ELSEVIER
Contents lists available at ScienceDirect
Toxicology and Applied Pharmacology
journal homepage: www.elsevier.com/locate/ytaap
H44A
Toxicology and Applied Pharmacology
Inhalation carcinogenicity study with nickel metal powder in Wistar rats
Adriana R. Oller a,*, Daniel T. Kirkpatrick b, Ann Radovsky b, Hudson K. Bates ª
a NiPERA, 2605 Meridian Parkway, Suite 200, Durham, NC 27713, USA
b WIL Research Laboratories, LLC, 1407 George Road, Ashland, OH 44805 8946, USA
ARTICLE INFO
Article history: Received 30 May 2008
Revised 21 August 2008
Accepted 23 August 2008 Available online 11 September 2008
Keywords:
Nickel Metal
Carcinogenicity Cancer Inhalation
ABSTRACT
Epidemiological studies of nickel refinery workers have demonstrated an association between increased respiratory cancer risk and exposure to certain nickel compounds (later confirmed in animal studies). However, the lack of an association found in epidemiological analyses for nickel metal remained unconfirmed for lack of robust animal inhalation studies.
In the present study, Wistar rats were exposed by whole-body inhalation to 0, 0.1, 0.4, and 1.0 mg Ni/m3 nickel metal powder (MMAD=1.8 um, GSD=2.4 um) for 6 h/day, 5 days/week for up to 24 months. A subsequent six-month period without exposures preceded the final euthanasia. High mortality among rats exposed to 1.0 mg Ni/m3 nickel metal resulted in the earlier termination of exposures in this group. The exposure level of 0.4 mg Ni/m3 was established as the MTD for the study. Lung alterations associated with nickel metal exposure included alveolar proteinosis, alveolar histiocytosis, chronic inflammation, and bronchiolar-alveolar hyperplasia.
No increased incidence of neoplasm of the respiratory tract was observed. Adrenal gland pheochromocy- tomas (benign and malignant) in males and combined cortical adenomas/carcinomas in females were induced in a dose-dependent manner by the nickel metal exposure. The incidence of pheochromocytomas was statistically increased in the 0.4 mg Ni/m3 male group. Pheochromocytomas appear to be secondary to the lung toxicity associated with the exposure rather than being related to a direct nickel effect on the adrenal glands. The incidence of cortical tumors among 0.4 mg Ni/m3 females, although statistically higher compared to the concurrent controls, falls within the historical control range; therefore, in the present study, this tumor is of uncertain relationship to nickel metal exposure.
The lack of respiratory tumors in the present animal study is consistent with the findings of the epidemiological studies.
@ 2008 Elsevier Inc. All rights reserved.
Introduction
For the carcinogenicity assessment of nickel substances, the consideration of speciation (i.e., chemical form of nickel) is of utmost importance. In a critical epidemiological study published in 1990 in which 80,000 workers from ten different cohorts were included, an association between increased respiratory cancer mortality risk (both for nasal and lung tumors) and exposure to the water soluble and insoluble nickel compounds generated during the process of sulfidic ore refining was found. By contrast, no such association was noted for nickel metal exposures (ICNM, 1990). The positive association of increased respiratory cancer mortality with high inhalation exposures to sulfidic and oxidic nickel compounds was later confirmed by the positive results in animal inhalation studies with crystalline nickel subsulfide and high temperature green nickel oxide (NTP 1996a,b). In the case of nickel metal, the lack of an association found in
epidemiological studies remained unconfirmed due to the lack of robust animal inhalation studies.
Existing animal studies with nickel metal powder included earlier studies in rats, mice, guinea pigs, and hamsters via inhalation (Hueper, 1958; Hueper and Payne, 1962). These studies were all negative for respiratory tumor induction. However, some of these studies had very high toxicity or lacked proper controls. Intratracheal instillation and injection studies (e.g., intramuscular, intraperitoneal, subcutaneous, etc.) using metallic nickel powders produced both negative and positive results (Ivankovic et al., 1988; Pott et al., 1987; Muhle et al., 1992; and references cited in IARC, 1990 or Sivulka, 2005). Intratracheal instillation studies can give false-positive results if physiological clearance mechanisms are bypassed (Driscoll et al., 2000). Therefore, questions still remained about the intrinsic potential of nickel metal to be carcinogenic via inhalation. It was important for a modern study with appropriate controls and design to be done to critically test the hypothesis that nickel metal powders are not carcinogenic.
Based on the above mentioned results in animal studies with nickel metal powder and the fact that the human cohorts exposed mainly to nickel metal were relatively small, the lack of a carcinogenic
* Corresponding author. Fax: +1 919 544 7724. E-mail address: aoller@nipera.org (A.R. Oller).
potential for metallic nickel to humans could not be completely ruled out. This prompted the European Union (EU, 2006) and Germany’s BK Tox (Beraterkreis Toxikologie)1 to request that a guideline-compliant study with nickel metal powder be conducted.
An inhalation carcinogenicity study with nickel metal powder in male and female Wistar rats was initiated in 2004 (Kirkpatrick, 2008). Rats were chosen for this study as they represent a sensitive animal species to detect the possible carcinogenic effects of nickel particu- lates after inhalation (NTP 1996a,b).
Methods
Materials
Nickel metal powder (CAS No. 7440-02-0) was received from a nickel producing company. This powder is available from NiPERA’s Sample Repository upon request. The powder was stored at room temperature and was proven to be stable under this condition. The purity of the powder at the beginning and completion of the study was 99.9% as determined by Instrumental Neutron Activation Analysis (Elemental Analysis Incorporated, Lexinton, KY).
Animals and animal husbandry
Animal housing and care followed currently accepted standards. Three hundred twenty-eight male and 328 female Wistar Crl:WI(G1x/ BRL/Han) IGS BR VAF/Plus® rats were purchased from Charles River Laboratories, Inc., Raleigh, NC. While the animals were acclimated to the laboratory conditions prior to initiation of exposure, they were observed twice daily and weighed twice. Following randomization, each animal was identified with a tail tattoo and subcutaneous implantation of an American Veterinary Identification Device (AVID microchip). These devices were encapsulated in glass and had no exposed metal components. The results of the pretest health screen and serological analyses conducted by BioReliance Corporation, Rockville, MD, indicated that the population of animals was healthy.
Diet and drinking water
PMI Nutrition International Certified Rodent Chow® #5002 and purified municipal water were provided to the animals ad libitum throughout the study except during the daily 6-hour exposure periods when both food and water were unavailable (to limit consumption of nickel and in accordance with standard practice). Water was supplied to the animals by an automatic watering system or by bottles when clinical signs warranted. Within generally accepted limits, there were no contaminants in the diet or drinking water which would interfere with the conduct or objectives of the study.
Environmental conditions
In the animal housing room, controls were set to maintain an average daily temperature between 19 ℃ and 25 ℃ and relative humidity between 30% and 70%. Temperature and relative humidity were monitored continuously. Fluorescent lighting provided illumina- tion for a 12-hour light/12-hour dark photoperiod.
Selection of exposure levels
The exposure levels for the carcinogenicity study were selected based on the results from the 13-week range-finding study. The 13-
week inhalation range-finding study with nickel metal powder was conducted using exposure levels of 0, 1, 4, 8 mg/m3 (Kirkpatrick, 2004).2 Based on the toxicity observed in the 13-week range-finding study, exposure levels of 0, 0.1, 0.4 and 1.0 mg/m3 were selected for the 30-month inhalation carcinogenicity study (two-year exposure period followed by a six-month period without exposure prior to final euthanasia).
Exposure method
The animal exposures were conducted in four 2.0-m3 stainless steel and glass whole-body exposure chambers. One chamber was dedicated to each group for the duration of the study. The control group was exposed to clean, filtered air under conditions otherwise identical to the nickel-exposed groups. The chambers were operated under dynamic conditions from a HEPA- and charcoal-filtered, temperature- and humidity-controlled air source. Temperature and humidity daily averages were between 20 ℃ and 26 ℃, and 30% and 70%, respectively. There were at least 12-15 chamber air changes/ hour. Chamber airflow rate, temperature, and relative humidity were monitored continuously. During exposures, the animals were housed individually in stainless steel wire-mesh cage batteries. During the conduct of the study, the cage batteries were rotated between the six chamber positions in a regular fashion.
Aerosol generation
Nickel metal atmospheres were generated as dust aerosols of elemental nickel using Fluidized Bed Aerosol Generators (Model no. 3400A, TSI, Inc., St. Paul, MN). The generation systems were optimized to produce an aerosol of 1.5-3.0 um MMAD (Mass Median Aero- dynamic Diameter). This particle size range is considered optimal for delivery of nickel aerosols to all regions of the respiratory tract of the rat.
Characterization of exposure atmospheres
Real-time aerosol monitoring was performed using light scattering aerosol monitors as a guide to adjust concentration during animal exposure. Actual individual exposure concentrations of nickel metal were determined by standard gravimetric methods. For each chamber, at least one set of filters per week were analyzed for nickel by atomic absorption spectroscopy in the Analytical Chemistry Department, WIL Research Laboratories, LLC. For the control chamber, the analytically determined nickel concentrations were below the limit of quantita- tion. Overall mean analyzed exposure concentrations were 0.10, 0.41 and 1.01 mg/m3 for the 0.1, 0.4 and 1.0 mg/m3 groups, respectively. Aerosol particle size determinations for the nickel metal aerosols were conducted weekly during study weeks 0-18 and monthly thereafter using cascade impactors. The overall mean aerosol particle size in micrometers corresponding to the MMAD (with geometric standard deviation, GSD, in parenthesis) was 1.8 (2.40), 1.7 (2.16) and 1.8 (2.09) for the 0.1, 0.4 and 1.0 mg/m3 groups, respectively.
