Hydroquinone

Epidemiological data

Administrative data

Endpoint:

epidemiological data

Type of information:

experimental study

Adequacy of study:

key study

Study period:

1942 - 1990

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Study meets basic scientific principles as an epidemiological, prospective cohort study in HQ exposed workers: study well designed and documented, with essentially complete follow-up, exposure data, and minimization of bias; study acceptable as key study

Cross-reference

Reason / purpose for cross-reference:

reference to same study

Data source

Reference

Reference 1

Reference Type:

publication

Title:

Mortality study of employees engaged in the manufacture and use of hydroquinone.

Author:

Pifer JW, Hearne FT, Swanson FA, O'Donogue JL

Year:

1995

Bibliographic source:

Int Arch Occup Environ Health 67, 267 - 280

Reference 2

Reference Type:

publication

Title:

Quinone vapors and their harmful effects. II. Plant exposures associated with eye injuries.

Author:

Oglesby FL, Sterner JH & Anderson B

Year:

1947

Bibliographic source:

J Ind Hyg Toxicol 29, 74 - 84

Materials and methods

Study type:

cohort study (retrospective)

Endpoint addressed:

repeated dose toxicity: inhalation

carcinogenicity

Principles of method if other than guideline:

Mortality causes were evaluated in a 1942-1990 cohort of workers employed in the manufacture and use of hydroquinone .

GLP compliance:

no

Remarks:

epidemiological study

Test material

Test material information

Constituent 1

Reference substance name:

Hydroquinone

EC Number:

204-617-8

EC Name:

Hydroquinone

Cas Number:

123-31-9

Molecular formula:

C6H6O2

IUPAC Name:

hydroquinone

Test material form:

other: workplace environment (dusts, vapours)

Details on test material:

- Name of test material (as cited in study report): hydroquinone, HQ

Method

Type of population:

occupational

Ethical approval:

not applicable

Details on study design:

HYPOTHESIS TESTED (if cohort or case control study): existence of excesses of kidney or liver cancer or leukemia

METHOD OF DATA COLLECTION
- Type: review of death certificates
- Details: Vital status was obtained from both internal (Tennessee Eastman Division (TED) mortality data base) and external (Social Security Administration) sources. Death certificates have been routinely collected at TED since the 1940s. In addition, the social security numbers of the study subjects were matched against the Social Security Administration's Death master file through 1990.

STUDY PERIOD: 1942 to 1990

SETTING: Tennessee Eastman Division of Eastman Chemical Company in Kingsport, Tennessee; production of hydroquinone since 1930

STUDY POPULATION
- Total population (Total no. of persons in cohort from which the subjects were drawn): 879
- Selection criteria: employes working at least 6 months between January 1930 and December 1990 in hydroquinone manufacturing and other areas, in which HQ was used; identified from eye examination lists and medical records (1942-1975) and from computer files (1967-1990)
- Total number of subjects participating in study: 879
- Sex/age/race: 858 men, 21 women (no further data)
- Smoker/nonsmoker: no data
- Total number of subjects at end of study: 876 (death certificates were unavailable for 3 decedents)

COMPARISON POPULATION
- Type: State registry / Reference group
- State comparison: Mortality experience of study subjects was compared with vital statistics from the general population of the State of Tennessee. The expected number of deaths by cause was calculated by applying quinquennial age-sex-cause-specific death rates, 1950-1990, to the appropriate survival years of follow-up in the study population.
- Reference group comparison: similar procedure used for 30,000 hourly wage personnel employed at Kodak's Rochester facilities from 1964 to 1992, few of whom had potential HQ exposure

HEALTH EFFECTS STUDIED
- Disease(s): malignant and non-malignant diseases specified as causes of death
- ICD No.: 9
- Year of ICD revision: 1977
- Diagnostic procedure: cause of death was coded by a nosologist

Exposure assessment:

measured

Details on exposure:

