Only noninfectious pathological conditions or lesions in the liver and kidney which appear to have a positive relationship with contaminant exposure (i.e., toxicopathic lesions) are described in this report. The criteria used to evaluate the relationship between a lesion and contaminant exposure included: a) morphological similarity to toxicopathic lesions inducible in fish or mammals by experimental exposure to toxicants; b) similarity to lesions in wild fish that have been previously associated with contaminant exposure; or c) high prevalences of a lesion type in fish from urban sites known to have high levels of sediment contaminants, as compared to those at nonurban sites. Pathological conditions in wild fish reported to exhibit associations with contaminant exposure include certain toxicopathic liver lesions (Malins et al. 1984, 1987b; Myers et al. 1987, 1990, 1991; Hinton et al. 1990a , 1990b; Murchelano and Wolke 1985, 1991; Moore 1991; Baumann et al. 1991), and kidney lesions (Rhodes et al. 1987), as outlined in the introduction. As stated earlier, lesion prevalences for a site are reported only if a total of at least 15 specimens of a target species were captured and examined over Cycles I-V.
The presentation of the pathology data begins with a general description of toxicopathic lesions detected in the liver and kidney. This is followed by: a) a description of the geographical distribution for, and intersite comparisons of, hepatic and renal lesion prevalences in the primary target species represented by flathead sole, English sole, starry flounder, hornyhead turbot, white croaker, and black croaker (data for other target species with low or homogeneous prevalences of toxicopathic hepatic or renal lesions are presented in the Appendix, Figures A-1 to A-10); b) a description of the risk factors (site, age, and gender) associated with hepatic and renal lesion occurrence in individual fish of these species; c) a description of the relationships shown between levels of biliary FACs and risk of hepatic and renal lesion occurrence in a subset of the above data set for individual fish of these species from which bile was collected and analyzed; and d) a separate section examining the relationships between hepatic and renal lesion prevalences in these species and concentrations of contaminants in sediment (measures of potential exposure), stomach contents (measures of dietary exposure), liver (measures of hepatic bioaccumulation), and mean levels of biliary FACs (estimated measure of exposure to and metabolism of AHs), as determined on a site basis for each year sampled, while adjusting for mean age and gender ratio. The overall presentation of this above information for liver and kidney lesions is organized by species.
Toxicopathic liver lesions were placed into six categories reflecting various pathologic processes. These categories and the main constituent lesion types were 1) neoplasms (hepatocellular adenoma, hepatocellular carcinoma, cholangioma, cholangiocellular carcinoma, mixed hepatobiliary carcinoma, and rarely, mesenchymal neoplasms); 2) putatively preneoplastic foci of cellular alteration (eosinophilic focus, clear cell focus, basophilic focus); 3) non-neoplastic proliferative lesions (proliferative lesions) (hepatocellular regeneration, biliary hyperplasia or proliferation, presumptive oval cell proliferation, cholangiofibrosis/adenofibrosis, and increased mitotic activity in hepatocellular/biliary epithelial cells); 4) unique or specific degenerative/necrotic conditions (SDN) (megalocytic hepatosis, hepatocellular nuclear pleomorphism, and rarely, spongiosis hepatis); 5) nonspecific necrotic lesions unassociated with visible infectious agents (necrosis) (hepatocellular and or biliary coagulative necrosis, hyalinization, pyknosis, and karyorrhexis); and 6) hydropic vacuolation of biliary and/or hepatocellular epithelial cells.
Certain noninfectious hepatic lesions that are not considered to be related to contaminant exposure were diagnosed in the fish examined in this study. These included cytoplasmic storage conditions (e.g., fatty change, hemosiderosis), vascular lesions (e.g., hemorrhage, congestion), and inflammatory/fibrotic conditions (e.g., pericholangiitis, parenchymal fibroplasia/fibrosis). Because these lesions appear unrelated to contaminant exposure, they are not reported or dealt with further in this study. However, these data have been transmitted in ASCI text format to the Coastal Monitoring and Bioeffects Assessment Division of the Office of Ocean Resources Conservation and Assessment of NOAA's National Ocean Services and are available to other investigators.