Experimental design
This study was conducted in general compliance with the OECD and EPA guidelines for carcinogenicity studies (OECD 451 adopted in 1981) and conformed with the Good Laboratory Practice (GLP) Standards promulgated by the EPA and OECD. The study design for the carcinogenicity study is shown in Fig. 1. Changes to the original protocol are noted. In the Core study, 50 female and 50 male animals were assigned to each exposure group using a computer
1 This group of toxicology experts evaluated the carcinogenic, mutagenic, and reproductive properties of commerce materials and proposed classification to the German Committee on Dangerous Substances or AGS (Ausschuß für Gefahrstoffe). The German Ordinance on Hazardous Substances was renewed in December 2004 and the tasks of the BK Tox Group were taken over in essence by the Subcommittee III of the AGS.
2 Findings from this study are described in the Supplementary information.
randomization program. At the initiation of exposure, animals were approximately six weeks of age with body weights ranging from 129 to 192 g for males and 107 to 155 g for females. Satellite A groups consisted of 22 animals/sex/group. Satellite A animals had exposures identical to the Core animals but were scheduled for interim euthanasia to measure nickel lung burdens, blood nickel levels and/ or to isolate bronchiolar-alveolar lung fluid (BALF) for analysis.
The exposure period was 6 h/day, 5 days/week for 104 consecutive weeks, with the exceptions noted in Fig. 1. After completing 24 months of exposure, Core study animals in Groups 1, 2, and 3 (males) were maintained without further nickel exposure for an additional six months before scheduled necropsies took place. Females in Group 3 (0.4 mg/m3) in the Core study were maintained for additional 11 months without nickel exposure, after completing 19 months of exposure.
Blood samples for hematology evaluation were collected from all animals euthanized in extremis. Otherwise, blood was collected from the retro-orbital sinus of animals anesthetized by inhalation of isoflurane (Fig. 1). Hematology parameters were measured including the white blood cell differential count (as required by OECD 451 guideline).
Except for Group 4 animals that were transferred to Satellite B group for evaluation of lung nickel levels, all Core study animals were subjected to a complete gross necropsy examination with collection of standard protocol-defined tissues at the time of death or euthanasia. All tissues were examined microscopically from all animals in Groups 1 and 3, while only target tissues were examined in Group 2 animals. For Group 4, microscopic exams were performed only for the animals that died prior to the exposure termination dates. Histopathological evaluation was conducted according to standard protocol by one of the authors (A.R.), a WIL Research Laboratories board-certified veterinary pathologist. A peer review of the initial histopathology results was performed by a pathologist at Charles River Laboratories- Pathology Associates in order to confirm potential treatment-related effects.
Biokinetic evaluation
Satellite groups were included in the study to evaluate nickel blood levels, nickel lung burden, and BALF parameters (Fig. 1). Blood nickel levels were determined from 5 animals/group, twenty to twenty-four hours after completing the last exposure in each of the 3- and 12-
Acclimation/pretest period
Assignment of animals to study groups
Carcinogenicity (Core Study) Groups
Satellite A Phase Groups
50 animals/sex/group (Groups 1-4) exposed for 104 consecutive weeks, with following exceptions; exposures terminated early for Group 4 males (study day 375), Group 4 females (study day 427) and Group 3 females (study day 580); surviving Group 4 males and females reassigned to Satellite B Phase on day following exposure termination
22 animals/sex/group (Groups 1A-4A) exposed for up to 104 consecutive weeks
Clinical observations recorded daily; detailed physical examinations recorded weekly; body weights and food consumption recorded weekly or biweekly
Observations for moribundity and mortality twice daily. Clinical observations recorded daily; detailed physical examinations for palpable masses recorded weekly; body weights and food consumption recorded weekly or biweekly
Necropsies performed on all animals found dead or euthanized in extremis
Necropsies performed on 5 animals/sex/group at 3, 6 and 12 months of exposure and all surviving Group 4 males and females during study week 79 or 87; selected tissues examined macroscopically; lung nickel burden analyses at 3, 6, 12 and 24 months; blood nickel level analyses at 3 and 12 months; bronchoalveolar lavage evaluations at 6 and 12 months
Blood samples collected from all rats euthanized in extremis and at study weeks 52, 78 and 131 (scheduled necropsy) for hematology evaluation
4 Group 3 females reassigned to Core Study
Necropsies performed on all animals found dead or euthanized in extremis; selected tissues examined microscopically
Satellite B Phase Animals
Necropsies performed on all surviving animals; selected organs weighed; selected tissues examined microscopically
Necropsies performed on all animals found dead or euthanized in extremis
Necropsies performed on 6 animals/sex at 0 (females), 1, 3 and 6 months after termination of exposure; selected tissues examined macroscopically; lung nickel burden analyses conducted
month exposure groups. Blood was collected at the time of necropsy in a polypropylene tube and frozen at approximately -70 ℃. The frozen samples were transferred to the WIL Research Analytical Chemistry Department for nickel analysis by electrothermal atomic absorption spectroscopy. Prior to analysis, rat blood samples were digested with nitric acid at 70 ℃ in a water bath and centrifuged. The results were compared to reference calibration curves.
The right lobes of the lungs from 5 animals/sex/group after 6 and 12 months, and from surviving satellite animals after 24 months of exposure were collected and analyzed for nickel using atomic absorption spectroscopy by the Analytical Chemistry Department, WIL Research Laboratories, LLC. The lungs were placed in crucibles, minced with a scalpel and ashed at 500 ℃ for at least 30 min in a Thermolyne Furnace 6000. HCl (1.5 ml) and HNO3 (0.5 ml) were added to the crucibles to dissolve any metallic nickel. After additional rinses, the supernatant was filtered. Samples were diluted as needed to bring their concentrations within the range of the calibration curve. Lung nickel levels were expressed as total mass of nickel per right lung.
After 6 and 12 months of exposure, the left lungs from 5 animals/ sex/group were subjected to bronchiolar-alveolar lavage using 3 sequential lavages with 3-7 ml of room temperature Hank’s Balanced Salt Solution (Sigma-Aldrich, Inc., St. Louis, MO, cat. no. H6648) per lavage. The BALF was centrifuged and the supernatant from the first lavage was used for biochemical assays (protein and lactate dehy- drogenase, LDH). The cell pellets from all three lavages were combined for hemacytometer cell counts and differential cell counts. BALF cell differential counts (of 200 cells per animal) were performed by Consulting Clinical Pathologist, Kalamazoo, MI.
Historical control data
Two historical control databases were used in this study. A historical control database for Wistar rats was reported by Poteracki and Walsh (1998). The data came from 5 different 24-month studies conducted between 1980 and 1990, comprising 465 males and 465 females. The second historical control database was compiled by Bomhard (1992). The data came from 9 different 30-month control groups from 7 different studies, comprising 445 male and 448 female Wistar rats. These studies were finalized between 1979 and 1987. As expected, the background tumor rates in the Poteracki and Walsh studies after 24 months are lower than those in Bomhard’s studies after 30 months. Some of the spontaneous tumors that were identified as having significant increases between 24 and 30 months were pituitary adenomas, pheochromocytomas, and tumors of the mam- mary gland in females. The effect of age on spontaneous incidence of cortical adenomas was not able to be assessed in the Bomhard (1992) study.
Statistical analyses
Analyses were conducted using two-tailed tests (except as noted otherwise) for a minimum significance level of 5% and/or 1%, comparing each nickel metal-exposed group solely to the control group by sex. The Bonferroni inequality was used to control the false- positive error rate at the overall 0.05 level when several pair-wise comparisons were made. Statistical analyses were not conducted if the number of animals was 2 or less.
In-life data. Statistical analysis of in-life data was performed by WIL Research Laboratories, LLC, using statistical algorithms in WTDMS™. Body weight, body weight change, food consumption, white blood cell absolute and differential counts, organ weight and BALF data were subjected to a parametric one-way analysis of variance (ANOVA) (Snedecor and Cochran, 1980) to determine intergroup differences. If the ANOVA revealed statistical significance (p<0.05), Dunnett’s test
(Dunnett, 1964) was used, except in the case of BALF parameters where the Tukey test (Zar, 1999) was used.
Survival data. A Tumor Data Set (TDS) created from the WTDMS™ data in accordance with FDA guidelines (United States FDA, 1999) was used to perform statistical analyses on tumor incidence. Survival data and tumor incidence was analyzed by BioSTAT Consultants, Inc. (Portage, MI). Kaplan-Meier estimates (Kaplan and Meier, 1958) of group survival rates were calculated by sex and presented graphically. The generalized Wilcoxon test (Gehan, 1965) for survival was used to compare the homogeneity of survival rates across the groups. If the rates were significantly different (p<0.05), the generalized Wilcoxon test (Gehan, 1965) was used to make pair-wise comparisons, using Bonferroni adjusted significance levels. A log-rank dose-response trend test of survival rates was also performed.
Tumor analyses. The linear trend analysis by the method of Peto et al. (1980) was used to evaluate tumor incidence and guide interpretation of possible oncogenic effects. The mortality- prevalence method of Peto was performed without continuity correction, incorporating the context (incidental or fatal) in which tumors were observed. The fixed intervals used for incidental tumor analyses were: study weeks 0-50, 51-82, 83-104, 105-end of study (up to, but not including, scheduled terminal euthanasia) and the scheduled terminal euthanasia. Tumors were characterized as malignant, benign or as a metastatic site, by tissue or organ affected and by cell of origin. Each diagnosed tumor type was analyzed separately; analysis of combined tumor types was performed as described by McConnell et al. (1986).
