- Email: [email protected]
Inhalation toxicity study of formamide in rats
13,702-713(1989)
FVNDAMENTALANDAPPLIEDTOXICOLOGY
Inhalation Toxicity Study of Formamide in Rats DAVID B. WARHEIT, LAURA A. KINNEY,
MICHAEL C. CARAKOSTAS,
AND PAUL E. Ross
E. I. du Pont de Nemours and Company, Incorporated, Haskell Laboratory for Toxicology and Industrial Medicine, P. 0. Box SO,Elkton Road, Newark, Delaware 19714 Received January IO, 1989; accepted May I&I989 Inhalation Toxicity Study of Formamide in Rats. WARHEIT, D. B., KINNEY, L. A., CARAKOSC., AND Ross, P. E. (1989). Fundam. Appl. Toxicol. 13, 702-713. Formamide is a widely used solvent for the manufacture and processing of plastics, and the possibility for inhalation exposure exists for workers. To assessthe toxicity of repeated inhalation of sublethal concentrations of formamide, three groups of 10 male Crl:CD BR rats each were exposed nose-only for 6 hr/day, 5 days/week for 2 weeks to design concentrations of 100, 500, or 1500 ppm of formamide vapor in air. A control group of 10 male rats was exposed simultaneously to air only. At the end of the exposure period, blood and urine samples were collected for clinical analyses, and 5 rats per group were killed for pathologic examination. The remaining 5 rats per group were retained for a 14-day postexposure observation (recovery) period and then subjected to the same clinical and pathologic examinations. Male rats exposed to I500 ppm had significantly depressed body weights and body weight gains during the exposure and recovery periods compared to controls. Clinical pathologic examinations revealed that decreased platelet and/or lymphocyte counts were observed in rats exposed to 500 or 1500 ppm of formamide. Pathologic examinations revealed compound-related microscopic changes in the kidneys of rats exposed to 1500 ppm formamide. Minimal to severe necrosis and regeneration of renal tubular epithelial cells were observed principally in the outer stripe of the outer medulla and in cortical medullary rays. Based upon the hematologic and clinical chemical parameters measured, the no-observedeffect exposure concentration for repeated inhalation of formamide was considered to be 100 ppm, under the conditions of this study. The findings of treatment-related microscopic lesions in the kidneys as well as increases in mean absolute kidney weights and kidney-to-body weight ratios R?fleCA the target organ tOXiCi@. 0 1989 Society of Toxicology. TAS, M.
Formamide is widely used as a solvent in the manufacture and processing of plastics (Kennedy, 1986; Barlow and Sullivan, 1982). Its toxicity has frequently been assumed to be identical to that of an analog, dimethylformamide (DMF). The target organ for DMF following all routes of exposure is the liver (Kennedy, 1986; Scailteur and Lauwerys, 1987). Formamide applied dermaliy to rats for 3 months resulted in a number of organ weight changes but did not produce liver alterations (BASF Corp., 1985). Since inhalation is the exposure route of industrial importance, it was necessary to investigate the inhalation toxicity of formamide following single and repeated exposures. 0272-0590/89 $3.00 Ccwrigbt 0 1989 by the Society ofToxicology. AU fig&a of reptvduction in any form naemd.
Formamide exposure results in low mammalian cell membrane toxicity (Eichhom et al., 1987) and is slightly toxic in male rats on an acute inhalation basis (4-hr ALC of 2 1 mg/ liter = 11,403 ppm) (Du Pont Co., 1987). Based on these data, design concentrations of 100,500, and 1500 ppm were selected for the subchronic study. The purpose of this study was to determine the toxic effects of repeated inhalation of sublethal concentrations of formamide. Inhalation was selected as a route which simulates possible human exposure. MATERIALS
AND
METHODS
Compound. Technical grade formamide (CAS Regi.+ try No. 75-12-7; !H% pure) was purchased !?om the Ald702
INHALATION
TOXICITY
rich Chemical Co. Formamide was assumed to be stable throughout the exposure phase of the study. Generalexperimentaldesign.Four groups of 10 male CrhCD BR rats (9 weeks old, Charles River Breeding Laboratories, Kingston, NY) were used to assessthe toxic effects of repeated formamide inhalation on body weights, clinical signs, clinical pathology, and pathology parameters. Three test groups were exposed to design concentrations of 100,500, or 1500 ppm of formamide vapor in air (groups II, III, and IV, respectively). A control group of age-, sex-, and weight-matched rats was exposed simultaneously to air only (group I). Rats were exposed nose-only 6 hr/day, 5 days/week for 2 weeks, and were retained for a 14day postexposure recovery period. All rats were monitored daily for body weight changes and clinical signs of toxicity (weekends excluded) throughout the study. At the end of the exposure period (i.e., test Day 12) blood and urine samples were collected from all surviving rats per group for clinical analyses, and five rats per group were killed for pathologic examination. After 14 days of recovery (i.e., test Day 26) the remaining five male rats per group were given similar clinical and pathologic examinations. Atmospheregeneration.For the low level chamber, the test atmosphere was generated by metering formamide into a 500-ml Instathetm IIask using a Harvard, Model 975, compact infusion pump. The flask was heated to 182°C which vaporized the formamide. For the intermediate level and high level exposure chambers, a I OOO-ml, three-neck, round-bottom flask was heated up to 24o’C inside a heating mantle. Nitrogen introduced into the system at the tlask swept the vapor into the 38liter cylindrical glass exposure chamber (Pesce Associates, Kennett Square, PA). Dilution air was introduced between the flask and chamber (approximately 42 liters/mitt). For all exposure chambers, the vapor/air mixtures were dispersed with a dispersion funnel as they entered the 38liter glass exposure chambers. The chamber aerosols were exhausted through scrubbers containing water, dry ice-cold traps, and MSA cartridge filters prior to being vented into the hoods. The control rats were exposed in the same type of exposure chamber to approximately 40 liters/min of air only. A representative schematic of the atmosphere generation system is presented in Fig. 1. Analysisof the testatmospheres.The atmospheric concentration of formamide in each exposure chamber was determined at approximately 45-min intervals during each exposure. Samples of the chamber atmospheres were collected in duplicate from the rats’ breathing zones with tandem midget impingers containing methanol as a trapping solvent. Each sample was injected into a Hewlett-Packard 5880 gas chromatograph equipped with a &me ionization detector. Samples were chromatographed isothemmlly at 7o’C on a 30 m X 0.53-mm i.d. polydimethylsiloxane Megabore column. The atmospheric concentration of formamide was determined by
OF FORMAMIDE
703
comparing peak areas with standard curves calculated daily. Standard samples were prepared at least weekly by diluting a known amount of liquid formamide in a volumetric flask containing methanol. During each exposure, chamber temperatures were measured with mercury thermometers, relative humidities were measured with a Vaisala HMI 3 1F temperature and humidity indicator or Bendix Model 566 psychrometer, and chamber oxygen contents were measured with a Biosystems Model 3 1OORoxygen monitor. Body weights and clinical observations.The means and standard deviations of the starting weights for the four groups, respectively, were 230.9 ? 8.1 (control), 230.7 + 7.0 (100 ppm group), 232.4 + 9.2 (500 ppm group), and 236.3 ? 8.5 (1500 ppm group). During the exposure period, all rats were weighed and observed for clinical signs of toxicity before each exposure. Rats could not be observed during exposure due to the physical design of the chambers. However, group observations for clinical signs were taken immediately following each exposure. During the recovery period, all rats were weighed and observed daily, weekends excluded, except when warranted by the rats’ condition. Clinical measurements.Urine samples were collected over a 16-hr period from all surviving rats after the ninth exposure and from remaining rats on the 13th day of recovery. Samples were analyzed for volume, osmolality, and standard biochemical parameters using a commercially available urine dipstick (Multi&x, Ames Division, Miles Laboratories, Elkhart, IN). The sediment from each sample was also examined microscopically. Blood samples were obtained from all surviving rats after the 10th exposure (test Day 12) and from the remaining rats on the 14th day of recovery (test Day 26). Whole blood samples were analyzed for erythrccyte count, hemoglobin concentration, hematocrit, platelet count, and total and differential leukocyte counts. Wintrobe indices were also calculated. Serum samples were analyzed for alkaline phosphatase, alanine aminotransferase, and aspartate aminotransferase activity and for concentrations of urea nitrogen, creatinine, total protein, and cholesterol. Pathology.The 10 male rats per group were each sub divided into groups of five based on computer-generated random number tables. The 6rst 5 rats per group were killed after the 10th exposure by sodium pentobarbital anesthesia and exsanguination (3 from group IV), and the remaining rats were killed on the 14th day of recovery for gross and histopathologic examinations (4 from group IV). The lungs, liver, kidneys, spleen, and testes were weighed at necropsy, and representative samples of the following tissues were obtained for microscopic examination: heart, lungs, mesenteric lymph nodes, nasal cavities (nose), trachea, liver, pancreas, esophagus, stomach, duodenum, jejunum, ileum, cecum, colon, rectum, kidneys, urinary bladder, bone marrow (sternal), spleen, thymus, thyroid gland, adrenal glands, brain, eyes,testes,
704
WARHEIT
ET AL.
