fluid proteins in men with chronic encephalopathy after exposure to organic solvents.

Cerebrospinal fluid proteins in men with chronicencephalopa thy after exposure to organic solvents. Scand J Work Environ Health 1990;16:423-7. Cerebrospinal fluid was examined for 23 patients with chronic toxic encephalopathy after heavy exposure to organic solvents and 23 healthyage-matched referents. No differences were found betweenthe patients and refer ents with respect to the levels of albumin, immunoglobulin, prealbumin, alpha-I-antitrypsin, beta-2 microglobulin, haptoglobin, or the astroglial cell proteins SI00 and glial fibrillary acidic protein in the cerebrospinal fluid. The albumin ratio was normal for both the patients and the referents. The patient group had had heavy exposure to organic solvents, but its members had not been exposed for at least one year beforethe study. It wasconcluded that, if exposureto organicsolventsaffectsproteinsincerebro spinal fluid, such effects are probably reversible.

Or gan ic solvents have widespread use in paints, glues, and deg reasing agents. The nervou s syste m is th e critical organ for most organic sol ven ts (1). In addition to the acute narcoti c effects of high exposure, chronic neu ropsychiatric disease has been ind icated by case reports and cohort and case -referent studies (2)(3)(4)(5) after lon g-term occupational exposure . In the 1980s the name "chro nic to xic encephalopathy" was coined for th is clinic al syndro me (6,7). The pathogenet ic mechan ism is uncl ear. There are no an imal expe riment al models for mo st o f the sol vents (8) . Sign s of ast rogliosis ha ve, howe ver, been reported in rod ents after longte rm expo sure to various sol vents (9, l O). Ad verse effects on membrane function have also been suggested (7). In two previous studies cerebros pinal flu id (CSF) has been exami ned in patients with chronic toxic encephalopathy, and also in healthy men expo sed to organ ic solvents (11,12). In these investigations signs of slight immunoactivation (11) and increased permeability of the blood-CSF barrier (II, 12) were reported.
The aim of the present investi gation was to study CS F pro te ins in a group of men with clear signs of chronic toxic enceph alopathy a fter heavy, alth ough not recent , expos ure to organic solvents. In the pro tein a na lyses we also included those astroglial markers found to be a ffected in earlier animal exper ime nts .

Subjects
Among all the patients examined bet ween 1976 and 1984 at the clin ic of occupational medicine , those patients were selected whose final dia gnosis was chronic to xic encephalopathy caused by exposure to organic solvent s. The pati ent s selected also had to meet the following criteria: (i) sex male, (ii) age not exceed ing 60 years at th e tim e of th e study, (iii) a cumulati ve expo su re to organ ic so lvents corresponding to at least 10 years of work at th e Swed ish hygieni c standard," 20 years of work at 0.5 times the standard , and so forth (for mixtures of solvents , the sum of th e fr actions of the hygienic standards were used), as evaluated by an expe rien ced occupational hygienist, (iv) no exp osure to solvents for at least one year at the time of the study, (v) symptoms and signs clearly indicating diffuse encephalopathy (eg, memory deterioration, concentration difficulties, irritability, and emotional instability) at the examination by specialists in occupational medic ine, neurology, and (in most case s) psych iatry, (vi) psychometric test performance [test battery (13) including ver ba l, log ical , spatial and perceptual fun ctions, memory, and manual dexterity) support ive of diffuse brain damage as evaluated by an experienced psycholo gist , (vii) exclusion of other diseases that might resul t in neuropsychiatric sympt oms and /or brain dam age (alcoholism, traumatic injuries, presenile Alzheim er ' s disease, cerebrov ascular disease, depr ession , etc) by necess ary examinations, (viii) exclu sion , Examples for the most common solvents used by patients in the present study (hygienic standard from 19' 84): white spirit (oS 22 <170 aromatic compounds) 500 mg/rn ' , xylene 350 mg/rn ' , methyl ethyl ketone 300 mg/rn ' . of diseases that might affect CSF proteins (diabetes, long-term hypertension, nerve root compression, etc). Out of the 122 patients with a diagnosis of chronic toxic encephalopathy only 34 met these rather strict criteria. The most common reasons for exclusion from the study were that the cumulative exposure had not been unequivocally high enough, that high alcohol consumption could not be excluded, or that the patient was too old. As nine of the 34 eligible patients refused to take part in the study and lumbar puncture was unsuccessful for two others, the final number of patients examined in the present study was 23. Most of them had been exposed to solvents as painters (seven house painter s, four ship painters , and four car painters). The mean exposure time was 27 (range 11-42) years .
Twenty-three healthy age-matched (± 4 years) men served as referents. They had to meet the following criteria: no previous or present exposure to organic solvents, no neurological or psychiatric disease or alcohol abuse , no disease that might affect CSF proteins (see the list given for the patients). Nineteen of the referents were selected from workplaces known by the authors. Nine were physicians, six had other whitecollar jobs, and four were blue-collar workers. Another four referents were patients who underwent spinal anesthesia in connection with urological surgery. In all, 30 subjects who met the inclusion criteria had to be asked before 23 agreed to participate in the study.

