c3ab doc_id: 6724 cord_uid: x22sl43t Because the effects of drug abuse on the cellular elements of the human brain have not been studied systematically, an investigation was performed using histology, immunohistochemistry, and morphometry. The main cortical and subcortical brain areas of 50 polydrug deaths were analyzed as compared with controls. In the brains of drug abusers, a significant neuronal loss was present. Interestingly, the number of glial fibrillary acidic protein (GFAP)-positive astrocytes was reduced. the numerical density of perivascular and parenchymal microglia was increased in the white matter and in most subcortical regions. In the white matter there were widespread β-amyloid precursor protein deposits. Furthermore, there was a prominent vascular hyalinosis, endothelial cell proliferation, and a loss of immunoreactivity for collagen type IV within the vascular basal lamina. The neuronal loss seems to be the result of a direct impairment of nerve cells and, indirectly, to a damage of astrocytes, axons, and the microvasculature. The reduction of GFAP-positive astrocytes is also indicative of a drug-induced damage. The axonal injury suggests a toxic-metabolic drug effect, whereas the concomitant activation of microglia is indicative of a long-standing progressive process. The noninflammatory vasculopathy can be considered as the morphological substrate of a disturbed blood-brain barrier. Our findings demonstrate that drugs of abuse initiate a cascade of interacting toxic, vascular, and hypoxic factors that finally result in widespread disturbances within the complex network of central nervous system cell-cell interactions. Drug abuse represents a significant forensic issue worldwide. Although no brain lesion specific for drug abuse exists, a broad spectrum of changes affecting the central nervous system (CNS) is seen in drug abusers (1, 2) . In addition, morphological, physiological, and neurochemical abnormalities have been demonstrated by using neuroradiological techniques such as computed tomography, magnetic resonance imaging, magnetic resonance spectroscopy, positron emission tomography, or single photon emission computed tomography (1, 3) . However, because the consequences of drug abuse on the cellular elements of the brain have not been studied systematically, an investigation was performed using histology, immunohistochemistry, and morphometry. The analysis included the spectrum, frequency, and topography of histopathological alterations in the brains of polydrug abusers as compared with controls through the determination of nerve cell density, the reaction pattern of astrocytes and microglia, as well as an examination of the white matter and the cerebral microvasculature. astrocytes, a monoclonal antibody (MAb) against glial fibrillary acidic protein (GFAP, dilution 1:100, DAKO, Germany) was used and for microglia, an MAb against the human lymphocyte antigens (HLA)-DP, DQ, DR antigen (CR 3/43, dilution 1:100, DAKO, Germany, predigestion with formic acid) was used. The vascular basal lamina was immunostained with an MAb against collagen type IV (dilution 1:200, DAKO, Germany, predigestion with protease). For the investigation of white matter changes, an MAb against ß-amyloid precursor protein (ß-APP 695a, dilution 1:200, Zytomed, Germany, microwave antigen-retrieval procedure with citrate buffer) was used. The antigens were detected with a Histostain ® -Plus Peroxidase Kit (Zytomed, Germany). The neurohistopathological analyses were performed using a semiquantitative scoring system with 0 being not present; 1, moderately present; and 2, strongly present. Forensic Science, Medicine, and Pathology V2-2 The determination of the numerical density (n/mm 2 ) of astrocytes and microglia was performed at 400 × magnification with a counting grid (4) . For the quantification of microglia, the perivascular and parenchymal microglia was counted separately, then summed up as total microglia. The gray matter was analyzed following the "systematic row sampling," whereas the white matter, the hippocampal formation, the subcortical, and the brainstem regions were analyzed following the "random systematic sampling" (4) . For the determination of ß-APP immunostaining, a semiquantitative scoring system was used with 0 being no ß-APP immunostaining; 1, moderate immunostaining, however focal; 2, scattered patches of ß-APP foci; and 3, extensive ß-APP deposits throughout large areas of the white matter. For collagen type IV, the intensity of the immunostaining was evaluated using a 3-point rating scale. Staining reactivity was denoted as 1 for mild immunoreactivity, 