RATS LOADED NOSE ONLY
I
TYPE K THERMOCOUPLE THERMOMETER FLOWMETER llm GENERATION NITROQEN
-I rnr i.xl VARIAC
51 .QY ’ -,--.-
II
DRY ICE CONDENSER
MSA PARTICULATE FILTER
II
THREE-FLASK EXHAUST
FIG. I. Schematic demonstrating the formamide vapor generation system. Briefly, a vapor atmosphere of formamide was generated by a syringe driving the test material into a heated flask. Nitrogen was introduced at the flask and swept the resulting vapor through a dispersion funnel and into the test chamber. Dilution air and oxygen were introduced to the system prior to entering the chamber. The chamber exhaust was drawn through a scrubber containing water, a dry ice-cold trap, and an MSA cartridge filter prior to being discharged into the hood.
epididymides, and any other organs or tissues with gross lesions, The scheme for microscopic grading of tissue lesions was minimal, mild, moderate, and severe. Statistical analyses.Mean body weights and body weight gains for test rats were compared to controls during the exposure and recovery periods. Data were statistically analyzed by one-way analysis of variance (ANOVA). Exposure group values were compared to controls by the least significant difference test and Dunnett’s test when the ratio of variance (F) indicated a significant among-to-within group variation. Significant differences were declared at the 0.05 probability level. For analysis of clinical measurements, ANOVA and Bartlett’s test were calculated for each sampling time. When the Ftest from ANOVA was significant, Dunnett’s test was used to compare means from the control groups and each of the groups exposed to formamide.
RESULTS Chamber atmosphere analysis. During the exposure phase of the study, the mean chamber temperatures for each of the groups were: control, 23.6 f 0.8”C; 100 ppm, 27.9 + O.YC; 500 ppm, 26.8 + 0.9”C; and 1500 ppm, 27.0
& 0.9% In the control and test chambers, relative humidities ranged from 3 1 to 5 I%, and chamber oxygen contents ranged from 20 to 2 1%. Overall mean atmospheric concentrations are presented in Table 1. In general, the determined exposure concentrations were in close agreement with the design concentrations and thus, for purposes of this report, exposures in each group were considered to be 100,500, and 1500 ppm. Body and organ weight analyses. No biologically significant differences in mean body weights were observed in rats exposed to 100 or 500 ppm of formamide (Fig. 2). A few instances of statistically significant reductions in body weight gains were observed in rats exposed during the first week of exposure to 100 and 500 ppm of formamide. Rats exposed to 1500 ppm had significantly depressed mean body weights which were evident following the fourth exposure and lasted until the sixth day of the postexposure recovery period (Fig. 2). Similarly, mean body weight gains were
INHALATION
TOXICITY
705
OF FORMAMIDE
TABLE 1 MEAN
ATMOSPHERIC
CONCENTRATIONS
OF FORMAMIDE” Analyzed
Group no. I II III IV
concentration (ppm)
Design concentration
Mean
SD
Range
0 mm 100 mm 500 Pm-n 1500 ppm
-b 113 500 1504
33.8 97.4 284
56.2-280 148-84 1 449-2215
LIRepresents the mean, standard deviation, and range of all samples from all exposures (approximately per exposure chamber). b Values which were not measured.
significantly reduced during the first week of exposure in rats exposed to 1500 ppm formamide (Table 2). Statistically and biologically significant reduced mean final body weights and increased
100 samples
kidney-to-body weight ratios were observed after the 10th exposure in rats exposed to 1500 ppm of formamide (data not shown). Mean absolute kidney weights were elevated and were considered to be biologically,
380.00
2
320.00
240.00
0.00 -
5.00
10.00
CONTROL LOW
DAYS 0 wm 100 ppm
16.00 ON TEST r-a -
20.00 INT HIGH
26.00 SO0 1500
30.00
ppm ppm
FIG. 2. Growth curve for male rats exposed to 0,100,500, and 1500 ppm formamide. Significant declines in mean body weights are obvious in animals exposed to 1500 ppm when compared to controls.
706
WARHEIT
ET AL.
TABLE 2 MEAN
BODY
WEIGHT GAINS (GRAMS)
FOR MALE
RATS EXPOSED
TO 0,100,500,
OR 1500 PPM
FORMAMIDE
Group (Concentration) Days on test
I (0 mm)
II (100ppm)
III (500 r-m)
IV (1500 mm)
l-5” 8-12’ 1-12’ 13-26d
19.4 (5.48) 9.8(10.1) 54.0 (15.6) 83.6(11.4)
22.2 (6.36) 13.1 (12.5)
15.1 (6.61) 6.2 (14.0) 43.7 (20.5) 84.3 (19.6)
-9.2 (19.0)* 13.0 (23.5) -12.2 (19.4)* 103.2 (17.0)
59.5(16.1)
89.5 (7.92)
Note. Values in parentheses are standard deviations. a First week of exposures. b Second week of exposures. ’ Exposure period (combined). d Recovery period. * Significantly different from controls by least significant difference test (P < 0.05).
though not statistically, significant. Following the 14day recovery period, rats had statistically and biologically significant reduced mean final body weights and elevated mean absolute kidney weights and kidney-to-body weight ratios. Clinical observations.Immediately following exposures, rats in all groups (including controls) frequently had colored discharges of the eyes and/or nose (i.e., chromodacryorrhea) and/or diarrhea, effects common in rats under restraint. No other clinical signs of toxicity were observed following exposure. During the exposure period, low incidences of diarrhea were observed in all groups. In addition, a low incidence of weakness and hunched posture was observed in rats exposed to 1500 ppm formamide. Following termination of the exposure period, chromodacryorrhea was observed in two rats from the control group and in two rats exposed to 1500 ppm formamide. Five rats exhibited diarrhea in the 1500 ppm group during the recovery period. Summaries of clinical observations for male rats are presented in Table 3. Clinical pathology.Relevant clinical laboratory results are given in Table 4. A statistically significant decrease in the platelet count was observed in rats exposed to 500 or 1500
ppm of formamide. The decreased platelet count was considered larger than could be accounted for by expected analytical or biologic variation alone, and therefore was considered due to formamide exposure. A decrease in circulating lymphocytes in rats exposed to 1500 ppm was also observed. Both the relative thrombocytopenia (i.e., compared to control rats) and lymphopenia were also present after the 1Cday recovery period. Results for urea nitrogen, creatinine, urine volume, and urine osmolality were within the expected range of normal biologic variation and did not indicate any significant change in renal function. Differences between control and treatment group results for the remaining clinical laboratory parameters measured were either not statistically significant or were considered not biologically significant based on this laboratory’s experience with male Crl: CD BR rats. Pathology.One 1500 ppm rat was found dead on study Day 3. Tissues from this rat were inadvertently discarded and thus were unavailable for microscopic examination. Another 1500 ppm rat was found dead on study Day 9. One moribund 1500 ppm rat was killed in extremis on Day 11 of the study. In addition, there were no treatment-related gross findings in study rats.