An exact permutation test was conducted for analyses with low tumor incidence. Trend tests were conducted at 0.005 and 0.025 significance levels for common and rare tumors, respectively (Lin, 1995, 1997; Lin and Rahman, 1998a,b). Pair-wise comparisons with the control group were conducted at the 0.01 and 0.05 significance levels for common and rare tumors, respectively (Haseman, 1983). Rare tumors were defined as those with a spontaneous rate of less than 1% in the concurrent control group and/or published historical control information. Neoplasms that were not related to nickel exposure based on comparison to incidence in control animals or statistical analysis were considered as spontaneous tumors.
Results
13-week range-finding study
The toxicity data from the 13-week inhalation study with nickel metal powder was used to select the exposure range in the carcinogenicity study (Kirkpatrick, 2004).3 The selection of exposure levels for the carcinogenicity study was based on recommendations provided by an Expert Group of toxicologists from academic, regulatory and industrial backgrounds.4 The Expert Group followed the conduct of the study from the design of the 4-week toxicity study, to the completion and interpretation of the carcinogenicity study results. The exposure levels selected for the carcinogenicity study were 0.1, 0.4 and 1.0 mg/m3.
Based on the results from the 13-week study, it was expected that the dose of 1.0 mg/m3 could exceed the Maximum Tolerated Dose
3 The Supplementary information provides a brief description of the results from the 13-week study.
4 Expert Group: G. Oberdörster (Rochester University, USA), D. Beyersman (University of Bremen, Germany), A. Gamer (BASF, Germany), V Schultz-Kemp (Thyssenkrupp Nirosta, Germany), D. Broeckman (WvMetalle, Germany), P. Koundak- jian (Eurofer, ISSF, Belgium), A. Oller (NiPERA, USA), H. Bates (NiPERA, USA), and D. Kirkpatrick (WIL Laboratories, USA). In 2006, D. Beyersman retired and U. Heinrich (Fraunhofer Institute, Germany) joined the Expert Group.
A
Male Rat Survival in an Inhalation Carcinogenicity Study of Respirable Metallic Nickel Powder
1.0
araaaa
0.9
Kaplan-Meier Estimate of Survival
0.8
0.7
0.6
0.5
0.4
0.3
Control
0.1 mg Ni/m3
0.2
0.4 mg Ni/m3
1.0 mg Ni/m3
0.1
0.0
0
10
20
30
40
50
60
70
80
90
100
110
120
130
Weeks of Exposure
B Female Rat Survival in an Inhalation Carcinogenicity Study of Respirable Metallic Nickel Powder
1.0
0.9
Kaplan-Meier Estimate of Survival
0.8
0.7
0.6
0.5
0.4
Control
0.3
0.1 mg Ni/m3
0.4 mg Ni/m3
0.2
1.0 mg Ni/m3
0.1
0
10
20
30
40
50
60
70
80
90
100
110
120
130
Weeks of Exposure
(MTD) in a 2-year carcinogenicity study. However, it was considered that this would assure the maximum sensitivity to detect any possible weak carcinogenic effect associated with inhalation of nickel metal powder. The selected low exposure concentration of 0.1 mg/m3 was
| Group 1 0 mg Ni/m3 | Group 2 0.1 mg Ni/m3 | Group 3 0.4 mg Ni/m3 | Group 4 1.0 mg Ni/m3 | |
|---|---|---|---|---|
| Males | ||||
| No. of animals in group | 25 | 32 | 27 | 13 |
| Lung proteinosis | 0 | 0 | 4 | 0 |
| Thrombosis | 0 | 0 | 1 | 3 |
| (heart or atrial) | ||||
| Females | ||||
| No. of animals in group | 28 | 31 | 47 | 9 |
| Lung proteinosis | 0 | 0 | 3 | 2 |
| Thrombosis | 0 | 0 | 3 | 2 |
| (heart or atrial) |
not expected to result in significant toxicity. The medium exposure concentration of 0.4 mg/m3 was chosen to be near the middle of the low-high range.
Two-year carcinogenicity study
Mortality and survival analyses
Disproportionate mortality in the 1.0 mg/m3 males and females began during study weeks 40 and 53, respectively and was considered related to nickel exposure (Fig. 2). This resulted in the early termination of exposures in this group (Fig. 1). The last day of exposure for the 1.0 mg/m3 males and females was study day 375 (week 53, ~12 months), and study day 427 (week 61, ~14 months), respectively. The first week of exposure for all animals in the study was study week zero. Due to the limited exposure duration (50% of scheduled exposure time) the surviving animals in this group were taken off the main study and assigned as Satellite B animals for the purpose of investigating lung clearance of nickel metal particles (Fig. 1). Clearly, the exposure level of 1.0 mg/m3 (MMAD=1.8) exceeded the MTD in this study for both males and females. The 0.4 mg/m3 exposure groups then became the high- exposure groups for carcinogenicity assessment.
Higher mortality was evident among the 0.4 mg/m3 exposure females after ~18 months of exposure (Fig. 2). Exposure of these animals was stopped but they were retained in the main study to evaluate carcinogenicity. The last day of exposure for the 0.4 mg/m3 females was study day 580 (week 82), completing ~19 months of exposure. The 0.4 mg/m3 males completed the scheduled exposure period of 24 months.
The twenty-four-month survival (i.e., through the end of study week 103) in the control, 0.1 and 0.4 mg/m3 groups was 82%, 82% and 72%, respectively for males, and 76%, 76% and 48%, respectively for females (Table 1). Also evaluated in the carcinogenicity assessment were 4 females from the 0.4 mg/m3 Satellite A exposure group (Fig. 1). Two of these 4 additional animals survived through 24 months, bringing the number of females in the 0.4 mg/m3 group surviving until
| Males Females | ||||||
|---|---|---|---|---|---|---|
| Group 1 0 mg Ni/m3 | Group 2 0.1 mg Ni/m3 | Group 3 0.4 mg Ni/m3 | Group 1 0 mg Ni/m3 | Group 2 0.1 mg Ni/m3 | Group 3 0.4 mg Ni/m3 | |
| 103 weeks | ||||||
| Number of animals | 41 | 41 | 36 | 38 | 38 | 24ª |
| Survival % | 82 | 82 | 72 | 76 | 76 | 48 |
| 130 weeks | ||||||
| Number of animals | 25 | 18 | 23 | 22 | 19 | 7 |
| Survival % | 50 | 36 | 46 | 44 | 38 | 14 |
Number and percentage of animals surviving (n=50/group).
a Four satellite Group 3A females were reassigned to the core study to increase the number of Group 3 females surviving until the end of the exposure period to 26 animals (2/4 animals survived).
the end of the planned exposure period to 26 animals. Lung proteinosis and thrombosis in the heart were causes of unscheduled death that were considered exposure-related (Table 2). Heart thrombosis was typically at the root of the aorta or in the atria and may have been related to the increased erythropoesis observed in the exposed animals.
Clinical observations
Increased respiratory rate and “blue” tinged (cyanosis) extremities, body, and/or facial area were the most common and persistent clinical findings noted in the nickel-exposed animals in the 0.4 and 1.0 mg/m3 groups. These findings establish the lung as the target organ for inhalation toxicity with nickel metal powder. Other exposure-related
clinical observations included cool/pale extremities and/or body, dermal atonia, thinness, decreased defecation, and labored respiration (data not shown). With the exception of cyanosis of the extremities and/or facial area, these findings were also noted in the 0.1 mg/m3 group with incidences that were similar, or slightly higher, than in the control group.
The clinical effects described above continued to be observed during the recovery period. However, increased respiration rate and blue extremities, body and/or facial area were observed in lower proportions of surviving animals as the 6-month non-exposure period progressed.
Body weights
Significant nickel metal exposure-related effects on body weights were noted in the 0.1 (males only), 0.4 and 1.0 mg/m3 groups (Fig. 3).
A
Inhalation Carcinogenicity Study of Metallic Nickel Powder Mean Male Body Weights (+ 2 X SEM)
625
600
575
550
525
500
475
Body Weight (g)
450
425
400
375
350
325
300
275
250
225
200
175
150
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Weeks on Study
Control — 0.1 mg/m3
0.4 mg/m3
1.0 mg/m3
B
Inhalation Carcinogenicity Study of Metallic Nickel Powder Mean Female Body Weights (+ 2 X SEM)
400
375
350
325
300
Body Weight (g)
275
250
225
200
175
150
125
100
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Weeks on Study
Control — 0.1 mg/m3 — 0.4 mg/m3
1.0 mg/m3
At the end of the scheduled exposure period (week 103), mean body weights in the 0.4 mg/m3 group males and females were 27% and 18% lower than controls, respectively. In the 0.1 mg/m3 exposure group, significant reduction in body weight (11%) was noted only for the males.
Food consumption
Exposure-related effects on food consumption were noted for the 0.4 and 1.0 mg/m3 males and females. Mean food consumption in the 0.4 mg/m3 group was lower than the control group from study weeks 58 to 59 through the end of the study for males, and from study weeks 66 to 67 through 86 to 87 for females; the differences were generally statistically significant (p<0.01 or p<0.05). In these groups, effects on food consumption were generally correlated with effects on body weight gain (data not shown).
