Long-term observations on tubular and glomerular function in cadmium-exposed persons.

Four groups of cadmium-exposed persons, from different workplaces and with different types of exposure, have been followed for periods of 9-20 years. In one group the total observation time is over 30 years, since they were included in Friberg's original study. The studies include determination of inulin or creatinine clearance, protein excretion and specific indicators of renal tubular dysfunction. The results indicate that once tubular dysfunction is established, it is irreversible, even when it is minor. In some persons it was noted that the development of renal dysfunction seemed to be a multistage process. The initial stage is characterized by an increased excretion of low molecular weight proteins like beta 2-microglobulin and ribonuclease. After a period of several years with no or low exposure, there was a relatively sharp increase in excretion of total proteins and albumin and a decrease in glomerular filtration rate. This is interpreted as being the result of further increases in renal concentration of cadmium and in spread of cadmium along the tubules. Metallothionein absorption in the tubules, its catabolism and synthesis must play an important role for the development and progress of the tubular dysfunction. It was not possible to show that a decrease in glomerular filtration rate occurs before low molecular weight proteinuria.


Introduction
The renal dysfunction caused by cadmium is generally regarded as irreversible. Workers from the group originally studied by Friberg (1) in 1946 were subsequently subjected to several reexaminations (2-6). The first follow-up study was carried out by Friberg and Nystrom in 1952 (2). These authors found that the proteinuria in the examined group had not disappeared; on the contrary, they found that proteinuria had developed in some subjects after cessation of exposure. In 1959,39 of these workers were again re-examined (3), and it was found that there was practically no change in the results of qualitative tests using nitric acid, but the number of trichloroacetic acidpositives had increased. Further studies on workers from this particular group, using quantitative protein determinations, showed that at intervals of 6 months and 3 years, respectively, no major changes had taken place in urine protein excretion (4,5). A new follow-up in 1977 (6) showed *Department of Environmental Hygiene, Karolinska Institute and the National Institute of Environmental Medicine, Stockholm, Sweden. that of 18 workers a few had a further increase of protein excretion, but in most cases excretion rate was unchanged. In no case was there a normalization of urine protein excretion or of the electrophoretic pattern. Thus, even cases with moderate increase in total protein (100-200 mg/day) remained at that level. It has been claimed (7) that the proteinuria is reversible, but this finding was based on qualitative tests only.
By determination of inulin clearance, Friberg (1) found a decrease in glomerular filtration rate (GFR) in 19 workers subject to heavy exposure. These workers and several others were examined in 1959 and inulin clearance determined (8). A relationship between exposure time and decrease in inulin clearance was found.
Since 1959, most Swedish studies have focused on the tubular dysfunction and measurement of protein excretion. However, in several groups of workers, glomerular function has been studied since 1969 by determination of creatinine clearance, serum creatinine and serum 2-microglobulin.
A number of separate studies have been performed in different industries where various types of cadmium exposure exist. In the present paper some preliminary data from these studies of interest to the discussion on mechanisms for and development of renal disease following cadmium exposure are discussed. An overview of these findings has recently been presented (9).

Materials and Methods
Four different groups of workers were studied. The first group consisted of seven workers exposed to cadmium oxide dust in a battery plant. They were first examined by Friberg in 1947 (1) and have subsequently been examined several times during the years 1959-1978. A second group consisted of 16 battery workers exposed to cadmium oxide dust, who were first examined in 1969 (10) and re-examined in 1982. The third group consisted of six workers in a copper-cadmium alloy producing plant. They were first examined in 1969 (10) and re-examined in 1981. The fourth group consisted of 11 men and 3 women. This group was exposed to cadmium dust during polishing of cadmium-plated candle holders for periods of a few months to 4 years. For all subjects exposure ceased in 1972, at which time the hazard was discovered. Repeated examinations were carried out in 1972 to 1975, and 1977 to 1979. During the latter period subjects were examined every fourth month.
Electrophoretic analysis of concentrated urine proteins on paper (4) has been performed routinely since 1959. Urinary cadmium was determined by atomic absorption spectrophotometry (14).

