of

Respiratory Effects Chloromethyl Methyl Ether

The
William Weiss, MD, Katharine R. Boucot, MD

\s=b\ A prospective study of 125 chemical workers was carried out for ten years to investigate the incidence of lung cancer. Some of the men were exposed to chloromethyl methyl ether containing bis(chloromethyl) ether as an impurity. Bronchogenic carcinoma was markedly increased among them, with a strong dose-response relationship. An unexpected inverse relationship was noted between smoking and the incidence of lung cancer. The neoplasms (all small-cell carcinomas) occurred in relatively young men. Symptoms of chronic bronchitis were reported more often among men exposed to chloromethyl methyl ether, and a dose-response relationship was apparent, with smoking a cofactor. Ventilatory function was not significantly affected by chemical exposure. Periodic screening over the first five years of the study showed a decrease in chronic coughing and an increase in dyspnea while chemical exposure was diminishing.

wheeze,
cm

or

weight loss of more than 4.5 kg

were

indications for 36x43 roentgenograms (posteroanterior and lateral) and interview by a physician. Ap propriate clinical action was recommended to the plant physician. Seventy percent of the 125 men were aged 30 to 49 years at the start of the study (Table 1). Ten percent had never smoked; 8% smoked cigars or pipes only; 5% were ex-smokers of cigarettes, and 78% were current cigarette smokers; 24% smoked more than one pack per day (Table
in six months

2).

(JAMA 234:1139-1142, 1975)

IN A CHEMICAL plant with approx imately 2,500 employees and a bian nual chest roentgenogram screening program, the physician and personnel manager noted that three men aged 33 to 37 years developed lung cancer during 1962. All three men worked as chemical operators in one building, with potential exposure to more than 100 substances. At the end of that year, 45 men who had worked at least three months in the area under suspi cion and 80 men from other areas of the plant were screened periodically for five years, in conjunction with the Philadelphia Pulmonary Neoplasm Re search Project.1 Four of the 125 men developed lung cancer, indicat ing an excessive risk.2 Analysis of the work histories in the lung cancer cases suggested that exposure to chloromethyl methyl
From the Division of Occupational Health, Hahnemann Medical College and Hospital (Dr Weiss), the Department of Preventive Medicine, the Medical College of Pennsylvania (Dr Boucot), and the American Lung Association of Philadelphia and Montgomery County, Phila-

ether (CMME)

its contaminant ether (BCME) was the common denominator. By 1971, several experimental animal stud ies310 had shown that BCME was a potent inhalant carcinogen and that CMME had slight carcinogenic poten tial. Relative exposure data have been developed for all the production workers in the plant, from 1948, when CMME was first manufactured, to 1972. We will describe the findings in the 125 chemical workers studied prospectively for ten years in relation to CMME (BCME) exposure. METHODS
or

bis(chloromethyl)

70-mm chest

delphia. Reprint requests to Division of Occupational Health, Hahnemann Medical College and Hospital, 917 Schaff Bldg, 1505 Race St, Philadelphia, PA 19102 (Dr Weiss).

recorded at ambient temperature and barometric pressure. Results for the forced expiratory volume test were compared with the predicted normal standards of Baldwin et al11 according to age and

were screened by means of photofluorograms and ques tionnaires regarding age, smoking habits, and respiratory symptoms, at intervals averaging 8.5 months. Weight was re corded at each visit. At a visit early in 1965, 120 men reported for screening, and spirograms were done on 103. A bellows spirometer was used; volumes were

