Exposure to Agent Orange is a significant predictor of prostate-specific antigen (PSA)-based recurrence and a rapid PSA doubling time after radical prostatectomy

Authors

  • Sagar R. Shah,

    1. Sections of Urology, Augusta Veterans Affairs Medical Center, and
    2. Medical College of Georgia, Augusta, Georgia,
    Search for more papers by this author
  • Stephen J. Freedland,

    1. Department of Surgery, Durham Veterans Affairs Medical Center,
    2. Division of Urologic Surgery, Departments of Surgery and Pathology and Duke Prostate Center, Duke University School of Medicine, Durham, North Carolina,
    Search for more papers by this author
  • William J. Aronson,

    1. Urology Section, Department of Surgery, Veterans Affairs Greater Los Angeles Healthcare System,
    2. Department of Urology, University of California, Los Angeles, School of Medicine, Los Angeles,
    Search for more papers by this author
  • Christopher J. Kane,

    1. Division of Urology, Department of Surgery, University of California San Diego, San Diego,
    Search for more papers by this author
  • Joseph C. Presti Jr,

    1. Department of Urology, Stanford University School of Medicine, and
    2. Urology Section, Department of Surgery, Veterans Affairs Palo Alto Healthcare System, Palo Alto, California, and
    Search for more papers by this author
  • Christopher L. Amling,

    1. Department of Urology, University of Alabama at Birmingham, Birmingham, Alabama, USA
    Search for more papers by this author
  • Martha K. Terris

    1. Sections of Urology, Augusta Veterans Affairs Medical Center, and
    2. Medical College of Georgia, Augusta, Georgia,
    Search for more papers by this author

Martha K. Terris, Medical College of Georgia, 1120 Fifteenth Street, BA 8414, Augusta, GA 30912–4050, USA.
e-mail: mterris@mcg.edu

Abstract

OBJECTIVE

To investigate and report the clinicopathological characteristics and outcomes after radical prostatectomy (RP) in patients with prostate cancer and previous exposure to Agent Orange (AO), particularly in relationship to race.

PATIENTS AND METHODS

In 1495 veterans who had undergone RP the clinicopathological characteristics, biochemical progression rates, and prostate-specific antigen (PSA) doubling time (DT) after recurrence between AO-exposed and unexposed men were compared using logistic and linear regression and Cox proportional hazards analyses, and stratified by race.

RESULTS

The 206 (14%) men with AO exposure were more likely to be black (P = 0.001), younger (P < 0.001), treated more recently (P < 0.001), have a higher body mass index (P = 0.001), have clinical stage T1 disease (P < 0.001), and have lower preoperative PSA levels (P = 0.001). After adjusting for several clinical characteristics, AO exposure was not significantly related to adverse pathological features but was significantly associated with biochemical progression risk (relative risk 1.55, 95% confidence interval 1.15–2.09, P = 0.004) and shorter PSADT (P < 0.001) after recurrence (8.2 vs 18.6 months). When stratified by race, these associations were present and similar in both races, with no significant interaction between race and AO exposure for predicting biochemical recurrence or mean adjusted PSADT (P interaction >0.20).

CONCLUSIONS

Patients with AO exposure and treated with RP were more likely to be black, present with lower risk features, have an increased risk of biochemical progression, and shorter PSADT after recurrence. When stratified by race, the association between AO exposure and poor outcomes was present in both races. These findings suggest that among selected men who choose RP, AO exposure might be associated with more aggressive prostate cancer.

Abbreviations
AO

Agent Orange

RP

radical prostatectomy

DT

doubling time

VA

Veterans’ Affairs

SEARCH

Shared Equal Access Research Cancer Hospital

BMI

body mass index

RR

relative risk.

