Cancer Risk Associated with Soil Distribution of Polycyclic Aromatic Hydrocarbons within three Environmental Justice Neighborhoods in Houston, Texas

Abstract

Residents and advocacy groups began voicing concerns over the environmental quality located in the neighborhoods of Kashmere Gardens, Fifth Ward, and Denver Harbor in Houston, TX, following the confirmation of a cancer cluster in 2019 and another in 2021. These neighborhoods are in close proximity to a railyard and former wood treatment plant known to have utilized coal tar creosote and contain polycyclic aromatic hydrocarbons (PAHs). This research took core soil samples in September and October 2020 from 46 sites to assess for the presence and concentration of the U.S. Environmental Protection Agency’s (USEPA) 7 Carcinogenic PAHs. Results showed the cumulative concentration of these PAHs in each sample was variable with a range of 13,767 ng/g to 328 ng/g and a mean of 2,517.2 ng/g ± 3,122. A regional soil screening evaluation revealed that 40 of the 46 soil samples were in excess of the USEPAs most conservative screening levels of 1.0 ×10⁻⁶ increased cancer risk, but none exceeding levels considered actionable for remediation. This study is a fundamental first step for quantifying the environmental pollutants in this minority-majority community. Findings revealed a low risk of cancer risk based current PAH concentrations alone but cannot assess contributions from other contaminants or from past, possibly higher, levels of contamination. Further research is needed to identify the potential casual pathways of the observed cancer cluster and to explore possible remediation needs.

Introduction

In the last several decades research has conclusively revealed that communities comprised predominantly of people of color (POC) are disproportionally burdened by environmental contaminants and ongoing health effects (Bullard, 2000). These environmental justice communities shoulder the brunt of urban pollution and proximity to industrial pursuits (Walker, 2012; Bryan & Mohai, 1991). The negative impact on the lives of those located within these neighborhoods is underscored by the reality that these economically disadvantaged communities find it financially difficult to leave these neighborhoods. The historical practice of placing polluting facilities in regions with predominantly low socioeconomic individuals and POC has created generations that have experienced toxic exposures in locales where individuals are less likely to object because of their lack of political leverage (Bullard, 2000). While numerous studies have assessed chronic exposures and their health implications, few have characterized the overall distribution and concentration of exposures within confirmed cancer clusters.

Kashmere Gardens, Fifth Ward, and Denver Harbor are geographically defined neighborhoods in proximity to one another in North East Houston, TX. These neighborhoods garnered national attention with the discovery of a cancer cluster within Fifth Ward and Kashmere Gardens in 2019, and then another confirmed cluster in 2021 in the same area. While Fifth Ward and Kashmere Gardens are home to two of twenty-one Superfund Sites located in Houston, it is the presence of a 359-acre site formerly named the Englewood Radar Yard, that has local environmental organizations and residents concerned. It was constructed in 1956 and is currently owned by Union Pacific, the Englewood Radar Yard functioned as a wood-treating plant for decades and was situated on the largest railyard in the Southern United States. An environmental assessment performed in 1988 identified groundwater contaminants of concern including polycyclic aromatic hydrocarbons (PAHs) and heavy metals (USEPA, 2021). At the peak of the Southern Pacific Railroad’s operations in 1979, 34% of residents in the Fifth Ward lived below the poverty level, and only 25% had a high school diploma. These disparities in education and financial situations left most residents unable to leave despite knowing about harmful contaminants. During the railyards most productive years it was shipping up to 1.7 million railroad ties annually and gained the nickname “el Creosote” from Hispanic community members due to the heavy smell in the air surrounding the site and within the neighborhoods themselves (West, 1979).

Coal tar creosote is a form of creosote that is not naturally occurring and is the most common form of creosote present in hazardous waste sites in the United States (CDC, 2014). Workers in fields related to wood preservation using coal tar creosote are at specific risk for exposure, as well as those living or working in areas where wood preservation sites used to exist. Coal tar creosote has been found in as many as 46 of the 1,613 National Priorities List (NPL) sites as reported by the United States Environmental Protection Agency (USEPA) (Gan et al., 2009). Coal tar creosote is made of many chemicals that carry risk of impactful health effects, including PAHs, phenol, and cresols. PAHs from coal tar creosotes occur as a result of incomplete combustion and pyrolysis of organic matter and can come from many sources, including but not limited to petrochemical production, coal production, wood preservation sites, and forest fires (Sansom, et al., 2021).

