PRELIMINARY HEALTH REVIEWS IN RANCHO CORDOVA, SACRAMENTO COUNTY, CALIFORNIA
RANCHO CORDOVA, SACRAMENTO COUNTY, CALIFORNIA
Preliminary reviews of the scientific literature were conducted by CDHS personnel located in the Environmental Health Investigations Branch looking for known health effects from exposure to perchlorate. Searches of the pertinent human and animal data bases found that the primary organ which is affected by perchlorate is the thyroid gland. However, there is a very small amount of information available related to perchlorate ingestion in healthy, human subjects; most of the information was found in studies of patients with Graves' Disease (hyperthyroidism), where perchlorate was used as an antagonist to decrease circulating thyroid hormone levels. Aside from producing hypothyroidism, perchlorate ingestion may also have produced goiter and hematologic abnormalities.
Several statewide databases were surveyed for these possible perchlorate-related health outcomes during the suspected years of contamination and limited to the likely areas of exposure (by zip code). Corresponding specific exposure information was not available at this time, so exposure was broadly defined as residing in a suspected contaminated zip code during the period from 1985 to 1996. Additionally, we have included in this health consultation, data analyses performed at the request of concerned citizens from Rancho Cordova regarding cancer incidence in suspected exposed areas.
Neonatal Thyroid Hormone Levels
Little medical information is available on perchlorate's developmental toxicity; however, it is believed that perchlorate acts in a fashion similar to other anti-thyroid medications (propyl thiouracil and methimazole) by crossing the placenta and inhibiting fetal thyroid synthesis, thus producing hypothyroidism in the newborn (9). Neonatal hypothyroidism can have tragic health consequences including mental retardation and cretinism, but can be treated, if problems are ascertained early. In California, thyroid hormone levels (T4) are drawn and maintained on file with the Genetic Disease Branch of the CDHS for all newborn infants. In a preliminary survey, thyroid hormone levels were obtained for all children born to mothers believed to have been exposed to perchlorate by residing in the following zip codes: 95670, 95742, 95655 and 95827. Additionally, data were abstracted for those children born to mothers residing in neighboring areas not believed to have been exposed to perchlorate from zip codes 95628, 95608, 95864 and 95662, as well as data from the remainder of California. The time period for this abstracted information was from 1985 through 1996.
There were 11,814 thyroid hormone screens in the potentially-exposed area with 4 cases of hypothyroidism observed. 3.76 cases would have been expected based on the statewide rate of 3.18 cases per 10,000 live births during this period. The non-exposed area found 6 cases of hypothyroidism out of 20,135 routine blood screens (6.41 cases were expected). Table 1 shows these figures for exposed and unexposed zip codes as well as for the remainder of the state. These data do not show strong evidence of an association between residence in the potentially-exposed zip codes and neonatal hypothyroidism. Additionally, using residence as a proxy for exposure places limitations on the validity of any findings of this preliminary survey as there is no well-documented exposure information for the birth mothers. In addition, there are numerous other causes for neonatal hypothyroidism which instead may be responsible, including: deficiencies of Thyrotropin Releasing Factor and Thyroid Stimulating Hormone, aplasia or hypoplasia of the thyroid gland and iodine deficiency (9).
In addition to the above preliminary analysis, values of thyroid hormone (T4) and thyroid stimulating hormone (TSH) were obtained for each group, looking for differences in mean and median values between the exposure groups(Tables 2 and 3, respectively). It should be noted that TSH levels are not routinely obtained in the initial screen of all live births, but are ascertained secondarily from neonates with low initial T4 levels (below 10 Iu/dl) who are then re-screened.
The mean neonatal thyroid hormone value in the potentially-exposed zip codes was statistically significantly lower than the non-exposed population (p=0.0001), yet statistically significantly higher than the population from the rest of the state (p=0.0001). If neonatal thyroid hormone levels were affected by maternal perchlorate ingestion, then one would expect the group levels in the exposed population to be lower than both populations. In addition, since exposures were not more accurately determined through dose-reconstruction analyses, it is not possible to more accurately assess whether perchlorate was responsible for the levels of T4 observed in the exposed group. For TSH values, the mean level observed in the suspected exposed group was not statistically significantly different from either the unexposed group (p=0.49) or the values found throughout the rest of the state (p=0.99).
