* BASF Department of Toxicology, Ludwigshafen/Rhein, Federal Republic of Germany;
Zeneca Central Toxicology Laboratory, Alderley Park, GB-Macclesfield, Cheshire, SK10 4TJ, United Kingdom; and
Center for Life Sciences and Toxicology, Research Triangle Institute, Research Triangle Park, North Carolina 27709
Received July 21, 1999; accepted December 3, 1999
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: polymeric methylenediphenyl diisocyanate; developmental toxicity; maternal toxicity; fetal skeletal variations and retardations; whole-body exposure to aerosols.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The primary toxicologic effect of MDI and polymeric MDI is respiratory sensitization or occupational asthma. Occupational exposure limits have been selected on the basis of a potential to cause respiratory sensitization and pulmonary function decrements. The 1998 ACGIH TLV (threshold limit value) is 0.005 ppm (approximately 0.05 mg/m3), which is near the theoretical saturated vapor concentration (ACGIH, 1998). This study on the prenatal toxicity of polymeric MDI was undertaken to fill a data gap. Since this study was conducted, Buschmann et al. (1996) have reported an embryotoxicity evaluation of monomeric MDI in Wistar rats.
Selection of target concentrations for this study was based on three previous studies with this test substance. A prenatal toxicity range-finding study (BASF, unpublished report) in rats was performed to assess the maternal toxicity of polymeric MDI aerosol and to establish concentrations for this full-scale study. In that study, pregnant female Wistar rats, eight/group, were exposed to HPLC-measured concentrations of 4.75, 10.4, and 15.0 mg/m3 polymeric MDI aerosol, with a particle size below 2.8 µm MMAD during gd 6 through 15. At 15.0 mg/m3, two out of eight dams died toward the end of the study. All animals showed marked respiratory distress, diminished feed and water consumption, reduced body weight gains, and increased lung weights. Changes in hematologic and clinicochemical parameters reflected the clinically detectable impairment of health. Increased postimplantation loss, which was predominantly caused by a high number of late resorptions, was also found at 15.0 mg/m3. The other gestational parameters examined were not adversely affected. Exposure to concentrations of 10.4 or 4.75 mg/m3 caused increased lung weights as a symptom of maternal toxicity, but no evidence of developmental toxicity.
These results are also consistent with the results of another prenatal inhalation toxicity range-finding study (TNO Nutrition and Food Research Institute, unpublished report), in which exposure to gravimetrically determined concentrations of MDI at 2.0, 7.9, and 12.0 mg/m3 led mainly to an increased lung weight at the high concentration. Further corroborative evidence can be obtained from a 14-day inhalation study (TNO, unpublished report), in which exposure to about 15 mg/m3 caused death and severe respiratory distress and growth retardation. Exposure to 2 or 5 mg/m3 led to increased lung weights.
Based on the results described above, which point towards a steep concentration-response curve in the range between 10 and 15 mg/m3, the target concentrations selected for the prenatal inhalation toxicity study were 12, 4, and 1 mg/m3. The high concentration was expected to cause overt maternal toxicity and possible prenatal toxicity; the mid concentration could have resulted in an increase in maternal lung weight, but no fetal effects should have been present; and the low concentration was anticipated to be a maternal and developmental no observed adverse effect concentration (NOAEC).
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
For each concentration, the test substance was supplied to a two-component atomizer with a continuous infusion pump (Perfusor, Braun) and heated with a circulating thermostat. By means of compressed air, the aerosol was generated into an aerosol mixing stage. In the mixing stage, the aerosol was mixed with conditioned air and passed through a cyclone separator into the inhalation chamber. The inhalation chamber at 0 mg/m3 was not equipped with a generation system but was directly supplied with conditioned air. The exposure systems were also connected to an exhaust air system. Uniform distribution of the inhalation atmosphere (test compound in air) was assured by adequate measurement and technical design of the horizontal-flow chambers, along with establishment of appropriate airflow characteristics.
The proportions of test substance were as follows: the 0 mg/m3 chamber used fresh air (and no atomizer). The 1 mg/m3 chamber used 0.27 ml/h of test chemical and an atomizer temperature of 45.0°C. The 4 mg/m3 chamber used 0.82 ml/h of test chemical and an atomizer temperature of 44.3°C. The 12 mg/m3 chamber used 5.08 ml/h of test chemical and an atomizer temperature of 45.4°C. Supply air (conditioned and compressed) and exhaust air, in m3/h, for the four chambers are described in Table 1.
|
Exposure System
The animals were kept singly in wire cages that were in a glass-steel inhalation chamber of 1.4 m3 volume (manufacturer, BASF Aktiengesellschaft) for the whole-body exposures. The animals did not have access to water or feed during the exposure.