Suspected toxicopathic kidney lesions clearly unassociated with visible infectious agents were placed into three separate diagnostic categories, representing differing pathologic processes, as follows: 1) proliferative lesions, including hemopoietic tissue hyperplasia or proliferation, glomerular hypercellularity, proliferation of the visceral or parietal layer of Bowman's capsule, and hyperplasia, regeneration or proliferation of the renal tubular epithelium in any segment of the nephron; 2) necrotic lesions (necrosis), including mesangiolysis, generalized glomerular degeneration and hyalinization, necrosis or hyalinization or hydropic degeneration or pyknosis of the tubular epithelium, necrosis or atrophy or hypoplasia of the hemopoietic tissue elements; and 3) sclerotic lesions (sclerosis), including mesangiosclerosis or mesangial fibrosis, glomerular sclerosis, hypermembranous renal tubules (thickened basal lamina), hypermembranous glomeruli (thickened peripheral basal lamina), and peritubular fibrosis and fibrosis/fibroplasia of the renal interstitium not associated with visible infectious agents.
The following sections present the site-specific prevalences of, and intersite prevalence comparisons for, toxicopathic hepatic and renal lesions in the primary target species exhibiting significant prevalences of these lesions (Figs. 3-14), followed by a desciption of the results of logistic regression analyses examining the significance of geographic (site of capture) and biological (age and gender) risk factors to disease in individual fish of each species. By convention, when lesion prevalences are identified in the text as higher than the expected value at the reference site (as indicated on the figures by an asterisk), this difference is statistically significant at p < 0.05; similarly, any stated associations of risk factors with increased risk of disease are defined as being statistically significant (p < 0.05). The presentation of species is organized according to geographic region of capture, from north to south.
Toxicopathic lesions in the liver and kidney were found infrequently in the following target species, at prevalences that were generally homogeneous among the sampling sites for a particular species; lesion prevalence data for these species are presented in the Appendix in Figures A-1 through A-10: fourhorn sculpin (Fig. A-1); Arctic flounder (Fig. A-2); yellowfin sole, Pleuronectes asper (Fig. A-3); Pacific staghorn sculpin, Leptocottus armatus (Fig. A-4); barred sand bass, Paralabrax nebulifer (Fig. A-5); spotted sand bass (Fig. A-6); spotted turbot, Pleuronichthys ritteri (Fig. A-7); diamond turbot (Fig. A-8); California tonguefish, Symphurus atricauda (Fig. A-9); and California halibut, Paralichthys californicus (Fig. A-10). By virtue of either inherent resistance to contaminant effects, obfuscatory migratory movements, restricted geographic distribution and/or low sample sizes, these species are not considered to be appropriate target species when utilizing hepatic or renal lesions as biomarkers of contaminant exposure effects.
Hepatic lesion prevalence and distribution--Prevalences of liver lesions in flathead sole were quite low at most of the sampling sites (Fig. 3), and although neoplasms and foci of cellular alteration were detected in fish from Elliott Bay, a hepatocellular adenoma was also detected in a single 11-year-old male from the Boca de Quadra site in southeast Alaska. Because of the normally broad variation in diameters of hepatocellular nuclei exhibited in specimens of this species from all sites, SDN was not interpreted as a lesion in this species. Hydropic vacuolation was not detected in this species. The only lesion prevalence higher than that at the reference site at Kamishak Bay was liver necrosis at Lutak Inlet (Fig. 3).
Risk factor analyses for hepatic lesions in individual fish--Logistic regression analysis of liver lesions in individual fish with respect to the risk factors of site, age or estimated age, and gender revealed no sites with an elevated relative risk for any of the hepatic lesion categories detected (Table 4), and neither age nor gender were risk factors for any hepatic lesion.
Renal lesion prevalence and distribution--Prevalences of renal lesions in this species are presented in Figure 4. In general, lesions were found infrequently, with higher prevalences detected at Commencement Bay (proliferative lesions), Elliott Bay (necrosis and sclerosis) and Boca de Quadra (sclerosis).