Hematology
Hematological evaluations were conducted at study week 52, 78 and 130. At the 78-week evaluation (18 months), higher (p<0.01) mean red blood cell counts, hemoglobin levels and hematocrit values were noted in the 0.1 and 0.4 mg/m3 group males and the 0.4 mg/m3 group females (Table 3). Ranges for percent difference from control group were approximately 7-8% in the 0.1 mg/m3 group males and 16- 22% and 21-31% in the 0.4 mg/m3 group males and females, respectively. Differences from control lost statistical significance during a 6-month post-exposure period prior to the terminal necropsy. Thus, partial recovery occurred during the post-exposure period.
Slightly higher total leukocyte counts were seen at one or more intervals (data not shown). These higher counts were interpreted to be a bone marrow compensatory response, probably related to inflam- mation in the lung. Thus, they were considered to be secondary effects and not relevant to the direct toxicity of nickel metal powder. None of the differences was suggestive of a hyperplastic (i.e., leukemia) response.
BALF analyses
After 6 and 12 months of exposure, BALF samples were collected. After six months of exposure, neutrophil numbers, LDH levels and protein content were increased in all exposed groups compared to control animals, but only reached statistical significance in the 0.4 and 1.0 mg/m3 groups (data not shown). Increases of up to 140-fold (males) and 500-fold (females) for neutrophils, 70-fold (males) and 65-fold (females) for LDH and 30-fold (males) and 40-fold (females) for protein were observed. With continued exposure for an additional
| Group 1 0 mg/m3 | Group 2 0.1 mg/m3 | Group 3 0.4 mg/m3 | |
|---|---|---|---|
| Males | n=48 | n=45 | n=47 |
| Red Blood Cells (mil/LLª) | 8.93±0.103 | 9.57±0.115 ** | 10.36±0.170 ** |
| % Different from control | 7.2 | 16 | |
| Hemoglobin (g/dLa) | 16.0±0.17 | 17.2±0.20 ** | 19.3±0.41 ** |
| % Different from control | 7.5 | 21 | |
| Hematocrit (%) | 51.7±0.57 | 56.0±0.74 ** | 62.9±1.43 ** |
| % Different from control | 8.3 | 22 | |
| Females | n=47 | n=47 | n=40 |
| Red blood cells (mil/uL2) | 8.25±0.065 | 8.59±0.086 | 9.95±0.223 ** |
| % Different from control | 4.1 | 21 | |
| Hemoglobin (g/dL2) | 15.8±0.10 | 16.4±0.15 | 19.9±0.55 ** |
| % Different from control | 3.8 | 26 | |
| Hematocrit (%) | 50.0±0.37 | 52.3±0.52 | 65.3±1.94 ** |
| % Different from control | 4.6 | 31 |
Parameter values presented as mean+ standard error.
a mil/uL = millions/microliter; g/dL = grams/deciliter.
** Significantly different from the control group at 0.01 using Dunnett’s test.
40
0-M
35
0-F
0.1-M
30
0.1-F
0.4-M
Blood ng Ni/mL
0.4-F
25
1.0-M
-1.0-F
20
15
10
7
5
0
0
2
4
6
8
10
12
14
Months of Exposure
six months, similarly increased values were observed. The BALF results paralleled the histopathological findings.
Biokinetic evaluation
Nickel levels in blood. Blood nickel levels were measured after 3 and 12 months of exposure. The shape of the dose-response suggests that steady state levels may have been reached by 12 months of exposure (Fig. 4). The blood nickel levels achieved in this carcinogenicity study are several-fold lower than those reached in an oral (gavage) carcinogenicity study with nickel sulfate in Fischer 344 rats (Heim
180
160
0-M
0-F
Micrograms Ni/right lung
140
0.1-M
0.1-F
120
0.4-M
0.4-F
100
1.0-M
1.0-F
80
60
40
20
0
0
4
8
12
16
20
24
Months of Exposure
| Organ | Direction and magnitude of change | Group(s) mg/m3 | Sex | Histologic correlates |
|---|---|---|---|---|
| Lung (absolute; | (250%ª; 275%; | 0.1 | M | Alveolar proteinosis; |
| relative to final body | 261%) | alveolar histiocytosis; | ||
| weight; | (147%b; 146%; | 0.1 | F | Chronic chronic |
| relative to brain weight) | 146%) | inflammation | ||
| (273%ª; 360%; | 0.4 | M | ||
| 291%) (214%b; 227%; 210%) | 0.4 | F | ||
| Adrenal glands | (55%"; 89%; 63%) | 0.4 | M | Pheochromocytomas |
| (absolute; relative to final body weight; relative to brain weight) | of the adrenal medulla |
* Significantly (p<0.05 or p<0.01) different from the control group using Dunnett’s test.
a Absolute lung weights for males were: 1.83±0.033 (0 mg Ni/m3), 6.41 ±0.478 (0.1 mg Ni/m3) and 6.83±0.307 (0.4 mg Ni/m3).
Absolute lung weights for females were: 1.47±0.047 (0 mg Ni/m3), 3.63±0.185 (0.1 mg Ni/m3) and 4.61±0.457 (0.4 mg Ni/m3).
” Absolute adrenal gland weights for males were: 0.085±0.0036 (0 mg Ni/m3), 0.069±0.0035 (0.1 mg Ni/m3) and 0.131±0.0215 (0.4 mg Ni/m3).
et al., 2007). After 24 months of inhalation exposure to the highest tolerated exposure level (0.4 mg/m3, MMAD=1.7 um), the blood nickel levels in present study were estimated at ≤10 ng/ml, based on the 12- month measured values. In comparison, in the oral study, the corresponding blood nickel values measured after 24 months of exposure to the MTD (50 mg Ni/kg/day) fell in the range of 52 to 75 ng/ml.
Nickel levels in lung tissues. Lung nickel levels were measured after 3, 12, and 24 months of exposure for control, 0.1 and 0.4 (males only) mg/m3. For 0.4 mg/m3 females and 1.0 mg/m3 males and females data are available only for 3 and 12 months of exposure (Fig. 5). The data indicate that steady state levels of nickel in the lung were reached by 12 months of exposure for the 0.1 and 0.4 mg/m3 groups. At the MTD exposure level in this study (0.4 mg/m3, MMAD=1.7 um), lung burdens in males and females did not exceed 40 µg Ni/right lung. Both the blood and the lung nickel levels appear higher in females than in males in the present study. Nickel metal lung clearance data obtained
| Selected Lesions | Males | Females | ||||
|---|---|---|---|---|---|---|
| Group 1 0mg/m3 | Group 2 0.1mg/m3 | Group 3 0.4mg/m3 | Group 1 0mg/m3 | Group 2 0.1mg/m3 | Group 3 0.4mg/m3 | |
| Kidney-brown pigmenta | ||||||
| Minimal | 0 | 1 | 0 | 5 | 19 | 15 |
| Mild | 0 | 0 | 2 | 0 | 4 | 11 |
| Moderate | 0 | 0 | 4 | 0 | 1 | 3 |
| Severe | 0 | 0 | 1 | 0 | 0 | 3 |
| Spleen-Extramedullary hematopoesisb | ||||||
| Minimal | 5 | 10 | 9 | 7 | 6 | 4 |
| Mild | 5 | 5 | 11 | 8 | 10 | 10 |
| Moderate | 6 | 1 | 8 | 7 | 10 | 19 |
| Severe | 0 | 0 | 0 | 4 | 2 | 1 |
| Femoral bone marrow-hypercellularityb | ||||||
| Minimal | 2 | 3 | 9 | 5 | 6 | 5 |
| Mild | 1 | 6 | 7 | 8 | 7 | 18 |
| Moderate | 5 | 2 | 9 | 1 | 9 | 12 |
| Severe | 2 | 0 | 1 | 4 | 1 | 1 |
a Brown pigment granular or globular. Incidence based on 50 animals per group.
b Incidence based on 50 animals per group.
post-exposure in the 13-week and in the 30-month studies (for animals in the 1.0 mg/m3) will be discussed in a separate paper (Oller, Kirkpatrick and Oberdörster, manuscript under preparation).
Gross necropsy
Gross necropsy observations of animals euthanized in extremis, found dead, and/or at scheduled necropsy, revealed exposure-related alterations including pale and mottled lungs which did not collapse upon excision in both genders; enlarged and/or swollen spleens in females; and enlarged adrenal glands in males. Increases in absolute and relative weight of lungs (males and females) and adrenal glands (males) among exposed animals were considered to be nickel exposure-related (Table 4). All other organ weight changes appeared to be a result of decreases in final body weight.