Finally, it was also noted that from 1985 through 1996 there were only four cases of clinically significant neonatal hypothyroidism in the suspected exposure zip code area (Table 1). One case was born in 1985 and the birth mother resided in zip code 95827. This case's mother may not have regularly ingested contaminated water as the zip code location is the furthest west from the Aerojet facility and the time period may be before the drinking water wells in the Arden Cordova service area were affected by perchlorate contamination. The other three cases occurred from 1990 through 1996 and were all located in zip code 95670, which is the approximate area where perchlorate contamination of the drinking water wells is located. However, further information was unavailable to better define exposure status, without interviewing the family or reviewing the child's medical record.
Perchlorate competitively binds to receptors on the surface of the thyroid and limits the uptake of iodide by the thyroid gland, with the subsequent inhibition of thyroid hormone release. Decreased circulating thyroid hormone exerts a feedback effect on the pituitary gland producing a compensatory increase in TSH. TSH stimulates the thyroid gland and may produce hypertrophy and thyroid gland enlargement (10). The OSHPD hospital discharge data base was reviewed, searching for the diagnosis of goiter among the first five reported diagnoses for each hospitalized individual residing in zip code 95670 from 1991 through 1995(1). [The information obtained in this and the remaining surveys searched only in the 95670 zip code because this area was viewed as the most likely area of possible perchlorate contamination of the drinking water supply.] There were approximately 3-5 cases per year of goiter listed in the top 5 diagnoses. There are however, several problems with this approach. First, there are many diseases or conditions which can produce a goiter other than perchlorate ingestion and the data base cannot differentiate this aspect well. Also, not all patients with goiter are admitted and treated in a hospital setting and so the OSHPD count is probably an undercount of the actual number of people with goiter. Because of these reasons, it was concluded that OSHPD data would not be helpful in determining the prevalence of thyroid enlargement in the affected water district.
Aplastic anemia is a condition affecting the development of early blood cell precursors in the bone marrow which results in pancytopenia (reduced red and white blood cells and platelets). Agranulocytosis is the reduction of one particular blood cell line (white blood cells). Both aplastic anemia and agranulocytosis have occurred in individuals with Graves' Disease treated with perchlorate, documented in a series of papers in the early 1960s, which ultimately led to the discontinuation of perchlorate for the treatment of this condition (10). However, it is thought that the aplastic anemia seen in the Graves' Disease patients may have been a hypersensitivity reaction and unrelated to the dose of perchlorate ingested. Also, there is no information which would suggest that individuals without Graves' Disease would react in a similar fashion to perchlorate. Finally, aplastic anemia and agranulocytosis both may be caused by exposure to a variety of drugs or conditions including: cytotoxic medicines used in cancer chemotherapy, anticonvulsants, antibiotics, benzene, radiation, viral infections and genetic syndromes (11).
The OSHPD data was reviewed from 1991-95 for individuals residing in zip code 95670, searching for agranulocytosis or aplastic anemia as one of the top five diagnoses. This zip code was chosen because it was presumed to be the most likely area possibly receiving perchlorate-contaminated drinking water. Table 4 displays OSHPD data for hospitalizations for agranulocytosis for individuals residing in zip code 95670 during this time. There were 76 total cases diagnosed with agranulocytosis who resided in this zip code from 1991 through 1995 for a rate of 35.8 cases/100,000 individuals per year. This figure is less than the statewide rate of 41.6 hospitalizations per 100,000 individuals per year generated from OSHPD data during 1994 and 1995. Because the data does not suggest any increase in risk of agranulocytosis for individuals residing in the suspected perchlorate-exposed area and there are other more likely causes for agranulocytosis, a further search of this data was not conducted.