INHALATION
TOXICITY
707
OF FORMAMIDE
TABLE 3
Exposure group (concentration) I (0 m-d
Observation
II (l~m-4
III (500 mm)
Number of rats exhibiting sign* Colored discharge eye(s) left Colored discharge eye(s) right Colored discharge nose Hunched over Diarrhea Weak Total group incidence
l(l5)’ l(l5) 0 3 (4) 0
0 0 0 0 2 (9) 0
0 0 0 0 l(3) 0
6
2
1
1(16)
1(16) 0
2(11) 3 (3) 5 (7 2 (3) 13
’ Excluding clinical signs observed during or immediately following exposure. b Ten rats per group. ’ The number in parentheses is the median day on test that the sign was first observed.
TABLE 4 CLINICALLABORATORYREWLTSFORMALERATSEXPOSEDTOFORMAMIDE Group exposure concentration (ppm) Test
Days on test
0
100
500
1500
Total leukocyte count (X 103/P1) Lymphocyte count (X 1031PI) Platelet count (X 1O’lPl) Serum urea nitrogen OwzM) Serum creatinine bw/dU Urine volume (ml/l6 hr) Urine osmolality (mOWKg)
12” 26’ 12 26 12 26 12 26 12 26 12 26 12 26
14.9 (3.3)C 16.8 (3.0) 11.1(2.1) 12.1 (2.4) 1142 (273) 1180 (268) 17(3) 19 (5) 0.6 (0.1) 0.7 (0.1) 7.2 (4.1) 10.9 (7.7) 1410 (452) 1499 (547)
14.1 (4.4) 14.6 (2.2) 10.1 (3.2) ll.O(l.3) lOSO(184) 1105(124)
13.6 (3.7) 16.0 (3.2) 9.8 (2.5) 11.2 (2.8) 882(119)* 83 1 (69)*
lO.l(l.2) 11.3 (3.6) 6.9 (0.6)* 7.9 (2.0) 847 (90)* 147 (294)*
14 (2) 17 (2) 0.6 (0.0) 0.6 (0.0) 5.1(1.3) 9.2 (1.6) 1497 (406) 1847 (405)
15(2)
21(4)
16(1)
16(l)
0.6 (0.0) 0.6 (0.1) 4.5 (1.2) 7.6 (0.9) 1870 (535) 2293 (161)
0.6 (0.0) 0.6 (0.0) 7.2 (4.2) 10.1 (2.6) 1786(1514) 1426 (257)
a Twelve days on test with exposure to formamide 6 hr per day for 10 of the 12 days, n = 10 for all groups, except 15OOppmwheren=7. b Twenty-six days on test with 10 days of exposure (6 hr/day) to formamide during the first 12 days followed by 14 days of recovery (no exposure), n = 5 for all groups, except 1500 ppm where n = 4. c Group mean and standard deviation. * Statistically significant difference from control group mean, 01= 0.05.
708
WARHEIT
ET AL.
TABLE 5 RENAL
Rat no.
1 2 3 4 5 6
LESIONS
IN RATS EXPOSED
TO
1500 ppm FORMAMIDE
Disposition
FD FD ICE KD KD KD
(Day (Day (Day (Day (Day (Day
3) 9) 11)
12) 12) 12)
FOR
2 WEEKS
Lesion
Grade
2-week exposure II Necrosis (tubular epithelium) Regeneration (tubular epithelium) Regeneration (tubular epithelium) Regeneration (tubular epithelium) Regeneration (tubular epithelium)
Severe Moderate Moderate Moderate Minimal
2-week exposure + 2-week recovery 1 2 3 4
KD KD KD KD
(Day (Day (Day (Day
26) 26) 26) 26)
Regeneration Regeneration Regeneration Regeneration
(tubular (tubular (tubular (tubular
epithelium) epithelium) epithelium) epithelium)
Mild Mild Mild Mild
Note. FD, found dead (study day); KE, killed in extremis (study day); KD, killed by design (study day). DTissues inadvertently discarded.
Compound - related microscopic lesions were present in the kidneys of male rats exposed to 1500 ppm of formamide (Table 5). Necrosis and regeneration of tubular epithelium were zonal, occurring in the outer medulla and radiating up medullary rays in the cortex. Lesions in affected rat kidneys were striking compared to control rat kidneys (Figs. 3 and 4). Necrosis of tubular epithelium was severe in one 1500 ppm rat found dead on study Day 9 (Figs. 5 and 6). All tubules in the outer stripe of the outer medulla were affected and many tubules in the inner stripe of the outer medulla and cortical medullary rats were also affected. In most tubules, the lining epithelial cell architecture was replaced by granular casts of desquamated epithelial cells with total karyolysis. Regeneration of tubular epithelium was present in one moribund rat (moderate) on study Day 11, in three rats (one minimal, two moderate) killed by design on study Day 12, and in four rats (all mild) killed on study Day 26. Regenerating tubules were lined by basophilic cuboidal cells, some of which were in mitosis, and lumens contained occasional desquamated necrotic cells. Peritubular fibrosis was a feature
of some regenerative tubules (Figs. 7 and 8). Other tubules were dilated and lined by flattened epithelium and some of these contained much mineralized necrotic debris. Occasionally, mineralized necrotic debris was circumscribed by granulomatous inflammation. The alterations in kidney weights and the histopathology findings reflect the target organ toxicity. Based on the renal lesion produced by formamide exposure under the conditions of this study, the no-observed-effect exposure concentration for histopathology was considered to be 500 ppm. DISCUSSION Formamide-induced necrosis and regeneration of renal tubular epithelial cells in male rats exposed to 1500 ppm via inhalation were zonal in the outer stripe of the outer medulla and in cortical medullary rays. This lesion distribution is consistent with damage to the pars recta (S3, P3) of the proximal tubule (Owen, 1986). The mechanism of formamide toxicity to cells of this segment of the tubule
INHALATION
TOXICITY
709
OF FORMAMIDE
FIG. 3. Normal morphology of a control rat kidney. C, cortex; 0, outer stripe of outer medulla; I, inn1er medulla. H&E X 16.
is UI nknown. According to a recent review of rem il morphology (Bulger, 1986), necrosis in this region of the tubule can be produced by exr3osure to mercuric chloride, uranyl nitrate, cis-1platinum, renal arterial occlusion in rats, hemorrhagic shock.
FIG.
At 0 days recovery, rats exposed to 1500 ppm formamide had statistically and bi ologitally significant decmased mean body weights and higher kidney-to-body weight rati
4. Normal morphology of outer stripe of outer medulla of a control rat kidney. H&E
X
160.
