Pulmonary effects of a polyisocyanate aerosol: Hexamethylene diisocyanate trimer (HDlt) or desmodur-N (DES-N)

TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

89,332-346

(1987)

Pulmonary Effects of a Polyisocyanate Aerosol: Hexamethylene Diisocyanate Trimer (HDlt) or Desmodur-N (DES-N) J. S. FERGUSON, The Toxicology

Laboratory,

M. SCHAPER, AND Y. ALARIE

Department of Industrial Environmental Health Sciences, Graduate Health, University ofpittsburgh. Pittsburgh, Pennsylvania 15261

School of Public

ReceivedNovember 18. 1986; accepted March 27, 1987 pulmonary Effects of a Polyisocyanate Aerosol: Hexamethylene Diisocyanate Trimer (HDIt) or Desmodur-N (DES-N). FERGUSON, J. S., SCHAPER, M., AND ALARIE, Y. Toxicol. Appl. Pharmacol. 89, 332-346. Desmodur-N (DES-N) or hexamethylene diisocyanate trimer (HDIt), a biuret structure of hexamethylene diisocyanate, is a viscous liquid used for durable coatings and is applied by brushing or spraying. DES-N aerosol has been shown to be primarily a pulmonary irritant following a single exposure in mice. To explore the pulmonary effects of this agent further, groups of guinea pigs were exposed to concentrations ranging from 8 to 12 1 mg/m3 of DES-N for 3 hr. Prior to and following exposure, each animal was challenged with 10% COz in 20% O2 and 70% N2 to evaluate their pulmonary performance. Following a single exposure, these animals displayed a concentration-dependent increase in respiratory rate and decrease in tidal volume, as well as coughing and apnea. Their ventilatory response to 10% CO2 was abnormal and characteristic of a lung restriction response. Some airflow limitation was seen during expiration but this occurred more often during air breathing than during CO* challenge. With daily exposures repeated for 1I consecutive days, guinea pigs began to adapt to the exposures as indicated by a return to a normal ventilatory response to CO*. This adaptation occurred within the first 5 days of exposures. From Days 6 to 11, there was a demonstrable effect, but the level of response was much lessthan that following the first exposure. No cumulative effect could be demonstrated with this polyisocyanate and the effect was found to be different than that for mono- or diisocyanates. Acceptable levels of exposure to this polyisocyanate for industrial workers are suggested. 0 1987 Academic Press, Inc.

Desmodur-N (DES-N), a biuret structure of hexamethylene diisocyanate (HDI) and also called hexamethylene diisocyanate trimer (HDIt), is a viscous liquid of no appreciable vapor pressure (7.5X 10e5 mm Hg at 20°C) used for durable coatings and is applied by brushing or spraying. Thus, there is the possibility of workers inhaling this material in an aerosol form. A commercial formulation of HDIt, DES-N, has been described as both a sensory and pulmonary irritant in mice (Weyel et al., 1982) and Karol (1986) recently summarized the evidence of possible immunological sensitization in workers exposed to this chemical. In 3-hr exposures of mice to 0041-008X/87

$3.00

Copyright 0 1987 by Academic press Inc. AI1 rights of repmdwtion in any form reserved.

this polyisocyanate, Weyel et al. (1982) observed changes in respiratory rate which differed from the response obtained in similar exposures to toluene diisocyanate (TDI) or to HDI monomer (Sangha et al., 1981). In the initial stages of each DES-N exposure, a response pattern characteristic of sensory irritation was observed. (Alarie, 1966, 198 1). This response disappeared and a pattern typical of pulmonary irritation was then seen (Alarie, 198 1). The presence of free HDI, up to 0.7% of the commercially available polyisocyanate, could be responsible for the sensory irritation at the beginning of exposure (Sangha et al., 198 1) but would not explain the observed

332

PULMONARY

333

EFFECTS OF POLYISOCYANATE

deep lung effects (Weyel et al., 1982). Valentini et al. (1983) confirmed that DES-N exposures increased alveolar and capillary epithelium permeability. This leakiness resulted in the pulmonary edema observed in mice by Weyel et al. (1982). The purpose of this study was to investigate further the pulmonary effects of DES-N exposure. In guinea pigs, the ventilatory response to 10% CO* has been demonstrated to be effective in evaluating the pulmonary toxicity of a variety of airborne irritants (Wong and Alarie, 1982; Wong et al., 1983, 1984, 1985; Ellakkani et al., 1984). The normal response of guinea pigs to 10% COZ can be altered by two types of abnormalities (Schaper and Alarie, 1985; Schaper et al., 1985; Alarie and Schaper, 1987). The first type, termed “obstruction,” is characterized by a failure to increase both tidal volume (Vr) and respiratory frequency (f) during COZ challenge, with a lengthening of the duration of expiration. Abnormal flow-volume loops are also seen, where there is a decrease in airflow rate, particularly during expiration. These effects are observed in guinea pigs during exposure to known bronchoconstrictors such as carbamylcholine, histamine, and sulfuric acid mist (Schaper et al., 1985). The second type of abnormal CO2 response is termed “lung restriction” or “restriction” and is indicated by a failure to increase VT but with an increase in fabove the normal increase seen during CO* challenge. While I’, is diminished, inspiratory and expiratory airflows are maintained at or near control levels resulting in a flowvolume loop that appears more rectangular than square as in control conditions. This rapid, shallow breathing pattern results primarily from an increased stimulation of vagal nerve endings restricting tidal volume from increasing normally. Other abnormalities such as lung stiffness may also result in such a breathing pattern (Rebuck and Slutsky, 1986; Alarie and Schaper, 1987). Lung restriction has been induced in guinea pigs primarily with agents inducing an inflammatory reac-

tion at the alveolar level (Alarie and Schaper, 1987; Burleigh-Flayer and Alarie, 1987). We have used a 10% COz challenge to further evaluate the pulmonary toxicity of DESN. Flow-volume loops were obtained to better characterize the type of pulmonary abnormalities involved. In previous studies in mice (Weyel et al., 1982; Valentini et al., 1983), the effect of single exposures were reported. Here, we investigated the effect of a single exposure as well as the effect of repeated exposures to detect a possible cumulative effect. MATERIALS