Histopathology observations: non-neoplastic lesions
Non-respiratory tract organs. Granular brown pigment in the kidneys, extramedullary hematopoiesis in the spleen, and hypercellu- larity of the sternum and femoral bone marrows were also considered to be related to nickel exposure (Table 5A). The granular brown pigment, typically at the base of renal cortical epithelium, stained blue with Perl’s iron stain which was consistent with iron-containing hemosiderin. In the adrenal glands, increased angiectasis in the adrenal cortex of the 0.4 mg/m3 group females was observed.
| Selected lesions | Males | Females | ||||
|---|---|---|---|---|---|---|
| Group 1 0mg/m3 | Group 2 0.1mg/m3 | Group 3 0.4mg/m3 | Group 1 0mg/m3 | Group 2 0.1mg/m3 | Group 3 0.4mg/m3 | |
| Lungª | ||||||
| Proteinosis | ||||||
| alveolar | ||||||
| Minimal | 0 | 6 | 0 | 8 | 2 | 2 |
| Mild | 0 | 25 | 10 | 0 | 26 | 14 |
| Moderate | 0 | 19 | 15 | 0 | 18 | 16 |
| Severe | 0 | 0 | 25 | 0 | 4 | 22 |
| Histiocytosis | ||||||
| alveolar | ||||||
| Minimal | 26 | 13 | 8 | 20 | 5 | 20 |
| Mild | 2 | 30 | 19 | 6 | 36 | 20 |
| Moderate | 0 | 7 | 15 | 0 | 9 | 10 |
| Severe | 0 | 0 | 2 | 0 | 0 | 0 |
| Chronic | ||||||
| inflammationb | ||||||
| Minimal | 13 | 20 | 8 | 14 | 7 | 16 |
| Mild | 1 | 23 | 11 | 2 | 28 | 6 |
| Moderate | 0 | 1 | 18 | 0 | 10 | 20 |
| Severe | 0 | 0 | 4 | 0 | 0 | 3 |
| Hyperplasia | ||||||
| bronchiolar-alveolar | ||||||
| Minimal | 1 | 1 | 1 | 0 | 1 | 1 |
| Mild | 1 | 3 | 6 | 0 | 8 | 5 |
| Moderate | 1 | 3 | 5 | 0 | 6 | 2 |
| Severe | 0 | 0 | 4 | 1 | 3 | 1 |
| Bronchial lymph node“ | ||||||
| Infiltrate | ||||||
| histiocyte | ||||||
| Minimal | 4 | 8 | 7 | 2 | 11 | 7 |
| Mild | 0 | 12 | 11 | 0 | 13 | 11 |
| Moderate | 0 | 4 | 7 | 0 | 7 | 4 |
| Severe | 0 | 0 | 2 | 0 | 1 | 0 |
a Incidence based on 50 animals per group, except in Group 3 (0.4 mg Ni/m3) that had 54 females.
b Chronic inflammation includes both chronic and chronic-active inflammation.
” Incidence based on the following number of animals per group: 34 and 39 for Group 1, (males and females, respectively), 37 and 42 for Group 2, (males and females, respectively), and 42 and 44 for Group 3, (males and females, respectively).
Respiratory tract. Respiratory tract non-neoplastic histopathological alterations associated with exposure to nickel metal were found in rats dying during the exposure phase, the recovery phase and at the scheduled terminal necropsy. Representative histopathological find- ings are shown in Table 5B. Lung lesions of alveolar proteinosis, alveolar histiocytosis and chronic or chronic-active inflammation were clearly exposure-related, as was bronchiolar-alveolar hyperpla- sia. Compared to normal lung tissue (Fig. 6, top panel), lung, alveolar proteinosis presented as intra-alveolar accumulations of eosinophilic granular to irregular solid deposits (Fig. 6, middle panel). Alveolar histiocytosis consisted of intra-alveolar macrophages with abundant foamy eosinophilic cytoplasm (Fig. 6 bottom panel). Chronic inflam- mation corresponded to mononuclear inflammatory cells and increased connective tissue. Bronchiolar-alveolar hyperplasia of the bronchial epithelium occurred in unusual sites, sometimes radiating
| Males | Females | |||||
|---|---|---|---|---|---|---|
| Group 1 0 mg/m3 | Group 2 0.1 mg/m3 | Group 3 0.4 mg/m3 | Group 1 0 mg/m3 | Group 2 0.1 mg/m3 | Group 3 0.4 mg/m3 | |
| Pheochromocytoma | ||||||
| Benign | 0/50 | 5/50 | 19/50 | 0/50 | 5/49 | 3/53 |
| (0)ª | (10) | (38)b | (0)ª | (10) | (6) | |
| Malignant | 0/50 | 0/50 (0) | 5/50 (10)b | 0/50 (0)" | 0/49 (0) | 0/53 (0) |
| (0)" | ||||||
| Combined benign | 0/50 | 5/50 | 21/50 | 0/50 | 5/49 | 3/53 |
| and malignant | (0) | (10) | (42)d,b | (0) | (10) | (6) |
| Adrenal Cortex | ||||||
| Adenoma | 1/50 | 3/50 | 2/50 | 1/50 | 2/49 | 4/54 |
| (2)e | (6) | (4) | (2)e | (4) | (7) | |
| Carcinoma | 0/50 | 0/50 (0) | 0/50 (0) | 1/50 (2)f | 0/49 (0) | 3/54 |
| (0)f | (6) | |||||
| Combined adenoma | 1/50 | 3/50 | 2/50 | 2/50 | 2/49 | 7/54 |
| and carcinoma | (2) | (6) | (4) | (4) | (4) | (13)b |
a Historical control benign pheochromocytoma. Range of values for 9 control groups: 10-54.3% (males) and 0-10% (females), Bomhard (1992).
b Statistically significant according to Peto method.
” Historical control malignant pheochromocytoma. Range of values for 9 control groups: 0-8% (males) and 0-2% (females), Bomhard (1992).
d Three Group 3 males had both benign and malignant pheochromocytomas, counted only once in the combined incidence.
e Historical control cortical adenoma. Range of values for 9 control groups: 0-38% (males) and 0-31% (females), Bomhard (1992).
f Historical control cortical carcinoma in 24-month studies. Range for control values in 5 studies: 0-1.0% (males) and 0-1.7% (females), Poteracki and Walsh (1998). No data on adrenal cortical adenomas was included in the 30-month studies reported by Bomhard (1992).
from a normal bronchiole, or more often forming along an inflammatory focus. Two exposed females had squamous cysts filled with keratin in their lungs. The epithelium lining these cysts had orderly organization, there were few mitoses and they were not considered neoplastic according to the criterion outlined in Boorman et al. (1996).
Histopathology observations: neoplastic lesions
Adrenal glands. A significant dose-related increase in pheochromo- cytomas (benign, malignant and combined) was observed in the adrenal medulla of male rats. The incidence of pheochromocytomas in the 0.4 mg/m3 exposure group males was statistically significant for total benign neoplasms (incidence=38%), malignant neoplasms (incidence=10%) and combined benign/malignant neoplasms (inci- dence=42%) (Table 6). In the historical control database (30-month studies) compiled by Bomhard (1992), the incidence of spontaneous pheochromocytomas ranged from 10-54.3% (males) and 0-10% (females) for benign tumors and 0-8% (males) and 0-2% (females) for malignant tumors. Therefore the incidences of pheochromocyto- mas in the males control group in the present study (0%) can be considered low.
In females, there was a significant dose-related increase in combined adenoma/carcinoma of the adrenal cortex. In the 0.4 mg/m3 exposure group, there was an statistically significant increase for the combined incidence of adenomas or carcinomas (13%), but not for adenomas or carcinomas when analyzed separately (7% and 6%, respectively) (Table 6). Bomhard (1992) noted a wide variation in the percentage of control animals with spontaneous adrenal cortical adenomas in 30-month studies (0-38% in males and 0-31% in females). The 7.4% incidence of adrenal cortical adenoma in the females in the present study falls within Bomhard’s range. Poteracki and Walsh (1998) report a 0-7% sponta- neous incidence of cortical adenomas and a 0-1.7% incidence of spontaneous cortical carcinomas for females in 24-month studies. The
animals in the current study were held for 30 months and, with the exception of one adenoma in a 0.1 mg/m3 female, all of the adrenal cortical neoplasms were found in animals that died or were euthanized after 24 months.
Respiratory tract. The carcinogenicity assessment included examina- tion of the complete respiratory tract including the nasal tissues, larynx, trachea and lungs (bronchial and alveolar regions) as well as all other tissues required in OECD 451 protocol. No increased incidence was observed for any neoplasm at any level of the respiratory tract.
Histopathology observations: peer review
The pathology findings were reviewed by CRL Pathology Associates following accepted review protocols as determined by the Society of Toxicologic Pathologists (STP, 1997). The results included in this paper are the consensus opinion of both pathologists. Because there were no disagreements in the interpretation of the slides by the primary and secondary pathologists, it was not deemed necessary to have a Peer Review Pathology Working Group.
Discussion
Respiratory toxicity effects
The LOAEL for respiratory effects associated with inhalation exposure to nickel metal powder (MMAD=1.8 um, GSD=2.4 um) was 0.1 mg/m3. Before these results are used in a risk assessment, the equivalent deposited doses in the pulmonary region of humans need to be calculated. The aerosols present in the workplace are usually much coarser and polydisperse than those used in experimental inhalation studies in rats; with the respirable size fraction (particles of <10 um diameter) usually comprising less than 10% of the mass of the aerosol (Werner et al., 1999; Tsai et al., 1995, 1996a,b). Consequently, the differences in particle size distribution need to be considered to estimate human equivalent deposited doses that may be associated with similar respiratory effects. An approach for conducting such an extrapolation will be described in Oller, Kirkpatrick and Oberdörster (manuscript under preparation).
The lung toxicity effects observed in rats after inhalation of nickel metal powder are comparable to those observed in the NTP studies after inhalation of water soluble and insoluble nickel compounds. Fibrosis in the present study was not separately diagnosed because the fibrous connective tissue was invariably accompanied by mono- nuclear inflammatory cells, but fibrotic lesions appeared less severe than in the nickel oxide/subsulfide studies.
Alveolar histiocytosis was increased in incidence and severity in nickel metal-exposed rats compared to controls, but was not increased in severity with higher exposure concentration. Alveolar histiocytosis was called macrophage hyperplasia in the NTP studies with nickel subsulfide (NTP, 1996a) and nickel sulfate hexahydrate (NTP, 1996c) and it was included with the diagnosis of chronic inflammation in the NTP study with nickel oxide (NTP, 1996b). Since alveolar macrophages play a primary role in clearance of inhaled particles from the alveolar region of the lung, an increased number of histiocytic alveolar macrophages is the expected adaptive response to repeated inhalation of particulates.
