Three-week-old weaned and colostrum-deprived neonatal (< 1 day old) pigs were inoculated to determine the pathogenicity of 2 enterotoxigenic Escherichia coli isolates that do not express K88, K99, F41, or 987P adhesins (strains 2134 and 2171). Strains 2134 and 2171 were isolated from pigs that had diarrhea after weaning attributable to enterotoxigenic E coli infection. We found that both strains of E coli adhered in the ileum and caused diarrhea in pigs of both age groups. In control experiments, adherent bacteria were not seen in the ileum of pigs < 1 day old or 3 weeks old that were noninoculated or inoculated with a nonpathogenic strain of E coli. These control pigs did not develop diarrhea. Antisera raised against strains 2134 and 2171 and absorbed with the autologous strain, grown at 18 C, were used for bacterial-agglutination and colony-immunoblot assays. Both absorbed antisera reacted with strains 2134 and 2171, but not with strains that express K99, F41, or 987P adhesins. A cross-reaction was observed with 2 wild-type K88 strains, but not with a K12 strain that expresses K88 pili. Indirect immunofluorescence with these absorbed antisera revealed adherent bacteria in frozen sections of ileum from pigs infected with either strain. We concluded that these strains are pathogenic and express a common surface antigen that may be a novel adhesin in E coli strains that cause diarrhea in weaned pigs.

Three-week-old weaned and colostrum-deprived neonatal (< 1 day old) pigs were inoculated to determine the pathogenicity of 2 enterotoxigenic Escherichia coli isolates that do not express K88, K99, F41, or 987P adhesins (strains 2134 and 2171). Strains 2134 and 2171 were isolated from pigs that had diarrhea after weaning attributable to enterotoxigenic E coli infection. We found that both strains of E coli adhered in the ileum and caused diarrhea in pigs of both age groups. In control experiments, adherent bacteria were not seen in the ileum of pigs < 1 day old or 3 weeks old that were noninoculated or inoculated with a nonpathogenic strain of E coli. These control pigs did not develop diarrhea. Antisera raised against strains 2134 and 2171 and absorbed with the autologous strain, grown at 18 C, were used for bacterial-agglutination and colony-immunoblot assays. Both absorbed antisera reacted with strains 2134 and 2171, but not with strains that express K99, F41, or 987P adhesins. A cross-reaction was observed with 2 wild-type K88 strains, but not with a K12 strain that expresses K88 pili. Indirect immunofluorescence with these absorbed antisera revealed adherent bacteria in frozen sections of ileum from pigs infected with either strain. We concluded that these strains are pathogenic and express a common surface antigen that may be a novel adhesin in E coli strains that cause diarrhea in weaned pigs.

Live, avirulent Escherichia coli vaccine strains were constructed and tested for efficacy in preventing colibacillosis in 4-week-old pigs. Either or both of 2 plasmids were inserted into avirulent E coli strain G58-1 (0101:NM). These plasmids were pPMC4, which encodes for LTb subunits of heat-labile enterotoxin, and pDHF1, which encodes for K88ac fimbriae. Litter- and weight-matched pigs were removed from sows when they were 10 days old and vaccinated orally with the constructed strains or with G58-1 (negative control vaccine) when they were 2 weeks old and 5 days later. All pigs were challenge-inoculated with virulent E coli strain 3030-2 (0157:K88, LT+, STb+) 2 weeks after the first vaccination. Only 1 pig vaccinated with G58-1/pPMC4/pDHF1 developed diarrhea and none died following challenge inoculation. Seventeen of 31 control pigs developed diarrhea and 11 died. Of 18 pigs vaccinated with G58-1/pDHF1 then challenge-inoculated with the virulent strain, 5 developed diarrhea and 2 died. Fifteen of 18 litter- and weight-matched controls developed diarrhea and 8 died. When compared with G58-1 (negative control), G58-1/pPMC4 afforded no protection to pigs challenge-inoculated with 3030-2.

