Table 4.

Effect of α-tocopherol on prostate carcinogenesis and nonneoplastic prostate lesions in 17β-estradiol plus testosterone–treated NBL rats

Group123
α-Tocopherol dose (mg/kg diet)02,0004,000
Effective no. of animals293030
Mortality (wk after start of hormone treatment)
    Mean wk of death ± SD42 ± 7.345 ± 6.642 ± 5.4
    Median wk of death (95% confidence interval)42 (39,45)43 (42,47)41 (39,43)
    Range (minimum to maximum)27-5428-5227-54
Lesion incidenceNo. (%) of rats with lesion
Adenocarcinoma
    Total cancer incidence26 (90%)26 (87%)25 (83%)
        Periurethral dorsolateral—anterior prostatic ducts alone22 (76)26 (87)24 (80)
        Distal anterior prostate ducts alone17 (59)22 (73)16 (53)
        Dorsolateral prostate lobe (glandular portion)1 (3)1 (3)0
    Tumor multiplicity
        One or two carcinomas9 (31)6 (20)11 (37)
        More than two carcinomas17 (57)20 (67)14 (47)
    Carcinoma in situ
        Anterior prostate (glands & ducts)1 (3)01 (3)
        Seminal vesicle (glands & ducts)5 (17)7 (23)4 (13)
Dysplasia in anterior prostate/seminal vesicle (no. of animals with lesion/evaluable animals)
    Anterior prostate (glands & ducts)6/20 (30)12/23 (52)12/26 (46)
    Seminal vesicle (glands & ducts)3/20 (15)2/23 (9)3/26 (12)
Inflammation & dysplasia in lateral prostate
Inflammation (no. of animals with lesion/evaluable animals)
    Total incidence29/29 (100)30/30 (100)30/30 (100)
        Marked inflammation5 (17)*3 (10)*0*
        Slight to moderate inflammation24 (83)*27 (90)*30 (100)*
Dysplasia associated with inflammation (no. of animals with lesion/evaluable animals)
    Total incidence23/23 (100)26/26 (100)26/26 (100)
        Marked dysplasia19 (83)15 (58)10 (38)
        Slight to moderate dysplasia4 (17)11 (42)16 (62)
Dysplasia not associated with inflammation (no. animals with lesion/evaluable animals)
    Total incidence21/22 (100)17/22 (100)17/26 (100)
        Marked dysplasia8 (36)1 (5)1 (4)
        Slight to moderate dysplasia13 (59)16 (72)16 (62)
        No dysplasia1 (5)5 (23)9 (34)

NOTE: Data were tested for significance using the Fisher's exact test and for linear trend with dose using a 2 × 3 χ2 test. Differences were considered statistically significant at P ≤ 0.05 (two-sided for Fisher's exact test). Note that for lesions that can only be assessed micropscopically, the number of evaluable animals (with nonautolytic tissue) was often lower than the effective number of animals; the prostate of one rat in the control group was lost to cannibalism, reducing the effective number of animals.

  • *P = 0.0667; Ptrend = 0.0205 (χ2 test).

  • P = 0.0074; Ptrend = 0.0018 (χ2 test).

  • P = 0.0017; Ptrend > 0.05, but 0.0017 when combining the slight to moderate and none categories (χ2 test).