710
WARHEIT
ET AL
FIG. 5. Kidney ofa rat exposed to 1500 ppm formamide via inhalation found dead on study Day 9. Note severe necrosis and loss of tubule architecture in medulla and in medullary rays in the cortex. H&E X 16.
logi tally, although not statistically significan t. Following the 1Cday recovery period, the alterations in these parameters persisted. FJormal urine volume and osmolality result .sindicate that at least one-third of the re-
nal tubules remained functional (Dun .can and Prasse, 1986). Normal serum urea ni trogen and creatinine concentrations indic xte that glomerular filtration was likewise not affected enough to cause reduction in rz:nal
FIG. 6. Kidney of rat exposed to 1500 ppm formamide via inhalation and found dead on study Day 9. Note coagulation necrosis of most tubular epithelial cells in the outer stripe of the outer medulla. H&E x160.
INHALATION
TOXICITY
OF FORMAMIDE
711
FIG. 7. Kidney of rat 14 recovery days after inhalation exposure to i 500 ppm formamide. Regenerating tubules in the medulla and medullary rays in the cortex are mildly dilated. H&E X 16.
function (i.e., less than a 75% reduction in GFR). While the statistically significant increase in urea nitrogen concentration is compatible with the renal lesions observed histo-
logically, the small difference in urea nitrogen results between the control and 1500 ppm groups was not considered to be predictive for renal disease.
FIG. 8. Kidney of rat 14 recovery days after inhalation exposure to 1500 ppm formamide. Some regenerating tubules in the outer stripe of the outer medulla are surrounded by excess fibrous connective tissue (arrowheads) H&E X 160.
712
WARHEIT
The statistically significant thrombocytopenia observed in the 500 and 1500 ppm groups was considered to have little clinical significance. Rats normally have a very high circulating platelet count (850-1200 X 103/ ~1) (Jain, 1986). Rats in the two affected groups were in no danger of hemorrhaging spontaneously given the large number of platelets still remaining. Therefore, the term “thrombocytopenia” is used in this paper to describe a decrease in platelet count relative to the control group and not to indicate a clinically significant coagulopathy. However, compared to the control group platelet count and reference ranges established in this laboratory, the platelet counts in rats in the 500 and 1500 ppm groups were decreased more than could be accounted for by analytical or normal physiologic factors. Large platelets (shift platelets), indicating enhanced thrombopoiesis, were not observed in blood smears from thrombocytopenic rats. Sections of bone marrow appeared to contain a normal number of megakaryocytes. So neither active platelet destruction (with a consequent increase in platelet production indicated by shift platelets and increased marrow megakaryocytes) nor decreased platelet production (indicated by decreased marrow megakaryocytes) was determined to be the cause for the thrombocytopenia. Conceivably, the effects on circulating platelet homeostasis were too mild to cause a visible change in thrombopoiesis. Other possibilities include platelet sequestration in the lung or spleen or a depression of only extramedullary thrombopoiesis (e.g., in the lung). However, neither of these two possible explanations could be confirmed. One of the four rats from the 1500 ppm group remaining at the end of the recovery period had a much lower platelet count than the other three rats (309,000/~1). This result was responsible for the large standard deviation in the mean platelet count at this time point. The mean platelet count for the other three rats in this group was 893,000/~1 (+44,000), which was still considered slightly
ET
AL.
low. The reason for the marked decrease in the platelet count in one rat after a relatively prolonged period of recovery is not clear. Other thrombocytopenic rats exposed to 500 or 1500 ppm did not exhibit this marked decrease in platelet count, but neither did they show much evidence of recovery toward control group counts. Whether this indicates irreversible damage to thrombopoiesis or a residual effect that requires more time to recover is unclear. The decrease in mean lymphocyte count, observed only in the 1500 ppm group, suggests a stress-induced increase in corticosteroid levels, secondary to treatment-related effects. The diarrhea and hunched posture observed in the 1500 ppm group rats are also compatible with increased stress. The mean lymphocyte count for the 1500 ppm group after the recovery period was still considered slightly decreased based on the laboratory reference range. There is a paucity of data regarding the inhalation toxicity of formamide. Previous reports from our laboratory have suggested that toxicity by inhalation exposure was low. In these studies, no signs of toxicity were detected in rats given a single exposure of formamide for 6 hr at 3900 ppm or in two rats exposed for 10 days duration at concentrations approximating 1500 ppm vapor (Du Pont Co., 1952) and no indications of organ damage were observed following pathologic evaluation. However, in another study, formamide was slightly toxic in male rats on an acute inhalation basis. In this study, the 4-hr approximate lethal concentration was determined to be 2 1 mg/liter ( 11,403 ppm). Clinical signs following acute inhalation included lethargy, hunched posture, ocular and/or nasal discharges, and slight to severe weight losses (Du Pont Co., 1987). The studies presented here indicate that 10 repeated exposures to formamide at concentrations of 1500 ppm produce nephrotoxicity, characterized by necrosis and regeneration of renal tubular cells. Moreover, formamide exposure re-
INHALATION
TOXICITY
sulted in a thrombocytopenia and lymphopenia. The current TLV for formamide is based essentially on information regarding the toxicity of an analog, dimethylformamide. The target organ for DMF is the liver. We have demonstrated here that the liver is not affected by repeated formamide exposures. Moreover, in addition, we have shown that compound-related lesions were observed in the kidneys of male rats exposed to 1500 ppm of formamide. Under the conditions of this study, the noobserved-effect exposure level for male rats is 100 ppm, based on the hematologic and clinical chemical parameters measured in the rats exposed to 500 ppm formamide. Clinical pathologic examinations revealed that rats exposed to 500 and 1500 ppm of formamide were thrombocytopenic after 10 days of exposure and following 14 days of recovery. The biological significance of this mild thrombocytopenia was considered to be equivocal. Based upon the histopathologic finding of renal lesions at 1500 ppm, as well as the decrease in circulating platelets and lymphocytes at 1500 ppm, 500 ppm appears to be the exposure concentration at which adverse effects begin to occur in male rats (i.e., 500 ppm is the no-adverse-effect-exposure level). ACKNOWLEDGMENTS This study was supported by the Central Research and Development Department, E. I. du Pont de Nemours
713
OF FORMAMIDE
and Co., Inc. The authors acknowledge J. Brent Rhodes, Mark Hartsky, and Lena Achinko for their technical expertise. We also thank Gerald L. Kennedy, Jr., David P. Kelly, and Dr. Rudolph Valentine for contributing important insights into the preparation of this manuscript.
REFERENCES S. M., AND SULLIVAN, F. M. (Eds.) (1982). Formamides. In Reproductive Hazards of Industrial Chemicals, pp. 346-359 Academic Press, London. BASF Corp. (1985). Unpublished data. BULGER, R. E. (1986). Kidney morphology: Update 1985. Toxicol. Pathol. 14, 13-25. DUNCAN, J. R., AND PRASSE, K. W. (1986). Veterinary Laboratory Medicine, 2nd ed., pp. 153-173. Iowa State Univ. Press, Ames. Du Pont Co. (1952). Unpublished data. Du Pont Co. (1987). Unpublished data. BARLOW,
EICHHORN, U., KL~~KING, KLOCKING, H. P. (1987).
R., S~HWEIZER,
H., AND
Cell membrane toxicity detected with the chromium-5 1 release test. Arch. Toxicol. Suppl. 11,334-337. JAIN, N. C. ( 1986). Schalm ‘s Veterinary Hematology, 4th ed., pp. 288-298. Lea and Febiger, Philadelphia. KENNEDY, G. L., JR. (1986). Biological effects of acetamide, formamide, and their monomethyl and dimethyl derivatives. CRC Crit. Rev. Toxicol. 17, 129182. OWEN, R. A. (1986). Acute tubular lesions, kidney, rat. In Monographs on Pathology of Laboratory Animals, Urinary System (T. C. Jones, U. Mohr, and R. D. Hunt, Eds.), pp. 229-239. Springer-Verlag, New York. SCAILTEUR, V., AND LAUWERYS, R. R. (1987). Dimethylformamide (DMF) hepatotoxicity. Toxicology 43, 231-238.