Live, avirulent Escherichia coli vaccine strains were constructed and tested for efficacy in preventing colibacillosis in 4-week-old pigs. Either or both of 2 plasmids were inserted into avirulent E coli strain G58-1 (0101:NM). These plasmids were pPMC4, which encodes for LTb subunits of heat-labile enterotoxin, and pDHF1, which encodes for K88ac fimbriae. Litter- and weight-matched pigs were removed from sows when they were 10 days old and vaccinated orally with the constructed strains or with G58-1 (negative control vaccine) when they were 2 weeks old and 5 days later. All pigs were challenge-inoculated with virulent E coli strain 3030-2 (0157:K88, LT+, STb+) 2 weeks after the first vaccination. Only 1 pig vaccinated with G58-1/pPMC4/pDHF1 developed diarrhea and none died following challenge inoculation. Seventeen of 31 control pigs developed diarrhea and 11 died. Of 18 pigs vaccinated with G58-1/pDHF1 then challenge-inoculated with the virulent strain, 5 developed diarrhea and 2 died. Fifteen of 18 litter- and weight-matched controls developed diarrhea and 8 died. When compared with G58-1 (negative control), G58-1/pPMC4 afforded no protection to pigs challenge-inoculated with 3030-2.

The vaccine efficacy of a genetically engineered deletion mutant strain of pseudorabies virus, strain 783, was compared with that of the conventionally attenuated Bartha strain. Strain 783 has deletions in the genes coding for glycoprotein I and thymidine kinase. In experiment 1, which had a 3-month interval between vaccination and challenge exposure, strain 783 protected pigs significantly (P < 0.05) better against virulent virus challenge exposure than did the Bartha strain. The growth of pigs vaccinated with strain 783 was not arrested, whereas that of pigs vaccinated with the Bartha strain was arrested for 7 days. Of 8 pigs given strain 783, 4 were fully protected against challenge exposure; none of the pigs given strain Bartha was fully protected. In experiment 2, which had a 3-week interval between vaccination and challenge exposure, the growth of pigs vaccinated with strain 783 was arrested for 3.5 days, whereas that of pigs vaccinated with the Bartha strain was arrested for 6 days. In experiment 3, pigs with moderate titer of maternal antibodies were vaccinated twice IM or once intranasally with either strain 783 or Bartha and were challenge-exposed 3 months after vaccination. Pigs given strain 783 twice IM were significantly (P < 0.05) better protected than were the other pigs. They had growth arrest of only 6 days, compared with 9 days for pigs of other groups, and shed less virus after challenge exposure. Results of this study indicate that the vaccine based on the deletion mutant strain 783 is more efficacious than is the Bartha strain of pseudorabies virus.

The vaccine efficacy of a genetically engineered deletion mutant strain of pseudorabies virus, strain 783, was compared with that of the conventionally attenuated Bartha strain. Strain 783 has deletions in the genes coding for glycoprotein I and thymidine kinase. In experiment 1, which had a 3-month interval between vaccination and challenge exposure, strain 783 protected pigs significantly (P < 0.05) better against virulent virus challenge exposure than did the Bartha strain. The growth of pigs vaccinated with strain 783 was not arrested, whereas that of pigs vaccinated with the Bartha strain was arrested for 7 days. Of 8 pigs given strain 783, 4 were fully protected against challenge exposure; none of the pigs given strain Bartha was fully protected. In experiment 2, which had a 3-week interval between vaccination and challenge exposure, the growth of pigs vaccinated with strain 783 was arrested for 3.5 days, whereas that of pigs vaccinated with the Bartha strain was arrested for 6 days. In experiment 3, pigs with moderate titer of maternal antibodies were vaccinated twice IM or once intranasally with either strain 783 or Bartha and were challenge-exposed 3 months after vaccination. Pigs given strain 783 twice IM were significantly (P < 0.05) better protected than were the other pigs. They had growth arrest of only 6 days, compared with 9 days for pigs of other groups, and shed less virus after challenge exposure. Results of this study indicate that the vaccine based on the deletion mutant strain 783 is more efficacious than is the Bartha strain of pseudorabies virus.