Table 5 lists the number of cases of aplastic anemia diagnosed in individuals who resided in the 95670 zip code from 1991 through 1995. For the five year time period, there were eight hospitalizations for 95670 zip code residents for a rate of 3.8 hospitalizations per 100,000 individuals per year, which is higher than the statewide rate of 2.2 hospitalizations per 100,000 individuals per year. However, all but one of the eight hospitalizations also had an additional diagnosis of cancer or chemotherapy or radiation which would seem to be the likely explanation for aplastic anemia, as chemotherapeutic agents and radiation treatments have been shown to produce bone marrow aplasia. It was not felt that reviewing the medical records of these individuals would provide meaningful information on the possible role of perchlorate.
Reviews of the medical literature suggest that long-term interference of the thyroid-pituitary axis can lead to thyroid follicular cell neoplasia. Perchlorate may cause similar changes in thyroid cells by decreasing the amount of circulating thyroid hormones, thereby increasing stimulation of the thyroid gland by TSH, with a concommitant hyperplasia of thyroid cells. Ultimately, with continued TSH stimulation, the diffuse hyperplasia may progresse to nodular proliferation of the follicular cells and eventually to benign and malignant tumors (10). These changes have been observed in animal testing studies, primarily in rats, who appear to be a more sensitive species for these effects; it is not clear whether humans are similarly affected.
Concerned citizens of Rancho Cordova contacted the CCR and requested data reviewing the incidence of all types of cancer, thyroid cancer and cancer of the blood in the potentially-affected area. This area was confined to zip code 95670 during the years 1988 through 1994. The analysis was further stratified by gender (male and female) and the zip code was deconstructed into its corresponding 1990 census tracts (Figure 2). It should be noted that there is not an exact 1:1 correspondance for each census tract and the entire zip code, but this estimation is fairly close. Analysis of the data for cancer cases, all sites combined, failed to demonstrate evidence of a statistically significant increase in the observed number of cancer cases for any of the census tracts relative to the number of cases of cancer that would be expected, based on the statewide rate for cancer (Table 4). Also, cases of thyroid cancer were analyzed (Table 5) and no statistically significant elevations of observed numbers of thyroid cancer cases were detected compared to the numbers of cases which would be expected, based on corresponding statewide thyroid cancer rates. It is interesting to note that higher numbers of observed cancer cases and thyroid cancer cases were seen in census tract 8904 among women, however it is not possible to infer a relationship to perchlorate exposure from this limited data.
Leukemia is cancer of the white blood cell lines. Acute Lymphoblastic Leukemia (ALL) is the most common malignancy among children less than 15 years of age (12). Acute Myeloblastic Leukemia (AML) is the second most common childhood malignancy with about 400 new cases occurring annually in the United States (13). It has been postulated that risk factors for childhood leukemia include maternal exposures during pregnancy to radiation, infectious agents, pesticides, and genetic factors (12,13). No information concerning perchlorate's relationship with leukemia was found in the medical literature, however information concerning all leukemias and childhood leukemia in particular was readily available in the Cancer Registry data base and was another cancer which concerned residents from Rancho Cordova requested information on from CCR. The CCR, Region 3, provided information on all childhood leukemia cases from their data files for individuals residing in zip code 95670 from 1987 through 1996.(2) During that time, there were four cases in children aged less than 14 years, consisting of three cases of ALL and one case of AML. Since the total child population of this zip code was 9,648 individuals by the 1990 Census, the crude rate of childhood leukemia was calculated to be 4.2 cases per 100,000 children per year [4 cases/(9,648 people X 10 years)]. This rate is less than the corresponding childhood leukemia rate for California from 1988 through 1992 of 4.68 cases per 100,000 children per year.
In addition, after reviewing the data on all cases of leukemia, the CCR found no evidence of a statistically significant increase in the observed number of leukemia cases in the affected census tracts in Rancho Cordova from 1988 through 1994, when compared to the number of cases expected, based on statewide rates of leukemia.
Based on this information and the fact that additional factors may be responsible for adult and childhood leukemia, it was not recommended that further study be performed reviewing leukemia statistics at this time.