2 for moderate immunoreactivity, and 3 for strong immunoreactivity. The numerical density of vessels for each staining intensity and the total numerical density of vessels was calculated by dividing the number of counted vessels by the total area of all measuring fields and expressed as n/mm 2 . The data were analyzed with the statistical program SPSS (Statistical Package for the Social Sciences, version 13.0 for Windows). The nonparametric Mann-Whitney U-test as well as the Spearman Rank correlation were used. All drug deaths were polydrug abusers who died owing to acute drug intoxication ( Table 2 ). Besides heroin in 30 cases, codeine/dihydrocodeine (n = 10), methadone (n = 6), cocaine (n = 2), and methylenedioxymethamphetamine (n = 2) could be detected as the major substance that caused death. Thus, 92% of all drug deaths were associated with the abuse of opioids. Concomitant use of other drugs and alcohol were frequently seen ( Table 2 ). The major findings of the peripheral organs of the drug fatalities are listed in Table 3 . None of the drug and control cases revealed signs of hepatic or pancreatic changes suggestive of long-term excessive alcohol abuse at general autopsy. None of the cases had a history of severe CNS trauma. The macroscopic data on brain weight, signs of brain edema, and increased intracranial pressure did not differ significantly between both groups. In none of the cases could significant herniation with secondary brain stem hemorrhages be detected. The histopathological examination of the brains of controls and of polydrug abusers did not demonstrate changes due to infectious agents or cerebrovascular lesions. The amount and distribution of leptomeningeal fibrosis, congestion and cellular infiltrates as well as brain edema, vascular congestion, perivascular cuffing, perivascular hemorrhages, gliomesenchymal nodules, and any hypoxic nerve cell damage, as assessed in the routinely stained sections, did not statistically differ between both groups. There was no significant correlation between the measured parameters and postmortem interval. There was also no significant difference for the measured parameters between both sexes. In the brains of polydrug abusers there was a significant nerve cell loss in the orbitofrontal, frontal, temporal, parietal and occipital cortex as well as in the caudate nucleus, putamen, globus pallidus, thalamus, substantia nigra, pons, inferior olivary nucleus, dentate nucleus, and the cerebellar granular cell layer as compared with controls. There was no significant difference between both groups for the neuronal number in the hippocampal formation and in the Purkinje cell layer of the cerebellum (Figs. 1 and 2 ). On LFB staining, pallor of the white matter could be detected in six polydrug abuse cases and in one control case, but the difference was not statistically significant. However, in the white matter of all regions examined, ß-APP-immunoreactivity was significantly increased in the brains of polydrug abusers as compared to controls (5 -, unremarkable, not present; ++, enlargement; R, right; L, left; ARDS, acute respiratory distress syndrome; asp, aspiration; cb, contraction bands; co, congestion; ed, edema; Fe+, iron-containing macrophages; fib, fibrosis; fl, fatty liver; hep, hepatitis; hy, hypertrophy; i.m., injection mark; mu, mushroom of blood-tingled foam; pl, portal lymph node; pneumo, bronchopneumonia; tracks, needle tracks. In the brains of polydrug abusers, the numerical density of GFAP-positive astrocytes was significantly reduced in the gray and white matter in all regions examined, with the exception of the hippocampal CA1 region, the pyramidal tract and the cerebellar cortex (Fig. 3) . In the brains of polydrug abusers, the numerical density of CR3/43-positive microglia (i.e., both perivascular and parenchymal) was significantly increased in the orbitofrontal, frontal, parietal, and occipital white matter as well as in the mesencephalon, pons, medulla oblongata, inferior olivary nucleus, and the cerebellum (Fig. 4) . The numerical density of perivascular microglia was significantly increased only in the inferior olivary nucleus, whereas in the other brain regions there was no significant difference between both groups. In contrast, the numerical density of parenchymal microglia was significantly increased in the orbitofrontal, frontal, temporal, parietal and occipital white matter as well as in the mesencephalon, pons, medulla oblongata, inferior olivary nucleus, and the cerebellum. The total extent and distribution of vascular congestion, perivascular cuffing, or perivascular hemosiderin deposits