Soil, water, air, and vegetation contaminated by creosote and PAHs present severe human health and environmental hazards (NTP, 2016). Some components of coal tar creosote are water soluble. Once dissolved in water, these chemicals can move through soil and reaching ground water, where they may take years to break down. Whereas other chemicals take a few months to breakdown, and others may take decades. The most common route for coal tar creosote exposure is dermal, through contact with soil or water, or through direct contact with wood treated with coal tar creosotes. Extended and repeated exposures to coal tar creosote increases the concentration of the creosotes that pass into the blood stream. Pregnant women and infants are particularly vulnerable groups, as some components of coal tar creosotes metabolize and can be stored in body fat for long periods of time, leading to potential oral exposure through breast milk. Certain coal tar creosote components, including PAHs, are absorbed at a higher-than-normal rate through the lungs, stomach, and intestines, making routes of exposure including oral and inhalation even more dangerous. The USEPA has determined that coal tar creosotes are probable human carcinogens, and seven of the priority 16 PAHs are reasonably anticipated to be human carcinogens (NTP, 2016). The objective of our study is twofold. First, to characterize the distribution and concentration of the 7 carcinogenic PAHs within these environmental justice communities and second to assess for the potential carcinogenic impacts associated with these exposures based on regional screening levels in soil samples.

Methods and Materials

Site Location

Kashmere Gardens, Fifth Ward, and Denver Harbor are geographically defined neighborhoods in North East Houston (Fig 1). Individuals living within this community experience chronic pollution (Chakraborty et al., 2014), nuisance flooding (Houston Chronicle, 2001), air pollution (Sexton et al., 2007), poor housing conditions, and contaminated water (Sansom et al., 2018). An Assessment of the Occurrence of Cancer published in March 2020 was conducted by the Texas Department of State Health Services (DSHS) which examined the amount of cancer diagnoses across 21 census tracts in Houston, Texas. The data covered the 17-year period from 2000–2016 and determined that the rates for acute myeloid leukemia, lung and bronchus, esophagus, larynx, and liver cancers were significantly greater than could be expected across all tracts investigated (Texas Department of State Health Services, 2019). The assessment determined that in the tract covering Fifth Ward and Kashmere Gardens, the observed number of adult cases of liver, and lung and bronchus cancers were found to be significantly greater than expected. The Supplemental Assessment, published in 2021, further indicated that the observed number of childhood acute lymphoblastic leukemia cases were statistically greater than expected during this time as well (Texas Department of State Health Services, 2021).

Figure 1.

Study location and sampling area for Greater Fifth Ward, Kashmere Gardens, and Denver Harbor in Houston, TX.

The neighborhoods are within close proximity to an inactive creosote processing facility owned by Union Pacific (Houston Health Department, n.d.). As creosote has been identified as a probable human carcinogen, this potential exposure pathway may be partially responsible for public health issues within this community (ATDSR, 2002). Despite the proven cancer clusters within Fifth Ward and Kashmere Gardens, little has been done to quantify potential hazards and see how the health and environmental issues may intertwine. The populations of Fifth Ward, Kashmere Gardens, and Denver Harbor are comprised primarily of African American and non-white Hispanic groups (USEPA, 2002). Household family income is approximately 37 percent lower than the city as a whole and only one third have received any college (U. S. Census Bureau, 2000).

Soil Sample Collection

Utilizing geographic information systems ArcMap (ESRI, Redlands, CA), 16 locations were selected within the neighborhood boundary of Kashmere Gardens, 15 locations were selected in both the neighborhoods of Fifth Ward and Denver Harbor for soil sampling locations to assess for the presence and concentration of PAHs. The local civic group the Coalition of Community Organizations (COCO), who were already engaged with the Environmental Defense Fund (EDF) and the community lead group Impact Greater Fifth Ward, assisted with data collection, and provided much needed local knowledge on sampling locations. This project and all related materials were approved by the Texas A&M University Institutional Review Board (IRB2020-1064).