FVNDAMENTALANDAPPLIEDTOXICOLOGY
Inhalation Toxicity Study of Formamide in Rats DAVID B. WARHEIT, LAURA A. KINNEY,
MICHAEL C. CARAKOSTAS,
AND PAUL E. Ross
E. I. du Pont de Nemours and Company, Incorporated, Haskell Laboratory for Toxicology and Industrial Medicine, P. 0. Box SO,Elkton Road, Newark, Delaware 19714 Received January IO, 1989; accepted May I&I989 Inhalation Toxicity Study of Formamide in Rats. WARHEIT, D. B., KINNEY, L. A., CARAKOSC., AND Ross, P. E. (1989). Fundam. Appl. Toxicol. 13, 702-713. Formamide is a widely used solvent for the manufacture and processing of plastics, and the possibility for inhalation exposure exists for workers. To assessthe toxicity of repeated inhalation of sublethal concentrations of formamide, three groups of 10 male Crl:CD BR rats each were exposed nose-only for 6 hr/day, 5 days/week for 2 weeks to design concentrations of 100, 500, or 1500 ppm of formamide vapor in air. A control group of 10 male rats was exposed simultaneously to air only. At the end of the exposure period, blood and urine samples were collected for clinical analyses, and 5 rats per group were killed for pathologic examination. The remaining 5 rats per group were retained for a 14-day postexposure observation (recovery) period and then subjected to the same clinical and pathologic examinations. Male rats exposed to I500 ppm had significantly depressed body weights and body weight gains during the exposure and recovery periods compared to controls. Clinical pathologic examinations revealed that decreased platelet and/or lymphocyte counts were observed in rats exposed to 500 or 1500 ppm of formamide. Pathologic examinations revealed compound-related microscopic changes in the kidneys of rats exposed to 1500 ppm formamide. Minimal to severe necrosis and regeneration of renal tubular epithelial cells were observed principally in the outer stripe of the outer medulla and in cortical medullary rays. Based upon the hematologic and clinical chemical parameters measured, the no-observedeffect exposure concentration for repeated inhalation of formamide was considered to be 100 ppm, under the conditions of this study. The findings of treatment-related microscopic lesions in the kidneys as well as increases in mean absolute kidney weights and kidney-to-body weight ratios R?fleCA the target organ tOXiCi@. 0 1989 Society of Toxicology. TAS, M.
Formamide is widely used as a solvent in the manufacture and processing of plastics (Kennedy, 1986; Barlow and Sullivan, 1982). Its toxicity has frequently been assumed to be identical to that of an analog, dimethylformamide (DMF). The target organ for DMF following all routes of exposure is the liver (Kennedy, 1986; Scailteur and Lauwerys, 1987). Formamide applied dermaliy to rats for 3 months resulted in a number of organ weight changes but did not produce liver alterations (BASF Corp., 1985). Since inhalation is the exposure route of industrial importance, it was necessary to investigate the inhalation toxicity of formamide following single and repeated exposures. 0272-0590/89 $3.00 Ccwrigbt 0 1989 by the Society ofToxicology. AU fig&a of reptvduction in any form naemd.
Formamide exposure results in low mammalian cell membrane toxicity (Eichhom et al., 1987) and is slightly toxic in male rats on an acute inhalation basis (4-hr ALC of 2 1 mg/ liter = 11,403 ppm) (Du Pont Co., 1987). Based on these data, design concentrations of 100,500, and 1500 ppm were selected for the subchronic study. The purpose of this study was to determine the toxic effects of repeated inhalation of sublethal concentrations of formamide. Inhalation was selected as a route which simulates possible human exposure. MATERIALS
AND
METHODS
Compound. Technical grade formamide (CAS Regi.+ try No. 75-12-7; !H% pure) was purchased !?om the Ald702
INHALATION
TOXICITY
rich Chemical Co. Formamide was assumed to be stable throughout the exposure phase of the study. Generalexperimentaldesign.Four groups of 10 male CrhCD BR rats (9 weeks old, Charles River Breeding Laboratories, Kingston, NY) were used to assessthe toxic effects of repeated formamide inhalation on body weights, clinical signs, clinical pathology, and pathology parameters. Three test groups were exposed to design concentrations of 100,500, or 1500 ppm of formamide vapor in air (groups II, III, and IV, respectively). A control group of age-, sex-, and weight-matched rats was exposed simultaneously to air only (group I). Rats were exposed nose-only 6 hr/day, 5 days/week for 2 weeks, and were retained for a 14day postexposure recovery period. All rats were monitored daily for body weight changes and clinical signs of toxicity (weekends excluded) throughout the study. At the end of the exposure period (i.e., test Day 12) blood and urine samples were collected from all surviving rats per group for clinical analyses, and five rats per group were killed for pathologic examination. After 14 days of recovery (i.e., test Day 26) the remaining five male rats per group were given similar clinical and pathologic examinations. Atmospheregeneration.For the low level chamber, the test atmosphere was generated by metering formamide into a 500-ml Instathetm IIask using a Harvard, Model 975, compact infusion pump. The flask was heated to 182°C which vaporized the formamide. For the intermediate level and high level exposure chambers, a I OOO-ml, three-neck, round-bottom flask was heated up to 24o’C inside a heating mantle. Nitrogen introduced into the system at the tlask swept the vapor into the 38liter cylindrical glass exposure chamber (Pesce Associates, Kennett Square, PA). Dilution air was introduced between the flask and chamber (approximately 42 liters/mitt). For all exposure chambers, the vapor/air mixtures were dispersed with a dispersion funnel as they entered the 38liter glass exposure chambers. The chamber aerosols were exhausted through scrubbers containing water, dry ice-cold traps, and MSA cartridge filters prior to being vented into the hoods. The control rats were exposed in the same type of exposure chamber to approximately 40 liters/min of air only. A representative schematic of the atmosphere generation system is presented in Fig. 1. Analysisof the testatmospheres.The atmospheric concentration of formamide in each exposure chamber was determined at approximately 45-min intervals during each exposure. Samples of the chamber atmospheres were collected in duplicate from the rats’ breathing zones with tandem midget impingers containing methanol as a trapping solvent. Each sample was injected into a Hewlett-Packard 5880 gas chromatograph equipped with a &me ionization detector. Samples were chromatographed isothemmlly at 7o’C on a 30 m X 0.53-mm i.d. polydimethylsiloxane Megabore column. The atmospheric concentration of formamide was determined by
OF FORMAMIDE
703
comparing peak areas with standard curves calculated daily. Standard samples were prepared at least weekly by diluting a known amount of liquid formamide in a volumetric flask containing methanol. During each exposure, chamber temperatures were measured with mercury thermometers, relative humidities were measured with a Vaisala HMI 3 1F temperature and humidity indicator or Bendix Model 566 psychrometer, and chamber oxygen contents were measured with a Biosystems Model 3 1OORoxygen monitor. Body weights and clinical observations.The means and standard deviations of the starting weights for the four groups, respectively, were 230.9 ? 8.1 (control), 230.7 + 7.0 (100 ppm group), 232.4 + 9.2 (500 ppm group), and 236.3 ? 8.5 (1500 ppm group). During the exposure period, all rats were weighed and observed for clinical signs of toxicity before each exposure. Rats could not be observed during exposure due to the physical design of the chambers. However, group observations for clinical signs were taken immediately following each exposure. During the recovery period, all rats were weighed and observed daily, weekends excluded, except when warranted by the rats’ condition. Clinical measurements.Urine samples were collected over a 16-hr period from all surviving rats after the ninth exposure and from remaining rats on the 13th day of recovery. Samples were analyzed for volume, osmolality, and standard biochemical parameters using a commercially available urine dipstick (Multi&x, Ames Division, Miles Laboratories, Elkhart, IN). The sediment from each sample was also examined microscopically. Blood samples were obtained from all surviving rats after the 10th exposure (test Day 12) and from the remaining rats on the 14th day of recovery (test Day 26). Whole blood samples were analyzed for erythrccyte count, hemoglobin concentration, hematocrit, platelet count, and total and differential leukocyte counts. Wintrobe indices were also calculated. Serum samples were analyzed for alkaline phosphatase, alanine aminotransferase, and aspartate aminotransferase activity and for concentrations of urea nitrogen, creatinine, total protein, and cholesterol. Pathology.The 10 male rats per group were each sub divided into groups of five based on computer-generated random number tables. The 6rst 5 rats per group were killed after the 10th exposure by sodium pentobarbital anesthesia and exsanguination (3 from group IV), and the remaining rats were killed on the 14th day of recovery for gross and histopathologic examinations (4 from group IV). The lungs, liver, kidneys, spleen, and testes were weighed at necropsy, and representative samples of the following tissues were obtained for microscopic examination: heart, lungs, mesenteric lymph nodes, nasal cavities (nose), trachea, liver, pancreas, esophagus, stomach, duodenum, jejunum, ileum, cecum, colon, rectum, kidneys, urinary bladder, bone marrow (sternal), spleen, thymus, thyroid gland, adrenal glands, brain, eyes,testes,
704
WARHEIT
ET AL.