and hemorrhages did not statistically differ between both groups. However, in the brains of polydrug abusers the number of vessels showing hyalinotic thickening was significantly increased in the gray and white matter of the orbitofrontal, frontal, temporal, parietal, and occipital cortex as well as in the caudate nucleus, putamen, globus pallidus, internal capsule, and medulla oblongata (Fig. 5) . The alterations consisted of concentric hyalinotic thickening, sometimes with luminal narrowing (Fig. 6A,B) . In the brains of polydrug abusers, the number of vessels showing endothelial proliferation was significantly increased in the gray and white matter of the orbitofrontal, frontal, temporal, parietal, and occipital cortex as well as in the caudate nucleus and substantia nigra (Fig. 7) . The alterations consisted of a marked endothelial swelling and endothelial cell hyperplasia (Fig. 8A,B) , but they never reached the extent of a true angiogenesis. The results for the alterations of collagen type IV in the vascular basal lamina of the frontal, temporal, parietal, and occipital lobe have been reported previously (6) . Briefly, in the cortical gray and white matter of the brains of polydrug abusers, the number of vessels showing strong immunoreactivity for collagen type IV was significantly reduced, whereas the number of vessels with mild and moderate immunoreactivity was increased as compared to controls. The total numerical density of vessels was not significantly changed. In the other regions examined, similar changes could be observed in the caudate nucleus, mesencephalon, and medulla oblongata but not in the putamen, globus pallidus, internal capsule, thalamus, and pons (Fig. 9A,B) . Although animal and neuroimaging research within the past number of years have clarified various aspects of CNS alterations in the context of drug abuse, systematic morphological studies of the human brain have been lacking. Therefore, the aim of this study was a systematic morphological analysis of the brains of polydrug abusers. The major finding was a widespread neuronal loss, a reduction of GFAP-positive astrocytes, an axonal damage with concomitant microglia activation, as well as reactive and degenerative vascular changes. Although this study was performed on a large well-documented group of polydrug deaths, some variables could not be excluded. The information about the duration and types of drug abused can only be estimated, since the first appearance of the individual in the police records does usually not reflect the actual beginning of drug abuse. Another problem consists of distinguishing between substance-specific effects related to the properties of the drug itself and secondary effects related to lifestyle (e.g., malnutrition, infections, and peripheral diseases). The illicit substances taken by the polydrug abusers in our study reflect the characteristic spectrum that are consumed worldwide, with opioids being the major substances abused (1,7) . Because 92% of our polydrug deaths were associated Forensic Science, Medicine, and Pathology V2-2 Forensic Science, Medicine, and Pathology V2-2 with long-standing opioid abuse, the observed alterations are most probably owing to these substances. Because all subtypes of opiate receptors are expressed on neurons, astrocytes, microglia, and endothelial cells (8, 9) , opioids exert their influence on almost all cell types of the CNS. There are only few reports on histopathological alterations in the brain of drug abusers, predominantly describing edema, vascular congestion, ischemic nerve cell damage, and neuronal loss (7, (10) (11) (12) . These changes have been attributed to toxic primary respiratory failure and are therefore considered as being nonspecific (7) . However, in most of these studies, there was no statistical comparison to a control group and systematic data on frequency or topography of the lesions are lacking. In this study, the findings mentioned above could be confirmed only partially. In polydrug abusers hypoxic nerve cell damage was significantly more often present only in the orbitofrontal, temporal, and occipital cortex. In contrast, there was a widespread neuronal loss in the brains of polydrug abusers, with the exception of the hippocampal formation and the Purkinje cell layer. In both regions there was also no significant reduction of GFAP-positive astrocytes. In addition, there were concomitant vascular alterations in the cortical regions and a reduction of GFAP-positive astrocytes, but no surrounding microglial activation. Within the subcortical regions there was a significant neuronal loss and a reduction