The area of skin supplied by the afferent fibers in one cutaneous nerve is called the cutaneous area (CA) for that nerve. The CA of peripheral branches of lumbar and sacral spinal nerves responsive to the stimulation of hair follicle mechanoreceptors were mapped in 27 dogs. The amount of overlap among the CA was similar to that found for other CA of the body. The CA of peripheral branches of the sciatic nerve were restricted to the lateral, cranial, and caudal aspects of the pelvic limb distal to the stifle. The CA of the saphenous nerve was located on the medial side of the limb, except for a small area located on the lateral side of the crus. The distal part of the CA of the saphenous nerve was completely overlapped in the hind paw by branches of the superficial peroneal nerve laterally and the medial plantar branch of the tibial nerve medially. The CA for the deep peroneal nerve was located on the dorsal surface of the webbing between digits 2 and 3 and the adjacent skin of these digits. The CA of the plantar branches of the tibial nerve were small in comparison with the diameter of the nerve, suggesting that these branches contained nerve fibers supplying other, deeper structures in the hindpaw and that damage to these nerves would interfere with cutaneous sensation in only a small region on the plantar surface of the hindpaw. Knowledge of the CA of the various branches of the sciatic nerve allows more accurate localization of injury to the sciatic nerve or its branches by using areas of anesthesia.

The area of skin supplied by the afferent fibers in one cutaneous nerve is called the cutaneous area (CA) for that nerve. The CA of peripheral branches of lumbar and sacral spinal nerves responsive to the stimulation of hair follicle mechanoreceptors were mapped in 27 dogs. The amount of overlap among the CA was similar to that found for other CA of the body. The CA of peripheral branches of the sciatic nerve were restricted to the lateral, cranial, and caudal aspects of the pelvic limb distal to the stifle. The CA of the saphenous nerve was located on the medial side of the limb, except for a small area located on the lateral side of the crus. The distal part of the CA of the saphenous nerve was completely overlapped in the hind paw by branches of the superficial peroneal nerve laterally and the medial plantar branch of the tibial nerve medially. The CA for the deep peroneal nerve was located on the dorsal surface of the webbing between digits 2 and 3 and the adjacent skin of these digits. The CA of the plantar branches of the tibial nerve were small in comparison with the diameter of the nerve, suggesting that these branches contained nerve fibers supplying other, deeper structures in the hindpaw and that damage to these nerves would interfere with cutaneous sensation in only a small region on the plantar surface of the hindpaw. Knowledge of the CA of the various branches of the sciatic nerve allows more accurate localization of injury to the sciatic nerve or its branches by using areas of anesthesia.

Canine erythrocytes were loaded with the antineoplastic drug doxorubicin and then treated with 0.16% glutaraldehyde. This procedure has been previously shown to slow down the efflux of doxorubicin from erythrocytes and to result in the selective targeting of the carrier erythrocytes to liver. Three dogs were treated each with 2 different schedules of IV bolus administration of doxorubicin (0.4 mg/kg of body weight): free drug and doxorubicin encapsulated in glutaraldehyde-treated erythrocytes. The 2 treatments yielded consistent differences in the plasma pharmacokinetic properties of doxorubicin and of its only metabolite, doxorubicinol. A triphasic exponential decay of doxorubicin plasma concentrations was observed on injection of the free drug. Conversely, in the case of erythrocyte-encapsulated doxorubicin, 4 phases of plasma concentrations of doxorubicin were found. The plasma concentrations of doxorubicinol, after a steady increase during the first hour, followed patterns of decay comparable to those of the parent drug. On the basis of the kinetic variables calculated with the 2 administration schedules, area under curve concentrations of plasma doxorubicin were 136 microgram.h/L (free infusion) and 734 microgram.h/L erythrocyte-encapsulated drug). Significant alterations of hematologic and hematochemical factors were not observed in the 3 dogs during and after the 2 treatments. On the basis of our findings, doxorubicin-loaded and glutaraldehyde-treated erythrocytes may potentially be used in the treatment of systemic and hepatic tumors in dogs.

Canine erythrocytes were loaded with the antineoplastic drug doxorubicin and then treated with 0.16% glutaraldehyde. This procedure has been previously shown to slow down the efflux of doxorubicin from erythrocytes and to result in the selective targeting of the carrier erythrocytes to liver. Three dogs were treated each with 2 different schedules of IV bolus administration of doxorubicin (0.4 mg/kg of body weight): free drug and doxorubicin encapsulated in glutaraldehyde-treated erythrocytes. The 2 treatments yielded consistent differences in the plasma pharmacokinetic properties of doxorubicin and of its only metabolite, doxorubicinol. A triphasic exponential decay of doxorubicin plasma concentrations was observed on injection of the free drug. Conversely, in the case of erythrocyte-encapsulated doxorubicin, 4 phases of plasma concentrations of doxorubicin were found. The plasma concentrations of doxorubicinol, after a steady increase during the first hour, followed patterns of decay comparable to those of the parent drug. On the basis of the kinetic variables calculated with the 2 administration schedules, area under curve concentrations of plasma doxorubicin were 136 microgram.h/L (free infusion) and 734 microgram.h/L erythrocyte-encapsulated drug). Significant alterations of hematologic and hematochemical factors were not observed in the 3 dogs during and after the 2 treatments. On the basis of our findings, doxorubicin-loaded and glutaraldehyde-treated erythrocytes may potentially be used in the treatment of systemic and hepatic tumors in dogs.