Data collection teams donned powder-free nitrile gloves and soil core samples were taken from the pre-identified geographic locations using disposal 5-gram coring devices. A square grid-sampling method was used to collect the top 2 to 3-cm of soil. A square of 2 meters was established and one core from each corner and the center provided the composite sample for each location. The core samples were returned to a pre-labeled glass container and placed into a cooler for transport. After returning to the laboratory the samples were stored at −20°C before analysis in accordance with EPA guidelines (USEPA, 2002).

Sample Extraction and Analysis

Sample extraction and analysis was performed in compliance with the standard operating procedure of B&B Labs/TDI Brooks International. Quality control samples (method blank, blank spike, matrix spike/matrix spike duplicate, and laboratory duplicate sample) and the standard reference material (SRM-1941b) were prepared with the requisite surrogate and spike volumes added. Extraction was achieved with a Dionex ASE 200 Accelerated Solvent Extractor (ASE) operated at a high temperature (100°C) and pressure (1500 psi) with 100% dichloromethane. The extracts were collected in 60 mL ASE vials and concentrated to – 3 mL in a thermally regulated water bath. Activated granular copper was added to each vial in increments to remove elemental sulpher. To determine the Extractable Organic Matter (EOM), the extracts were transferred to clean class A K-D tubes and concentrated to 3mL in a water bath. Next, 100 uL was drawn from each sample extract and transferred unto a dry and pre-weighed glass fiber filter using a micropipette. The filter was then placed in an oven maintained at a temperature range of 38 – 42 °C for 2 minutes after which the change in mass is determined as EOM. Post extraction cleanup was performed on the remaining sample in a long glass column packed with deactivated alumina and silica. Sample extracts were loaded unto columns sequentially equilibrated with dichloromethane and pentane and eluted with 50/50 v/v mixture of dichloromethane and pentane. The elute was concentrated to –1mL and internal standard was added to before the gas chromatography–mass spectrometry (GC-MS) analysis.

Quantitative analysis of the PAHs was executed using an Agilent 6890N gas chromatograph (Agilent 6890N) coupled with an Agilent 5973 mass spectrometer (Agilent 5973, Agilent Technologies, Santa Clara, CA). A HP-5MS capillary column (Agilent HP- 5MS, 60 m long with an interior diameter of 0.25 mm and 0.25 μm film thickness, Agilent Technologies, Santa Clara, CA) was used to chromatographically separate PAHs and n- alkanes analytes. The oven temperature was programmed to 60^°C, after the injection the oven increased by 7°C/min until it reached the final holding temp of 310°C with a final hold temperature for 22 minutes. To establish retention times, organics in marine sediment standard reference material (SRM-1941b), were used. The GC-MS measurements included a tune system, a six-level initial calibration (ICal), an independent calibration verification solution (ICV), continuing calibration checks (CCC), a reference oil (SRM 1582), a method blank, a laboratory duplicate (DUP), a blank spike (BS), a matrix spike (MS), and a matrix spike duplicate (MSD).

Data Analysis

The mean concentration and deviation of the 7 carcinogenic PAHs identified by the USEPA were calculated across the 46 samples. The total concentration of the PAHs in each soil sample were analyzed by summing the PAH concentrations. Next, the toxicity equivalent (TEQ) method was used to determine the carcinogenic risks of these locations. The total Benzo[a]pyrene (BaP) equivalent concentration (BaP_eq) was calculated by the sum of BaP_eq for each PAH using toxicity equivalent factors (USEPA, 2016; USEPA, 1991). To evaluate the potential implication of soil PAHs from these communities for human health, a total RSL was calculated included dermal, ingestion, and inhalation exposure routes (eq 1.) (USEPA, 2016).