Risk factor analyses for renal lesions in individual fish--The significance of the geographic and biological risk factors to kidney lesion occurrence in individual flathead sole is shown in Table 5. An increased risk of disease was demonstrated for both necrotic and sclerotic lesions at several of the sites of capture, including Elliott Bay (necrosis and sclerosis), Port Moller (necrosis), and Boca de Quadra (sclerosis). No risk factors were identified for proliferative lesionserative lesions. Sclerosis was the only lesion type for which age was a risk factor.
Due to the broad geographical distribution of this species among the sampling sites (Alaska to central California), sites were grouped into northern and southern areas for the purposes of data display. Consequently, lesion prevalences at sites of capture for English sole are shown in Figs. 5A (northern sites from Coos Bay north to Boca de Quadra) and 5B (southern sites from Bodega Bay south to west Santa Monica Bay).
Hepatic lesion prevalence and distribution--Among the northern sites, differences in lesion prevalences were computed using Nisqually Reach as the reference site (Fig. 5A). Prevalences of neoplasms, foci of cellular alteration, proliferative lesions, SDN, and necrosis in this species were higher at either the Commencement Bay or Elliott Bay sites. Hydropic vacuolation did not occur in any English sole examined. Prevalences of neoplasms and necrotic lesions were below 10%, with higher prevalences of foci of cellular alteration (13.3%), proliferative lesions (13.3%), and SDN (42.5%) detected among these two urban sites. Specifically, prevalences of neoplasms, proliferative lesions, and SDN were higher in sole from Commencement Bay, and prevalences of all five lesion categories were higher in fish from Elliott Bay. No hepatic lesions in sole were detected at the Boca de Quadra site in southeast Alaska, and Coos Bay sole were rarely affected, and only with necrotic lesions.
For the southern sites, Bodega Bay served as the reference site (Fig. 5B). Prevalences of all lesions were low (<10%) at all sites, except for the Monterey Bay site where higher prevalences of proliferative lesions, SDN, and necrosis were detected.
Risk factor analyses for hepatic lesions in individual fish--Several risk factors were associated with an increased risk of hepatic lesion occurrence in individual English sole (Table 6). In this analysis, odds ratios/relative risks were determined relative to the combined data for both reference sites. Substantially elevated relative risks of neoplasms, foci of cellular alteration, proliferative lesions, and SDN were calculated for sole from the urban site in Elliott Bay. For example, an English sole from Elliott Bay was 34.190 times more likely to be affected by a hepatic neoplasm than a fish of comparable age and gender from the reference sites. Fish age was a risk factor only for hepatic neoplasms. For necrosis, an elevated risk was established only at Monterey Bay. In addition, females were less likely to be affected by necrotic lesions. Gender was not a risk factor for any other hepatic lesion.
Renal lesion prevalence and distribution--Figures 6A and 6B present the prevalences of renal lesion categories in English sole from northern and southern stations, respectively. In general, higher prevalences were detected at the urban sites. Prevalences of each lesion category were relatively low at all sites, but higher prevalences, as compared to those at the reference sites of Nisqually Reach (northern stations) or Bodega Bay (southern stations), were shown for necrosis and sclerosis at Commencement Bay and west Santa Monica Bay, proliferative lesions and necrosis at Elliott Bay, and necrosis at both Monterey Bay and Moss Landing.
Risk factor analyses for renal lesions in individual fish--Increased risks for certain renal lesions were associated with residence at two of the sampling sites (Table 7). Fish at Commencement Bay were about seven times more likely to be affected by sclerotic lesions than fish of similar age and gender from the reference sites. The site at Boca de Quadra also showed an increased relative risk for proliferative kidney lesions. No kidney lesions showed an increased risk in fish captured at Elliott Bay. Age was a risk factor for both necrotic and sclerotic lesions. The only lesion showing a gender-related effect was necrosis, which preferentially affected males.
