Calvine-Florin Cancer Cluster Investigation
Technical Assistance Provided to the Concerned Residents’ Initiative
Richard Clapp, DSc
Molly Jacobs, MPH
2005
Introduction/Background
Beginning in 1995, concerns regarding the health of Calvine-Florin —an unincorporated area of southern Sacramento County in California— residents emerged after two children living houses away from one another on Auberry Drive were diagnosed with acute lymphocytic leukemia (ALL) in the same year. During that same year, another resident living on the same street was diagnosed with non-Hodgkin’s lymphoma (NHL). This apparent clustering of cancer cases prompted concerned residents of Calvine-Florin to contact the California Department of Health Services (CDHS) and request an assessment of cancers in the neighborhood. CDHS analyzed cases of cancer diagnosed while living in the census tract associated with Auberry Drive and determined that rates of childhood leukemia were not statistically significantly elevated compared to the number of cases that would be expected in the area if they were occurring in the similar proportions as in the statewide childhood cancer incidence data at the same rate as the state. In the following years, additional children living on Auberry Drive were diagnosed with leukemia. Although CDHS continued to monitor the area and updated their analyses in 1996, 1998, 2002 and 2003, the pattern of cancer rates in this neighborhood was not considered in excess of expected rates. Frustrated by the state’s conclusion, a grass-roots effort, the Concerned Residents’ Initiative (CRI), launched their own community health investigation in order to uncover possible environmental risks in the community and to determine whether health conditions in the neighborhood were associated with these risks.
CRI discovered that the Calvine-Florin area was dominated by agriculture including vineyards, dairy farms and a variety of agricultural crops from 1930-1980 and became a primarily residential neighborhood in the 1980s.
Possible exposure to pesticides from former agricultural activities was a concern, although no knowledge of levels of pesticides in air or water have been documented. CRI also identified a number of industries whose activities may have resulted in toxic exposures to the community (Appendix A). In 1997, AmerPride, an industrial dry cleaner, discovered that tetrachloroethylene (PCE) had contaminated the ground water. Ground water monitoring wells near Wilbur 1, a municipal water head, showed PCE levels as high as 9600 ppb. In December 2002, elevated concentrations of PCE in drinking water from the plume were detected in Wilbur 1 (detected=79 ppb; EPA’s MCL=5 ppb). Although the municipal water purveyor concluded that the water was safe for consumption, the impacted well was taken off-line. A second well was also closed a few months later.
In addition to inventorying publicly available documents about contamination and possible toxic exposures from industrial activities in the community, CRI also tested the soil, indoor air and tap water of a number of residences. Results from the testing showed elevated concentrations of n-nitrosodimethylamine (NDMA) in tap water (detected=0.0022-0.0081 ppb; California Department of Health Service’s action level= 0.0010 ppb).
CRI launched a community health survey in 2002 in order to ascertain the health status of individuals diagnosed with various conditions between 1990 and 2000. With the help of 200 volunteers, CRI surveyed the 25,000 residents of Calvine-Florin. The results indicated a wide range of reported
health conditions among residents including cancers, auto-immune diseases and birth defects. Cases identified in the CRI survey were used to repeat the CDHS analysis to determine whether or not cancer was elevated in the Calvine-Florin area. However, the CRI method was different from the CDHS’s assessment because all cases that lived in Calvine-Florin were included in the analysis as compared to the CDHS analysis that only included cases that lived in the community at the time of diagnosis. The CRI results suggested that among individuals age 0-24, ALL, AML (acute myelogenous leukemia) and NHL were all elevated in the community (standardized incidence ratio= 1.90, 1.57 and 3.39 respectively) although there were only two cases of both AML and NHL.
In 2003, CRI contacted Dr. Richard Clapp and Molly Jacobs at Tellus Institute to help CRI analyze their community health survey in order to determine whether or not cancer rates are elevated in the community and if warranted, to work with CRI and provide recommendations to CDHS regarding additional analyses that should be completed in order to address the community’s environmental health concerns.