AND

METHODS

Animals Male English short-haired guinea pigs were purchased from Hiitop Lab Animals, Inc. (Scottdale, PA). These animals weighed 300-425 g prior to DES-N exposure. The animals had free accessto guinea pig chow and water and were maintained on an 8-hr/16-hr light-dark cycle. Exposure to DES-N HDIt, produced commercially as DES-N, was obtained from Mobay Chemical Corp. Plastics and Coatings Division (Pittsburgh, PA). Exposure concentrations were generated by dissolving various amounts of DES N in 50 ml of pesticide-grade water-free acetone. These solutions were fed into a Pitt No. 1 glass aerosol generator (Wang and Alarie, 1982) at 0.22 ml/min using a syringe pump to produce a concentration range of 8 to 12 1 mg/ m3. Dried air, fed into the generator at 15 psi, was used to produce the aerosol. The DES-N aerosol and acetone vapor produced by the generator were mixed with makeup air entering the exposure chamber. A 33-liter all-glass aquarium served as the exposure chamber and was exhausted at 25-28 liters/min. As in previous studies with DES-N, the syringe pump and exhaust flow settings were selected to maintain the acetone vapor concentration below 3000 ppm and to minimize possible sensory or pulmonary irritation during exposure to this agent (Weyel et al., 1982; Valentini et al., 1983). An absolute filter was placed in the chamber exhaust line to collect the DES-N particles. Exposure concentrations of DES-N were determined gravimetrically by pulling 2 liters/min of chamber air through a 0.2~Nm-pore-size polytetrafluoroethylene filter (Schleicher & Scheull, Inc., Keene, NH) for 40 to 60 min. Two or three samples were collected per exposure. The

334

FERGUSON,

SCHAPER. AND ALARIE

particle size distribution of the aerosol was determined using a cascade impactor (DCl-5) and an Andersen mini-impactor. The aerodynamic equivalent diameter by weight varied between a mass median diameter of 0.38 pm at low DES-N concentrations and 0.73 pm at high exposure concentrations. The size distribution indicated that over the range of concentrations generated 98% of the particles by weight were less that 3 Km. A Miran Model I infrared analyzer was used to determine acetone concentrations in the chamber. The acetone vapor concentrations during exposures was found to be less than 2000 ppm in all cases. Evaluation ofPulmonary Eflects Measurement of tidal volume and respiratory fiequency during air breathing and during 10% CO, challenge, The pulmonary response to DES-N exposure was evaluated by an indirect measurement of Vr andfof the guinea pigs during inhalation of room air and during a challenge with 10% CO1 in 20% Oz, and with 70% Nz. The system used has been described previously in detail (Wong and Alarie, 1982). Briefly, when an animal was placed inside a whole body plethysmograph, changes in pressure (A P) could be recorded with each breath. These pressure fluctuations have been shown to be proportional to VT (Wong and Alarie, 1982; Ellakkani et al., 1985). Calibration of A P was performed using a Harvard Apparatus small animal respirator attached to one port of the plethysmograph. Four milliliters of air was pushed into the chamber over a range of frequencies (60-2OO/min), providing a dynamic calibration for each animal (Schaper et al., 1985). Respiratory frequency was determined by counting the number of pressure waves per minute. The following protocol was carried out for all CO2 challenges. Each guinea pig was placed in a whole body plethysmograph and allowed to acclimatize for 10 min during which time the A?’ calibration was performed. During a 5-min baseline period, AP and fwere monitored as animals breathed room air. A 7-min challenge period followed, in which the 10% CO* mixture was introduced into the system. Changes in AP andfwere recorded throughout the CO* period and during a 5-min recovery period. The guinea pigs were challenged with CO2 prior to exposure, immediately following exposure, and at 7 and 24 hr after exposure. They were then periodically challenged for several days following their final exposure. Mean values for A Pandfobtained postexposure were compared to the preexposure value of the same animal or to values of a control group as in the series of repeated exposures. Measurements offlow-volume (G-V,) loops during air breathing and during CO, challenge. In addition to evaluation of the ventilatory response to CO2 as given above,

(G-V,) loops were obtained using a system previously described (Alarie et al., 1987). These measurements were performed on only one group of animals following the highest exposure concentration of DES-N. Each animal was fitted with a head chamber that was attached to a pneumotachograph and pressure transducer (Statham PM-l 5). As air, or the 10% CO2 mixture, was directed through the head chamber, airflow (v) could be obtained and integrated over time to obtain V,. Calibration of e was performed by passing known airflows through the pneumotachograph as detailed by Schaper et al. (1985). The airflow signal was digitized (200 sample&c) with the values stored on floppy disk. A computer program integrated V to obtain VT. V-V? loops were displayed on a video terminal as were the V signal and V, obtained from digital integration. From these parameters, the duration of inspiration (Ti) and expiration (Ps) as well asfwere calculated. Series I: Single Exposure to DES-N (a) Time-response and concentration-response relationships for a single exposure to DES-N. Groups of four guinea pigs were exposed to various concentrations of DES-N (from 8 to 121 mg/m’) for a single 3-hr period. Ventilatory response to 10% COz (A P and f) was determined for each group prior to exposure and at various time intervals following exposure until AP andfvalues returned to preexposure levels. Mean values for A P and fwere plotted against time for each exposure concentration. Concentration-response relationships were then obtained by plotting the mean change in AP or f from preexposure values. One group of eight guinea pigs was exposed to 1870 ppm ofacetone for 3 hr in order to ascertain the possible effect of this solvent which was used for aerosolizing DES-N. (b) Measurement ofpow-volume loops during air and CO, challenge. Four animals were exposed to 12 1 mg/ m3 DES-N for 3 hr and their ventilatory responses to 10% CO* were determined at 0, 3,7, and 24 hr post exposure using the system to obtain V-V, loops. The animals were killed following the measurements made 24 hr postexposure. Lungs were removed and inflated with buffered formalin and held at 20 cm HI0 for 2 hr prior to being placed in buffered formalin for further fixation. Slides were prepared for microscopic identification of lung pathology. A group of four control animals was subjected to an identical sham exposure, as well as ventilatory measurements. Lungs from these animals were also fixed and sectioned for microscopic examination. (c) Determination of lung weight following a single DES-N exposure. Following measurement of ventilatory responses to 10% CO*, eight guinea pigs were exposed to 22.0 mg/m3 DES-N for 3 hr. The animals were challenged with COz immediately following exposure and at