INTRODUCTION

Agent Orange (AO), a defoliant used during the Vietnam War, is a mixture of 2,4- dichloro-phenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid, and has been reported to be contaminated, from <0.05 to almost 50 parts per million with 2,3,7,8-tetrachlorodibenzo-p-dioxin (dioxin) [1]. In 1998, the National Academy of Science concluded ‘limited/suggestive evidence’ of an association between AO and prostate cancer [2]. Most of the data linking AO exposure to an increased risk of prostate cancer have come from farmers and forestry workers exposed to herbicides [3–5]. Initial studies on AO exposure and prostate cancer in Vietnam veterans yielded conflicting results and were limited by the lack of quantification of AO exposure [6–8]. A study in US Air Force veterans of the Vietnam era with confirmed exposure to AO showed a greater risk of prostate cancer in white veterans in the AO-exposed group, and in veterans serving in the same region, than in age-matched controls [9]. That study also reported a greater risk of prostate cancer in those white veterans with higher serum levels of dioxin than in age-matched controls and comparison veterans, lending further support to a direct causal relationship.

There is very little information on the clinical and pathological characteristics of prostate cancer in men with AO exposure. Therefore, our goal was to further characterize the clinicopathological features of prostate cancer biochemical progression rates, and the PSA doubling time (PSADT) after recurrence, in veterans with previous exposure to AO who had RP, and to determine if there is any variation with race.

PATIENTS AND METHODS

Institutional Review Board approval was obtained to abstract and combine data from patients treated with RP between 1988 and 2007 at the Shared Equal Access Research Cancer Hospital (SEARCH) database Veterans Affairs (VA) Health Care Facilities in West Los Angeles and Palo Alto, California, Augusta, Georgia, and Durham, North Carolina [10]. Data included patient age at RP, race, height, weight, clinical stage, grade of cancer on diagnostic biopsies, preoperative PSA levels, surgical specimen pathology (specimen weight, tumour grade, stage, and surgical margin status), and follow-up PSA levels. Patients treated with preoperative androgen deprivation or radiotherapy were excluded. Of the 1747 men enrolled in SEARCH, AO exposure status was available on 1592 (91%). We excluded men who were neither black nor white (94), because the few men in this group with AO exposure (five) prevented meaningful analyses. We also exclude three men with missing data for race, resulting in a study population of 1495. Previous AO exposure, a routine component of demographic data in VA healthcare facilities, was abstracted from the electronic medical records system. Body mass index (BMI) was calculated as usual. Biochemical progression was defined as one PSA level of >0.2 ng/mL, two of 0.2 ng/mL, or secondary treatment for an elevated PSA level after RP. The RP specimens were sectioned according to each institution’s protocol [10].

Patients were grouped by the presence or absence of exposure to AO. The distribution of clinical and pathological characteristics was compared between the groups using rank-sum analysis for continuous variables and the chi-squared test for categorical variables. Age at RP, preoperative PSA level and year of surgery were examined as continuous variables. Gleason grade (3 + 4, ≥4 + 3 vs 2–6), clinical stage (T2/3 vs T1), race (black vs white), and BMI (25.0–29.9, 30.0–34.9, ≥35.0 vs <25.0 kg/m2) were examined as categorical variables. Because PSA was not normally distributed, values were evaluated after logarithmic transformation. Because of our a priori hypothesis that the associations between AO and prostate cancer outcomes might vary by race, all analyses were also examined separately among black and white men. We tested for interactions between race and AO exposure by including both main-effects terms along with a cross-product term into the models, and evaluated the coefficient using the Wald test.

The association between AO exposure and the adverse pathological features of high-grade disease (Gleason grade ≥4 + 3), positive surgical margins, extracapsular extension, and seminal vesicle invasion was examined using a logistic regression analysis, mutually adjusting for the clinical variables of age at RP, race, preoperative PSA level, year of RP, biopsy Gleason grade, clinical stage, and centre.

Time to biochemical progression was compared for AO exposed and unexposed patients using Kaplan-Meier analysis and the log-rank test. The relative risk (RR) of biochemical progression associated with AO exposure was determined using a Cox proportional hazards regression model adjusting for age at RP, race, preoperative PSA level, year of RP, biopsy Gleason grade, clinical stage, and centre.

For men with a biochemical recurrence, PSADT was calculated by natural log 2 (0.693) divided by the slope of the linear regression line of natural log of all PSA values obtained within the first 2 years after biochemical recurrence. Patients (53) with no increase in PSA were assigned a PSADT of 100 months. All PSA values were obtained before any secondary therapy. Log-transformed PSADT was compared between men with and without AO exposure using linear regression analysis, adjusting for clinical and pathological characteristics. Back transformation to the geometric mean was used for ease of interpretation.