$${\text{SL}}_{\text{res-sol-totc}}\left(\frac{\text{mg}}{\text{kg}}\right)=\frac{1}{\frac{1}{{\text{SL}}_{\text{res-sol-ingc}}}+\frac{1}{{\text{SL}}_{\text{res-sol-inhc}}}+\frac{1}{{\text{SL}}_{\text{res-sol-derc}}}}$$ Eq.1

The ratio of BaP_eq soil concentrations with the cancer RSL yields a screening level risk characterization in units of excess cancer risk per million. Similar methods have been performed in other neighborhoods in Houston, TX (Sansom et al., 2021), as well a multi-community assessment (Ali, 2019; Li et al., 2018; Zheng et al., 2018), in vegetation (Tesi et. al., 2021) and in occupational settings (Kamal et al., 2014; Inam et al., 2016). A complete list of equations can be viewed within the USEPA PAH equations (ESEPA, 2016). A box and whisker plot were produced to represent the RSL values (Microsoft Excel, Redmond, WA).

Results

The cumulative concentration of the 7 carcinogenic PAHs in each sample was variable with a range of 13,767 ng/g to 328 ng/g and a mean of 2,517.2 ng/g ± 3,122 standard deviation. Differences in PAH concentrations were observed across the 46 sites. The USEPA provide interpretive guidelines of RSL values, and ratios yielding values below 10⁻⁶ are considered to be without concern, values above 10⁻⁴ risk considered actionable, and intermediate values as the “target range” for remedial actions (USEPA, 2002). All 40 of the 46 soil samples were found to have a value in excess of the USEPAs screening level for increased cancer risk of 1.0 ×10⁻⁶ The two sample sites with the highest values of 2.75×10⁻⁶ and 2.92×10⁻⁶ respectively were locations in the neighborhood of Denver Harbor. (Fig 2). Thus, despite being above screening levels, none of the values represented actionable levels for PAH carcinogenic risks. The site-specific cancer risk assessment utilized the TEQ method for each of the 7 carcinogenic PAHs (Table 1).

Figure 2.

Box and whisker plot (N=46) of Regional Screening Levels (RSL). The area delineated in red indicates a value in excess of the lowest screening level risk

Table 1.

Mean concentrations of polycyclic aromatic hydrocarbons in soil samples

Chemical	TEF*	Concentration (ng/g)
Benzo(a)anthracene	0.1	2.27E+02
Benzo(a)pyrene	1	3.40E+02
Benzo(b)fluoranthene	0.1	7.08E+02
Benzo(k)fluoranthene	0.01	4.66E+02
Chrysene	0.001	3.55E+02
Dibenzo(a,h)anthracene	1	8.08E+01
Indeno(1,2,3-c,d)pyrene	0.1	3.40E+02

^*TEF = Toxic Equivalency Factor

To characterize the accumulation of PAHs in the study area, spatial interpolation was performed by mapping the site values onto the sampling locations to identify hotspots (Fig 3). For all PAHs and selected analytes, a concentration was revealed in areas near the railyard, highways, and the former location of the creosote plant. All three neighborhoods had hotspot locations with the majority appearing in the southern section with BaP and pyrene presenting the greatest variability throughout the study area. The BaP-TEQ was calculated for each sample, location specific BaP-TEQ values varied between samples and neighborhoods. Values were calculated ranging from 79.5 BaP-TEQ to 3,596.9 BaP-TEQ (ng/g). There was a mean value of 618.1 BaP-TEQ and a standard deviation of 782.5 BaP-TEQ.

Figure 3.

Spatial distribution of PAHs illustrates hotspots within study location. Concentration of benzo(a)pyrene, pyrene, naphthalene, and total PAHs, with all concentrations in ng/g.

Discussion

Overcoming health disparities witnessed within communities of color requires an empirical understanding of the environmental conditions within these neighborhoods. Within our study location, the majority of samples exceeded lowest limits, while this should be characterized as a relatively low overall risk, this could be a partial explanation for the increased cancer rates in this area. According to soil contamination classification proposed by Maliszewska-Kordybach (1996), every sample revealed heavy contamination levels throughout all three neighborhoods. While PAH distribution was seen throughout the three neighborhoods there were several local hotspots centered in the sampling location. Areas in close proximity to the railyard, major highway, and the former creosote plant had the highest concentrations.