Medical examination
Each participant's medical history was taken and a new general and neurological examination (except for the four subjects who underwent spinal anesthesia) was carried out. Blood samples were analyzed for serum proteins, liver enzymes, and levels of vitamin B 12 and folic acid (other screening tests had been performed at the initial examination). A lumbar puncture was performed with the subject in the supine position. CSF and blood were stored at -20 CC until analyzed.

Analyses of the cerebrospinal fluid
Protein content was determined as described by Lowry et al (14). Albumin, immunoglobulin (lg) G, prealbumin, alpha-l-antitrypsin, and haptoglobin were mea-sured with the use of electroimmunoassay (15). Beta-2-microglobulin and the astroglial cell proteins SIOO and glial fibrillary acidic protein were determined with enzyme-linked immunosorbent assay (10,16,17). The albumin ratio [CSF albumin/S albumin (S =serum)] was calculated and used as an indicator of the blood-CSF barrier function. The IgG index (CSF IgG/S IgG)/(CSF albumin/S albumin) was used to estimate the local production of IgG within the central nervous system. Furthermore, erythrocytes, polymorphonuclear leukocytes, monocytes, and phagocytes were counted in the CSF. In cases of contamination by blood, the ratios of red blood cells in the CSF versus tho se of the whole blood were used to correct the CSF protein levels for the fractions of plasma proteins that could have contaminated the CSF .

Electrophoretic examinations
Absorption chromatography was performed with the use of the IgG fractions of rabbit antiserum against human serum proteins (Dako, Copenhagen, Denmark, product A 208) (18).
Unconcentrated CSF was passed through the columns and eluted in two fractions, one containing enriched CSF proteins (fraction I) and the other containing proteins antigenically identical to serum proteins (fraction II). Fraction I was eluted with 0.15 M tris-hydrochloric acid in 0.5 M sodium chloride (pH 8.0) and fraction II with 0.15 M glycin hydrochloric acid in 0.5 M sodium chloride (pH 2.8). The samples were concentrated with the use of Pellicon molecular filters (nominal molecular weight limit _10 4 , Millipore) to 3-4 gil.
Isoelectric focusing of the CSF and serum was performed in flat beds of polyacrylamide gel, II x 25 x 0.1 em (19). The gel contained a total of 5 rr /o (g/ 100 ml of gel) polyacrylamide (4.8 % acrylamide and 0.2 % N,N-methylene bisacrylamide), and the pH range was 3.5-9.5. Ampholine (LKB products, Bromrna, Sweden) was added to a final concentration of 2 % and had the following composition: pH 4-6 (0.7 ml), 5-7 (0.7 ml), 9-11 (0.4 m!) , and 3.5-10.0 (1. 8 ml). Twenty microliter s of un concentrated CSF and 20 III of diluted seru m (1 1200) were added to application clot s at the most cathodal part of the gel. Silver staining was performed essentially according to the method of Oakley et al (20) .

Statistical anal yses
In the comparisons of the expo sed gro up with the referents , Student ' s t-test for pa ired ob servations was used. For correlation bet ween variables, linear and multiple linear regression analyses were used.