TYPE OF EXPOSURE: inhalation of HQ dust and BQ vapours

TYPE OF EXPOSURE MEASUREMENT: Area air sampling / Personal sampling; short-term measurements since late 1940s until to 1984, full-shift samples since 1985

EXPOSURE LEVELS: estimated annual exposure to HQ dust as estimated 8-h time-weighted averages: mean values ranging from 0.1 to 2.2 mg/m3 depending on working operation (see Table 1), range from <0.1-6.0 mg/m3; higher levels of HQ dust reported before 1942 with up to 35 mg/m3 e.g. in the packaging area; benzoquinone vapour 0.1-0.3 mg/m3.
Career exposure estimates of HQ dust levels for five periods: 1930-1945, 5.0 mg/m3; 1946-1951, 2.5 mg/m3; 1952-1956, 6.0 mg/m3; 1957-1966, 2.0 mg/m3; 1967-1990, 0.4 mg/m3

EXPOSURE PERIOD: 1942-1990, 22,895 person-years during the 50-year observation period
- Person-years of risk: initiated after the subject had completed 6 months of sevice were accumulated for the study subjects by 5-year age group, sex and calendar year category beginning January 1942, or on the date hired (between 1942 and 1990) and ending at death, the date of termination, or 31 December 1991.
Distribution of cohort members according to year of first employment: 1930-1939, 5%; 1940-1949, 15%,; 1950-1959, 22%; 1960-1969, 21%; 1970-1979, 27%; 1980-1990, 10%
Mean duration of tenure in a HQ environment was 13.7 years, mean follow-up from first exposure was 26.8 yrs; 2205 person-years (10%) were represented by subjects with ages of 65 and older
Termination of exposure by 10% or N=91 subjects with average age of 28.8 yrs and average length of service of 3.8 yrs.

POSTEXPOSURE PERIOD: no data for subjects on inactive status (35%) or died subjects (19%), 46% were on active status on 31 December 1991

DESCRIPTION / DELINEATION OF EXPOSURE GROUPS / CATEGORIES:
Low exposure (LE): < 3mg/m3-years; medium exposure (ME): 3 to <15 mg/m3-years; high exposure (HE): >= 15 mg/m3-years

Statistical methods:

Statistical significance of obseved-to-expected differences was tested according to two sets of criteria depending on whether the cause of death was hypthesized or nonhypothesized. A priori hypotheses for deaths due to kidney and liver cancer and leukemia were evaluated using one-tailed t-test; all other causes assessed using two-tailed t-test, both with P<=0.05; Poisson probability distribution applied to test statistical significance

Results and discussion

Results:

OBSERVED DEATHS/ EXPECTED DEATHS (OD/ED vs. Tennessee and Kodak Rochester control populations, respectively)
No statistically significant increased for any cause of death
statistically signifcantly decreased OD/ED-Ten/ED-KR: all deaths 168 / 256.7 / 204.3; all maligant neoplasms 33 / 57.9 / 62; respiratory system malignancies 14 / 24.5 (Ten); digestive system malignancies (KR) 7 / 15.4; cardiovascular diseases 92 / 118.3 (Ten); digestive diseases 3 / 9.1 (Ten)
OD/ED for hypothesized causes of death: kidney tumours 2 / 1.3 / 1.3 (not significant); liver 0 / 1.3 /0.8; leukemia 0 / 2.3 / 2.7

STATISTICAL RESULTS (for details see Table 2)
- SMR (Standard mortality ratio vs. Tennessee and Kodak Rochester control populations, respectively):
No statistically significantly increased SMRs;
statistically signifcantly decreased SMRs: all observed deaths 65 and 82, all malignant neoplasms 57 and 62;
SMRs for hypothesized causes of death: kidney tumours 150 and 159 (not significant); liver 0; leukemia 0

EVALUATION OF DOSE RESPONSE (for details see Table 3): cumulative exposure categories < 3, 3 to <15, >=15 mg/m3-years
No evidence of a dose-related effect for any cause of death category;
statistically signifcantly decreased SMRs vs. Tennessee control population: all observed deaths 55, 75, 66; all malignant neoplasms 35, 77, 60
OD/ED for hypothesized causes of death: kidney tumours OD 0, 1, 1, ED 0.4-0.5; liver OD 0, ED 0.1 (for all categories); leukemia OD 0, ED 0.7-1.0 (for all categories)