The association between AO, adverse pathology, and biochemical progression was not meaningfully affected by adjusting for BMI. Therefore, given that BMI data were only available on 1307 men (87%), BMI was not included in multivariate models. The distribution of clinicopathological variables was similar among the centres in the SEARCH Database therefore data were combined for analyses.

RESULTS

Of the 1495 men with available AO-exposure status, 206 (14%) had documented exposure to AO. Relative to men with no AO exposure, men with exposure to AO were more likely to be black (P = 0.001), be younger at the time of RP (P < 0.001), be treated in more recent years (P < 0.001), have a higher BMI (P = 0.001), more likely to have stage T1 disease (P < 0.001), and have a lower preoperative PSA level (P = 0.001) (Table 1). There were no significant differences between the groups in biopsy or pathological tumour grade. Associations between AO exposure and clinical and pathological characteristics were similar when men were examined separately by race.

Table 1.  The clinical and pathological features of men undergoing RP, segregated by their history of AO exposure
VariableExposure to AOP
NoYes
No. patients1289 206 
Mean (sd) age, years  62.0 (6.5)  58.8 (4.8)<0.001
Median year of surgery19992003<0.001
Race, n (%)   0.001
 White 728 (56)  91 (44) 
 Black 561 (44) 115 (56) 
BMI, kg/m2, n (%)   0.001
 <25.0 302 (27)  42 (22) 
 25.0–29.9 494 (44)  80 (42) 
 30.0–34.9 235 (21)  39 (20) 
 ≥35.0  84 (8)  31 (16) 
PSA, ng/mL   0.001
 Mean (sd)  10.3 (10.5)   8.4 (8.7) 
 Median   7.4   6.1 
n (%):
TNM clinical T stage  <0.001
 T1 591 (50) 136 (70) 
 T2/T3 583 (50)  59 (30) 
Biopsy Gleason score   0.84
 2–6 783 (63) 131 (64) 
 3 + 4 268 (21)  40 (20) 
 ≥4 + 3 200 (16)  33 (16) 
Pathological Gleason score   0.66
 2–6 504 (40)  84 (41) 
 3 + 4 469 (37)  69 (34) 
 ≥4 + 3 288 (23)  50 (25) 
Positive margins 642 (49) 104 (48) 0.86
Extracapsular extension 307 (25)  54 (27) 0.50
Seminal vesicle invasion 144 (12)  17 (8) 0.20
Lymph node metastasis  20 (2)   3 (1) 0.99

After adjusting for clinical variables, exposure to AO was not significantly associated with the odds of a pathological Gleason sum of ≥4 + 3 (P = 0.19), positive surgical margins (P = 0.45), extracapsular extension (P = 0.45), or seminal vesicle invasion (P = 0.99) (Table 2). The patterns were similar when black and white men were examined separately.

Table 2.  Multivariable logistic regression analysis of association between AO exposure and adverse pathological findings among men in the SEARCH database
VariableOdds ratio (95% CI)P
  1. Analysis were adjusted for preoperative PSA level, age at surgery, biopsy Gleason sum, year of surgery, clinical stage, race, and centre.

High-grade disease (Gleason sum ≥4 + 3)1.33 (0.86–2.06)0.19
Positive surgical margins1.15 (0.81–1.64)0.45
Extracapsular extension1.17 (0.78–1.74)0.45
Seminal vesicle invasion1.00 (0.51–1.94)0.99

During a median and mean (sd) follow-up of 49 and 60 (46) months, respectively, 501 (34%) men had biochemical progression. Men who were exposed to AO were more likely to progress than men who were not exposed, although this was not quite statistically significant (log-rank test, P = 0.07; Fig. 1). However, AO exposure was significantly associated with increased risk of biochemical progression, after adjusting for clinical characteristics (P = 0.004) or both clinical and pathological characteristics (P = 0.02, Table 3). The trends were similar when black and white men were examined separately (P interaction = 0.67, Table 3).

Figure 1.

The 10-year Kaplan-Meier estimates of biochemical recurrence rates of patients with or without exposure to AO (log-rank test, P = 0.07).