of GFAP-positive astrocytes, but only scarce vascular alterations. Therefore, a vascular mechanism of the neu-ronal loss cannot be assumed. An age-associated neuronal reduction can also be excluded because the mean age of the control group was higher than that of the polydrug abusers. Although the neuronal loss could be induced by recurrent hypoxic-ischemic episodes owing to respiratory depression during the intoxicated state, the reduction of GFAP-positive astrocytes argues against such a phenomenon as the sole cause. Furthermore, CNS lesions after global hypoxic-ischemic damage are predominantly seen in the Purkinje cell layer of the cerebellum and the hippocampal formation (13) . Because the nerve cell density in these regions was not significantly changed between both groups, other factors must be of pathogenetic significance. Forensic Science, Medicine, and Pathology V2-2 Besides hypoxic-ischemic nerve cell damage, apoptotic processes can result in neuronal cell death (14) . It has been shown for nearly all drugs of abuse that apoptosis can occur (15) (16) (17) (18) (19) (20) (21) . In these studies it was shown that drugs of abuse can induce neuronal apoptosis by an increased expression of proapoptotic factors (e.g., p53, bax, caspases, endonucleases) and by a decreased expression of the anti-apoptotic oncoprotein bcl-2. Furthermore, a drug-induced decrease of the astrocytic glutamate transporter resulted in an increase of extracellular glutamate with subsequent excitatory nerve cell damage (22) . However, data on the human brain are lacking so far. Another possibility for the neuronal loss might be drug-induced alterations of neurofilament (NF) proteins (23) . In opioid deaths, a marked reduction in total NF proteins, immunoreactive NF proteins, and an aberrant hyperphosphorylation of NF has been demonstrated (24, 25) . Although there were no significant white matter changes seen on myelin-stained sections, our results show a widespread axonal damage in the brains of polydrug abusers. Because this group did not significantly differ from the controls in the pres-Forensic Science, Medicine, and Pathology V2-2 ence of brain edema or signs of increased intracranial pressure, the mechanism of a secondary phenomenon owing to global hypoxia-ischemia cannot explain our findings. Furthermore, the microglial activation selectively in the white matter of polydrug abusers argues against an acute or agonal phenomenon and is indicative of a long-standing progressive process. Based on these findings, it seems likely that drugs of abuse might induce direct toxic-metabolic axonal damage, which might be induced or enhanced by cerebral hypoxia. Besides these direct mechanisms, the activated microglia could increase the axonal damage by the release of cytotoxic substances. The extent of the axonal damage is likely to be underestimated, because ß-APP only detects relatively recently damaged axons (26) , whereas the duration of the abuse of drugs often lasts several years. Therefore, in conjunction with the microglial activation, a chronic-progressive process has to be considered, which might be initiated and supported by drugs of abuse. The alterations might be the morphological correlate of the observed demyelination and hyperintense areas seen on magnetic resonance imaging (2, 27, 28) . Although an increased expression of GFAP in astrocytes (astrogliosis) has been described in various CNS lesions (29) , there are only few data on toxic CNS damage (30) . In an older study, widespread fragmentation and a numerical depletion of astrocytes in the white matter have been reported in the brains of drug deaths (12) . In HIV-positive drug abusers, HIVnegative drug abusers, and non-drug-using controls there was no statistical difference between the three groups in relation to astrocytes (31, 32) . In the prefrontal cortex of opioid deaths, the immunodensity of GFAP was found unchanged (24) . Morphine inhibited astrocytic proliferation in murine cell cultures, and there was a dysregulation of calcium homeostasis and an increase of reactive oxygen species (33) . After cocaine cytes could be associated with our observation and might represent a specific long-term effect of drugs of abuse on astrocytes. Because astrocytes are of eminent importance for the maintenance of the blood-brain barrier (BBB) integrity (38) our findings might also be indicative of an indirect impairment of the BBB. Because GFAP expression is also essential for normal white matter architecture (39), its reduction might contribute to the white matter changes seen in polydrug deaths. The activation of microglia predominantly