The workers

height. Photofluorographic abnormality, wor sening cough, hemoptysis, unilateral

Fourteen men left the investigation at various intervals during the five-year screening study because of job termi nation. However, the survival status of 120 men among the 125, as of Dec 31,1972, was determined through follow-up. Determining exposure data was difficult because CMME had been manufactured and used in three different buildings at different times, changes of employees in volved in the chemical processes were fre quent, and amount of exposure diminished over time as a result of process improve ments. It was found that workers other than the 45 who were employed in the area under suspicion in December 1962 had been exposed to CMME in two other areas of the plant. Since suitable chemical deter minations of CMME or BCME concentra tions in the work environment had not been made during the study period, super visory personnel associated with the manu facture or use of CMME during the period 1948 to 1972 were asked to rank the expo sure of all job classifications retrospec tively on an arbitrary scale from 0 to 6. Work histories and exposure ratings for each job at every important stage in the development of the process were used to calculate time-weighted average exposure ratings and total exposure times. An expo sure index was calculated for each man by multiplying his time-weighted average rating by the number of years of exposure as of Dec 31,1964, just prior to the time the spirograms were made (two years after the beginning of the five-year screening study), and as of Dec 31, 1972, at the end of the ten-year observation period. Both in-

Downloaded from www.jama.com by guest on November 4, 2010

Table

1.Age Distribution (No [%]) in 1963 by Degree of Exposure to Chloromethyl Methyl Ether
Exposure Index*
Total
0.1-12.9

ship held among men in the moderate and heavy exposure groups, which in cluded all the men with lung cancer:
six of the 11
>25.0
cancer cases

occurred

Age, yr
20-29 30-39 40-49 50-59 Total

13.0-24.9

among 13

men

who

were

not current

15(12) 43(34) 45(36) 22(18) 125(100)

exposure

6(16) 10(27) 11(30) 10(27) 37(100)

6(15) 17(44) 11(28) 5(13) 39(100)

8(28) 15(52) 5(17) 29(100)

1(3)

2(10) 8(40) 8(40) 2(10) 20(100)

"Time-weighted

rating multiplied by years of exposure (1948-1972).

Table 2.Smoking Habits (No [%]) in 1963 by Degree of Exposure to Chloromethyl Methyl Ether

Exposure Index*
Habit Nonsmokers and noncigarette smokers Nonsmokers

Smoking

Total

0.1-12.9

13.0-24.9

>25.0

smokers <1 pack/day >1 pack/day Total

Cigar or pipe only Ex-cigartte smokers Current cigarette

29(23) 13(10) 10(8)

6(5)

8(22) 3(8) 4(11) 1(3) 29(78) 23(62)

8(21) 2(5) 4(10) 2(5) 31(79) 16(41) 15(38) 39(100)

6(21) 3(10) 1(3) 2(7) 23(79) 15(52) 8(28) 29(100)

7(35) 5(25) 1(5) 1(5)

96(78) 66(53) 30(24) 125(100)

6(16) 37(100)

13(65) 12(60) 1(5) 20(100)

"Time-weighted

exposure rating

multiplied by years of exposure (1948-1972).

dices were highly correlated (Pearson coef ficient, .83; i 13.15; P<.001). The men were then divided into four groups accord ing to the degree of exposure to CMME: no exposure, light, moderate, and heavy expo
=

Table 3.Risk of Lung Cancer, 1963-1972, by Degree of Exposure to Chloromethyl Methyl Ether

Exposure
Index*
0 0.1-12.9 13.0-24.9 > 25.0

No at Risk 37 39 29 20

Developed Cancer, No(%)

0 0

sure.

RESULTS

5(17.2) 6(30.0)

Total_125_11(8.8)f
by
years of exposure (1948-1972). tx2=21.11;df=3;P<.001.

'Time-weighted exposure rating multiplied

Table 4.Incidence* of Lung Cancer by Smoking Habits

Age distributions at the beginning of the study were similar in all groups, except for a slight trend toward smaller proportions of men in the oldest age-group with increasing exposure (Table 1). Smoking habits differed among exposed groups: the proportion of men who never smoked increased and the proportion of heavy cigarette smokers decreased with in
creasing exposure (Table 2). Lung cancer developed in four men during the first five years of the study and in seven more during the subse

Smoking Habitt Nonsmokers and noncigarette smokers

Nonsmokers

Lung Cancer, No(%) 6/29(20.7) 3/13(23.1) 1/10(10.0)