EVALUATION ACCORDING TO TIME FROM FIRST EXPOSURE (for details see Table 4): categories < 20 yrs, 20 to 34 yrs, >= 35 yrs
No consistent mortality pattern;
statistically signifcantly decreased SMRs vs. Tennessee control population: all observed deaths 63, 62, 70; all malignant neoplasms 53, 34, 77
OD/ED for hypothesized causes of death: kidney tumours OD 0, 0, 2, ED 0.3-0.6 (no significant effect); liver OD 0, ED 0.1-0.2; leukemia OD 0, ED 0.7-1.0

EVALUATION ACCORDING TO CAREER EXPOSURE AND TIME FROM FIRST EXPOSURE (for details see Table 5):
No significant effects with the exception of respiratory cancer deaths (statistically significant heterogeneity, P<0.007; but no evidence of a dose-response relationship)

CHARACTERISTICS OF KIDNEY CANCER DEATHS (see also confounding factors)
Case 1: subjects worked for 35 yrs in various jobs in the chemical industry with exposure to several chemicals; estimated HQ career exposure of 4.5 mg/m3-years based on less than 2 years of working as crude process operator in the early 1950s; diagnosis of renal cell carcinoma in 1985, death in 1987 at age of 62 yrs
Case 2: 38-year career in chemical industry with HQ-career of 22 yrs from 1933 to 1953 as operator in oxidation, filtration, and crystallization of HQ; lifetime HQ exposure 86.8 mg/m3-years; transferred to a non-HQ environment in 1953 because of progressive visual loss from corneal involvement; retirement in 1968, death in 1984 of a malignant hypernephroma at age of 78 yrs

Confounding factors:

CHARACTERISTICS OF KIDNEY CANCER DEATHS
possible exposure to other chemicals
History of smoking: both subjects had long-term tobacco usage with one pack a day for at least 20 and 40 yrs, respectively.

Strengths and weaknesses:

Strengths:
- use of both occupational and general population reference groups
- occupational reference group selected from large imaging and health products company (Eastman Kodak Company) provided comparability with respect to type of industry, compensation, health insurance coverage, medical services, and safety practices, including on-site smoking restrictions.
- general population reference group: State of Tennessee comparison group with similar geography and rate stability associated with large population base
- cohort definition and case finding: based on detailed review of employment files, medical records, personal correspondence, reports prepared from the company's personnel database, and most importantly, lists of employess scheduled for eye examinations. Starting in the 1940s and continuing to the present, all employees (including suppport personnel) who manufactured and used HQ at TED have received periodic eye examinations, which have also been available to retirees. Therefore the study group presumably included essentially all individuals who were exposed to HQ and BQ at TED during the past 50 yrs.
- minimization of selectivity bias due to exclusion of pre-1942 person-years: large proportion of early HQ production personnel included in cohort; stability employment at TED and the lack of job opportunities during the 1930s presumably reduced turnover rates. Exclusion of pre-1942 person-years represents a small bias in favor of positive findings.
- sufficient time from first exposure (TTFE) for development of carcinogenic effects: more than 20% of subjects followed for at least 40 yrs; median TTFE > 25 yrs;
subjects with longest follow-up experienced higest mean cumulative exposure: TTFE >= 35 yrs, 20.0 mg/m3-yrs; TTFE 20-34 yrs, 4.1 mg/m3-yrs; TTFE < 20 yrs, 2.3 mg/m3-yrs
- period of follow-up covers early exposures to HQ and BQ: exposures were excessive and significantly greater than current workplace standards with detectable biomarkers of exposure as demontrated by the presence of ocular pigmentation and corneal damage (for details see Oglesby et al, 1947).