Risks of exposure to environmental contaminants, as well a possible distribution of PAHs, were heightened in August 2017, when 38 counties in Texas were greatly affected by Hurricane Harvey, a Category 4 hurricane (Kaplan et al., 2017). In the wake of the devastating flooding caused by Hurricane Harvey, the Environmental Protection Agency (EPA) confirmed that of the 41 Superfund Sites in Texas, 13 were flooded. The EPA reported that these sites could be experiencing storm damage due to the flooding, and that floodwaters had the potential to spread contaminants of concern from these Superfund Sites to surrounding communities (CNBC, 2017). Similar research in Houston revealed possible distribution of PAHs within environmental justice communities (Sansom et al., 2021).

Interpreting these environmental exposures in regard to human health is not straight forward. Personal and/or historical exposure levels were outside the scope of this research and are a necessary component to effectively understand the influence that past contamination could have on the health of this community. Further, there is a need for better sensitivity analysis that suggests a more conservative interpretation of environmental exposures, particularly dermal contact, which should be undertaken (Chen and Liao, 2006). There are additional uncertainties in risk assessment involving PAHs, specifically, the approach taken here is limited to the subset of analytes measured in this and prior (Hu et al., 2007) research and does not consider the toxicity of all PAHs that may be present, nor of other carcinogenic contaminants to which residents may have been exposed in the past.

Plants and animals are capable of absorbing portions of the creosote present in the soil. While present in the groundwater, there is high risk of oral exposure through drinking of contaminated water, especially from private wells. Components of coal tar creosote that are not water soluble are likely to remain where left, coagulating into a tar-like gel. While migration of the tar-like substance left behind is normally not expansive, external factors are much more likely to cause an increase in the spread. Due to the identified underground plume of contaminates associated with creosote underneath Fifth Ward, the Houston Health Department has initiated research attempting to identify public health issues within homes in the community (Houston Health Department, 2020).

Research conducted in the wake of Hurricane Harvey demonstrated that PAHs were diluted and distributed due to the heavy precipitation associated with the tropical storm (Horney et al., 2018). Some components of coal tar creosote are also capable of evaporating, although this amount is small in comparison to the amounts that enter the soil and water. PAHs as a component of coal tar creosote are hydrophobic, and as such will attach themselves to solid particles, becoming extremely persistent micropollutants that have the capability to spread through vegetation, soil, water, and air (Gan et al., 2009).

Some strengths with this study include the participation and collaborative nature between the research team and local organizations. Residents and civic groups offered feedback and advice on sampling locations and provided the baseline for the inquiry. Samples were collected throughout the neighborhoods and offer the ability to reveal the distributions of PAHs in a large geographic area. Important limitations of the project include the large time lapse between the creosote plant being active and when sampling occurred. As many PAHs degrade over time this analysis is likely to be a conservative estimate of exposure risks compared to when the site was active. The RSL values have several assumptions on the characteristics of the individuals who live within these neighborhoods that, while intended to be relatively conservative, may not be accurate for all residents.

Conclusion

This study was a critical first step in identifying the conditions that were leading to an excess of several separate cancers impacting adults and children within these environmental justice neighborhoods. There are a number of uncertainties with this research that should be pursued before any conclusive statements can be made on the environmental conditions in relation to human health. Among these are the need for fine scale environmental and public health research targeting individuals and their specific exposures to understand the impact more fully. There is a need to conduct a more extensive characterization of exposures during such research, this would require the collection of a broader set of potential exposure routes and contaminants.

Targeted interventions are vital in mitigating the poor public health conditions within this area, and environmental characterization of this region provides a steppingstone for this and future assessments. Research has shown that operating with the community and responding to the identified needs of residents can increase public health outcomes and future adoptions of approached poised to improve the life and wellbeing of those within contaminated regions (Houston Health Department, n.d.).