Results
Age , exposure time, and neurological findings are shown for the subjects in table I. There were slight signs of cerebellar dysfunction and sligh t sensory peripheral neuropathy in several patients but no abnormalitie s in the referents. Two patient s had pathologically increased levels of liver enzyme s, and one pat ient and on e referen t had concentrations o f vita min B 12 that were below the normal limit.
Fo r th ree pat ient s the CSF was conta mina ted with blood . For all the other patients and referents the cell findin gs were normal, with 0-4 monocytes!ml, and no differenc es were found between the pati ents and th e re fer ent s.
Th e seru m an d CSF protein levels ar e show n in table 2. Th e to ta l seru m proteins and the IgG index were higher in the patient gro up . Th eir CSF-lgG was, however, not significantly increased , a nd their albumin ratios were lower than those of the referents. Th ere were no differenc es bet ween the patient s with neurological findings and the other patients with resp ect to CSF al-bumin, the albumin ratio , or the IgG ind ex. Isoelectric focusing of the total CSF proteins showed slight signs o f blood-CSF barrier dysfunction in two pa tients and three referents. There were no abnormalities in the CSF-enr iched fr actions of either group.
The correlations between age , exposure time, and the various proteins in the combined group of expo sed patients and une xposed referents are shown in table 3. Total CSF protein, CSF albumin, albumin ratio, CSF IgG and CSF alpha-I-antitryps in were highl y intercorrelated , as could be expe cted. In th e exposed group there were no significant correlations between the exposure tim e and the various protein levels . The S I00 level was slightly negatively corr elated with age and CSF protein. When the multiple linear regression analysis wa s used, it was shown th at th e variation in the albumin ratio wa s almost completel y a result of the vari ation in CSF albumin. , Patient s N =21, referent s N =22.
• P<0.05. Table 3. Correlations between age, proteins in th e cerebrospinal fluid (CSF), and prote ins in serum in the comb ined group of patients with chronic toxic encephalopath y and unexposed referents (N = 46), Preal bumin, haptoglobin and beta -z-rnlcroglobulin were not significantly correlated with any other variable (not shown in the

Discussion
No evidence of changes in the CSF protein pattern was seen in the patients with chronic toxic encephalopathy after long-term exposure to organic solvents. In an earlier Finnish study (11), there were, similarly, no significant differences in the group mean values of CSF proteins between 33 solvent-exposed subjects and 46 healthy referents. However in 22 of the subjects with recent exposure to solvents, but no disease caused by such exposure, five had a slightly increased CSF albumin level, possibly indicating a blood-CSF barrier dysfunction.
In a Swedish study (12) no differences were seen in the mean CSF protein levels between solvent-exposed patients and referents. However, three of the 17 patients had increased albumin ratios. There was also a significant positive correlation between exposure time and albumin ratio. For seven out of 17 patients a relative increase in a protein band with Ip 4.7 was seen in the isoelectric focusing of the CSF-enriched proteins. The patient group included workers with ongoing exposure and also workers whose exposure to solvents had ceased many years earlier.
In the present study we used strict inclusion criteria to make the diagnosis chronic toxic encephalopathy as certain as possible. We were also able to differentiate between possible permanent effects on the CSF protein pattern and transient effects from solvent exposure, as the patients were examined at least one year after exposure ceased. No signs of blood-CSF barrier dysfunction were found in the patient group as compared with the referents. The power of the present study permits an 80 070 chance of detecting a 20 % increase in albumin concentration or albumin ratio in the patients as compared with the values of the referents (a = 0.05, one-tailed). In our opinion a reasonable conclusion is that, if solvents affect the blood-CSF barrier function, such effects are probably reversible.
In the present study no correlation was seen between exposure time and the albumin ratio. This finding is not in agreement with the results of a previous Swedish study (12). However, exposure time is a crude measure of total exposure to solvents. With the classification of that study, all of our 23 patients would be considered to have been "heavily exposed." The previous finding of a relative increase in a protein band with Ip 4.7 could not be reproduced in the present study.
No increase in astroglial proteins was shown in the CSF of the solvent-exposed group. Such an increase has been demonstrated in brain tissue in animal studies after exposure to neurotoxic agents (9-10). Similar changes might occur in the human brain in toxic encephalopathy caused by organic solvents. Even so, in this study these proteins were not significantly increased in the CSF. In addition, in other diseases with marked astrogliosis, the determination of CSF astroglial protein concentrations has not been a very sue-426 cessful method of differentiation between patients and referents (21). This lack of success might be explained by the limited leakage of these proteins into the CSF from astrogliotic brain tissue and the substantial normal variation between individuals.
The referents in our study, including, for example, nine physicians, were not socioeconomically comparable with the patients. If anything, they should, however, be expected to have a better general health status than solvent-exposed workers, and therefore the selection of referents could hardly explain the "negative" results.
In conclusion, CSF proteins do not seem to be affected one year or more after heavy exposure to organic solvents. CSF analyses could, however, be justified as a method of detecting other neurological diseases.