Air flow rates, pressure conditions, relative humidities, and temperatures in the inhalation system were measured continuously by an automated measuring system and were monitored against preset limits and partially regulated. The generator parameter compressed air was also recorded by means of this system.
The temperatures in the generator systems were measured continuously (digital thermometer) and recorded one time/exposure. No surveillance of the oxygen content in the inhalation system was performed. The air change (about 20/h) within the inhalation systems was judged to be sufficient to prevent oxygen depletion by the breathing of the animals, and the concentrations of the test substance used could not have a substantial influence on oxygen partial pressure.
Characterization of Inhalation Atmospheres
The nominal concentrations were calculated from the mean of the supplied substance and the amount of the supply air. The concentrations of the inhalation atmospheres of the test groups were real-time monitored with noncalibrated, scattered-light photometers (RAM 1, MIE, USA) and quantified by HPLC after precolumn derivatization with 1-(2-pyridyl)-piperazin.
For each concentration, samples were taken from the breathing zones of the animals in approximately hourly intervals. Amounts of 1090 l were sampled at a flow of 3 l/minute, with an array of a glass sampling probe with quartz wool plug and two absorption flasks filled with a 26-mM solution of 1-(2-pyridyl)-piperazin in acetonitrile. MDI content of sampling probe, quartz wool plug, and first flask was determined after each sampling. The content of the second flask was analyzed at the end of daily exposure to control absorption efficiency of the array. After derivatization at 60°C for 20 min, the samples were prepared for analysis.
High-performance liquid chromatography (HPLC; Hewlett Packard 1050) was used for analyses with a 250 mm (L) x 4 mm (ID) column; separation material was Nucleasil C8, 5 µm particle size. The detector was a photometric detector. The wavelength was 254 nm; injection volume 20 µl; column temperature was room temperature isocratic; the mobile phase 900 ml acetonitrile, 775 ml bidistilled water, 30 ml acetic acid glacial (100%), 18 g ammonium acetate; the flow rate was 1 ml/min.; and the retention time was 7.5 min.
Mass values of polymeric MDI were obtained from the area integrals using the calibration point. The concentrations of the test groups were calculated from these mass values in relation to the injected volume and the sample volume of the inhalation atmosphere.
The particle size analyses were carried out with Anderson Mark III and Sierra Marple 298 impactors. Aerosol volumes of 12150 or 12315 l were sampled from the breathing zones of the animals in the three MDI exposure chambers with the Mark III or Marple 298 impactor, respectively. The mass deposits inside the impactors were analyzed by the HPLC method described above. The calculation of particle size distributions from the masses deposited by the impactor stages was based on mathematical methods for evaluation of particle measurements (DIN 66141 and 66161).
Animals and Husbandry
Sexually mature, virgin Wistar rats (Chbb:THOM [SPF]), supplied by Karl Thomae, Biberach an der Riss, FRG, that were free from clinical signs of disease were used for the investigations. This strain was selected as extensive experience is available on Wistar rats, and this strain has proved to be sensitive to substances with a teratogenic potential. The female rats were 6170 days old when supplied. At the beginning of the study (gd 0, detection of sperm), the rats were 7686 days old. Their mean weight was approximately 238.9 g. The animals were free from any clinically evident signs of disease prior to the beginning of the study.
During the period when the rats were not exposed, they were housed singly in wire cages (type DK III of Becker & Co., Castrop-Rauxel, FRG). The animals were kept in air-conditioned rooms in which temperature was maintained in the range of 2024°C and relative humidity in the range of 3070%. A light/dark cycle of 12 h (light: 06001800 h) was maintained.
The animals were offered ground KLIBA laboratory diet rat/mouse/hamster maintenance GLP 343, Klingentalmuhle AG, Kaiseraugst, Switzerland, and tap water ad libitum, except during exposures when feed and water were withdrawn. The feed used in the study was assayed for chemical as well as for microbiologic contaminants. The drinking water was regularly assayed for chemical contaminants by the municipal authorities of Frankenthal and the Technical Services of BASF Aktiengesellschaft, as well as for the presence of germs by a contract laboratory. Both feed and water were suitable, based on the pertinent guidelines for animal feed and the German requirements for drinking water; their use has no effect on the design, conduct, or conclusions of this study.