PULMONARY

335

EFFECTS OF POLYISOCYANATE

0.7

0.6

23 zi-

0.5

N 8

0.4

iz z

0.3

0 18

2 %

0.2

n 0.1

0.0

a.B

250

8

200

i2 E 2

150

c z

100

E I!

50

mg/m3

0 28 mg/m3 33mg/m3

0 79 mg/m3 I

,

0

24

”

I

‘.

46

1..

72

I

06

”

1”.

I20

1

144

”

I

166

0

FIG. 1. Time-response relationships for AP (ml; top panel) andf(breaths/min; bottom panel) measured in whole body plethysmographs during a 10% CO2 challenge. Groups of four guinea pigs received a single 3-hr exposure to various concentrations of DES-N and their respiratory responses were measured periodically until the values returned to preexposure levels. Each point represents the mean of four animals. Coefficients of variation for these mean values were below 20%.

7 hr postexposure. Four of the animals were killed by sodium pentobarbital injection following the 7-hr postexposure CO2 challenge. The lungs and trachea were removed, trimmed, and weighed. The remaining four guinea pigs were challenged with CO2 at 24 hr postexposure and then lolled as above for lung weight measurment. This same procedure was conducted on eight guinea pigs exposed to 84.0 mg/m3 DES-N and eight control animals receiving a 3-hr sham exposure. Series II: Multiple Exposures to DES-N (a) Daily exposure to DES-N for 5 consecutive days. This series of exposures was conducted to assessthe pos-

sibility of a cumulative effect. Four guinea pigs were exposed to DES-N for 3 hr per day for 5 consecutive days. CO2 challenges were performed prior to and following each exposure and periodically following the last exposure to observe recovery of the animals. (b) Daily exposure to DES-N for 1I consecutive days. Analysis of the data in Series IIa revealed a fade in the response following consecutive exposures, not a cumulative effect. Therefore, another series of exposures (IIb) was performed using a higher DES-N concentration to ensure a significantly high initial effect in an attempt to better characterize the adaptive response and how long it would last. The guinea pigs were exposed to DES-N for 3 hr/day for I1 consecutive days. On Days 12 and 13, the animals received no exposure or COa challenges and

336

FERGUSON,

SCHAPER, AND ALARIE

were then exposed for another 3 hr on Day 14. Ventilatory responses to 10% CO2 were determined on each day prior to exposure, immediately after exposure, and at 7 hr postexposure. These animals were not exposed but were challenged with 10% CO2 on Days 15. 16, 18,2 I, 23, and 24. A final exposure was performed on Day 25 with ventilatory responses to CO2 determined prior to and following exposure and at 7,24, and 48 hr postexposure. A group of four control animals was subjected to the same regimen using sham exposures and CO* challenges as described above. Statistical Analysis Time-response and lung weight data. Mean values for each time interval were compared to control values using one-way analysis of variance (Steel and Torie, 1960). When significant (p < 0.05) differences were obtained, comparison of exposure and control means were performed using Dunnett’s test (Steel and Torie, 1960). Concentration-Response analysis. The means of the COZ-induced changes in APandffrom preexposure values were determined for the various time intervals after exposure. These values were plotted against the logarithm of the DES-N exposure concentration. Linear regression analysis was perfomed on each of the curves. If linearity was found, the slope was obtained and tested for significance (p < 0.05) from zero by the method of least squares (Armitage, 197 1). Growth analysisforseriesllb. The ACOVSM program on the University of Pittsburgh’s DEC-10 System (Seth, 1983) was used to analyze differences in body weight changes between the exposed and control groups. Polynomials of various degrees were fitted to the data (Pothoff and Roy, 1964). Analysis of covariance was performed in order to detect significant differences in growth between the two groups.

RESULTS I. Single Exposures to DES-N (a) Time-response and concentration-response relationships using APandffollowing a single 3-hr exposure. Immediately following exposure or at 7 or 24 hr postexposure, there were no detectable changes in AP or f in guinea pigs breathing air except at 79 and 121 mg/m3. At these concentrations, there was a decrease in AP with an increase in f: Thus, no significant concentration-response

relationship was obtained. Coughing, however, was noted at all exposure concentrations while these measurements were being made and apneic periods were also seen at the higher exposure concentrations. No change was observed in the group exposed to acetone. The changes obtained during CO2 challenges were much greater than those obtained during air breathing. The results are presented in Fig. 1. The pulmonary effects of DES-N consisted of a decrease in AP and an increase in f from preexposure values. At exposure concentrations less that 28 mg/m3, the largest change occurred immediately postexposure or at 7 hr postexposure. At concentrations of 28 mg/m3 or above, the effects on AP and f continued to increase and reached a maximum at 24 hr postexposure. A return to control values occurred in 6 days as can be seen in Fig. 1. No change from preexposure was observed for the group exposed to acetone and similarly challenged with C02. The mean values for A P and f for each time interval and for all exposure concentrations were plotted against the logarithm of the DES-N exposure concentration. Linear leastsquares regression analysis for each time interval revealed that a significant concentration-response relationship was obtained with measurements of AP and f during CO* at 7 and 24 hr postexposure. The mean values at each exposure concentration were used to calculate the percentage decrease in AP and percentage increase in f for each concentration. The concentration-response curves (for AP and-f) obtained at 24 hr postexposure had significantly greater slopes than those for 7 hr postexposure. These curves indicated that a maximum respiratory response to DES-N inhalation occurred approximately 24 hr following exposure. The concentration-response relationships obtained at 24 hr postexposure are shown in Fig. 2. (b) Measurements of flow-volume loops during air and CO, challenge. A 3 hr exposure at a concentration of 12 1 mg/m3 DESN with a separate group of four guinea pigs