2/6

All the cancers were small-cell car cinomas. Resection was done in two cases. All 11 died within two years of diagnosis. Of the 114 men without lung cancer, six died in the ten years from December 1962 to December 1972: four of cardiovascular disease (one was exposed to CMME), one of pancreatic carcinoma (exposed to CMME), and one of "natural causes." A total of 16 men died in the ten years, including ten of the 11 men who died of lung cancer (the 11th died after the ten-year study period) and six who died of other causes. The ex pected number of deaths in a tenyear period was calculated to be 5.9 from age-specific probability data for white males for I96012 and 1965 (writ ten communication from L.M. Axtell, Feb 2, 1973). Thus, lung cancer was responsible for the entire excess of deaths. We examined the prevalence of symptoms in relation to CMME expo sure as of 1965 because a spirogram was available at that time. The expo sure index used for this purpose was calculated for the period from 1948 to Dec 31, 1964. One hundred twenty men answered questionnaires in early 1965. The common symptoms (Table 5) were chronic cough (cough for months or years) and expectoration. Both symptoms showed a statistically significant dose-response relation ship, rising to 50% and 54% in the heavy-exposure group. The relation ship of the index to these symptoms at the time of the first screening was similar. Wheeze and dyspnea were less common and unrelated to expo
sure. was

cigarette smokers, while five devel oped among 36 men who were.

Ex-cigarette

Cigar/pipe only

smokers Current cigarette smokers <1 pack/day >1 pack/day

5/96(5.2) 4/66(6.1) 1/30(3.3)

Total_11/125(8.8)
number in whom lung cancer developed; percent is not calculated when denominator is less than 10. fComparlng current cigarette smokers with all other groups combined, x2=4-86 with Yates correction; P < .05.
are

"Numerators

quent five years. The ten-year inci dence (Table 3) showed a strong doseresponse relationship, the incidence rising to 30% among 20 men in the heavy exposure group. In marked contrast, there was an inverse rela tionship between smoking habits and the ten-year risk of lung cancer (Table 4). The highest incidence of lung cancer occurred in 13 nonsmok ers and the lowest in current heavy cigarette smokers. The same relation-

The prevalence of chronic cough examined in relation to both smoking habits and CMME exposure (Table 6). Because the numbers of

nonsmokers, cigar or pipe-only smokers, and ex-smokers of cigarettes

were

small when divided into

unex-

posed and exposed groups, these men were combined into a single category. There was a direct relationship be tween chronic cough and cigarette smoking, but cough was consistently

Downloaded from www.jama.com by guest on November 4, 2010

prevalent among exposed men smoking categories. Ventilatory function was tested in 103 men early in 1965. The mean forced-expiratory volume (FEV) was 100.7% of predicted and the one-sec
more

in all

Table 5.Prevalence of Symptoms

to

(No [%]) in 1965 by Degree of Exposure Chloromethyl Methyl Ether

Exposure Index*
0.1-4.9 5.0-12.9

ond FEV was 80.9% of the observed FEV. The results did not vary signifi

Symptom Any symptom

Wheeze Total

Total

>13.0

Chronic cough

Expectoration

cantly

the observation period (December 1962 to March 1963) and again five years later in January to March 1968. The change in prevalence of chronic cough during this period was studied in relation to CMME exposure calcu lated near the midpoint of this period, Dec 31,1964 (Table 7). The unexposed group had a slight decrease in the prevalence of chronic cough while the exposed men showed larger de
creases.

among exposure groups. seven men answered both at the start of questionnaires

One hundred

Dyspnea Employees

56(47) 36(30) 38(32) 11(9) 14(12)

120

13(35) 8(22) 4(11) 4(11) 3(8)

37

10(36) 4(14) 6(21)

0

3(11)
28

17(59) 11 (38) 14(48) 3(10) 5(17)

29

16(62) 13(50)t 3(12)