This report describes activities undertaken by Tellus in support of CRI’s health investigation.
Activities
Review of Environmental Sources of Concern
CRI and Miller, Sher & Sawyer LLP provided Tellus with a variety of documents on environmental sources of concern in Calvine-Florin. The documents consisted of a number of environmental site assessment reports for the industrial sites outlined in appendix A, a list of causes of deaths among employees at Chinet/Keyes Fiber, and a report by the municipal water purveyor, California-American Water Company ’s evaluation of the detection of PCE in the well head Wilbur 1. The documents were reviewed to: 1) become familiar with the industrial activities of concern that took/take place in Calvine-Florin; 2) to identify any toxic exposures of potential concern and 3) to recommend additional chemicals that should be tested for during a 2nd round of residential testing by CRI.
The documents reviewed did not demonstrate additional chemicals of concern that were not previously identified by CRI and colleagues at Miller, Sher & Sawyer LLP. Solvents, metals and likely dioxin were all suspected and/or documented toxic exposures of concern given the industrial practices that took place in the area. However, the review highlighted the complexity of the exposure issue facing CRI: there were several industries that used chemicals known or suspected to cause cancer, but no documented information about known exposure pathways based on available data. Our review was overall consistent with the CDHS’s Environmental Health Investigations Branch’s review of potential pollutant sources affecting Calvine-Florine. No additional chemicals were recommended for testing based on our review of the reports provided. However, Tellus recommended that chloroform and/or trihalomethanes should be examined in the tap water given that they are by-products of
water chlorination and the scientific literature has demonstrated links with childhood cancer.
Health Survey Review
CRI provided Tellus with the Access database containing the results of the community health survey as well as a copy of the original survey instrument. According to CRI, volunteers attempted to survey all households within census tracts 93.03 & 93.04 (1990 census tract designations). The majority of surveys were completed in-person and some were mailed. All completed surveys were re-contacted to collect additional data on variables such as race/ethnicity and to verify diagnoses. The survey queried whether individuals living in the household were diagnosed with cancers, autoimmune diseases (such as lupus and M.S.), had experienced pregnancy complications such as miscarriages, fetal death, preterm birth and still births) or experienced other chronic illness. Additional questions queried the type of illness, date of diagnoses, age of diagnosis, whether the individual lived at the place of residence at the time of diagnosis and whether any child was born with a birth defect. Given that the review of environmental data did not uncover a specific exposure source and pathway of concern and that the CRI survey did not collect information about individual exposure to environmental risk factors, Tellus agreed to use the survey to examine whether individuals with opportunity to exposure (as defined based on distance of residence from industrial sources of concern) were experiencing health conditions more frequently
than those individuals with less opportunity for exposure.
Tellus quickly learned that the survey results would not permit the desired analysis. The first question on the survey asked, “Has anyone living in the household been diagnosed with: leukemia, lymphoma, Hodgkin’s disease, aplastic anemia, other cancers M.S., lupus…” If an individual answered “no”, data were not generally recorded and no data on these individuals were entered into the survey database. As a result, the CRI database only consisted of those individuals diagnosed with the health conditions of concern and did not represent the experience of the entire study population. Without data on the entire study population, it was impossible to analyze the survey results to answer questions about whether Calvine-Florin residents were suffering from excesses of
diseases possibly associated with toxic exposures in the community.
In October 2003, Tellus met with CRI and attorney Jan Schlichtmann to plan how best to move forward given what was known about possible environmental risks in the community and the inability to use the community health survey as originally planned. A number of questions in the meeting were voiced by CRI which helped to focus Tellus’ technical assistance strategy. These concerns included: 1) whether or not the CDHS’s cancer analyses of Calvine-Florine included all individuals who lived in the area and were later diagnosed with cancer and 2) whether cancers observed in the community were caused by poor drinking water–given the known PCE contaminated ground water and municipal water head and the elevated concentrations of NDMA found in tap water—and whether the population is still at risk. Based on these concerns Tellus agreed to:
- Use the CRI survey results to examine the status of childhood cancer and whether there were opportunities for exposure to drinking water while residing in homes within the service area of the contaminated well field;
- Use the CRI survey results for childhood cancer to compare the list of cases used in the CDHS’s analyses; and
- Recommend additional analyses that the CDHS should conduct that would better examine the cancer experience among the population in the water service area of the contaminated well field.