RATS LOADED NOSE ONLY
I
TYPE K THERMOCOUPLE THERMOMETER FLOWMETER llm GENERATION NITROQEN
-I rnr i.xl VARIAC
51 .QY ’ -,--.-
II
DRY ICE CONDENSER
MSA PARTICULATE FILTER
II
THREE-FLASK EXHAUST
FIG. I. Schematic demonstrating the formamide vapor generation system. Briefly, a vapor atmosphere of formamide was generated by a syringe driving the test material into a heated flask. Nitrogen was introduced at the flask and swept the resulting vapor through a dispersion funnel and into the test chamber. Dilution air and oxygen were introduced to the system prior to entering the chamber. The chamber exhaust was drawn through a scrubber containing water, a dry ice-cold trap, and an MSA cartridge filter prior to being discharged into the hood.
epididymides, and any other organs or tissues with gross lesions, The scheme for microscopic grading of tissue lesions was minimal, mild, moderate, and severe. Statistical analyses.Mean body weights and body weight gains for test rats were compared to controls during the exposure and recovery periods. Data were statistically analyzed by one-way analysis of variance (ANOVA). Exposure group values were compared to controls by the least significant difference test and Dunnett’s test when the ratio of variance (F) indicated a significant among-to-within group variation. Significant differences were declared at the 0.05 probability level. For analysis of clinical measurements, ANOVA and Bartlett’s test were calculated for each sampling time. When the Ftest from ANOVA was significant, Dunnett’s test was used to compare means from the control groups and each of the groups exposed to formamide.
RESULTS Chamber atmosphere analysis. During the exposure phase of the study, the mean chamber temperatures for each of the groups were: control, 23.6 f 0.8”C; 100 ppm, 27.9 + O.YC; 500 ppm, 26.8 + 0.9”C; and 1500 ppm, 27.0
& 0.9% In the control and test chambers, relative humidities ranged from 3 1 to 5 I%, and chamber oxygen contents ranged from 20 to 2 1%. Overall mean atmospheric concentrations are presented in Table 1. In general, the determined exposure concentrations were in close agreement with the design concentrations and thus, for purposes of this report, exposures in each group were considered to be 100,500, and 1500 ppm. Body and organ weight analyses. No biologically significant differences in mean body weights were observed in rats exposed to 100 or 500 ppm of formamide (Fig. 2). A few instances of statistically significant reductions in body weight gains were observed in rats exposed during the first week of exposure to 100 and 500 ppm of formamide. Rats exposed to 1500 ppm had significantly depressed mean body weights which were evident following the fourth exposure and lasted until the sixth day of the postexposure recovery period (Fig. 2). Similarly, mean body weight gains were
INHALATION
TOXICITY
705
OF FORMAMIDE
TABLE 1 MEAN
ATMOSPHERIC
CONCENTRATIONS
OF FORMAMIDE” Analyzed
Group no. I II III IV
concentration (ppm)
Design concentration
Mean
SD
Range
0 mm 100 mm 500 Pm-n 1500 ppm
-b 113 500 1504
33.8 97.4 284
56.2-280 148-84 1 449-2215
LIRepresents the mean, standard deviation, and range of all samples from all exposures (approximately per exposure chamber). b Values which were not measured.
significantly reduced during the first week of exposure in rats exposed to 1500 ppm formamide (Table 2). Statistically and biologically significant reduced mean final body weights and increased
100 samples
kidney-to-body weight ratios were observed after the 10th exposure in rats exposed to 1500 ppm of formamide (data not shown). Mean absolute kidney weights were elevated and were considered to be biologically,
380.00
2
320.00
240.00
0.00 -
5.00
10.00
CONTROL LOW
DAYS 0 wm 100 ppm
16.00 ON TEST r-a -
20.00 INT HIGH
26.00 SO0 1500
30.00
ppm ppm
FIG. 2. Growth curve for male rats exposed to 0,100,500, and 1500 ppm formamide. Significant declines in mean body weights are obvious in animals exposed to 1500 ppm when compared to controls.
706
WARHEIT
ET AL.
TABLE 2 MEAN
BODY
WEIGHT GAINS (GRAMS)
FOR MALE
RATS EXPOSED
TO 0,100,500,
OR 1500 PPM
FORMAMIDE
Group (Concentration) Days on test
I (0 mm)
II (100ppm)
III (500 r-m)
IV (1500 mm)
l-5” 8-12’ 1-12’ 13-26d
19.4 (5.48) 9.8(10.1) 54.0 (15.6) 83.6(11.4)
22.2 (6.36) 13.1 (12.5)
15.1 (6.61) 6.2 (14.0) 43.7 (20.5) 84.3 (19.6)
-9.2 (19.0)* 13.0 (23.5) -12.2 (19.4)* 103.2 (17.0)
59.5(16.1)
89.5 (7.92)
Note. Values in parentheses are standard deviations. a First week of exposures. b Second week of exposures. ’ Exposure period (combined). d Recovery period. * Significantly different from controls by least significant difference test (P < 0.05).
though not statistically, significant. Following the 14day recovery period, rats had statistically and biologically significant reduced mean final body weights and elevated mean absolute kidney weights and kidney-to-body weight ratios. Clinical observations.Immediately following exposures, rats in all groups (including controls) frequently had colored discharges of the eyes and/or nose (i.e., chromodacryorrhea) and/or diarrhea, effects common in rats under restraint. No other clinical signs of toxicity were observed following exposure. During the exposure period, low incidences of diarrhea were observed in all groups. In addition, a low incidence of weakness and hunched posture was observed in rats exposed to 1500 ppm formamide. Following termination of the exposure period, chromodacryorrhea was observed in two rats from the control group and in two rats exposed to 1500 ppm formamide. Five rats exhibited diarrhea in the 1500 ppm group during the recovery period. Summaries of clinical observations for male rats are presented in Table 3. Clinical pathology.Relevant clinical laboratory results are given in Table 4. A statistically significant decrease in the platelet count was observed in rats exposed to 500 or 1500
ppm of formamide. The decreased platelet count was considered larger than could be accounted for by expected analytical or biologic variation alone, and therefore was considered due to formamide exposure. A decrease in circulating lymphocytes in rats exposed to 1500 ppm was also observed. Both the relative thrombocytopenia (i.e., compared to control rats) and lymphopenia were also present after the 1Cday recovery period. Results for urea nitrogen, creatinine, urine volume, and urine osmolality were within the expected range of normal biologic variation and did not indicate any significant change in renal function. Differences between control and treatment group results for the remaining clinical laboratory parameters measured were either not statistically significant or were considered not biologically significant based on this laboratory’s experience with male Crl: CD BR rats. Pathology.One 1500 ppm rat was found dead on study Day 3. Tissues from this rat were inadvertently discarded and thus were unavailable for microscopic examination. Another 1500 ppm rat was found dead on study Day 9. One moribund 1500 ppm rat was killed in extremis on Day 11 of the study. In addition, there were no treatment-related gross findings in study rats.