14(54)$ 4(15)
26

"Time-weighted exposure rating multiplied by years of exposure (1948-1964). tx2=10.39; df=3;P<.025. tX2=18.40; df=3;P<.0005.
Table 6.Prevalence (No [%]) of Chronic Cough in 1963 by Smoking Habits and Exposure to Chloromethyl Methyl Ether

Smoking Habit Nonsmokers and noncigarette smokers Current cigarette smokers <1 pack/day >1 pack/day Total

Total

Unexposed
0/8

Exposed

2/29(6.9) 38/96(39.6) 24/66(36.4) 16/30(53.3) 40/125(32.0)

9/29(31.0) 6/23(26.1)
3/6

9/37(31.0)

2/21(9.5) 31/67(46.3) 18/43(41.9) 13/24(54.2) 33/88(37.5)

In contrast, there was a more im pressive increase in the prevalence of dyspnea among the exposed groups, and this seemed to be dose-related

"Numerators are number with chronic nator is less than 10.

cough; percent is

not calculated when denomi

Table

(Table 8).
There were 41 men who had a chronic cough at the first visit, and 84 who did not. The ten-year incidence of lung cancer was almost the same in both groups: 10% and 8.3%. A break down of these groups by CMME-exposure index showed no new findings.
COMMENT

7.Change* in Frequency of Chronic Cough From 1963 to 1968 by Degree of Exposure to Chloromethyl Methyl Ether
No Of Men
30 27

Exposure
Indext 0 0.1-4.9 5.0-12.9 >13.0

Frequency
No(%) in 1963

Total

27 23 107

7(23) 6(22) 12(44) 9(39)

34(32)

No(%) in 1968 5(17) 3(11) 4(14) 6(26) 18(17)

Change, %
-7 -11

-30 -13 -15

Chloromethyl methyl ether is a use ful chemical in the manufacture of ion-exchange resins, pesticides, bac
tricides, dispersing agents, waterrepellents, flame-proofing agents, and other products. Bis(chlorbmethyl)

"Comparing the change in prevalence of cough among all exposed groups with the change in the unexposed group, Z=1.82; .06 < P < .07 (two-tailed test). jTime-weighted exposure rating multiplied by years of exposure (1948-1964).
Table 8.Change* in Frequency of Dyspnea From 1963 to 1968 by Degree
of Exposure to

Chloromethyl Methyl Ether

in 1963

Exposure
Indext 0

ether is a contaminant in CMME. Both substances are volatile and in moist air break down to hydrochloric acid, formaldehyde, and, in the case of CMME, methanol. These chemicals are quite toxic, BCME more so than CMME.9 Longterm lifetime inhalation of CMME at 1 ppm showed carcinogenicity of a low order.10 Bis(chloromethyl) ether was much more potent8 and the inci dences were dose-related. Among 200 rats with limited numbers of expo sures to BCME in a concentration of only 0.1 ppm, 40 cancers developed, mainly squamous-cell carcinoma of the lung and esthesioneuroepitheliomas of the nose. The CMME ex periments showed development of

0.1-4.9
5.0-12.9 >13.0 Total

7(7) "Comparing change in prevalence of dyspnea among all exposed groups with change unexposed group, Z=3.50; P < .001 (two-tailed test). fTime-weighted exposure rating multiplied by years of exposure (1948-1964).

No of Men 30 27 27 23 107

No(%)
0
0

5(17)
2(7)

Frequency No(%) in 1968 6(20) 6(22) 7(26) 7(30) 26(24)

Change, %
+3 4-22 4-19 4-30 +17

trachal squamous metaplasia and bronchial mucosal hyperplasia in rats. These carcinogenic effects have been demonstrated in human beings. In 1972, there was a report of lung cancer in four workers exposed to BCME in a western ion-exchange resin plant.13 In 1973, Figueroa and associates2 described 14 cases of lung cancer in workers at the same chem ical plant where the present study