Hepatic lesion prevalence and distribution--Sufficient numbers of starry flounder for inclusion into this study were captured at nine sampling sites, ranging from the Chukchi Sea in the north to several sampling sites in San Francisco Bay in the south, with Bodega Bay representing the reference or comparison site (Fig. 7). Prevalences of hepatic lesions other than hydropic vacuolation were generally below 10% at all the sampling sites, and neoplasms were not detected in any of the 851 starry flounder livers examined (Fig. 7). Significant intersite differences in prevalence of lesions other than hydropic vacuolation were demonstrated only for foci of cellular alteration and SDN in starry flounder from Hunters Point in San Francisco Bay. For hydropic vacuolation, prevalences ranging from approximately 3 to 50% were found at eight of the nine sampling sites; however, higher prevalences were shown only at three urban sites within San Francisco Bay (Castro Creek, Southampton Shoal, and Hunters Point).
Risk factor analyses for hepatic lesions in individual fish--Both site-related and biological risk factors were associated with increased risk of hepatic lesion occurrence in starry flounder (Table 8). For foci of cellular alteration, increased risk was indicated in fish from the urban sites of Castro Creek and Hunters Point in San Francisco Bay, while the relative risk for SDN was higher only at Hunters Point. Other sites with increased relative risks of hepatic disease were shown only for hydropic vacuolation, and with the exception of Youngs Bay in Oregon, all of these sites were urban sites within San Francisco Bay. In general, the highest relative risks for the hepatic lesions detected, especially hydropic vacuolation, were shown for the urban sites of Castro Creek and Hunters Point. Age was a risk factor for foci of cellular alteration, proliferative lesions, and hydropic vacuolation. Neither gender was disproportionately affected by any hepatic lesion category.
Renal lesion prevalence and distribution--Figure 8 presents site-specific prevalences of renal lesions detected in starry flounder. Higher prevalences of necrotic and sclerotic lesions are generally shown at the urban sampling sites, while low prevalences of proliferative lesions were found at all of the sampling sites. Higher prevalences of necrotic lesions, as compared to those at Bodega Bay, were shown for the San Pablo Bay, Southampton Shoal and Hunters Point sites within San Francisco Bay, as well as at the Coos Bay, the Columbia River Estuary, and Youngs Bay sites. Sclerotic lesions were higher only in flounder from the Southampton Shoal site in San Francisco Bay.
Risk factor analyses for renal lesions in individual fish--Increased relative risks in two diagnostic categories of renal lesions were associated with residence at a number of the sampling sites, primarily at sites within urban embayments (Table 9). Necrotic renal lesions were significantly more likely to affect flounder from Youngs Bay, Columbia River Estuary, and Hunters Point and Southampton Shoal within San Francisco Bay. Increased relative risks of sclerotic lesions were also demonstrated at Southampton Shoal and the Columbia River Estuary. No sites with increased relative risk for proliferative lesions were identified. In parallel with the results for English sole, age was also a risk factor for both the necrotic and sclerotic kidney lesions in starry flounder. Gender was not a risk factor for any kidney lesion in this species.
Hepatic lesion prevalence and distribution--Hepatic lesions other than necrosis were detected at very low prevalences (< 3%) among the six sites on the coast of southern California where sufficient numbers of hornyhead turbot were captured (Fig. 9). The prevalence of necrosis ranged from about 3% at the west Santa Monica Bay site to 12% at the San Pedro Canyon site. No hepatic lesions at any of the sites showed higher prevalences than those detected at the reference site at Dana Point. In fact, foci of cellular alteration, neoplasms, and hydropic vacuolation were detected at low prevalences in turbot from this site; the only other site where foci of cellular alteration were detected was west Santa Monica Bay.
Risk factor analyses for hepatic lesions in individual fish--The only risk factors identified for hepatic lesions were age, which was associated with an increased risk of neoplasms and necrotic lesions, and capture at the outer San Diego Bay site, which was associated with increased risk of necrosis (Table 10). Gender was not a risk factor for any hepatic lesion category.