PULMONARY

. % INCREASE o % DECREASE

EFFECTS OF POLYISOCYANATE

IN f IN AP

CONCENTRATION

MS-N

(mglm3)

FIG. 2. Concentration-response relationships for f (percentage increase) and AP (percentage decrease) during 10% CO2 challenge measured 24 hr following a single exposure to DES-N. Each point represents a percentage change of the mean of four animals from the corresponding preexposure value. The points for AP andfat 18 mg/ m3 overlap. The curves were fitted by least-squares linear regression. From these relationships, the exposure concentration of DES-N associated with an increase infduring CO2 of 50% above preexposure was calculated to be 40.8 mg/m3 with 95% CL of 3 1 and 52 mg/m’. The concentration to decrease A P during CO* by 50% from preexposure was calculated to be 53.1 m&m3 with 95% CL of 38 and 73 mg/m3.

produced changes in respiration during air and CO*, consistent with the responses observed in the previous series. The mean values for V,, T,, and TE measured during air and 10% COz prior to exposure, immediately postexposure, and at 7 and 24 hr postexposure are presented in Table 1. As described above, changes in V, (measured as A P in the series reported above) andfwere evident following DES-N exposure. Here, a significant decrease in V, was also measured immediately following exposure and at 7 and 24 hr postexposure while the animals were breathing air. A significant increase in f was not detected until 24 hr postexposure. The effectes of DES-N exposure were again much more dramatic during 10% CO2 challenge as shown in Table 1 and Fig. 3. For exposed guinea pigs, values for V, were significantly lowered immediately following exposure and further

337

reduced at 24 hr postexposure. A significant increase in f was measured at both 7 and 24 hr postexposure. Furthermore, the frequency at 24 hr was significantly higher than that recorded at 7 hr. V-V, loops obtained for each animal (one example shown in Fig. 3) showed obvious airflow limitation toward the end of expiration during air breathing. Airflow changes were not as apparent during the 10% CO;? challenges, although there was evidence of airflow limitation 24 hr postexposure. The V-V, loops appeared rectangular in shape following exposure, in contrast to normal preexposure loops which were rather square. Histopathological examination of the lungs of these animals revealed a patchy, acute interstitial pneumonitis with infiltration of inflammatory cells and perivascular edema. An increase in alveolar macrophages, extravasation of red blood cells, and multiple foci of acute congestion were also observed. Areas of mild pleuritis and pleural cell hypertrophy could also be identified in association with DES-N inhalation. (c) Measurement of lung weights. For the groups exposed to DES-N (22 and 84 mg/ m3), mean values of AP and f were found to be significantly different from the control group at 7 and 24 hr during CO2 challenge as expected from the above results. The means for AP and f of the acetone-exposed group were not different from those for the controls. While the ventilator-y response to CO2 was affected by exposure to DES-N, no significant lung weight difference was seen when comparing the exposure groups, DES-N or acetone, to the controls. II. Repeated Exposures to DES-N (a) Daily exposure to DES-Nfor 5 consecutive days. The exposure concentration varied between 27.5 and 34.4 mg/m3 during the 5 consecutive, 3 hr daily exposures. The changes in A P and f during COz measured at

FERGUSON.

SCHAPER,

AND

ALARIE

7 hr postexposure and on the second day were similar to those found for a single exposure at this concentration in Series I. However, with subsequent exposures no cumulative effect was observed. In fact, a fade in the response was observed as better illustrated below. (b) Daily exposure to DES-Nfor 11 consecutive days. Exposure concentrations ranged from 65.1 to 74.4 mg/m3 during the 11 consecutive, 3-hr daily exposures and the results are presented in Figs. 4 and 5. The mean value for AP decreased significantly at 7 hr and the maximum decrease occurred 24 hr after the first exposure. Subsequent exposures over the next 5 days produced a diminished AP response while breathing 10% C02. By Day 6, values for A P had returned to near preexposure levels (Day 0). From Days 6 to 11, significant differences between preexposure and postexposure values were measured, but this effect was not significant when compared to a similar pattern displayed by the controls. The f response during 10% CO2 increased maximally 24 hr following the exposure given on Day 1. Similar to A P, a fade in the f response occurred during the first 5 days of exposure, with an attenuated effect produced by subsequent exposures up to Day 11. For both A P andJ; a response was elicited by DES-N on Day 14, as shown in Figs. 4 and 5, following a 3-day interval with no exposure to this agent. However, the response was significantly less than the response following the first exposure to DES-N. Following an 11 -day period without exposure to DES-N, exposure to this agent on Day 25 induced a response similar in magnitude (for AP) to that observed with the first exposure. Throughout the study, growth rates for the control and exposed animals for this series were significantly different from zero. However, there was no difference between the growth rates of the two groups. The overall increase in AP during both air and CO, breathing observed in the exposed and con-