Any other information on results incl. tables

Legend Table 2 to 4:

OD, observed deaths; ED, expected deaths; SMR, standardised mortality ratio; TN, Control population State of Tennessee; KR, Control population Kodak Rochester

* statistically significant decrease, P<0.05 two-tailed t-test; ** statistically significant increase, P<0.05 two-tailed t-test

Hypothesized causes of death

^a There was no statistically significant change in HQ-exposed workers for any other cause of death: selection is either based on a statistically significantly decreased number of deaths in the HQ cohort compared to the respective control population, or on data for causes of death related to the tested hypothesis

^b expected numbers based on Tennessee white death rates 1950, 1955, 1960, 1965, 1970, 1975, 1980, 1985, and 1990

^c expected numbers based on Kodak Rochester hourly death rates 1964-1967, 1968-1972, 1973-1977, 1978-1982, 1983-1987, and 1988-1992

^d X² test for linear trend

^e test for difference among observed/expected ratios

^f there were no liver or leukaemia deaths

^g expected numbers based on the within cohort person-years distribution by age, sex, and calendar year strata

Table 2: Number of observed and expected deaths, and SMRs and 95% confidence intervals by selected causes^a

ICD-9	Cause of death category	HQ cohort	TN			KR
		OD	ED ^b	SMR	95% CI	ED ^c	SMR	95% CI
140-208	Malignant neoplasms	33	57.9*	57	39-80	53.5*	62	542-87
150-159	Digestive system	7	12.5	56	23-116	15.4*	46	18-94
155	Liver and intrahepatic bile ducts	0	1.3	0	0-292	0.8	0	0-434
160-165	Respiratory system	14	24.5	57	31-96	16.8	83	45-140
189	Kidney and other urinary organs	2	1.3	150	17-541	1.3	159	18-574
204, 208	Leukemia	0	2.3	0	0-159	2.7	0	0-134
390, 459	Diseases of circulatory system	92	118.3*	78	64-95	99.6	92	74-113
460-519	Diseases of respiratory system	11	19.0	58	29-104	12.2	90	45-162
520-579	Diseases of digestive system	3	9.1*	33	7-97	6.6	45	9-133
580-629	Diseases of genitourinary system	1	3.2			2.5
001-999	All causes of death	168	256.7*	65	56-76	204.3*	82	70-96

Table 3: Number of observed and expected deaths by selected ^a underlying cause and career hydroquinone exposure (mg/m³-years)

Cause of death category		< 3 mg/m^3-years			3 - 15 mg/m^3-years			>= 15 mg/m^3-years
		OD	ED		OD	ED		OD	ED
			TN ^b	KR ^c		TN	KR		TN	KR
Malignant neoplasms	All	7	20.2*	18.7*	15	19.4	18.0	11	18.3	16.8
	Trachea, bronchus and lung	1	8.0*	5.6*	10	7.4	5.5	3	7.0	5.2
	Kidney	0	0.5	0.4	1	0.4	0.4	1	0.4	0.4
	Liver	0	0.1	0.1	0	0.1	0.1	0	0.1	0.1
	Leukemia	0	0.9	1.0	0	0.7	0.9	0	0.7	0.8
All causes of death		50	90.5*	69.6*	65	86.3*	69.8	53	79.9*	64.9
Standardized mortality ratio (SMR)			55*	72*		75*	93		66*	82

Table 4: Number of observed and expected deaths by selected ^a underlying cause and time from first exposure