Renal lesion prevalence and distribution--Prevalences of all lesion types in hornyhead turbot were generally 5% or less at all sampling sites, and the geographical distribution of prevalences for each of the three lesion types was statistically homogeneous (Fig.10).
Risk factor analyses for renal lesions in individual fish--Neither site of capture, age, nor gender were significant risk factors for any of the renal lesions in the 387 individual hornyhead turbot composing the data set analyzed (data not shown).
Because of the broad geographic distribution of this species, sampling sites were separately grouped into northern and southern zones for the purposes of data display. Lesion prevalences at sites of capture for white croaker are shown in Figures 11A (northern sites including Bodega Bay and all sites within San Francisco Bay) and 11B (southern sites in the Los Angeles vicinity south to San Diego Bay).
Hepatic lesion prevalence and distribution--All six hepatic lesion categories were detected in white croaker captured among the eight northern sampling sites, generally at prevalences less than 10% (Fig. 11A). Lesion prevalences at the reference site of Bodega Bay were very low, with foci of cellular alteration, hydropic vacuolation, and necrosis not found at this site. The only neoplasms (both cholangiocellular carcinomas) diagnosed among the northern sites were in croaker from Bodega Bay; however, the two affected fish were at least 12 years of age (Varanasi et al. 1989a). Several lesion types occurred at higher prevalences at urban sites than at the reference site: the prevalence of foci of cellular alteration, though less than 3%, was elevated at both urban sites of Hunters Point and Redwood City; proliferative lesions and SDN were higher in croaker from the Oakland Estuary; hydropic vacuolation was higher at all of the urban sites within San Francisco Bay except Redwood City, with the highest prevalence found at Castro Creek; and necrosis was higher at San Pablo Bay, Castro Creek, Southampton Shoal, Hunters Point, and Oakland Estuary.
All six hepatic lesion categories were also diagnosed in croaker captured among the eight southern sampling sites (Fig. 11B), with prevalences of lesions other than SDN ranging no higher than 15% at any site. Prevalences at the reference site of Dana Point were low, and neoplasms, foci of cellular alteration, proliferative lesions, and necrosis were not detected in croaker sampled here. Neoplasms were found at low (<5%), but significantly higher prevalences in croaker from San Pedro Outer Harbor, Cerritos Channel, and Long Beach; prevalences of foci of cellular alteration (<5%) were higher only at the latter two sites. Proliferative lesions were higher at the west Harbor Island site in San Diego Bay, Cerritos Channel, and at the Outer Harbor, and Seal Beach sites in San Pedro Bay. Higher prevalences of SDN were detected at Cerritos Channel (~35%) and at San Pedro Outer Harbor (~17%). In contrast to the pattern shown for the northern sites, hydropic vacuolation was rarely detected in croaker from the southern sites, and there were no sites showing prevalences higher than those at the reference site. Prevalences of necrosis were higher at four of the eight southern sampling sites. Sites showing the most lesion categories with elevated prevalences were Cerritos Channel (5 lesion types) and San Pedro Outer Harbor (4 lesion types), both in the Los Angeles vicinity. Within San Diego Bay, croaker from the west Harbor Island site appeared to be the most frequently impacted by hepatic lesions.
Risk factor analyses for hepatic lesions in individual fish--Risk factors associated with lesion occurrence in white croaker and their calculated odds ratios/relative risks are presented in Table 11. Odds ratios for these risk factors (site of capture, age, and gender) are interpreted relative to the combined data for the northern and southern reference sites at Bodega Bay and Dana Point. All lesion types, with the exceptions of SDN and necrosis, showed a higher risk of lesion occurrence with increasing fish age. Aside from neoplasms, where males were more likely to be affected, no other lesion category disproportionately affected either gender. Capture at many of the urban sampling sites, especially in the Los Angeles vicinity and San Francisco Bay, was associated with increased risk of hepatic disease. For example, croaker from San Pedro Outer Harbor were 12.4 times more likely to be affected by neoplasms, 26.7 times more likely to be affected by proliferative lesions, 11.8 times more likely to have SDN, and 9.2 times more likely to have necrotic lesions than a comparable fish from the reference sites. A fourteenfold increase in the risk of foci of cellular alteration was associated with capture at the highly contaminated Cerritos Channel. Capture at Cerritos Channel also dramatically increased the risk of proliferative lesions, SDN, and necrosis. A substantially increased risk of hepatic lesion occurrence was also linked with residence at another highly contaminated urban site, the Oakland Estuary; fish from this site were far more likely to have proliferative lesions, SDN, hydropic vacuolation, and necrotic lesions. Other sites showing increased risk of hepatic disease were Long Beach (SDN and hydropic vacuolation), Castro Creek (hydropic vacuolation), and San Pedro Canyon and Hunters Point (necrosis).