The decision to only focus on childhood cancers was driven by the apparent elevation of cancer (particularly leukemia) among children in Calvine-Florin as well as by evidence in the scientific literature that suggest increased cancer risk to children when exposed to toxic contaminants in drinking water while in utero and in early life.,,
Review of Childhood Cancers
Tellus used the CRI health survey database to query all childhood cancers (diagnosed between 0-19 years of age). The original, completed surveys were obtained for all cases in order to check the data and to obtain any additional information that was not recorded in the database. A timeline was created for each case in order to examine when the child lived within the water service area and when the child was diagnosed with cancer (Appendix B). Based on these results, the CRI health investigation identified 8 cases of ALL, 2 cases of AML and 6 other cancers (2 cases of Wilm’s tumor, 2 cases of rhabdomyosarcoma, 1 other case of soft tissue sarcoma, 1 case of bone and 1 case of cervical).
All cases seem plausibly associated with some exposure while living in Calvine-Florin considering the timing of residence and that the general induction period for childhood cancers is 1-5 years. Among the cases of leukemia (both ALL and AML), all were diagnosed after living in Calvine-Florin. Two cases were diagnosed after moving away from Calvine-Florin although both lived there while in utero and through the first year of life. For the other cancers, all were also diagnosed after moving away from Calvine-Florin. The two Wilm’s tumor cases both lived in Calvine-Florin while in utero. With the exception of cervical cancer, all cancer types observed in Calvine-Florin are suggested to be associated with environmental risk factors.6
Dick Clapp met with Dr. William Wright of CDHS’s Cancer Surveillance Section in order to compare the cancer cases identified through the CRI health investigation with those used in the CDHS’s cancer analysis (appendix C). Since only childhood leukemias were examined by CDHS, only leukemias were compared. Of the 10 leukemias that were identified by CRI, only 5 were included in the CDHS’s cancer analysis. Of those not included by CDHS, four were diagnosed while living outside of Calvine-Florin, two were diagnosed after 2000, and one was over 14 years of age (California defines childhood cancer as 0-14). This exercise demonstrated that the CDHS included the appropriate cases based how the data are reported to the registry and the time period of the analysis. According to CDHS, “Even if the seven cases on the south Sacramento list that were not included in the CCR study had been included, the total number of cases would only have reached 19, only half the number required to reach a statistical conclusion of a significantly elevated number of cases.” CRI acknowledges that given resource limitations, a large proportion of the Calvine-Florin population was not surveyed. CRI’s counts of leukemia cases, that had an opportunity for exposure while living in Calvine-Florin, were likely an under-representation of the true frequency. However, it is impossible to speculate whether a higher response rate would have resulted in the required number of cases needed to reach statistical significance.
Although CRI asked that CDHS use all cases that ever lived in Calvine-Florin prior to their cancer diagnosis in the state’s cancer analysis, it is methodologically infeasible to do so for two reasons. First, the California Cancer Registry does not collect residential histories on cases reported to the registry, only the residential address at time of diagnosis. Thus, there are no available data for the state to determine the full population that ever lived in Calvine-Florin (or any area) prior to diagnosis. Secondly, even if the California Registry collected residential history data and could identify all individuals who ever lived in a particular area of concern prior to diagnosis, comparable data for the rest of California
would not be available. Equivalent cancer data and population data are needed to run the calculations needed to determine whether or not cancer is elevated in a particular community. CRI’s way of thinking about who are the appropriate people to include in a cluster investigation where the hypothesis is one of an environmental exposure in a particular community is valid. However, initial
descriptive epidemiological analyses such as those used in cluster investigations are conducted by using available data.