Erythropoesis
Increases in erythropoesis were observed in rats exposed to 0.1 and 0.4 mg/m3 of nickel metal for 12 or 18 months in the present study. These increases were reflected in statistically significant elevation of red blood cell levels and hemoglobin (up to 20% males, 30% females), as well as hematocrit values (up to 25% males, 40% females). There was also accumulation of iron-positive pigment in the kidneys and evidence of extramedullary hematopoiesis in the spleen of exposed
animals. These effects are considered primarily due to the hypoxia that ensues when the lung toxicity (resulting from inhalation of nickel metal powder) impairs proper gas exchange. Increased erythropoietin production and secondary increases in red blood cells are known to occur with chronic pulmonary disease (Boorman and Eustis, 1990).
Direct nickel effects on gene expression of erythropoietin (e.g., mediated by HIF-inducible factor) have been reported in some model systems (Bunn et al., 1998; Salnikow et al., 2000; Maxwell and Salnikow, 2004). Increases in red blood cells, hematocrit and/or hemoglobin were also noted in some of the animals exposed to 30 and 50 mg/kg/day nickel sulfate hexahydrate in a recent oral carcinogeni- city study in which no lung toxicity was observed (Heim et al., 2007). However, in the oral study, the blood changes were small and did not follow a consistent exposure-related pattern. For example, there was a 10% elevation in red blood cells in the mid-exposure group males after two years, but not in hemoglobin or hematocrit, and there was no significant elevation in red blood cells in the high-exposure group. Therefore, the data from both the inhalation and oral studies suggest that although nickel may exert some direct effects on erythropoietin through the HIF pathway, most of the observed changes in the present inhalation study are secondary to the toxicity effects in the lung.
Carcinogenicity results: present study
The purpose of this study was to determine whether nickel metal powder was carcinogenic to Wistar rats when administered by inhalation 6 h/day, 5 days/week over a two-year period. This treatment did not produce an exposure-related increase in tumors anywhere in the respiratory tract, including the nose.
Significant dose-related increases in incidence of adrenal gland pheochromocytomas in males and combined cortical tumors in females were observed. Pheochromocytomas in males and combined cortical adenoma/carcinoma in females were statistically increased in the 0.4 mg/m3 but not in the 0.1 mg/m3 exposure groups compared to controls. The incidence of spontaneous pheochromocytomas in rats is fairly high, particularly among males (Haseman et al., 1998). Multiple factors can contribute to the induction of pheochromocytomas, but the prolonged stimulation of catecholamine release by the endocrine cells may be a common step. An association of pheochromocytomas in male rats and lung pathology in inhalation studies has been reported (Ozaki et al., 2002). A review of the data from several inhalation studies with metal compounds (including nickel oxide and nickel subsulfide) as well as talc, led Ozaki et al. to conclude that lung pathology can reduce gas exchange areas and chronically stimulate release of catecholamines from the adrenal medulla. This chronic stimulation can lead to hyperplasia and neoplasia. Fibrosis with inflammation, as well as hypoxemia were found to be the two events linked to the induction of pheochromocytomas.
In a recent carcinogenicity study in which Fischer 344 rats were exposed to nickel sulfate hexahydrate by gavage (i.e., no respiratory toxicity was observed), no increase incidence of neoplasms of the adrenal glands was detected (Heim et al., 2007). In the oral study, the blood nickel levels were several-fold higher than in the present study. Therefore, these results suggest that the pheochromocytomas in the 0.4 mg/m3 (MTD), while related to the nickel metal exposure in a dose-dependent manner, occurred as a consequence of the lung toxicity associated with the exposure and were not due to the presence of nickel ions in the adrenal glands.
The statistically significant increase in adrenal cortical tumors observed in this study in the 0.4 mg/m3 exposure female group (above the MTD) falls within historical control values. The relationship between these tumors and nickel metal exposure is therefore unclear. Since an increased incidence of adrenal cortical tumors was not observed in the oral study with nickel sulfate (nor in any of the inhalation studies with nickel compounds), it is considered likely that the cortical tumor incidence in this metallic nickel study resulted from
variations in background incidence levels rather than a direct effect of nickel ion in the adrenal glands.
An important factor in conclusively determining the lack of respiratory carcinogenic potential of nickel metal powder is to demonstrate that in this study: (1) there was an exposure group in which the MTD was achieved, and (2) mortality due to nickel exposure was low enough to preserve a sufficient number of animals for tumor analyses. The increased mortality seen in this study at the 1.0 mg/m3 and 0.4 mg/m3 (females only) exposure levels indicates that the MTD was reached and even exceeded. Despite the observed lethality, the number of males and females in the 0.4 mg/m3 exposure group surviving at the end of the scheduled exposure period (24 months) and available for tumor examination was 36 and 26, respectively. These numbers meet the OECD criterion for sufficient power (50% survival of the OECD recommended 50 animals per group) to detect a positive effect if one exists.
Carcinogenicity results: integrated analysis
The NTP studies showed that two water insoluble nickel compounds, crystalline nickel subsulfide and high calcining tempera- ture green nickel oxide were capable of inducing respiratory tumors in rats after inhalation. In contrast, water soluble nickel compounds, and now water insoluble nickel metal powders, did not. How can these different results be reconciled? The exact direct or indirect effects of Ni (II) ions needed for the generation of respiratory tumors are still the subject of much research (Costa et al., 2005; Salnikow and Kasprzak, 2005). Yet, it is clear that water solubility alone will not determine the carcinogenic potential of Ni-containing substances. Rather, it may be the bioavailability of Ni (II) ion at nuclear sites of target epithelial cells that determines carcinogenicity (Oller et al., 1997; Costa et al., 2003). This bioavailability will result from the interaction of different factors such as:
1. Respiratory toxicity. The intrinsic respiratory toxicity after inhala- tion of each Ni-containing substance will limit the exposure level corresponding to the MTD and therefore the lung deposited dose.
2. Deposition. For a given exposure level, the MMAD and GSD of the aerosol will affect the overall dose delivered to specific target areas of the respiratory tract.
3. Clearance. The clearance of particles will depend on their location, their size and their physico-chemical properties. Clearance will determine the retained dose.
4. Target cell uptake. Particle size as well as the intrinsic properties of Ni-containing particles will determine the ability of particles to enter epithelial cells in the respiratory tract.
5. Intracellular dissolution. Once inside the target cells, the physico- chemical properties of Ni-containing particles will determine the extent to which Ni (II) ion is released and becomes available at nuclear sites. In the case of nickel metal particles, an oxidation reaction has to occur (rather than simple dissolution) for Ni (II) ion to be released.
One way to integrate all these factors is to consider that the bioavailability of Ni (II) ions at nuclear sites of target cells and subsequent tumorigenicity will be a function of the MTD; i.e., the maximum tolerated dose (exposure) leading to (i) the maximum deposited dose (based on the particle size distribution of the aerosol) and (ii) the maximum retained dose (based on the clearance rate); as well as a function of the intracellular uptake and dissolution of the Ni-containing particles.
Information on MTD and particle size distribution (PSD) is available from the chronic NTP (1996a,b,c) and present studies with various nickel substances. Clearance (retention half-time) data are provided by Benson et al. (1994, 1995) and Oller, Kirkpatrick and Oberdörster (manuscript under preparation). In vivo information on cellular uptake and intracellular dissolution is not available for Ni-containing substances, but in vitro data on relative cellular uptake, intracellular dissolution
(Costa and Mollenhauer, 1980; Abbracchio et al., 1982; Ke et al., 2007), and cell transformation - as a surrogate for both uptake and intracellular dissolution - are available for various Ni-containing substances (e.g., Lin and Costa, 1994; Miura et al., 1989; S. Seilkop, personal communication). Fig. 7 uses these data to describe the hypothetical effects of the above listed factors on the predicted Ni (II) ion bioavailability and tumor- igenicity for four very different Ni-containing substances.
Water soluble nickel compounds cause relatively high lung toxicity. The low MTD, together with rapid lung clearance and very poor intracellular uptake predict low Ni (II) ion bioavailability at nuclear sites in vivo. This is consistent with a lack of respiratory tumorigenicity in animal studies (NTP, 1996c).5 Similarly, the max- imum levels of nickel metal powder (MMAD=1.7 um) that can be tolerated are limited by respiratory toxicity. The resulting relatively low retained dose, combined with very poor intracellular uptake and low intracellular dissolution (i.e., the particles need to be oxidized as reflected in its relatively low in vitro cell transformation potency) results in a low predicted nuclear bioavailability in vivo. This is also consistent with the lack of respiratory tumors in the present study.
In contrast, the highest tumor induction with nickel subsulfide observed in rats (NTP, 1996a) correlates with high predicted nuclear bioavailability based on the combination of slightly higher MTD, much higher cellular uptake, and higher intracellular dissolution (i.e., relatively high cell transformation potency in vitro). In the case of green nickel oxide, the MTD in the NTP (1996b) study is much higher due to its lower toxicity. At the MTD, particle clearance is impaired resulting in a much higher retained dose as evident by high lung burden. This is balanced by a lower cellular uptake and intracellular dissolution (i.e., relatively lower cell transformation potency) compared to nickel subsulfide. The predicted nuclear bioavailability of Ni (II) ions from green nickel oxide appears to be sufficient for direct tumor induction in rats but only under conditions of impaired particle clearance (NTP, 1996b; Oller et al., 1997).