Renal lesion prevalence and distribution--Among the northern sampling sites, kidney lesion prevalences in white croaker were quite low, generally below 5% for a particular lesion category. Proliferative and sclerotic lesions were especially rare, except at the Oakland and Southampton Shoal sites, respectively, where higher prevalences than at the reference site (Bodega Bay) were found (Fig. 12A).
Kidney lesions were detected more frequently among the southern sampling sites, but they were still lower than 10% at most sites (Fig. 12B). Prevalences higher than at the reference site (Dana Point) were evident for only proliferative lesions at San Pedro Outer Harbor, Seal Beach and San Pedro Canyon. Prevalences of other renal lesions were distributed homogeneously among the sites.
Risk factor analyses for renal lesions in individual fish--Risk factors associated with renal lesion occurrence and their respective odds ratios/relative risks are presented in Table 12. Odds ratios for the site of capture were interpreted relative to the combined prevalence data for croaker from Bodega Bay and Dana Point. No lesion type disproportionately affected either gender, and age was a risk factor only for the necrotic lesions. Capture at a number of the sites was associated with an increased relative risk of renal disease. Specifically, fish from Long Beach and San Pedro Outer Harbor showed a higher relative risk for necrotic lesions; croaker from San Pedro Canyon, Oakland, and the two former sites showed dramatically increased relative risks for proliferative lesions.
Hepatic lesion prevalence and distribution--Prevalences of hepatic lesions in black croaker are presented in Figure 13. Sufficient numbers of fish were captured at only three sites; two in San Diego Bay and one outside Mission Bay (reference). Well-differentiated nodular lesions that were histologically classified as benign hepatocellular adenomas (neoplasms), as well as foci of cellular alteration (McCain et al. 1992) were detected at significantly higher prevalences at the north San Diego and Shelter Island sites as compared to the reference site, where neither lesion type was found. Proliferative lesions were also present at higher prevalence at the north San Diego site than at outer Mission Bay. Neither SDN nor hydropic vacuolation were detected in black croaker.
Risk factor analyses for hepatic lesions in individual fish--In the sample of 196 black croaker analyzed by logistic regression, the only liver lesion associated with any risk factor was neoplasms; black croaker from the north San Diego site showed an increased relative risk of having a hepatocellular adenoma, relative to comparable fish from outer Mission Bay, of 15.570 (p = 0.009, grand mean = 0.917E-2)). Neither age (as estimated by length) nor gender were risk factors for any detected lesion category.
Renal lesion prevalence and distribution--The only kidney lesions diagnosed in black croaker were within the necrosis category, with detection at very low prevalences (<5%), and only in fish from the Shelter Island and north San Diego Bay sites (Fig. 14). Croaker from the reference site at outer Mission Bay were unaffected.
Risk factor analyses for renal lesions in individual fish--The only risk factor associated with kidney necrosis in this sample of black croaker (n = 185) was age, as estimated by length (p = 0.02, grand mean = 0.432E-5). The relative risk for this lesion type increased by 1.036 times for each additional millimeter in total length of the fish. After adjusting for age (length) and gender, no sites were identified as having increased relative risks for this lesion category. Consequently, the higher prevalences of necrotic lesions detected in fish from the Shelter Island and north San Diego Bay sites were interpreted primarily as being due to the sampling of larger fish.