In a follow-up case-control study where complete residential histories of cases and controls can be gathered, it would be possible to consider exposures in those who used to live in a community and then moved away. This is rarely done and requires additional staff and funding to identify and contact participants and gather the necessary historical information.
CDHS Request for Additional Analyses
Given the need to work within the methods available to state health departments when assessing a potential cancer cluster, Tellus and CRI spoke with Dr. Raymond Neutra, Chief of the Division of Environmental and Occupational Disease Control at CDHS to voice continued concerns in the community regarding a potential cancer cluster associated with drinking water quality. If drinking water among households within the water service area of concern was associated with cancer among those residents, the 2003 analysis by CDHS would likely not be able to uncover this link because the population analyzed was not appropriate (the true population at risk). Instead a much larger population was analyzed and any actual elevation of cancer would have been diluted. Tellus requested that CDHS repeat their analysis of childhood leukemia but only examine the population residing within the water service area.
Dr. Neutra completed the analysis as requested (appendix D). Five cases of leukemias residing within the water service area were identified through the cancer registry. These five cases were consistent with those validated by Drs. Clapp and Wright as previously described. The results from the analysis did not demonstrate that cancer was elevated in the Calvine-Florin area of concern. There were 27.8% leukemias in the water service area vs. 32.5% in Sacramento and 36.5% in the Calvine-Florin census tracts minus the water service area.
During the conversation with Dr. Neutra, CRI expressed concern about the elevated levels of NDMA in tap water as identified by CRI’s environmental testing. Dr. Neutra was unaware of the NDMA findings and later met with CRI, representatives from the drinking water purveyor and representatives from CDHS’s Drinking Water Program. Subsequent testing also identified NDMA in tap water and after maintenance was completed on the water storage tanks (including new liners), the raw tap water tested free of NDMA. Had it not been for CRI’s testing, this problem would never have been identified.
Concluding Comments
Although Tellus could not confirm the existence of a cluster in Calvine-Florin, we did work with CRI to: 1) validate multiple cases of the same or similar cancer among children who had opportunity to exposure to environmental risk factors while living in the same neighborhood; and 2) outline and submit a request to CDHS to conduct a more appropriate descriptive epidemiologic analysis that would address the community’s concerns regarding cancer among residents within the water service district associated with the Wilbur 1 wellhead. Although the CDHS’s conditions were not met in order to launch an epidemiologic study of Calvine-Florin, the patterns of cancer seen in the community and the environmental risks that have been identified warrant CRI’s continued vigilance to protect and ensure the public health.
Appendix A. Primary Industries of Concern in Calvine-Florin
|
Name
|
Activity
|
Time-period
|
Exposures of Concern
|
|
AmeriPride/Valley Industries
|
Industrial Dry Cleaning
|
1963-current
|
Tetrachloroethylene (PCE)
|
|
Chinet/Keyes Fibre/Huhtamake
|
Paper products manufacturing
|
1960-current
|
Dioxins
|
|
Sacramento Army Depot
|
Stored, issued and maintained electronics, supplies and commodities
|
1940s-1994
|
Area 22 waste pit: burned toxic chemicals including radium; trichloroethylene, cadmium and chromium identified in ground water
|
|
Union Carbide
|
Rocket fuel manufacturing
|
1964-1970
|
Liquid hydrogen, nitrogen tetrazide
|
|
Western Ceramics/ Chromalloy/General Radiator
|
Ceramics manufacturing & later Radiator manufacturing
|
Mid 1960’s-1980’s
|
Lead, chromium, zinc, solvents, waste oil in soil
|
Appendix B: Childhood Cancer Analysis
Appendix C: Comparison of Concerned Residents’ Initiative and California Department of Health Services cancer cases
Appendix C: Cont.
Appendix D: California Department of Health Services: Cancer Analysis of Water Service Area in Calvine-Florin
Appendix D: cont.