PULMONARY a

PRE EXPOSURE

EFFECTS OF POLYISOCYANATE

POST EXPOSURE ~--

7 HRS

24 HRS

339

CO2

----.-.-

pre exposure pm, exposure 7 hrs 24 hrs

FIG. 3a. Flow-volume ( p- Vr) loops during air and during 10% CO* challenge for one guinea pig (4 17 g body wt) prior to and at various intervals following a 3-hr exposure to 121 mg/m3 DES-N. Each q-k’, loop is separated by a horizontal line with inspiration and expiration above and below the line, respectively. The scales for V and VT apply to all the loops. Respiratory frequency (f) is given in breaths per minute above each loop and represents the average value of eight breaths. The overlapping loops at the right of the figure represent the tracings obtained at each time interval during 10% CO* challenge. The scales for both q and VT have been doubled and the origins of each loop, representing the start of each breath, have been superimposed for the purpose of comparing changes in V and VT at each time interval. Q-VT loops are presented for a single animal; however, the same changes were found for all exposed animals.

trol animals corresponds to body growth over the course of the experiment. DISCUSSION Single Exposure Following a single 3-hr exposure of guinea pigs to aerosolized DES-N, the response was characterized by coughing, frequent apneic periods, and rapid, shallow breathing. The increase in respiratory frequency and decrease in tidal volume were concentrationdependent when these variables were measured during CO2 challenge and reached a maximum 24 hr following exposure. According the S&eider and Hutchens (1979), the aerosol generated over the range of concentrations used in this study was respirable for guinea pigs. The change in mass median diameter found between the low and high exposure concentrations was too small to influence the percentage distribution (Schreider and Hutchens, 1979). Thus, the concentra-

tion-response relationship presented in Fig. 2 was not influenced by the changing particle size. Neither can the effect be ascribed to only the physical presence of the aerosol since guinea pigs exposed to celhrlose dust for 6 hr at 79 mg/m3 showed no effect during air or CO, challenges (Ellakkani, 1985). Challenges with 10% COZ magnified the effects of exposure to DES-N by exaggerating the increase in f and decrease in AP. Such increases in f and decreases in AP were evident during air breathing but at only the highest exposure concentrations (79 and 121 mg/m3). Recovery was complete within 6 days after a single exposure to each concentration of DES-N. Measurements of p and Vr indicated that the increase infwas due to a shortening of Tr and T, , both decreasing by about 50% during the 10% COz challenge performed 24 hr postexposure. The V-V, loops revealed that airflow limitation occurred toward the end of expiration during air breathing. This was particularly noticeable immediately following exposure and at 7 hr postexposure. In com-

340

FERGUSON, b

PRE EXPOSURE

POST EXPOSURE ~-

SCHAPER, AND ALARIE 24 HRS

T-20 HRS __

COP

----.-.-

pre exposure pas, exposure 24 h,S 120 hrs

FIG. 3b. Flow-volume (k-VT) loops during air and during 10% COz challenge for one guinea pig (302 g body wt) prior to and at various intervals following a 3-hr exposure to 37 ppm methyl isocyanate (MIC). The presentation of e-VT loops is the same as that for DES-N. Modified from Alarie et al. (1987).

paring y--VT loops (Fig. 3), it is desirable to relate V to absolute lung volume (Hyatt and Black, 1973). If such information is not available, it may be reasonable to make comparisons by superimposing the loops of each animal such that each has the same origin. Grimby et al: (197 1) have made similar comparisons of V- Vr loops with different Vr obtained in humans during various levels of exercise. When this was done for guinea pigs exposed to DES-N (Fig. 3), there was a clear indication that as V, decreased following exposure, airflow was maintained at or near normal levels during both inspiration and expiration while breathing 10% CO,. This resulted in a more rectangular-shaped loop when compared to the control. As observed during air breathing, airflow limitation was indicated by the appearance of a check-valve during CO, challenge at 24 hr postexposure. This suggests that the location of the effect for airflow limitation may have been the small airways. However, this remains to be elucidated. The effects observed with DES-N were different than those obtained with other isocyanates, i.e., toluene diisocyanate as reported by Wong et al. ( 1985) and methyl isocyanate (MIC) as reported by Alarie et al.

(1987). Some I’-V, loops obtained from guinea pigs exposed to MIC are also shown in Fig. 3. With MIC, severe airflow limitation was obvious during both air breathing and during COz challenge. Respiratory frequency was very low under both conditions. Thus, the characteristic pattern of obstruction was evident (Schaper et al., 1985; Alarie and Schaper, 1987). Also to be noted is the fact that the response with DES-N was maximum at 24 hr postexposure with recovery to normal within 5-6 days while the response with MIC was still evident 5 days postexposure and persisted beyond 8 months after exposure (studies in progress). The effects observed with DES-N were very similar to the effects obtained following inhalation of cotton dust (Ellakkani et al., 1984, 1985) or paraquat aerosols (Burleigh-Flayer and Alarie, 1987). These agents induced concentration-dependent changes in vr (or AP) and fduring CO2 challenge and V-V, loops during CO2 challenge which were rectangular with little or no evidence of airflow limitation. These are characteristics of lung restriction (Schaper et al., 1985; Alarie and Schaper, 1987). Histopathological examination following a single exposure to these three agents also revealed similar inflammatory re-

PULMONARY

341

EFFECTS OF POLYISDCYANATE

I’~““,“‘~“,~“~“,,~~~~~,

0

7

14

21

28

a EXPOSED o CONTROL

FIG. 4. A P (ml) measurements for exposed and control guinea pigs obtained while breathing room air or during CO* challenge inside a whole body plethysmograph. The animals were exposed to room air or DES-N for 3 hr/day for 11 consecutive days with subsequent DES-N exposures on Days 14 and 25. CO2 challenges were performed before and 7 hr after each exposure. Each point represents the mean of four animals. Coefficients of variation for these mean values were below 20%. a, Significant decrease (p = 0.05) in AP from the preexposure value; b, significant decrease (p = 0.05) in AP from the value obtained prior to exposure on that day; c, significant decrease (p = 0.05) in AP from the value obtained prior to the latest exposure.

actions at the alveolar level. The changes, ineluding perivascular edema, congestion, etc. are known to be an adequate stimulus to type J vagal nerve endings which when stimulated induce rapid, shallow breathing (Paintal, 1969, 1981). Thus, the restriction pattern (i.e., rapid, shallow breathing with rectangu-

lar Q-V, loops) was most probably induced reflexively. Another agent shown to induce similar rapid, shallow breathing, exaggerated by CO2 challenge, is the well-known pulmonary irritant, ozone (Lee et al., 1979, 1980). Here, vagal cooling was shown to abolish the response.