The bioavailability model described above provides one way to explain the different animal carcinogenicity results obtained for various nickel-containing substances and emphasizes the need to take speciation into account. When assessing the carcinogenicity of nickel substances, bioavailability and animal data should be considered together with the results of epidemiological human studies, in a weight of evidence approach.
Carcinogenicity assessment based on animal and human data: regulatory implications
A recent review of post-1990 epidemiological studies in which larger cohorts of refinery and non-refinery workers were included (mean “inhalable fraction” exposures to metallic nickel ranging from 0.01 to 6 mg Ni/m3) (Sivulka, 2005), confirms the lack of an association between metallic nickel exposures and increased respiratory cancer risks first described in the 1990 ICNCM study. The most relevant cohorts to study nickel metal exposures include workers involved in stainless steel and alloy production (due to the large number of workers employed) and workers in the production of barrier material for use in uranium enrichment (due to their exposure solely to relatively high levels of a high purity metallic nickel powder). None of
5 In the case of nickel sulfate hexahydrate, the relevance of the lack of respiratory tumors in animals inhalation studies has been previously questioned given that epidemiological studies of sulfidic ore refinery workers have shown increased respiratory cancer risk associated with complex exposures containing soluble nickel compounds (ICNCM, 1990; Andersen et al., 1996; Grimsrud et al., 2002). Studies of nickel platers, who lack the confounding exposures found in refineries, have been negative, but the cohorts have been too small for the studies to be conclusive (Pang et al., 1996). For nickel metal, animal and human studies are in agreement, perhaps because data from cohorts other than sulfidic ore nickel refining were included in the epidemiological analyses.
2.5
Exposure (MTD)
2
MTD
1.5
1
Deposited Dose f (MTD, PSD)
0.5
0
SUL
MET
SUB
OXI
140
120
Retained Dose f (Deposited Dose, Clearance)
Retention T1/2
100
80
60
40
20
0
SUL
MET
SUB
OXI
Bioavailability at Nuclear Sites f (Retained Dose, Cellular Uptake, Intracellular Dissolution
Cellular Uptake-Dissolution
3000
2500
2000
1500
1000
500
0
SUL
MET
SUB
OXI
Respiratory Tumors f (Bioavailability at Nuclear Sites of Target Cells)
Bioavailability Nuclear Sites
7000
6000
5000
?
4000
3000
2000
1000
0
SUL
MET
SUB
OXI
these studies provides evidence that metallic nickel increases the risk of respiratory cancer among exposed workers.
The combined animal (present study) and human data strongly indicate that nickel metal powder is not expected to be a human carcinogen via inhalation. In addition, because the absorption of nickel (II) ions from oral exposure to nickel metal has been shown to be lower than that of nickel sulfate hexahydrate (Ishimatsu et al., 1995; Hayman et al., 1984), the negative oral carcinogenicity results for the latter compound would indicate that nickel metal is not expected to be a human oral carcinogen. The results of the present study are expected to influence the carcinogenicity classification for nickel metal in Germany and also in Europe, under the new chemicals management system (REACH). In the United States, the EPA has not previously considered the evidence for carcinogenicity of nickel metal sufficient to assign it any carcinogenic classification (EPA, 1986). The present study also supports a similar position adopted in 1998 by the ACGIH (ACGIH,
1998) and in 2005 by the Agency for Toxic Substances and Disease Registry (ATSDR, 2005) ([n]o evidence was found that metallic nickel causes respiratory cancer).
In summary, we suggest that the results of this chronic inhalation carcinogenicity study in Wistar rats are definitive and do not show an association between nickel metal exposure and respiratory tumors. This study fills in an important existing data gap and together with the results from the oral nickel sulfate study and the epidemiological findings indicates that nickel metal powder is not likely to be a human carcinogen by relevant routes of exposure.
Funding
This study was funded by the Nickel Producers Environmental Research Association (NiPERA), Eurofer, the International Stainless Steel Federation (ISSF), VDEh, and WV Metalle.
Conflict of interest statement
This study was sponsored by industry funded metal associations and was monitored by NiPERA staff. In addition, the design, conduct and data interpretation of the carcinogenicity study were overseen by an Expert Group consisting of regulatory, academic and industry toxicologists. This study was conducted in an independent manner by WIL Research Laboratories. The pathology results from this study underwent a thorough peer review process prior to finalization of the report. The complete final report from this study is available from NiPERA upon request.
Acknowledgments
The authors thank the technicians, pathologists, and statisticians at WIL Research Laboratories and associated laboratories for their professional contributions to the conduct of this study. Critical reviews of this article by G. Oberdorster and P. De Marco are greatly appreciated.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the online version, at doi: 10.1016/j.taap.2008.08.017.
References
Abbracchio,, M.P., Simmons-Hansen,, J., Costa,, M., 1982. Cytoplasmic dissolution of phagocyized crystalline nickel sulfide particles: a prerequisite for nuclear uptake of nickel. J. Toxicol. Environ. Health 9, 663-676.
ACGIH, 1998. American Conference of Governmental Industrial Hygienists: threshold limit values and biological exposure indices for chemical substances and physical agents. Cincinnati, OH, p.51.
Andersen, A., Berge, S.R., Engeland, A., Norseth, T., 1996. Exposure to nickel compounds and smoking in relation to incidence of lung and nasal cancer among nickel refinery workers. Occup. Environ. Med. 53, 708-713.
ATSDR (Agency for Toxic Substances and Disease Registry), 2005. Toxicological Profile for Nickel. U.S. Department of Health and Human Services, Atlanta, Georgia. PB2006-100005.
Bomhard, E., 1992. Frequency of spontaneous tumors in Wistar rats in 30-months studies. Exp. Toxicol. Pathol. 44, 381-392.
Boorman, G.A., Eustis, S.L., 1990. The lung. Pathology of the Fischer Rat. Academic Press, San Diego, p. 348.
Boorman, G.A., Brockmann, M., Carlton, W.W., Davis, J.M., Dungworth, D.L., Hahn, F.F., Mohr, U., Reichhelm, H.B., Turusov, V.S., Wagner, B.M., 1996. Classification of cystic keratinizing squamous lesions of the rat lung: report of a workshop. Toxicol. Pathol. 24 (5), 564-572 Sep-Oct.
Benson, J.M., Barr, E.B., Bechtold, W.E., Cheng, Y.S., Dunnick, J.K., Eastin, W.E., Hobbs, C.H., Kennedy, C.H., Maples, K.R., 1994. Fate of inhaled nickel oxide and nickel subsulfide in F344/N rats. Inhal. Toxicol. 6, 167-183.
Benson, J.M., Chang, I .- Y., Cheng, Y.S., Hahn, F.F., Kennedy, C.H., Barr, E.B., Maples, K.R., Snipes, M.B., 1995. Particle clearance and histopathology in lungs of F344/N rats and B6C3F1 mice inhaling nickel oxide or nickel sulfate. Fundam. Appl. Toxicol. 28, 232-244.
Bunn, H.F., Gu, J., Huang, E., Park, J .- W., Zhu, H., 1998. Erythropoietin: a model system for studying oxygen-dependent gene regulation. J. Exp. Biol. 201, 1197-1201.
Costa, M., Mollenhauer, H.H., 1980. Phagocytosis of nickel subsulfide particles during the early stages of neoplastic transformation in tissue culture. Cancer Res. 40, 2688-2694.
Costa, M., Yan, Y., Zhao, D., Salnikow, K., 2003. Molecular mechanisms of nickel carcinogenesis: gene silencing by nickel delivery to the nucleus and gene activation/inactivation by nickel-induced cell signaling. J Environ Monit. 5 (2), 222-223 2003 Apr, Review.
Costa, M., Davidson, T.L., Chen, H., Ke, Q., Zhang, P., Yan, Y., Huang, C., Kluz, T., 2005. Nickel carcinogenesis: epigenetics and hypoxia signaling. Mutat Res. 30 (592(1-2)), 79-88.
Driscoll, K.E., Costa, D.L., Hatch, G., Henderson, R., Oberdörster, G., Salem, H., Schlesinger, R.B., 2000. Intratracheal instillation as an exposure technique for the evaluation of respiratory tract toxicity: uses and limitations. Toxicol. Sciences 55, 24-35.
Dunnett, C.W., 1964. New tables for multiple comparisons with a control. Biometrics 20, 482-491.
European Union Existing Substances Risk Assessment Report for Nickel Metal (May 2006, Chapter 4: Human Health). Danish Environmental Protection Agency. http://ecb.jrc.it/documents/Existing-Chemicals/RISK_ASSESSMENT/EURATS/ RAR_STATUS/eurats_MS_311.pdf.
EPA 1986. Health assessment document for nickel and nickel compounds (Final Report, September 1986. Pages 8-113.
Gehan, E.A., 1965. A generalized Wilcoxon test for comparing arbitrarily singly-censored samples. Biometrika, 52, 203-223.
Grimsrud, T.K., Berge, S.R., Haldorsen, T., Andersen, A., 2002. Exposure to different forms of nickel and risk of lung cancer. Am. J. Epidemiol. 156, 1123-1132.
Haseman, J.K., 1983. A reexamination of false-positive rates for carcinogenesis studies. Fundam. Appl. Toxicol. 3, 334-339.
Haseman, J.K., Hailey, J.R., Morris, R.W., 1998. Spontaneous neoplasm incidences in Fischer 344 rats and B6C3F1 mice in two-year carcinogenicity studies: a National Toxicology Program update. Toxicol. Pathol. 26, 428-441.