Epidemiologic data were evaluated from all dogs admitted to the University of Minnesota, Veterinary Teaching Hospital (UMVTH) between June 1981 and November 1989. Of 69,890 admissions, 2,077 were Miniature Schnauzers. Uroliths were retrieved from 63 of the 2,077 Miniature Schnauzers admitted. In 20 of the 63 urolith episodes, calcium oxalate was the predominant mineral identified. By comparison, calcium oxalate uroliths were identified in only 56 of the remaining 67,813 non-Miniature Schnauzer canine admissions. The odds that uroliths from Miniature Schnauzers were composed of calcium oxalate was 11.8 times greater than for other canine breeds evaluated at the UMVTH (95% confidence interval = 6.8 to 20.1). Data also were evaluated from files of uroliths retrieved from dogs and submitted to the Minnesota Urolith Center for quantitative mineral analysis between June 1981 and November 1989. Of 3,930 uroliths analyzed, 615 (15.6%) uroliths were obtained from Miniature Schnauzers. Of the 615 uroliths, 175 (28.4%) were calcium oxalate. By comparison, only 550 (16.6%) of the remaining 3,315 from dogs of breeds other than Miniature Schnauzers were calcium oxalate. The odds that uroliths submitted for analysis were composed of calcium oxalate was 2 times greater for Miniature Schnauzers than for dogs of other breeds (95% confidence interval = 1.6 to 2.4). Calcium oxalate uroliths were retrieved more frequently in males than females. The risk for males developing calcium oxalate uroliths was > 3 times the risk for females in both groups of data evaluated. The mean age of all Miniature Schnauzers admitted to the UMVTH with calcium oxalate uroliths was 9 years. Calcium oxalate uroliths were not detected in Miniature Schnauzers younger than 1.7 years.

Epidemiologic data were evaluated from all dogs admitted to the University of Minnesota, Veterinary Teaching Hospital (UMVTH) between June 1981 and November 1989. Of 69,890 admissions, 2,077 were Miniature Schnauzers. Uroliths were retrieved from 63 of the 2,077 Miniature Schnauzers admitted. In 20 of the 63 urolith episodes, calcium oxalate was the predominant mineral identified. By comparison, calcium oxalate uroliths were identified in only 56 of the remaining 67,813 non-Miniature Schnauzer canine admissions. The odds that uroliths from Miniature Schnauzers were composed of calcium oxalate was 11.8 times greater than for other canine breeds evaluated at the UMVTH (95% confidence interval = 6.8 to 20.1). Data also were evaluated from files of uroliths retrieved from dogs and submitted to the Minnesota Urolith Center for quantitative mineral analysis between June 1981 and November 1989. Of 3,930 uroliths analyzed, 615 (15.6%) uroliths were obtained from Miniature Schnauzers. Of the 615 uroliths, 175 (28.4%) were calcium oxalate. By comparison, only 550 (16.6%) of the remaining 3,315 from dogs of breeds other than Miniature Schnauzers were calcium oxalate. The odds that uroliths submitted for analysis were composed of calcium oxalate was 2 times greater for Miniature Schnauzers than for dogs of other breeds (95% confidence interval = 1.6 to 2.4). Calcium oxalate uroliths were retrieved more frequently in males than females. The risk for males developing calcium oxalate uroliths was > 3 times the risk for females in both groups of data evaluated. The mean age of all Miniature Schnauzers admitted to the UMVTH with calcium oxalate uroliths was 9 years. Calcium oxalate uroliths were not detected in Miniature Schnauzers younger than 1.7 years.