in the white matter of polydrug abusers is in accordance with the findings of other authors (10, 11, 32) . As mentioned earlier, this activation is most probably induced by the axonal damage and not the result of a direct drug-associated effect. The activation of microglia might lead to the release of cytotoxic substances and a disruption of synaptic integrity. administration, modifications in astrocytic numbers, cell size, and shape complexity were seen (34) . The reduction of GFAP-positive astrocytes suggests druginduced cell damage. One possible mechanism could be the interference of drugs with the GFAP-gene transcription inducing an altered GFAP-phosphorylation (35) . Furthermore, the induction of a cytochrome P450 isoform by drugs of abuse with the generation of free radicals (36) and subsequent damage of astrocytes might be conceivable. In addition, a drug-mediated effect on the astrocytic cytoskeleton, similar to the neuronal neurofilament proteins (23) (24) (25) might contribute to the decrease of GFAP-immunoreactivity. I2-imidazoline receptors are involved in the regulation of the GFAP expression (23) . In the frontal cortex of heroin deaths the density of I2-imidazoline receptors and the immunoreactivity of the related imidazoline receptor protein were decreased (37) . Therefore, a downregulation of I2-imidazoline receptors in astro-Forensic Science, Medicine, and Pathology V2-2 Central nervous system alterations in drug abuse Karch's Pathology of Drug Abuse Brain Imaging in Substance Abuse: Research, Clinical, and Forensic Applications Morphometry in the neurosciences Widespread axonal damage in the brain of drug abusers as evidenced by accumulation of ß-amyloid precursor protein (ß-APP): An immunohistochemical investigation Alterations of the vascular basal lamina in the cerebral cortex in drug abuse: A combined morphometric and immunohistochemical investigation Neuropathology in non-human immunodeficiency virus-infected drug addicts: hypoxic brain damage after chronic intravenous drug abuse Anatomy of CNS opioid receptors Presence of the µ 3 opiate receptor in endothelial cells. Coupling to nitric oxide production and vasodilation Neuropathologic analysis of postmortal brain samples of HIV-seropositive and -seronegative i.v. drug addicts Comparative brain pathology of HIV-seronegative and HIVinfected drug addicts Addiction to opiates: neurologic aspects Perivascular anoxia-ischemia lesions in the human brain Apoptosis in the nervous system Chronic morphine induces up-regulation of the pro-apoptotic Fas receptor and down-regulation of the anti-apoptotic Bcl-2 oncoprotein in rat brain Tetrahydrocannabinol-induced apoptosis of cultured cortical neurones is associated with cytochrome c release and caspase-3 activation Methamphetamine neurotoxicity: necrotic and apoptotic mechanisms and relevance to human abuse and treatment Morphine induces apoptosis of human microglia and neurons Up-regulation of murine double minute clone 2 (MDM2) gene expression in rat brain after morphine, heroin, and cocaine administrations Neuronal apoptosis associated with morphine tolerance: evidence for an opioid-induced neurotoxic mechanism Amphetamines induce apoptosis and regulation of bcl-x splice variants in neocortical neurons Regulation of the glial glutamate transporter GLT-1 by glutamate and δ-opioid receptor stimulation Attenuation of tolerance to opioid-induced antinociception and protection against morphine-induced decrease of neurofilament proteins by idazoxan and other I 2 -imidazoline ligands Regulation of nonphosphorylated and phosphorylated forms of neurofilament proteins in the prefrontal cortex of human opioid addicts Marked decrease of immunolabelled 68 kDa neurofilament (NF-L) proteins in brains of opiate addicts Axonal damage: a key predictor of outcome in human CNS diseases Brain imaging studies of drug abuse: therapeutic implications Radiological and neurological changes in the drug abuse patient. A study with MRI Astrocytes and brain injury Quantitative features of reactive gliosis following toxicant-induced damage of the CNS Relationship of Nef-positive and GFAP-reactive astrocytes to drug use in early and late HIV infection Upregulation of microglia in drug users with and without presymptomatic HIV infection Opioids disrupt Ca 2+ homeostasis and induce carbonyl oxyradical production in mouse astrocytes in vitro: transient increases and adaptation to sustained exposure Astroglial in vivo response to cocaine in mouse dentate gyrus: a quantitative and qualitative analysis by confocal microscopy A selective regional response of cultured astrocytes to