11 cases reported among the 14 cases previ ously reported). In the same year, Sakabe14 in Japan reported five cases among 32 people exposed to BCME over a 15-year period, and Von Thiess et al15 in Germany reported six cases among 18 workers exposed to BCME in a plant research laboratory over a 13-year period and two cases among 50 workers exposed in the plant itself.
was

done

here

(10 of the

were

Downloaded from www.jama.com by guest on November 4, 2010

The dose-related

lung

cancer

leaves little doubt that CMME or its BCME contaminant is carcinogenic for humans. In addition, there are differences between the cancers in the men exposed to CMME and lung cancers in the general population. Most striking are the inverse rela tionship of lung cancer incidence to smoking and the lack of relation to chronic cough, although the figures are based on small numbers. In the general population, smoking habits are the strongest predictor of lung cancer risk,1 and coughing cigarette smokers are at higher risk than noncoughing smokers.16 Among the chemical workers, self-selection may explain these differences in part. Since CMME is a bronchial irritant, chemical workers who smoked, espe cially those who also coughed, may have avoided working with CMME or may have become exsmokers or light smokers if they were heavy smokers prior to the start of our observation period. However, the peculiar rela tionship to smoking persists in the moderately and heavily exposed groups, so self-selection is not an ade quate explanation for this observa tion. All of the cancers in CMME-exposed men in this study appeared by age 55 or younger, while a majority of lung cancers in the general popu lation occur after age 55.J The aver age age of the five men with lung cancer in the moderate-exposure group was 50, while that of the six men in the heavy exposure group was only 45. Since the average age at on set of exposure was 32 in both groups, this suggests shortening of the induc tion period with increasing CMME exposure and provides evidence that CMME or its contaminant is a more potent carcinogen than cigarette smoke, because the decrease in age at diagnosis of lung cancer with increas ing cigarette dosage is smaller.17 The frequency of small-cell carci noma in this series of cases is in marked contrast to the proportion of this histologie type in the lung can cers of the general population. In the Philadelphia Pulmonary Neoplasm Re search Project,18 only 20% of the 67

high incidence of reported in our study

epidemiologic study of the ship between industrial

relation CMME (BCME) exposure and symptoms of chronic bronchitis. The bronchial re action was not severe enough in 1965 to be reflected in abnormalities of ventilatory function. The dose-re sponse relationship between CMME exposure and chronic cough or expec toration was consistent with the changes in the tracheobronchial mu cosa recently reported in animals.910 At autopsy in one of our cases, the pa thologist described severe, extensive tracheobronchial squamous meta plasia. Our data show that smoking and CMME exposure were cofactors in the prevalence of symptoms of chronic bronchitis, although, sur prisingly, they were not cofactors in the incidence of bronchogenic carci
noma.

Stanton C. Kelton, Jr, PhD, and Lester DeFonso of Rohm and Haas Co provided the data on exposure of employees to CMME.

References

project: Basic risk factors of lung cancer in older men. Am J Epidemiol 95:4-16, 1972. 2. Figueroa WG, Raszkowski R, Weiss W: Lung cancer in chloromethyl methyl ether workers. N Engl J Med 288:1096-1097, 1973. 3. VanDuuren BL, Goldschmidt BM, Katz C, et al: Alpha haloethers: A new type of alkylating carcinogen. Arch Environ Health 16:472-476,

Philadelphia Pulmonary Neoplasm

1. Boucot

KR, Weiss W, Seidman H, et al: The

Research

The gradient in the prevalence of chronic cough in relation to CMME exposure was present in 1963 and in 1965 but had almost disappeared in 1968 as exposure diminished. The de creased frequency of chronic cough with time was expected, but the in crease in dyspnea was not. The sig nificance of the latter observation is not clear because there were no con comitant changes in the chest roentgenograms to explain it, and serial

pulmonary-function tests were not done. The frequency of dyspnea was determined by our questionnaire, which asked, simply, "Are you short of breath?" There being no quantifi
cation of the affirmative response, the severity of the symptom could not be assessed. Aging over the five-year pe riod would not explain the increase in frequency of dyspnea because the in
crease was

posed

negligible

among

unex

men.