To evaluate underlying causes of calcium oxalate urolithiasis, 24-hour excretion of urine metabolites was measured in 6 Miniature Schnauzers that formed calcium oxalate (CaOx) uroliths during periods when they were fed a standard diet and during periods when food was withheld. Serum concentrations of parathyroid hormone and 1,25-dihydroxyvitamin D also were evaluated. Serum calcium concentrations were normal in all 6 affected Miniature Schnauzers; however, during diet consumption, mean 24-hour urinary excretion of calcium was significantly (P = 0.025) higher than calcium excretion when food was withheld. In 1 dog, urinary calcium excretion was lower during the period of food consumption, compared with the period when food was withheld. Compared with clinically normal Beagles, Miniature Schnauzers that formed CaOx uroliths excreted significantly greater quantities of calcium when food was consumed (P = 0.0004) and when food was withheld (P = 0.001). Miniature Schnauzers that formed CaOx uroliths excreted significantly less oxalate than clinically normal Beagles during fed (P = 0.028) and nonfed (P = 0.004) conditions. Affected Miniature Schnauzers also excreted abnormally high quantities of uric acid. Excretion of citrate was not different between Miniature Schnauzers with CaOx urolithiasis and clinically normal Beagles. In 5 of 6 Miniature Schnauzers with CaOx urolithiasis, concentrations of serum parathyroid hormone were similar to values from age- and gender-matched Miniature Schnauzers without uroliths. The concentration of serum parathyroid hormone in 1 dog was > 4 times the mean concentration of clinically normal Miniature Schnauzers. Mean serum concentrations of 1,25-dihydroxyvitamin D in Miniature Schnauzers with calcium oxalate urolithiasis were similar to concentrations of clinically normal Miniature Schnauzers.

To evaluate underlying causes of calcium oxalate urolithiasis, 24-hour excretion of urine metabolites was measured in 6 Miniature Schnauzers that formed calcium oxalate (CaOx) uroliths during periods when they were fed a standard diet and during periods when food was withheld. Serum concentrations of parathyroid hormone and 1,25-dihydroxyvitamin D also were evaluated. Serum calcium concentrations were normal in all 6 affected Miniature Schnauzers; however, during diet consumption, mean 24-hour urinary excretion of calcium was significantly (P = 0.025) higher than calcium excretion when food was withheld. In 1 dog, urinary calcium excretion was lower during the period of food consumption, compared with the period when food was withheld. Compared with clinically normal Beagles, Miniature Schnauzers that formed CaOx uroliths excreted significantly greater quantities of calcium when food was consumed (P = 0.0004) and when food was withheld (P = 0.001). Miniature Schnauzers that formed CaOx uroliths excreted significantly less oxalate than clinically normal Beagles during fed (P = 0.028) and nonfed (P = 0.004) conditions. Affected Miniature Schnauzers also excreted abnormally high quantities of uric acid. Excretion of citrate was not different between Miniature Schnauzers with CaOx urolithiasis and clinically normal Beagles. In 5 of 6 Miniature Schnauzers with CaOx urolithiasis, concentrations of serum parathyroid hormone were similar to values from age- and gender-matched Miniature Schnauzers without uroliths. The concentration of serum parathyroid hormone in 1 dog was > 4 times the mean concentration of clinically normal Miniature Schnauzers. Mean serum concentrations of 1,25-dihydroxyvitamin D in Miniature Schnauzers with calcium oxalate urolithiasis were similar to concentrations of clinically normal Miniature Schnauzers.

In mammalian species studied previously, pepsinogen consisted of biochemically different groups of isozymogens. By use of gel filtration chromatography and electrophoresis, we isolated a predominant pepsinogen from the gastric mucosa of a horse. Peptide mapping with V8 protease revealed differences with its porcine homologue. However, porcine and equine pepsinogens, when activated to pepsin, had a similar pattern of activity when hemoglobin was used as substrate. Those results suggest that differences must exist in the primary structure of the pepsinogens of the 2 species.

In mammalian species studied previously, pepsinogen consisted of biochemically different groups of isozymogens. By use of gel filtration chromatography and electrophoresis, we isolated a predominant pepsinogen from the gastric mucosa of a horse. Peptide mapping with V8 protease revealed differences with its porcine homologue. However, porcine and equine pepsinogens, when activated to pepsin, had a similar pattern of activity when hemoglobin was used as substrate. Those results suggest that differences must exist in the primary structure of the pepsinogens of the 2 species.