Experimental Procedures
Mating.
After an acclimatization period of about 14 days, up to two untreated female rat(s) were mated with one untreated fertile male of the same breed. Mating took place from about 4 p.m. to about 7:30 a.m. of the following day. The day that sperm were detected microscopically in the vaginal smear in the morning was designated gd 0 (beginning of the study) and the following day "day 1" postcoitum (p.c.).
From day 1 p.c. onward, the animals were placed in inhalation chambers for adaptation on 5 days over 6 h/day and were exposed to a stream of fresh air under similar conditions as during exposure.
Exposure and postexposure observations.
The animals were exposed in the inhalation chambers, using wire cages of the above-mentioned type, for 6 h daily, from days 615 p.c. From day 16 p.c. to the day of sacrifice (day 20 p.c.), the animals were subjected to a postexposure observation period under the housing conditions mentioned above.
Maternal examinations.
Dams were examined daily for mortality. In addition, the behavior and state of health of the test animals were checked at least three times on exposure days and, as a rule, once daily during the preflow period (chamber adaptation) and the postexposure observation period. During exposure, a groupwise examination was carried out.
The body weights of the animals were recorded on day 0 (day of detection of sperm) and on days 2, 6, 9, 12, 15, 17, and 20 p.c., and body weight gains were calculated. Body weight gain was also calculated for the three different study periods: preflow period (chamber adaptation; days 06 p.c.), exposure period (days 615 p.c.), and observation period (days 1520 p.c.). These values were defined as body weight change. The corrected body weight gain was calculated after terminal sacrifice (terminal body weight on day 20 p.c. minus weight of the gravid uterus minus body weight on day 6 p.c.). With the exception of day 0, the consumption of feed and water was determined on the same days as body weight.
Dams that died intercurrently (prior to scheduled sacrifice), as well as the contents of the uterus from these animals, were examined, if possible, in the same way as at terminal sacrifice (except that uterine weight was not recorded).
On day 20 p.c., the surviving dams were sacrificed and necropsied in random order by cervical dislocation, and the fetuses were dissected from the uterus. Maternal head, liver, kidneys, larynx, and paired lungs were retained in 4% formaldehyde; no histopathology was performed. The uterus and the ovaries were removed and the following data were recorded: gravid uterine weight, liver and paired lung weights, number of ovarian corpora lutea, number and distribution of implantation sites classified as resorptions (early, late), dead fetuses, and live fetuses. Uteri from apparently nonpregnant females or from apparent single-horn pregnancies were stained by the method of Salewski (1964) to detect early resorption sites.
Conception rate and pre- and postimplantation losses per dam, summarized within groups, were calculated as follows:
![]() |
![]() |
![]() |
Fetal examinations.
At necropsy, each fetus was weighed, sexed, and examined macroscopically for any external findings. The sex was determined by anogenital distance and was later confirmed in all fetuses fixed in Bouin's solution by internal examination. The viability of the fetuses and the condition of the placentae, the umbilical cords, the fetal membranes, and fluids were examined. Individual placental weights were recorded. After these examinations, approximately one-half of the fetuses per dam were placed in ethyl alcohol and the other half were placed in Bouin's solution.
After fixation in Bouin's solution, the fetuses were examined for visceral alterations according to the method of Barrow and Taylor (1969). After these examinations, these fetuses were discarded. After fixation in ethyl alcohol, the skeletons of the remaining fetuses were stained with alizarin red S according to the modified method of Dawson (1926) and examined for skeletal alterations. These fetuses were retained by litter.
Evaluation criteria for assessing the fetuses.