Results
In the first group of seven workers, inulin clear- ance was determined for four workers in 1947. As can be seen in Table 1, both inulin clearance and creatinine clearance were determined in 1959, whereas in 1969 and 1978 only creatinine clearance was determined. Five laboratories used five different methods for the determination of glomerular filtration rate. Therefore, is not possible to make a quantitative evaluation of these data, but it is obvious that there must have been a considerable decrease in glomerular filtration rate over the years much more than can be accounted for by age. Table 2 shows the results of quantitative determination of protein. Quantitative data on protein excretion are not available from 1947. However, in two cases, subjects 27 and 35, free electrophoresis of undiluted urines was performed, which indicates that protein excretion must have been considerable (1). With regard to electrophoretic patterns, there were no major changes in 1959 to 1978, except in case 16, where in 1959 the percentage of albumin was 10%, but during all later examinations was 22-23%. This change in electrophoretic pattern coincided with a large increase in protein excretion and probably also a considerable reduction in GFR, as is shown in Table 1.  1959  1963  1969  1978  1969  1972  1975  1978  16  67  280  790  1230  1160  880  440  530  660  18  72  350  430  750  370  600  390  460  27  70  570  760  640  920  410  540  350  620  32  73  890  850  1400  960  840  35  82  1090  770  730  1110  630  -790  1080  36  71  1100  1080  880  1280  810  910  980  1210  38  69  1300  1090  1400  750  870  740  850 aThe same as in previous papers (3)(4)(5)(6). bIn 1978.
Of the 16 workers in the second group, four had cardiovascular diseases with a considerable intake of drugs, e.g., cardiac glycosides and hypertensive drugs. These four workers were treated separately. In six workers, mean age 65 years, with normal clearance of 02-microglobulin in 1982, it was seen that there had been virtually no change in serum creatinine or creatinine clearance over a period of 13 years (9.2-9.5 mg/L and 120-124 mL/min, respectively). These workers also had a normal excretion of ribonuclease in 1969 and of total protein in 1969 and 1982. The urinary excretion of cadmium was, on the average, 6 jig/g creatinine in 1982.
Six workers, mean age 64 years, had an increased relative clearance of 32-microglobulin in 1982. In this group there were already in 1969 signs of tubular disease since both total protein and ribonuclease excretion were increased. However, the glomerular filtration rate also was already lower than in the normal group (106 mL/ min) in 1969. The urinary excretion of cadmium was on an average 10.7 ug/g creatinine in 1982. Over a period of 13 years there was an increase in serum creatinine, from 10.1 to 12 mg/L, indicating a decrease in glomerular filtration rate, but clearance determinations did not verify this, that, being 106 and 99 mL/min, respectively. The largest changes were seen in the group with cardiovascular disease with a considerable reduction in glomerular filtration rate, from 116 to 81 mL/ min. However, the clearance of ,2-microglobulin was only sightly elevated in this group. In Figure 1, the change of the reciprocal of serum creatinine, which is proportional to the decrease in glomerular filtration rate, is plotted against the relative clearance of 02-microglobulin. It will be seen that only in two cases, with a marked increase in clearance of 02-microglobulin, is there a considerable reduction in glomerular filtration rate, whereas in cases with moderate increase in relative clearance of 02-microglobulin it is not possible to show any trend for a decrease in filtration rate. More details on these findings ** are to be found elsewhere (9). In one case data were also obtained on renal function in 1977, 1978 (two observations) and 1980. Figure 2 shows that, whereas there was a CIB2-m/CICr% considerable reduction in blood and urine cadmium from 1969  In 1969, electrophoresis showed moderate tubular proteinuria, and in 1977 tubular proteinuria with low albumin and a large 02-microglobulin peak. By 1978 the pattern had changed to a "classic" tubular pattern with about 20% albumin.
The third group from the cadmium-copper alloy factory consisted of six persons only. Four of these had normal clearance of 02-microglobulin, and two had an increased clearance rate. The two latter cases have shown a significant reduction in creatinine clearance since 1969. The largest reduction was, however, seen in one person with normal 02-microglobulin clearance and normal protein excretion and a normal electrophoretic pattern. The urin-ary cadmium in this case was 6 gg/g creatinine. It has not been possible yet to obtain more data on this person which would make it possible to evaluate whether there is a relationship to cadmium exposure or some other cause for the change in clearance.
In the fourth group, repeated examinations were made over a period of 28 months. Since analysis was repeated on stored samples after termination of the study, quite reliable estimates were obtained for creatinine and P2-microglobulin in serum and urine, as well as for cadmium in blood and urine. Some data from these studies have been reported elsewhere (9,15). Three persons who displayed the highest cadmium excretion also had small but significant increases in the relative clearance of 02-microglobulin. In two of these cases there was also an increase in total protein excretion. The dose-response curve was of the hockey-stick type, i.e., no effect could be seen at cadmium levels below 5 ig/g creatinine in urine. The increases in 032-microglobulin excretion appeared to occur at about 6 gg/g creatinine.
During the study period there was no general trend toward increases or decreases in serum levels of creatinine or 02-microglobulin. In one woman an increase in serum creatinine was demonstrated. However, in this particular case no increase was seen in serum 2-microglobulin. A slow but continuous increase in urine protein and urinary P2-microglobulin was seen, and this woman also began to excrete more cadmium when protein excretion increased.

Conclusions and Summary
These preliminary data indicate that there is a slow progression of the renal dysfunction in cad-mium workers. It was not possible to find evidence of glomerular function being affected earlier than tubular function. In some cases with originally moderate proteinuria, more rapid changes were observed later, as can be seen in Table 2 and Figure 2. In these cases the protein excretion as well as the albumin content of the urine proteins increased after peak exposure. This fact stresses the need for examination even after cessation of exposure or a decrease in exposure. It is conceivable that the initial defect is at that site in the proximal tubule where metallothionein is initially reabsorbed. High local concentrations of metallothionein and cadmium in the tubular cells will eventually influence the reabsorption of 02-microglobulin. Cadmium will spread upwards or downwards in the tubule, and eventually the reabsorption sites for more anionic proteins like albumin or cationic proteins like lysozyme will be affected. This means that there is a multistage process, depending on at what stage the local critical concentrations are reached in the different segments of the proximal tubule. It is also possible that, depending on intensity of exposure, the metallothioneins delivered to the renal tubules may vary in charge, which may lead to small changes in absorption rates.
It is obvious that more knowledge about metallothionein metabolism and cadmium transport is needed to validate these speculations.