The thiamylal sparing effect of midazolam was studied in 30 healthy Beagle and mixed-breed dogs. Using a replicated Latin square design, all dogs were given placebo (saline solution) and 0.025, 0.05, 0.1, and 0.2 mg of midazolam/kg of body weight prior to IV administration of thiamylal sodium. The 0.1 and 0.2 mg/kg dosages significantly decreased the amount of thiamylal required to obtund swallowing reflex and easily achieve endotracheal intubation. Midazolam at 0.1 and 0.2 mg/kg reduced thiamylal requirement by 16.4% and 18.9%, respectively, whereas the 0.05 mg/kg dosage decreased thiamylal requirement by only 6.8%. The 0.2 mg/kg dosage did not further decrease thiamylal requirement beyond that achieved with the 0.1 mg/kg dosage of midazolam. This study demonstrates that the preanesthetic IV administration of midazolam reduces the thiamylal dose necessary to accomplish intubation. The optimal preanesthetic dosage (lowest dosage with significant effect) was 0.1 mg/kg.

The thiamylal sparing effect of midazolam was studied in 30 healthy Beagle and mixed-breed dogs. Using a replicated Latin square design, all dogs were given placebo (saline solution) and 0.025, 0.05, 0.1, and 0.2 mg of midazolam/kg of body weight prior to IV administration of thiamylal sodium. The 0.1 and 0.2 mg/kg dosages significantly decreased the amount of thiamylal required to obtund swallowing reflex and easily achieve endotracheal intubation. Midazolam at 0.1 and 0.2 mg/kg reduced thiamylal requirement by 16.4% and 18.9%, respectively, whereas the 0.05 mg/kg dosage decreased thiamylal requirement by only 6.8%. The 0.2 mg/kg dosage did not further decrease thiamylal requirement beyond that achieved with the 0.1 mg/kg dosage of midazolam. This study demonstrates that the preanesthetic IV administration of midazolam reduces the thiamylal dose necessary to accomplish intubation. The optimal preanesthetic dosage (lowest dosage with significant effect) was 0.1 mg/kg.

The association between ocular squamous cell tumors and various measures of solar radiation was evaluated for 2,775 contemporaneous Hereford cattle. The animals were from 34 herds located in 21 states in the United States and in one Canadian province. Examinations and photographs were made of the eyes of all cows and heifers in a herd; 33 herds were examined between June and August 1957 and 1 herd was examined in 1958. Solar radiation was measured by altitude and mean annual hours of sunshine applying to an individual herd, and global radiation, a cloudiness index, and elevation of nearby National Weather Service stations. All measures of solar radiation provided evidence of a significant association between increasing risks of developing cancer eye and increasing levels of radiation; adjustments were made for possible confounding effects of age and corneoscleral pigmentation. Associations were evident whether affliction was defined as the occurrence of any type of tumor (ie, plaque, papilloma, or carcinoma), or as the occurrence of only papilloma or carcinoma. Average ages of affected cattle also tended to be lower at high radiation levels than at low levels, further supporting the hypothesis that solar radiation has an important role in the development of this disease. In view of the heritability of eye pigmentation and its inhibitory effect on lesions, cancer eye may be controlled by selective breeding for increased pigmentation in and around the eye to afford protection against the harmful effects of incident UV radiation.

The association between ocular squamous cell tumors and various measures of solar radiation was evaluated for 2,775 contemporaneous Hereford cattle. The animals were from 34 herds located in 21 states in the United States and in one Canadian province. Examinations and photographs were made of the eyes of all cows and heifers in a herd; 33 herds were examined between June and August 1957 and 1 herd was examined in 1958. Solar radiation was measured by altitude and mean annual hours of sunshine applying to an individual herd, and global radiation, a cloudiness index, and elevation of nearby National Weather Service stations. All measures of solar radiation provided evidence of a significant association between increasing risks of developing cancer eye and increasing levels of radiation; adjustments were made for possible confounding effects of age and corneoscleral pigmentation. Associations were evident whether affliction was defined as the occurrence of any type of tumor (ie, plaque, papilloma, or carcinoma), or as the occurrence of only papilloma or carcinoma. Average ages of affected cattle also tended to be lower at high radiation levels than at low levels, further supporting the hypothesis that solar radiation has an important role in the development of this disease. In view of the heritability of eye pigmentation and its inhibitory effect on lesions, cancer eye may be controlled by selective breeding for increased pigmentation in and around the eye to afford protection against the harmful effects of incident UV radiation.