Cause of death category		< 20 years			20 – 34 years			>= 35 years
		OD	ED		OD	ED		OD	ED
			TN^b	KR^c		TN	KR		TN	KR
Malignant neoplasms	All	6	11.3	11.9	7	20.7*	18.5*	20	25.9	23.1
	Trachea, bronchus and lung	3	3.1	2.7	2	8.7*	6.1	9	10.6	7.6
	Liver	0	0.1	0.1	0	0.1	0.1	0	0.2	0.1
	Kidney	0	0.3	0.4	0	0.5	0.4	2	0.6	0.5
	Leukemia	0	0.7	0.9	0	0.7	1.0	0	0.9	0.9
Diseases of circulatory system	All	16	23.2	20.0	38	42.0	34.0	38	53.1*	45.6
	Cerebrovascular disease	5	2.7	1.6**	2	5.0	3.3	8	8.1	5.7
Symptoms and ill-defined conditions		2	1.7	0.2**	0	1.8	0.4	1	2.1	0.6
All causes of death		41	62.5*	47.2	53	85.8*	66.1	74	105.7*	91.0
Standardized mortality ratio (SMR)			63*	87		62*	80		70*	81

Further table with results presented in Overall remarks, attachments

Applicant's summary and conclusion

Conclusions:

The prospective cohort study showed no evidence of excess mortality in workers employed principally in the production of HQ at a large chemical plant during 1942 to 1991. In contrast, the number of deaths from all causes of death, from all forms of malignant cancer was statistically significantly below expectation compared with both general population and occupational controls. There was also a lower than expected death rate from circulatory diseases. Additionally, there were fewer than expected deaths from respiratory cancer, non-malignant pulmonary diseases, and genitourinary system illness, including nephrotoxicity. There was no evidence of a dose-relationship with respect to lifetime exposure or latency. This study did not support the findings of animal cancer bioassays, that had shown an increased prevalence of predominately benign neoplasms of the kidney and liver and of mononuclear cell leukemia in F344 rats or B6C3F1 mice (Kari et al., 1992; NTP 1989; Shibata et al., 1991). In the cohort of hydroquinone workers, cancer mortality was not statistically significantly different from expectation, and no liver tumour or leukaemia deaths were observed. The two observed cases of malignant kidney tumors (1.3 expected, no significant effect) were confounded by an excessive smoking history, which is recognized as major cause of renal cancer in humans, and by workplace exposure to other chemicals.

Executive summary:

A retrospective study was performed in a well-characterized 1942-1990 cohort of 879 workers (858 men and 21 women) employed principally in the manufacture and use of HQ at a large chemical plant comprising 22,895 person-years of exposure during the 50-year observation period. Mean duration of tenure in a HQ environment was 13.7 years, mean follow-up from first exposure was 26.8 yrs; 2205 person-years (10%) were represented by subjects with ages of 65 and older. Average exposure concentrations, 1949-1990, ranged from 0.1 to 6.0 mg/m³ for HQ dust and less than 0.1 to 0.3 mg/m³ for benzoquinone vapour. 20% of cohort members had presumably experienced higher exposures to HQ dust, e.g. up to 35 mg/m³ in the packaging area, at the period up to 1949, before measures to reduce exposure had been installed. Causes of mortality were followed up to 1991 and were compared with vital statistics from the general population of the State of Tennessee, and to an occupational reference group of 30,000 hourly wage personnel employed at Kodak's Rochester facilities from 1964 to 1992, few of whom had potential HQ exposure.

There was no evidence of excess mortality in the investigated cohort. In contrast, the number of deaths from all causes of death, from all forms of malignant cancer was statistically significantly below expectation compared with both general population and occupational controls. There was also a lower than expected death rate from circulatory diseases. Additionally, there were fewer than expected deaths from respiratory cancer, non-malignant pulmonary diseases, and genitourinary system illness, including nephrotoxicity. There was no evidence of a dose-relationship with respect to lifetime exposure or latency. This study did not support the findings of animal cancer bioassays, that had shown an increased prevalence of predominately benign neoplasms of the kidney and liver and of mononuclear cell leukemia in F344 rats or B6C3F1 mice (Kari et al., 1992; NTP 1989; Shibata et al., 1991). In the cohort of hydroquinone workers, cancer mortality was not statistically significantly different from expectation, and no liver tumour or leukaemia deaths were observed. The two observed cases of malignant kidney tumors (1.3 expected, no significant effect) were confounded by an excessive smoking history, which is recognized as major cause of renal cancer in humans, and by workplace exposure to other chemicals.