There are differing opinions on the classification and assessment of changes in fetuses (Beltrame and Giavini, 1990). In the present investigations, the following terms (definitions) were used for describing a change:
Statistics
The data were evaluated statistically using the computer systems of the Department of Toxicology of BASF Aktiengesellschaft. The Dunnett Test (Dunnett, 1955, 1964
) was used for a simultaneous comparison of several dose groups with the control. The hypothesis of equal means was tested. This test was performed two-sided and was used for the statistical evaluation of the following parameters: maternal feed and water consumption, body weight, body weight change, corrected body weight gain (maternal gestational body weight change minus the weight of the gravid uterus), gravid uterine weight, number of corpora lutea, number of implantations, number of resorptions and number of live fetuses; proportion of preimplantation loss, postimplantation loss, resorptions, and live fetuses in each litter; litter mean fetal body weight and litter mean placental weight. For the parameter maternal feed and water consumption, the mean of means was calculated and is presented in the relevant summary tables. The mean of means values allow a rough estimation of the total feed and water consumption during the different time intervals (preflow, exposure, and postsobservation period); they are not exactly precise values, because the size of the intervals taken for calculation differed. For the mean of means values, no statistical analysis was performed. Fisher's Exact Test (Siegel, 1956
) was used for a pairwise comparison of each dose group with the control for the hypothesis of equal proportions. This test was performed one-sided and was used for female mortality, females pregnant at terminal sacrifice, and the number of litters with fetal findings. The Wilcoxon Test (Hettmansperger, 1984
; Nijenhuis and Wilf, 1978
) was used for a comparison of each dose group with the control for the hypothesis of equal medians. This test was performed one-sided and was used for the proportion of fetuses with malformations, variations, retardations, and/or unclassified observations in each litter. Statistically significant differences for the control group values are designated as * for p < 0.05, and ** for p < 0.01 on the summary tables.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Maternal Toxicity
The distribution and fate of all study females are presented in Table 2. Two females (of 24 pregnant) died at 12.0 mg/m3 (one on gd 15 during exposures and one on gd 18 in the postexposure period). No females were removed from study or delivered early. Two females, one each at 4.0 and 12.0 mg/m3, were not pregnant at scheduled sacrifice. Of the pregnant females, one each at 1.0 and 12.0 mg/m3 carried a fully resorbed litter at termination; all remaining pregnant females had live litters (i.e., one or more live fetuses). The number of litters (and fetuses) examined were 25 (336 live fetuses), 24 (338 live fetuses), 24 (345 live fetuses), and 21 (274 live and 5 dead fetuses) at 0.0, 1.0, 4.0, and 12.0 mg/m3, respectively (Table 2
).
|
|
Gestational parameters are presented in Table 3. There were no effects of treatment on preimplantation or postimplantation loss, although one dam at 12.0 mg/m3 had 5 dead fetuses (and 10 live fetuses), the only dead fetuses in the study. Litter size and sex ratio (percent males/litter) were also unchanged across groups. Fetal body weight per litter and placental weight per litter (separately by sex or total) were significantly reduced at 12.0 mg/m3 (Table 3
).
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The obvious compromised status of the dams at 12 mg/m3 may have resulted in reduced gravid uterus weight. The observed developmental toxicity, including reduced fetal body weights (and related reduced placental weights) and an increased incidence of fetuses/litters with skeletal variations and skeletal retardations may have been due, at least in large part, to the reduced fetal body weights (Hood and Miller, 1997). No substance-induced teratogenic effects were observed up to and including the highest concentration (12 mg/m3).
Concerning fetal morphology, the historical control data shown in Table 4 for total malformations indicate that the 2-fold increase in the percent of litters with malformations in the high exposure group (38.1% vs. 20.0% in the control) is still within the historical control range (0.043.0%). The same applies to the 2.5-fold increase in the percentage of fetuses with malformations (5.4% vs. 2.1% in the control; historical control range is 0.05.7%). The incidences and the types of fetal external visceral and skeletal malformations observed (see Results section) were scattered across groups and at low incidence. There was a significant increase in the incidence of fetal visceral variations (particularly dilated renal pelvis) on a fetus per litter basis at 1 mg/m3 (but not at 4 or 12 mg/m3); in the incidence of bipartite sternebra(e), in terms of litters with one or more affected fetuses per litter at 1 mg/m3 (but not at 4 or 12 mg/m3) and on a fetus per litter basis at 1 and 12 mg/m3 (but not at 4 mg/m3); and of irregular-shaped sternebra(e) in terms of litter incidence at 12 mg/m3 (but not at 1 or 4 mg/m3). The incidences of total fetal variations, in terms of affected fetuses per litter, were significantly increased in all polymeric MDI-exposed groups at 1, 4, and 12 mg/m3. These fetal effects at exposure concentrations other than at 12 mg/m3 were most likely due to the unexpectedly low number of fetal variations in the concurrent control group and occurred in the absence of a concentration relationship; they were therefore considered spontaneous in nature. The fetal effects observed at 12 mg/m3, involving exclusively increases in fetal skeletal variations (mostly increases in indications of delayed ossification) and therefore increases in total fetal variations, exceeded the historical control range in the performing laboratory (Table 4
), were associated with compromised maternal status and reduced fetal body weights, and were therefore considered treatment related (Gamer, unpublished report).