INHALATION
TOXICITY
707
OF FORMAMIDE
TABLE 3
Exposure group (concentration) I (0 m-d
Observation
II (l~m-4
III (500 mm)
Number of rats exhibiting sign* Colored discharge eye(s) left Colored discharge eye(s) right Colored discharge nose Hunched over Diarrhea Weak Total group incidence
l(l5)’ l(l5) 0 3 (4) 0
0 0 0 0 2 (9) 0
0 0 0 0 l(3) 0
6
2
1
1(16)
1(16) 0
2(11) 3 (3) 5 (7 2 (3) 13
’ Excluding clinical signs observed during or immediately following exposure. b Ten rats per group. ’ The number in parentheses is the median day on test that the sign was first observed.
TABLE 4 CLINICALLABORATORYREWLTSFORMALERATSEXPOSEDTOFORMAMIDE Group exposure concentration (ppm) Test
Days on test
0
100
500
1500
Total leukocyte count (X 103/P1) Lymphocyte count (X 1031PI) Platelet count (X 1O’lPl) Serum urea nitrogen OwzM) Serum creatinine bw/dU Urine volume (ml/l6 hr) Urine osmolality (mOWKg)
12” 26’ 12 26 12 26 12 26 12 26 12 26 12 26
14.9 (3.3)C 16.8 (3.0) 11.1(2.1) 12.1 (2.4) 1142 (273) 1180 (268) 17(3) 19 (5) 0.6 (0.1) 0.7 (0.1) 7.2 (4.1) 10.9 (7.7) 1410 (452) 1499 (547)
14.1 (4.4) 14.6 (2.2) 10.1 (3.2) ll.O(l.3) lOSO(184) 1105(124)
13.6 (3.7) 16.0 (3.2) 9.8 (2.5) 11.2 (2.8) 882(119)* 83 1 (69)*
lO.l(l.2) 11.3 (3.6) 6.9 (0.6)* 7.9 (2.0) 847 (90)* 147 (294)*
14 (2) 17 (2) 0.6 (0.0) 0.6 (0.0) 5.1(1.3) 9.2 (1.6) 1497 (406) 1847 (405)
15(2)
21(4)
16(1)
16(l)
0.6 (0.0) 0.6 (0.1) 4.5 (1.2) 7.6 (0.9) 1870 (535) 2293 (161)
0.6 (0.0) 0.6 (0.0) 7.2 (4.2) 10.1 (2.6) 1786(1514) 1426 (257)
a Twelve days on test with exposure to formamide 6 hr per day for 10 of the 12 days, n = 10 for all groups, except 15OOppmwheren=7. b Twenty-six days on test with 10 days of exposure (6 hr/day) to formamide during the first 12 days followed by 14 days of recovery (no exposure), n = 5 for all groups, except 1500 ppm where n = 4. c Group mean and standard deviation. * Statistically significant difference from control group mean, 01= 0.05.
708
WARHEIT
ET AL.
TABLE 5 RENAL
Rat no.
1 2 3 4 5 6
LESIONS
IN RATS EXPOSED
TO
1500 ppm FORMAMIDE
Disposition
FD FD ICE KD KD KD
(Day (Day (Day (Day (Day (Day
3) 9) 11)
12) 12) 12)
FOR
2 WEEKS
Lesion
Grade
2-week exposure II Necrosis (tubular epithelium) Regeneration (tubular epithelium) Regeneration (tubular epithelium) Regeneration (tubular epithelium) Regeneration (tubular epithelium)
Severe Moderate Moderate Moderate Minimal
2-week exposure + 2-week recovery 1 2 3 4
KD KD KD KD
(Day (Day (Day (Day
26) 26) 26) 26)
Regeneration Regeneration Regeneration Regeneration
(tubular (tubular (tubular (tubular
epithelium) epithelium) epithelium) epithelium)
Mild Mild Mild Mild
Note. FD, found dead (study day); KE, killed in extremis (study day); KD, killed by design (study day). DTissues inadvertently discarded.
Compound - related microscopic lesions were present in the kidneys of male rats exposed to 1500 ppm of formamide (Table 5). Necrosis and regeneration of tubular epithelium were zonal, occurring in the outer medulla and radiating up medullary rays in the cortex. Lesions in affected rat kidneys were striking compared to control rat kidneys (Figs. 3 and 4). Necrosis of tubular epithelium was severe in one 1500 ppm rat found dead on study Day 9 (Figs. 5 and 6). All tubules in the outer stripe of the outer medulla were affected and many tubules in the inner stripe of the outer medulla and cortical medullary rats were also affected. In most tubules, the lining epithelial cell architecture was replaced by granular casts of desquamated epithelial cells with total karyolysis. Regeneration of tubular epithelium was present in one moribund rat (moderate) on study Day 11, in three rats (one minimal, two moderate) killed by design on study Day 12, and in four rats (all mild) killed on study Day 26. Regenerating tubules were lined by basophilic cuboidal cells, some of which were in mitosis, and lumens contained occasional desquamated necrotic cells. Peritubular fibrosis was a feature
of some regenerative tubules (Figs. 7 and 8). Other tubules were dilated and lined by flattened epithelium and some of these contained much mineralized necrotic debris. Occasionally, mineralized necrotic debris was circumscribed by granulomatous inflammation. The alterations in kidney weights and the histopathology findings reflect the target organ toxicity. Based on the renal lesion produced by formamide exposure under the conditions of this study, the no-observed-effect exposure concentration for histopathology was considered to be 500 ppm. DISCUSSION Formamide-induced necrosis and regeneration of renal tubular epithelial cells in male rats exposed to 1500 ppm via inhalation were zonal in the outer stripe of the outer medulla and in cortical medullary rays. This lesion distribution is consistent with damage to the pars recta (S3, P3) of the proximal tubule (Owen, 1986). The mechanism of formamide toxicity to cells of this segment of the tubule
INHALATION
TOXICITY
709
OF FORMAMIDE
FIG. 3. Normal morphology of a control rat kidney. C, cortex; 0, outer stripe of outer medulla; I, inn1er medulla. H&E X 16.
is UI nknown. According to a recent review of rem il morphology (Bulger, 1986), necrosis in this region of the tubule can be produced by exr3osure to mercuric chloride, uranyl nitrate, cis-1platinum, renal arterial occlusion in rats, hemorrhagic shock.
FIG.
At 0 days recovery, rats exposed to 1500 ppm formamide had statistically and bi ologitally significant decmased mean body weights and higher kidney-to-body weight rati
4. Normal morphology of outer stripe of outer medulla of a control rat kidney. H&E
X
160.
710
WARHEIT
ET AL
FIG. 5. Kidney ofa rat exposed to 1500 ppm formamide via inhalation found dead on study Day 9. Note severe necrosis and loss of tubule architecture in medulla and in medullary rays in the cortex. H&E X 16.
logi tally, although not statistically significan t. Following the 1Cday recovery period, the alterations in these parameters persisted. FJormal urine volume and osmolality result .sindicate that at least one-third of the re-
nal tubules remained functional (Dun .can and Prasse, 1986). Normal serum urea ni trogen and creatinine concentrations indic xte that glomerular filtration was likewise not affected enough to cause reduction in rz:nal
FIG. 6. Kidney of rat exposed to 1500 ppm formamide via inhalation and found dead on study Day 9. Note coagulation necrosis of most tubular epithelial cells in the outer stripe of the outer medulla. H&E x160.