342

FERGUSON.

SCHAPER.

Repeated Exposures Repeated exposures to DES-N resulted in adaptation as opposed to a cumulative effect demonstrated for toluene diisocyanate (Sangha and Alarie, 1979). A maximal change in fand AP was measured 24 hr after the first exposure, The response diminished with consecutive exposures with control levels being reached by the 11th day. After 2 days rest, a subsequent exposure produced a significant but smaller effect which peaked immediately following exposure. After a longer rest period a larger response was elicited than was observed on Day 14. Thus, adaptation was not a permanent phenomenon. The response pattern to multiple exposures of DES-N by inhalation is very similar to the adaptation and “Monday effect” described in guinea pigs exposed repeatedly to cotton dust (Ellakkani et al., 1985) and akin to the adaptive response observed in humans and animals exposed to ozone (Folinsbee et al., 1980; Hackney et al., 1977; Farrell et al., 1979). The concentration selected to explore a possible cumulative effect, 65-74 mg/m3, was very high in comparison to the 1 mg/m3 maximum exposure previously recommended for industrial exposure by Weyel et al. ( 1982). Since no cumulative pulmonary effect was observed, and in fact adaptation was seen, there is little chance that a cumulative respiratory response would develop in humans exposed to a much lower level.

D@erences between Mice and Guinea Pigs Previous studies by Weyel et al. (1982), who exposed Swiss-Webster mice to DES-N at concentrations between 25 and 131 mg/ m3, also showed that this agent acted slowly but the level of reaction measured 2 hr after exposure was the same as that at 24 hr postexposure. This differs from what we found here

AND

ALARIE

in guinea pigs. Another difference is the fact that a significant increase in lung weight occurred in mice at the above exposure concentration range, while we did not observe such an effect in guinea pigs exposed to 22 or 84 mg/m3. The work by Valentini et al. (1983) showed that this increase in lung weight in mice was due to an effect of DES-N on both the capillary endothelium and alveolar epithelium. This work also showed that complete recovery was achieved within 5 days following a single exposure as found here in guinea pigs. Therefore, mice seem to be more susceptible to pulmonary edema than guinea pigs following acute exposure to this agent. In guinea pigs, as shown from microscopic evaluation of lung tissure following a single exposure at 121 mg/m3, there were extensive changes but still no frank pulmonary edema. When evaluating the effect of DES-N on the breathing pattern in mice, Weyel et al. (1982) found that a concentration of 57 mg/ m3 produced 50% of the maximal effect. In guinea pigs, the exposure concentrations to increasejby 50% during CO2 challenge or decrease V, by 50% during CO2 were found to be 40.8 and 53.1 mg/m3, respectively. The concentrations evoking a 50% level of response in mice and guinea pigs are quite similar, but the mechanism by which DES-N acts upon each species may be very different. This conclusion is supported by the finding that a significant increase in lung weight was present in mice but was absent in guinea pigs at comparable exposure levels.

Efects Reported in Humans and Acceptable Industrial Exposure Levels In humans, one report of an adverse reaction at 4.2 mg/m3 of DES-N (Nielsen et al., 1985) indicated that chills, fever, dyspnea, wheezing, headache, arthralgia, and leukocytosis occurred a few hours after exposure. The authors favored a direct, nonallergic mecha-

PULMONARY

EFFECTS OF POLYISOCYANATE

343

. EXPOSED o CONTROL

FIG. 5. f(breaths/min) measurements for exposed and control guinea pigs obtained while breathing room air or during CO2 challenge inside a whole body plethysmograph. The animals were exposed to room air or DES-N for 3 hr/day for 11 consecutive days with subsequent DES-N exposure on Days 14 and 25. CO2 challenges were performed before and 7 hr alter each exposure. Each point represents the mean of four animals. Coefficients of variation for these mean values were below 20%. a, Significant increase (p = 0.05) inffrom the preexposure value; b, signihcant increase (p = 0.05) inffrom the value obtained prior to exposure on that day; c, significant increase (p = 0.05) inffrom the value obtained prior to the latest exposure.

nism as the explanation for this adverse reaction. A second report of human reaction to this agent included fever and dyspnea, but no wheezing. Leukocytosis or eosinophilia was also found (Berlin et al., 1981). The exposure concentration in this case was not known. A third report, also with unknown exposure

concentrations to DES-N, indicated several episodes of shortness of breath, wheezing, malaise, and chills. These symptoms occurred late in the afternoon on working days and lasted several hours. The subject was a foreman of a garage in which spray painting was done with polymeric HDI enamel (Malo

344

FERGUSON.