A previous developmental toxicity study in Wistar rats exposed to monomeric MDI as the aerosol, 6 h/day, gd 6 through 15 at 0, 1, 3, or 9 mg/m3 resulted in concentration-dependent reductions in maternal feed consumption during the exposure period in all MDI-exposed groups, and increased maternal lung weights only at 9 mg/m3. There were no effects on any other parameters of maternal toxicity and no effects on gestational or fetal parameters (including no effects on pre- or postimplantation loss, litter size, fetal body or placental weights, or on the incidences of fetal external or visceral anomalies or of the extend of fetal ossification). There was a slight but significant increase in litters with one or more fetuses at 9 mg/m3 with asymmetric sternebra(e); the incidence was considered within the limits of biological variability (Buschmann, 1994; Buschmann et al., 1996
). As the current study evaluated polymeric MDI at 0, 1, 4, and 12 mg/m3 with effects only at 12 mg/m3, the results of the two studies are consistent. The results of the present study are also consistent with the results from the two developmental toxicity range-finding studies on polymeric MDI (BASF, TNO, unpublished reports) and with results of a 14-day study on polymeric MDI (TNO, unpublished report), all of which indicated increased lung weights at exposure concentrations
9 mg/m3, adult mortality at
12 mg/m3, and increased postimplantation loss only at 15 mg/m3.
The NOAEC for maternal and developmental toxicity therefore is 4 mg/m3, and the NOAEC for teratogenic effects is at or above 12 mg/m3 in rats under the conditions of this study.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Barrow, M. V., and Taylor, W. J. (1969). A rapid method for detecting malformations in rat fetuses. J. Morphol. 127, 291305.[ISI][Medline]
Beltrame, D., and Giavini, E. (1990). Morphological abnormalities in experimental teratology: Need for a standardization of current terminology? Cong. Anom. 30(3), 187195.
Buschmann, J., Koch, W., Fuhst, R., and Heinrich, U. (1996). Embryotoxicity study of monomeric 4,4'-methylene diphenyl diisocyanate (MDI) aerosol after inhalation exposure in Wistar rats. Fundam. Appl. Toxicol. 32, 96101.[ISI][Medline]
Buschmann, J. (1994). Teratogenicity study of monomeric 4,4'-methylene-diphenyl diisocyanate (MDI) aerosol after inhalation exposure in Wistar rats. Poster presented at the 22nd Annual Conference of the European Teratology Society, Prague, Czechoslovakia, September 1215, 1994.
Dawson, A. B. (1926). A note on the staining of the skeleton of cleared specimens with Alizarin red S. Stain Technol. 1, 123.
DIN 66141: Darstellung von Korngrößenverteilungen. DIN 66161: Partikelgrößenanalyse (Beuth-Vertrieb GmbH, Berlin und Köln).
Dunnett, C. W. (1955). A multiple comparison procedure for comparing several treatments with a control J. Amer. Statist. Assoc. 50, 10961121.[ISI]
Dunnett, C. W. (1964). New tables for multiple comparisons with a control. Biometrics 20, 482491.[ISI]
Hettmansperger, T. P. (1984). Statistical inference based on ranks. John Wiley & Sons, New York, NY, 132142.
Hood, R. D., and Miller, D. B. (1997). Maternally mediated effects on development. In: Handbook of Developmental Toxicology (R. D. Hood, Ed.). CRC Press, Boca Raton, FL.
Nijenhuis, A., and Wilf, H. S. (1978). Combinatorial Algorithms, Academic Press, New York.
OECD (1981). OECD Guidelines for Testing of Chemicals, Section 4, Health Effects, Method No. 414. Organization for Economic Cooperation and Development, Paris.
Reuzel, P. G. J., Kuper, C. F., Feron, V. J., Appelman, L. M., and Löser, E. (1994). Acute, subacute, and subchronic inhalation toxicity studies of respirable polymeric methylene diphenyl diisocyanate (polymeric MDI) aerosol in rats. Fundam. Appl. Toxicol. 22, 186194.[ISI][Medline]
Salewski, E. (1964). Färbemethode Zum makrosckopischen nachweis von implantations-stellen am uterus der ratte. Naunyn-Schmiedebergs, Arch. Exp. Pathol. Pharmakol. 247, 367368.
Siegel, S. (1956). Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York.
U. S. Environmental Protection Agency (U.S. EPA) (1983). Part III, EPA Toxic Substances Control; Good Laboratory Practice Standards; Final Rule. Fed. Reg. 48(230), 5392253944.