tragic increased incidence of lung cancer, especially in a histologie type that is essentially untreatable and rapidly fatal, means that the only solution of the problem of CMME tox-

The

lung

cancers

among

cigarette-only

smokers were small-cell carcinomas. This is, to our knowledge, the first

icity today is prevention. Periodic screening with chest roentgenograms and questionnaires are inadequate to detect lung cancer early enough to improve the "cure" rate.19-20 If similar episodes of industrial lung cancer are to be prevented, chemicals must be subjected to appropriate testing to define safe working conditions before they are introduced into the work place.

1968. 4. VanDuuren BL, Sivak A, Goldschmidt BM, et al: Carcinogenicity of haloethers. J Natl Cancer Inst 43:481-486, 1969. 5. Gargus JL, Reese WH Jr, Ritter HA: Induction of lung adenoma in newborn mice by bis(chloromethyl) ether. Toxicol Appl Pharmacol 15:92-96, 1969. 6. Leong BKJ, MacFarland HN, Reese WH: Induction of lung adenomas by chronic inhalation of bis(chloromethyl) ether. Arch Environ Health 22:663-666, 1971. 7. Laskin S, Kuschner M, Drew RT, et al: Tumors of the respiratory tract induced by inhalation of bis(chloromethyl) ether. Arch Environ Health 23:135-136, 1971. 8. Kuschner M, Laskin S, Drew RT, et al: Inhalation carcinogenicity of alpha haloethers: III. Lifetime and limited period inhalation studies with bis(chloromethyl) ether at 0.1 ppm. Arch Environ Health 30:73-77, 1975. 9. Drew RT, Laskin S, Kuschner M, et al: Inhalation carcinogenicity of chloromethyl methyl ether and bis(chloromethyl) ether: I. The acute inhalation toxicity of chloromethyl methyl ether and bis(chloromethyl) ether. Arch Environ Health 30:61-69, 1975. 10. Laskin S, Drew RT, Capiello V, et al: Inhalation carcinogenicity of alpha haloethers: II. Chronic inhalation studies with chloromethyl methyl ether. Arch Environ Health 30:70-72, 1975. 11. Baldwin E deF, Cournand A, Richards DW Jr: Pulmonary insufficiency: I. Physiological classification, clinical methods of analysis, standard values in normal subjects. Medicine 27:243\x=req-\ 278, 1948. 12. Axtell LM: Computing survival rates for chronic disease patients. JAMA 186:1125-1128, 1963. 13. Fishbein G: Occupational Safety and Health Letter 2:1, 1972. 14. Sakabe H: Lung cancer due to exposure to bis(chloromethyl) ether. Ind Health 11:145-148, 1973. 15. Von Thiess AM, Hey W, Zeller H: Zur Toxikologie von Dichlordimethyl\l=a"\ther:Verdacht auf kancerogene Wirkung auch beim Menschen. Zentralbl Arbeitsmed 23:97-102, 1973. 16. Boucot KR, Cooper DA, Weiss W, et al: Cigarettes, cough, and cancer of the lung. JAMA 196:985-990, 1966. 17. Weiss W: Cigarette smoke as a carcinogen. Am Rev Respir Dis 108:364-366, 1973. 18. Weiss W, Boucot KR, Seidman H, et al: Risk of lung cancer according to histologic type and cigarette dosage. JAMA 222:799-801, 1972. 19. Boucot KR, Weiss W: Is curable lung cancer detected by semiannual screening? JAMA 224:1361-1365, 1973. 20. Weiss W, Seidman H, Boucot KR: The Philadelphia Pulmonary Neoplasm Research Project: Thwarting factors in periodic screening for lung cancer. Am Rev Respir Dis 111:289-297, 1975.

Downloaded from www.jama.com by guest on November 4, 2010