SCHAPER. AND ALARIE

et al., 1983). This individual was challenged by recreating a working environment, painting a board with an aerosolized liquid mixture. The concentration of the polymeric HDI was not measured during this challenge which lasted 5 min. Sixty minutes after the end of exposure, the subject reported a buming sensation in his chest and coughing was present. This was followed by a drop of FVC, chills, headache, malaise, increase in body temperature, and bibasal inspiratory crackles. The individual was found to have a restrictive breathing defect, together with marked reduction in arterial blood oxygen partial pressure. There was also some effect on the conducting airways; thus, the effect was described as a mixed restrictive and obstructive breathing defect. This is similar to our findings in guinea pigs where a pattern of restriction was observed but with the presence of airflow limitation as found with measurement of P-V, loops during air breathing. Field sampling of DES-N in spray painting operations indicated that the concentration can vary widely, from about 0.3 to 4 mg/m3 (Rosenberg and Tuomi, 1984). While DESN is being used in industry, no threshold limit value (TLV) has been established by the American Conference of Governmental Industrial Hygienists (ACGIH) nor has a permissible exposure limit (PEL) been promulgated by the Occupational Safety and Health Administration (OSHA). Clearly, this polyisocyanate induced different toxicological effects than those reported for mono- or diisocyanates for which TLVs or PELs have been established. Thus, setting a TLV or PEL by analogy is not possible. From the potency of DES-N to induce pulmonary irritation in mice, Weyel et al. (1982) suggested that the maximum concentration permitted in industry be no higher than 1 mg/m3. Using the COZ challenge model as used here, Alarie and Schaper (1987) suggested that the exposure concentration which reduces V, during CO2 by 50% from the normal response or increasesfduring CO2 by 50% above the nor-

ma1 response could be divided by 60 to yield an acceptable exposure concentration which would prevent pulmonary irritation. From the data presented in Fig. 2, this would be 0.7-0.8 mg/m3. This is close to the recommendation of Weyel et al. (1982) and certainly should not be exceeded. This should then prevent a pulmonary reaction. It should also be pointed out that while other mono- or diisocyanates have been found to be potent sensory irritants (Sangha and Alarie, 1979; Sangha et al., 1981; Alarie et al., 1987) and thus provide some warning properties in cases of overexposure, this is not so for DESN. Furthermore, the pulmonary reaction with this polyisocyanate is delayed. In summary, because there are few warning properties for an exposed worker, it is imperative to provide adequate sampling procedures to monitor exposure concentrations and to develop appropriate worker education programs. ACKNOWLEDGMENTS This work was supported under Grant l-ROIES02747 from the National Institute of Environmental Health Sciences. We thank Dr. Beverly Co&e11 of Experimental Pathology Laboratory, Hemdon, VA, for histopathological evaluation.

REFERENCES ALARIE, Y. (I 966). Irritating properties of airborne materials to the upper respiratory tract. Arch. Environ. Health 13,433-449. ALARIE, Y. (198 1). Toxicological evaluation of airborne chemical irritants and allergens using respiratory reflex reactions. In Proceedings of the Inhalation Toxicology and Technology Symposium (B. J. K. Leong, Ed.), pp. 207-23 1. Ann Arbor Science Pub., Ann Arbor, MI. ALARIE, Y., FBRGUSON, J. S., STOCK, M. F., WEYEL, D. A., AND SCHAPER,M. (1987). Sensory and pulmonary irritation of methyl isocyanate in mice and pulmonary irritation and possible cyanide-like effects of methyl isocyanate in guinea pigs. Eviron. Health Perspect., in press. ALARIE, Y., AND SCHAPER,M. (1987). Pulmonary performance in laboratory animals exposed to toxic

PULMONARY

EFFECTS OF POLYISOCYANATE

agents and correlations with lung diseases in humans. In Lung Biology in Health and Disease: Inhalation Toxicologv(J. Loke, Ed.), Dekker, NY. ARMITAGE, P. (1971). Statistical Methods in Medical Research. Wiley, New York. BERLIN, L., HJORTSBERG, U., AND WASS, U. (1981). Life-threatening pulmonary reaction to car paint containing a prepolymerized isocyanate. &and. J. Work. Environ. Health. 7,3 10-3 12. BURLEIGH-FLAYER, H., AND ALARIE, Y. (1987). Concentration-dependent respiratory response of guinea pigs to paraquat aerosol. Arch. Toxicol., in press. ELLAKKANI, M. A. ( 1985). An Animal Model for Byssinosis. Ph.D. Dissertation, University of Pittsburgh, Pittsburgh, PA. ELLAKKANI, M. A., ALARIE, Y. C., WEYEL, D. A., AND KAROL, M. H. (1985). Concentration-dependent respiratory response of guinea pigs to a single exposure to cotton dust. Toxicol. Appl. Pharmacol. 80,357-366. ELLAKKANI, M. A., ALARIE, Y., WEYEL, D. A., MAZUMDAR, S., AND KAROL, M. (1984). PU1mOnary reactions to inhaled cotton dust: An animal model for byssinosis. Toxicol. Appl. Pharmacol. 74,267-284. FARRELL, B. P., KERR, H. D., KULLE, T. J., SAUDER, L. R., AND YOUNG, J. L. (1979). Adaptation in human subjects to the effects of inhaled ozone after repeated exposure. Amer. Rev. Respir. Dis. 119,725-730. FOLINSBEE, L. J., BEDI, J. F., AND HORVATH, S. M. (1980). Respiratory responses in humans repeatedly exposed to low concentrations of ozone. Amer. Rev. Respir. Dis. 121,431-439. FOLINSBEE, L. J.. SILVERMAN, F., AND SHEPHARD, R. J. (1975). Exercise responses following ozone exposure. J. Appl. Physiol. 38(6), 996- 100 1. GRIMBY, G., SALTIN, B., AND WILHELMSON, L. (1971). Pulmonary flow-volume and pressure-volume relationship during submaximal and maximal exercise in young well-trained men. Bull. Eur. Physiopathol. Respir. 7, 157-168. HACKNEY, J. D., LINN, W. S., MOHLER, J. G., AND COLLIER, C. R. (1977). Adaptation to short-term respiratory effects of ozone in men exposed repeatedly. J. Appt. Physiol.: Respir. Environ. Exercise Physiol. 43(l), 82-85. HYA~, R. E., AND BLACK, L. F. ( 1973). The flow-volume curve: A current perspective. Amer. Rev. Resp. Dis. 107, 191-199. KAROL, M. H. (1986). Respiratory effects of inhaled isocyanates. CRC Crit. Rev. Toxicol. 16,349-379. LEE, L.-Y., DJOKIC, T. D., DUMONT, C., GRAF, P. D., AND NADEL, J. A. (1980). Mechanism of ozone induced tachypneic response to hypoxia and hypercapnia in conscious dogs. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 48( 1), 163-l 68. LEE, L.-Y ., DUMONT, C., DJOKIC, T. D., MENZEL, T. E., AND NADEL, J. A. (1979). Mechanism of rapid, shal-