Hayman, P.B., Goodgame, D.M.L., Snook, R.D., 1984. Studies of nickel absorption in rats using inductively coupled plasma atomic-emission spectrometry and liquid scintillation counting. Analyst 109, 1593-1595.
Heim, K.E., Bates, H.K., Rush, R.E., Oller, A.R., 2007. Oral carcinogenicity study with nickel sulfate hexahydrate in Fischer 344 rats. Fundam. Appl. Toxicol. 224, 126-137.
Hueper, W.C., 1958. Experimental studies in metal carcinogenesis: IX. Pulmonary lesions in guinea pigs and rats exposed to prolonged inhalation of powdered metallic nickel. AMA Arch. Pathol. 65, 600-607.
Hueper, W.C., Payne, W.W., 1962. Experimental studies in metal carcinogenesis. Arch. Environ. Health 5, 445-462.
IARC, 1990. International agency for research on cancer and World Health Organization monographs on the evaluation of carcinogenic risks to humans: chromium, nickel and welding. International Agency for Research on Cancer, 49. Lyon, France, p. 90. 677.
ICNCM, 1990. International committee on nickel carcinogenesis in man. Report of the International Committee on Nickel Carcinogenesis in Man, pp. 1-82.
Ishimatsu, S., Kawamoto, T., Matsuno, K., Kodama, Y., 1995. Distribution of various nickel compounds in rat organs after oral administration. Biol. Trace Elem. Res. 49 (1), 43-52.
Ivankovic, S., Zeller, W.J., Komitowski, D., Edler, L., Lehman, E., Frohlich, N., 1988. Carcinogenesis of nickel alloys in the hamster following intratracheal instillation. Schriftenreihe der Bundesanstalt für Arbeitsschutz, Dortmund, pp. 1-58.
Kaplan, E.L., Meier, P., 1958. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53, 457-481.
Ke, Q., Davidson, T., Kluz, T., Oller, A., Costa, M., 2007. Fluorescent tracking of nickel ions in human cultured cells. Toxicol. Appl. Pharmacol. 219 (1), 18-23 Electronic publication 2006 Oct 28.
Kirkpatrick, D.T., 2004. A 13-week inhalation toxicity study (with recovery) of nickel metal in albino rats (Study No. WIL-437002). WIL Research Laboratories, Inc., Ashland, Ohio, 2004.
Kirkpatrick, D.T., 2008. A 24-months inhalation carcinogenicity study of nickel metal (with a six-months senescence period) in Wistar rats (Study No. WIL-437003). WIL Research Laboratories, Inc., Ashland, Ohio, April, 2008.
Lin, K.K., 1995. A regulatory perspective on statistical methods for analyzing new drug carcinogenicity study data. Bio/Pharam. Q. 1 (2), 18-20.
Lin, K.K., 1997. Control of Overall False Positive Rates in Animal Carcinogenicity Studies of Pharmaceuticals, presented at the FDA Forum on Regulatory Science, Bethesda, MD, December 8-9, 1997.
Lin, K.K., Rahman, M.A., 1988a. Overall false positive rates in tests for linear trend in tumor incidence in animal carcinogenicity studies of new drugs. J. Pharm. Stat. 8 (1), 1-22.
Lin, K.K., Rahman, M.A., 1988b. False positive rates in tests for trend and differences in tumor incidence in animal carcinogenicity studies of pharmaceuticals under ICH guidance SIB, unpublished report. Division of Biometrics 2, Center for Drug Evaluation and Research, Food and Drug Administration.
Lin, X., Costa, M., 1994. Transformation of human osteoblasts to anchorage-independent growth by insoluble nickel particles. Environ. Health Perspect. 102 (Suppl 3), 289-292 Sep.
Maxwell, P., Salnikow, K., 2004. HIF-1: an oxygen and metal responsive transcription factor. Cancer Biol. Ther. 3 (1), 29-35 2004 Jan, Electronic publication 2004 Jan 10. Review.
McConnell, E.E., Solleveld, H.A., Swenberg, J.A., Boorman, G.A., 1986. Guidelines for combining neoplasms for evaluation of rodent carcinogenesis studies. J. Natl. Cancer Inst. 76, 283-289.
Miura, T., Patierno, S.R., Sakuramoto, T., Landolph, J.R., 1989. Morphological and neoplastic transformation of C3H/10T1/2 Cl 8 mouse embryo cells by insoluble carcinogenic nickel compounds. Environ. Mol. Mutagen. 14 (2), 65-78.
Muhle, H., Bellmann, B., Takenaka, S., Fuhst, R., Mohr, U., Pott, F., 1992. Chronic effects of intratracheally instilled nickel-containing particles in hamsters. In: Nieboer, E., Nriagu, J.O.E. (Eds.), Nickel and Human Health: Current Perspective. John Wiley & Sons, Inc., pp. 467-479.
NTP (National Toxicology Program), US Department of Health and Human Services, 1996a. National Toxicology Program Technical Report. Toxicological and carcino- genesis studies of nickel subsulfide in F344/N rats and B6C3F1 mice. NTP TR 453, NIH Publication Series No. 96-3369.
NTP (National Toxicology Program), US Department of Health and Human Services, 1996b. National Toxicology Program Technical Report. Toxicological and carcino- genesis studies of nickel oxide in F344/N rats and B6C3F1 mice. NTP TR 451, NIH Publication Series No. 96-3363.
NTP (National Toxicology Program), US Department of Health and Human Services, 1996c. National Toxicology Program Technical Report. Toxicological and carcino- genesis studies of nickel sulfate hexahydrate in F344/N rats and B6C3F1 mice. NTP TR 454, NIH Publication Series No. 96-3370.
Oller, A.R., Costa, M., Oberdörster, G., 1997. Carcinogenicity assessment of selected nickel compounds. Toxicol. Appl. Pharmacol. 143, 152-166.
Ozaki, K., Haseman, J.K., Hailey, J.R., Maronpot, R.R., Nyska, A., 2002. Association of adrenal pheochromocytoma and lung pathology in inhalation studies with
particulate compounds in the male F344 rat - the National Toxicology Program Experience. Toxicol. Pathol. 30, 263-270.
Pang, D., Burges, D.C., Sorahan, T., 1996. Mortality study of nickel platers with special reference to cancers of the stomach and lung, 1945-93. Occup. Environ. Med. 53, 714-717.
Peto, R., Pike, M.C., Day, N.E., Gray, R.G., Lee, P.N., Parish, S., Peto, J., Richards, S., Wahrendorf, J., 1980. Guidelines for simple, sensitive significance tests for carcinogenic effects in long-term animal experiments. Long-Term and Short-Term Screening Assays for Carcinogenesis: a Critical Appraisal. IARC Monographs, Supplement 2. Lyon, France, pp. 311-426.
Poteracki, J., Walsh, K.W., 1998. Spontaneous neoplasms in control Wistar rats: a comparison of reviews. Toxicol. Sci. 45, 1-8.
Pott, F., Ziem, U., Reiffer, F.J., Huth, F., Ernst, H., Mohr, U., 1987. Carcinogenicity studies on fibers, metal compounds, and some other dusts in rats. Exp. Pathol. 32, 129-152. Salnikow, K., Kasprzak, K.S., 2005. Ascorbate depletion: a critical step in nickel carcinogenesis? Environ Health Perspect 113 (5), 577-584.
Salnikow, K., Su, W., Blagosklonny, M.V., Costa, M., 2000. Carcinogenic metals induce hypoxia-inducible factor-stimulated transcription by reactive oxygen species- independent mechanism. Cancer Res. 60 (13), 3375-3378.
Sivulka, D.J., 2005. Assessment of respiratory carcinogenicity associated with exposure to metallic nickel: a review. Regul. Toxicol. Pharmacol. 43, 117-133.
Snedecor, G.W., Cochran, W.G., 1980. One way classifications; analysis of variance, Statistical Methods, 7th ed. The Iowa State University Press, Ames, IA, pp. 215-237. STP, 1997. Commentary. Documentation of the pathology peer review position of the Society of Toxicologic Pathologists. Toxicol. Pathol. 25 (6), 655.
Tsai, P.J., Vincent, J.H., Wahl, G., Maldonado, G., 1995. Occupational exposure to inhalable aerosol in the primary nickel production industry. Occup. Environ. Med. 52, 793-799.
Tsai, P.J., Vincent, J.H., Wahl, G.A., Maldonado, G., 1996a. Worker exposures to inhalable and ‘total’ aerosol during nickel alloy production. Ann. Occup. Hyg. 40 (6), 651-659.
Tsai, P.J., Werner, M.A., Vincent, J.H., Maldonado, G., 1996b. Worker exposure to nickel- containing aerosol in two electroplating shops: comparison between inhalable and total aerosol. Appl. Occup. Environ. Hyg. 11 (5), 484-492.
United States FDA, 1999. Guidance for industry: providing regulatory submissions in electronic Format - NDA’s. U.S. Department of Health and Human Services, Center for Drug Evaluation and Research (CDER). IT3. 74 pages.
Werner, M.A., Thomassen, Y., Hetland, S., Norseth, T., Berge, S.R., Vincent, J.H., 1999. Correlation of urinary nickel excretion with observed ‘total’ and inhalable aerosol exposures of nickel refinery workers. J. Environ. Monit. 1, 557-562.
Zar, J.H., 1999. Multiple comparisons, Biostatistical Analysis, 4th ed. Prentice Hall, Upper Saddle River, NJ, pp. 208-214.