INHALATION
TOXICITY
OF FORMAMIDE
711
FIG. 7. Kidney of rat 14 recovery days after inhalation exposure to i 500 ppm formamide. Regenerating tubules in the medulla and medullary rays in the cortex are mildly dilated. H&E X 16.
function (i.e., less than a 75% reduction in GFR). While the statistically significant increase in urea nitrogen concentration is compatible with the renal lesions observed histo-
logically, the small difference in urea nitrogen results between the control and 1500 ppm groups was not considered to be predictive for renal disease.
FIG. 8. Kidney of rat 14 recovery days after inhalation exposure to 1500 ppm formamide. Some regenerating tubules in the outer stripe of the outer medulla are surrounded by excess fibrous connective tissue (arrowheads) H&E X 160.
712
WARHEIT
The statistically significant thrombocytopenia observed in the 500 and 1500 ppm groups was considered to have little clinical significance. Rats normally have a very high circulating platelet count (850-1200 X 103/ ~1) (Jain, 1986). Rats in the two affected groups were in no danger of hemorrhaging spontaneously given the large number of platelets still remaining. Therefore, the term “thrombocytopenia” is used in this paper to describe a decrease in platelet count relative to the control group and not to indicate a clinically significant coagulopathy. However, compared to the control group platelet count and reference ranges established in this laboratory, the platelet counts in rats in the 500 and 1500 ppm groups were decreased more than could be accounted for by analytical or normal physiologic factors. Large platelets (shift platelets), indicating enhanced thrombopoiesis, were not observed in blood smears from thrombocytopenic rats. Sections of bone marrow appeared to contain a normal number of megakaryocytes. So neither active platelet destruction (with a consequent increase in platelet production indicated by shift platelets and increased marrow megakaryocytes) nor decreased platelet production (indicated by decreased marrow megakaryocytes) was determined to be the cause for the thrombocytopenia. Conceivably, the effects on circulating platelet homeostasis were too mild to cause a visible change in thrombopoiesis. Other possibilities include platelet sequestration in the lung or spleen or a depression of only extramedullary thrombopoiesis (e.g., in the lung). However, neither of these two possible explanations could be confirmed. One of the four rats from the 1500 ppm group remaining at the end of the recovery period had a much lower platelet count than the other three rats (309,000/~1). This result was responsible for the large standard deviation in the mean platelet count at this time point. The mean platelet count for the other three rats in this group was 893,000/~1 (+44,000), which was still considered slightly
ET
AL.
low. The reason for the marked decrease in the platelet count in one rat after a relatively prolonged period of recovery is not clear. Other thrombocytopenic rats exposed to 500 or 1500 ppm did not exhibit this marked decrease in platelet count, but neither did they show much evidence of recovery toward control group counts. Whether this indicates irreversible damage to thrombopoiesis or a residual effect that requires more time to recover is unclear. The decrease in mean lymphocyte count, observed only in the 1500 ppm group, suggests a stress-induced increase in corticosteroid levels, secondary to treatment-related effects. The diarrhea and hunched posture observed in the 1500 ppm group rats are also compatible with increased stress. The mean lymphocyte count for the 1500 ppm group after the recovery period was still considered slightly decreased based on the laboratory reference range. There is a paucity of data regarding the inhalation toxicity of formamide. Previous reports from our laboratory have suggested that toxicity by inhalation exposure was low. In these studies, no signs of toxicity were detected in rats given a single exposure of formamide for 6 hr at 3900 ppm or in two rats exposed for 10 days duration at concentrations approximating 1500 ppm vapor (Du Pont Co., 1952) and no indications of organ damage were observed following pathologic evaluation. However, in another study, formamide was slightly toxic in male rats on an acute inhalation basis. In this study, the 4-hr approximate lethal concentration was determined to be 2 1 mg/liter ( 11,403 ppm). Clinical signs following acute inhalation included lethargy, hunched posture, ocular and/or nasal discharges, and slight to severe weight losses (Du Pont Co., 1987). The studies presented here indicate that 10 repeated exposures to formamide at concentrations of 1500 ppm produce nephrotoxicity, characterized by necrosis and regeneration of renal tubular cells. Moreover, formamide exposure re-
INHALATION
TOXICITY
sulted in a thrombocytopenia and lymphopenia. The current TLV for formamide is based essentially on information regarding the toxicity of an analog, dimethylformamide. The target organ for DMF is the liver. We have demonstrated here that the liver is not affected by repeated formamide exposures. Moreover, in addition, we have shown that compound-related lesions were observed in the kidneys of male rats exposed to 1500 ppm of formamide. Under the conditions of this study, the noobserved-effect exposure level for male rats is 100 ppm, based on the hematologic and clinical chemical parameters measured in the rats exposed to 500 ppm formamide. Clinical pathologic examinations revealed that rats exposed to 500 and 1500 ppm of formamide were thrombocytopenic after 10 days of exposure and following 14 days of recovery. The biological significance of this mild thrombocytopenia was considered to be equivocal. Based upon the histopathologic finding of renal lesions at 1500 ppm, as well as the decrease in circulating platelets and lymphocytes at 1500 ppm, 500 ppm appears to be the exposure concentration at which adverse effects begin to occur in male rats (i.e., 500 ppm is the no-adverse-effect-exposure level). ACKNOWLEDGMENTS This study was supported by the Central Research and Development Department, E. I. du Pont de Nemours
713
OF FORMAMIDE
and Co., Inc. The authors acknowledge J. Brent Rhodes, Mark Hartsky, and Lena Achinko for their technical expertise. We also thank Gerald L. Kennedy, Jr., David P. Kelly, and Dr. Rudolph Valentine for contributing important insights into the preparation of this manuscript.
REFERENCES S. M., AND SULLIVAN, F. M. (Eds.) (1982). Formamides. In Reproductive Hazards of Industrial Chemicals, pp. 346-359 Academic Press, London. BASF Corp. (1985). Unpublished data. BULGER, R. E. (1986). Kidney morphology: Update 1985. Toxicol. Pathol. 14, 13-25. DUNCAN, J. R., AND PRASSE, K. W. (1986). Veterinary Laboratory Medicine, 2nd ed., pp. 153-173. Iowa State Univ. Press, Ames. Du Pont Co. (1952). Unpublished data. Du Pont Co. (1987). Unpublished data. BARLOW,
EICHHORN, U., KL~~KING, KLOCKING, H. P. (1987).
R., S~HWEIZER,
H., AND
Cell membrane toxicity detected with the chromium-5 1 release test. Arch. Toxicol. Suppl. 11,334-337. JAIN, N. C. ( 1986). Schalm ‘s Veterinary Hematology, 4th ed., pp. 288-298. Lea and Febiger, Philadelphia. KENNEDY, G. L., JR. (1986). Biological effects of acetamide, formamide, and their monomethyl and dimethyl derivatives. CRC Crit. Rev. Toxicol. 17, 129182. OWEN, R. A. (1986). Acute tubular lesions, kidney, rat. In Monographs on Pathology of Laboratory Animals, Urinary System (T. C. Jones, U. Mohr, and R. D. Hunt, Eds.), pp. 229-239. Springer-Verlag, New York. SCAILTEUR, V., AND LAUWERYS, R. R. (1987). Dimethylformamide (DMF) hepatotoxicity. Toxicology 43, 231-238.