methamphetamine Metabolic bioactivation reactions potentially related to drug toxicities Decreased density of I 2 -imidazoline receptors in the postmortem brains of heroin addicts Astrocyte-endothelial interactions and blood-brain barrier permeability GFAP is necessary for the integrity of white matter architecture and long-term maintenance myelination Vasculitis owing to substance abuse Cerebral infarction in adult AIDS patients: observations from the Edinburgh HIV autopsy cohort Cocaine enhances brain endothelial adhesion molecules and leukocyte migration Cocaine activates redox-regulated transcription factors and induces TNF-α expression in human brain endothelial cells Brain capillary lesions produced by cocaine in rats Methamphetamine induces AP-1 and NF-kappaB binding and transactivation in human brain endothelial cells Abnormal cerebral metabolism in polydrug abusers during early withdrawal: a 31 P MR spectroscopy study Cerebral glucose utilization in polysubstance abuse Expression and adhesive properties of basement membrane proteins in cerebral capillary endothelial cell cultures The authors are grateful to Susanne Ring and Miriam Finelli for their excellent technical assistance. The authors thank Dr. Ida C. Llenos for correcting the manuscript. The authors thank Veronika Paetzold, Florence Payrhammer, Katharina Rohrmoser, and Claus Kroehling for performing some of the investigations. This study has been supported by the Friedrich-Baur-Stiftung of the Medical Faculty of the Ludwig-Maximilians-University Munich (Grant No.: 0017/2000).The authors have stated that they do not have a significant financial interest or other relationship with any product manufacturer or provider of services discussed in this article. Forensic Science, Medicine, and Pathology V2-2 1 . Widespread alterations can be detected in the brains of drug abusers using histology, immunohistochemistry, and morphometry. 2. There is a significant nerve cell loss in nearly all brain regions. 3. The numerical density of GFAP-positive astrocytes is reduced in the gray and white matter. 4. The numerical density of CR3/43-positive microglia is increased in the white matter as well as in the brainstem, inferior olivary nucleus, and the cerebellum. 5. There is a reactive endothelial cell proliferation, degenerative hyalinotic thickening, and a loss of immunoreactivity for collagen type IV within the vascular basal lamina. Cerebrovascular accidents are frequently seen in the context of drug abuse, but to date there is no consensus about the cause (1,2). As one possible pathogenetic mechanism, a drug-induced vasculitis has been postulated (40) . However, in our study, vasculitic changes could not be observed, so that we cannot confirm the regular occurence of a vasculitis in drug abusers.In a study of AIDS patients, small-vessel thickening, perivascular space dilatation, pigment deposition, vessel wall mineralization, and perivascular inflammatory cell infiltrates were seen in 50% of then former drug abusers (41) .It has been shown that cocaine enhances the expression of inflammatory cytokines, adhesion molecules, and chemokines on brain endothelial cells with consequent increase in permeability of the BBB (42, 43) . After chronic cocaine administration, thrombosis, endothelial thickening, vascular wall fibrosis, and rupture of the basement membrane of brain capillaries have been observed in rats (44) . Methamphetamine induced disturbances in cellular redox status and upregulation of inflammatory genes in brain endothelial cells (45) .On neuroimaging, small areas of demyelination, focal perfusion deficits, a reduction of the cerebral glucose metabolism, and hyperintense areas have been described in the brains of drug abusers (3, 27, 28, 46, 47) , but to date there is no consensus about the cause or possible morphological substrates. Our study demonstrated profound alterations in the cerebral microvasculature. There was reactive endothelial cell proliferation, degenerative hyalinotic thickening, and a decrease of the collagen type IV content of the vascular basal lamina. A degeneration of the vascular basal lamina can result in an increased permeability and a reduced electrical resistance of the BBB (48) . Therefore, this noninflammatory vasculopathy can be considered as the morphological substrate of a disturbed BBB and might be associated with the alterations seen on neuroimaging.In conclusion, our findings demonstrate that drugs of abuse initiate a cascade of interacting toxic, vascular and hypoxic factors which finally result in widespread disturbances within the complex network of CNS cell-cell interactions.