345

low breathing after ozone exposure in conscious dogs. J. Appl. Physiob: Respir. Environ. Exercise Physiol. 46(6), 1108-l 114. MALO, J. L., OUIMET, G., CARTIER, A., LEVITZ, D., AND ZEISS, C. R. (1983). Combined alveolitis and asthma due to hexamethylene diisocyanate (HDI) with demonstration of crossed respiratory and immunologic reactivities to diphenylmethane diisocanate (MDI). J. Allergy Clin. Immunol. 72,4 13-4 19. NIELSEN, J., SANGO, C., WINROTH, G., HALLBERG, T., AND SKERIVING, S. (1985). Systemic reactions associated with polyisocyanate exposure. &and. J. Work Environ. Health 11,5 l-54. PAINTAL, A. S. (1969). Mechanism of stimulation oftype J pulmonary receptors. J. Physiol. 203,s 1 l-532. PAINTAL, A. S. (198 1). Effects of drugs on chemoreceptots., pulmonary and cardiovascular receptors. In Znternational Encyclopedia ofPharmacology and Therapeutics, Section 104, Respiratory Pharmacology (J. Widdicombe, Ed.), pp. 217-239, Pergamon, New York. POTHOFF, R. F., AND ROY, S. N. (1964). A generalized multivariate analysis of variance model useful especially for growth curve problems. Biometrika 51,3 13326. REBUCK, A. S., AND SLUTSKY, A. S. (1986). Control of breathing in diseases of the respiratory tract and lungs. In Handbook of Physiology, Vol. 2, Control of Breathing. (A. F. Fishman, Ed.), pp 771-791. American Physiological Society, Bethesda, MD, ROSENBERG,C., AND TUOMI, T. (1984). Airborne isocyanates in polyurethane spray painting: Determination and respirator efficiency. Amer. Ind. Hyg. Assoc. J. 45, 117-121. SANGHA, G. K., AND ALARIE, Y. ( 1979). Sensory irritation by toluene diisocyanate in single and repeated exposures. Toxicol. Appl. Pharmacol. 50,533-547. SANGHA, G. K., MATLJAK-SCHAPER, M., AND ALARIE, Y. (198 1). Comparison of some mono- and diisocyanates as sensory irritants. Toxicol. Appl. Pharamcol. 57,24 l-246. SCHAPER, M., AND ALARIE, Y. (1985). The effects of aerosols of carbamylcholine, serotonin, and proprano101on the ventilatory response to CO* in guinea pigs and comparison with the effects of histamine and sulfuric acid. Acta Pharmacol. Toxicol.56,244-249. SCHAPER,M., AND ALARIE, Y. (1986). New approaches for evaluating the effectiveness of drugs administered to prevent or reverse acutely induced pulmonary responses to aerosols. Pharmacologist 28, 14 1. SCHAPER, M., THOMPSON, R. D., AND ALARIE, Y. (1985). A method to classify airborne chemicals which alter the normal ventilatory response induced by CO*. Toxicol. Appl. Pharmacol. 79,332-341.

346

FERGUSON.

SCHAPER. AND ALARIE

SCHREIDER,J. P., AND HUTCHENS, J. 0. (1979). Particle deposition in the guinea pig respiratory tract. J. AerosolSci.

106,599-607.

SETH, A. K. ( 1983). Some Statistical Analysis children.

of Incomplete

Data

Considerations on Weight of Clej

in the palate

Doctoral Thesis, University of Pittsburgh. STEEL, R. G. D.. AND TORIE, J. H. ( 1960). Principles and Procedures of Statistics. McGraw-Hill, New York. VALENTINI, J. E., WONG, K. L., AND ALARIE, Y. (1983). Single-tracer technique to evaluate pulmonary edema and its application to detect the effect of hexamethylene diisocyanate trimer aerosol exposures. Toxicol. Appl. Pharmacol.

69,461-470.

WEYEL, D. A., RODNEY, B. S., AND ALARIE, Y. (1982). Sensory initiation, pulmonary irritation, and acute lethality of polymeric isocyanate and sensory irritation of 2,6-toluene diisocyanate. Toxicol. Appl. Pharmacol. 64,423-430.

WONG, K. L.. AND ALARIE, Y. (1982). A method for repeated evaluation of pulmonary performance in unanesthetized, unrestrained guinea pigs and its application to detect effects of sulfuric acid mist inhalation. Toxicol.

Appl. Pharmacol.

63,72-90.

WONG, K. L. KAROL, M. H.. AND ALARIE. Y. (1985). Use of repeated CO2 challenges to evaluate the pulmonary performance of guinea pigs exposed to toluene diisocyanate. J. Toxicol. Environ. Health 15, 137- 148. WONG, K. L., STOCK, M. F., AND ALARIE, Y. (1983). Evaluation of the pulmonary toxicity of plasticized polyvinylchloride thermal decomposition products in guinea pigs by repeated COZ challenges. Toxicol. Appl. Pharmacol.

70,236-248.

WONG, K. L., STOCK, M. F., MALEK, D. E., AND ALARIE, Y. (I 984). Evaluation of the pulmonary effects of wood smoke in guinea pigs by repeated COz challenges. Toxicol. Appl. Pharamcol. 7569-80.