Table 3.  The RR (95% CI) of biochemical progression after RP and with a history of AO exposure
VariableRR (95% CI)P
  • *

    Adjusted for age only;

  • †Adjusted for age, race, biopsy Gleason sum, clinical stage, preoperative PSA, year of surgery, and centre;

  • ‡Adjusted for age, race, biopsy Gleason sum, clinical stage, preoperative PSA, year of surgery, centre, pathological Gleason sum, margin status, extraprostatic extension, seminal vesicle invasion, and lymph node metastasis.

All men  
 Crude1.24 (0.94–1.64)0.13
 Age-adjusted*1.34 (1.00–1.78)0.047
Adjusted for clinical characteristics1.55 (1.15–2.09)0.004
Adjusted for clinical and pathological characteristics1.47 (1.08–2.00)0.02
White men  
 Crude1.11 (0.71–1.72)0.66
 Age adjusted*1.24 (0.79–1.96)0.35
Adjusted for clinical characteristics1.49 (0.93–2.41)0.10
Adjusted for clinical and pathological characteristics1.73 (1.05–2.87)0.03
Black men  
 Crude1.31 (0.91–1.89)0.14
 Age adjusted*1.38 (0.96–1.99)0.09
Adjusted for clinical characteristics1.56 (1.06–2.29)0.02
Adjusted for clinical and pathological characteristics1.40 (0.94–2.10)0.10

Among the 501 men with a PSA recurrence, the PSADT was available for 298 (59%). After adjusting for multiple clinical characteristics, there was no significant association between AO exposure and missing data for PSADT among all men (P = 0.82). However, after adjusting for clinical characteristics, among white men, those with AO exposure were significantly less likely to have PSADT data (P = 0.03), while among black men, AO-exposed patients were more likely to have PSADT data (P = 0.07; P interaction = .006). After adjusting for multiple clinical characteristics, AO exposure was associated with a significantly shorter mean adjusted PSADT (8.2 vs 18.6 months, P < 0.001). The results were similar when further adjusting for multiple clinical and pathological characteristics (8.4 vs 18.2 months, P = 0.001). When stratified by race, AO exposure was associated with significantly shorter adjusted PSADT values in both black and white men, with no significant interaction between race and AO exposure for predicting mean adjusted PSADT (P interaction ≥0.20).

DISCUSSION

The purported link between AO exposure and prostate cancer received confirmatory evidence in a study of US Air Force Veterans involved in Operation Ranch Hand, showing an increased risk of prostate cancer among veterans with higher levels of AO exposure than in veterans with lower levels of exposure and control veterans serving in South-east Asia but not in Vietnam [9,11]. As these veterans exposed to AO age, and enter the age range in which they are more likely to be screened and diagnosed with prostate cancer, there might be further studies clarifying the link between AO and risk of prostate cancer. It is not only important to know that such a link exists but also how such an exposure might affect the clinicopathological characteristics, rates of biochemical progression after treatment, and PSADT after recurrence of any prostate cancer that might develop.

In previous case-control studies there was no significant difference in age and preoperative PSA levels between veterans reporting AO exposure and those with no history of exposure [7,8]. In the present study, men with AO exposure and prostate cancer were more likely to be younger at the time of RP, had a lower preoperative PSA level, and were more likely to have stage T1 disease. This might be partly explained by the fact that AO-exposed veterans might be more aggressively screened for prostate cancer at VA healthcare facilities. By contrast to this, Kamradt et al.[6] reported that patients with AO exposure presented when younger but with more advanced disease. We found no difference in biopsy Gleason grade between those with and without AO exposure, similar to previous studies [7,8].

The component of AO felt to be the carcinogenic agent is dioxin. The exact mechanism by which this compound is carcinogenic is unknown. However, most theories involve its interaction and binding with the aryl hydrocarbon receptor, transforming it and allowing it to translocate into the nucleus, where it attaches to the aryl hydrocarbon nuclear translocator protein and ultimately binds to regulatory regions of target genes [2,12]. Through this mechanism, dioxin is more of a tumour promoter than a mutagen [13]. The oncogenic response in laboratory animals has been shown to depend on age, sex, species, dose and route of administration [14–16].

There are few reports on the pathological characteristics of prostate cancer in men with a history of AO exposure. By contrast with a report by Giri et al.[7] we found no greater incidence of extraprostatic extension (P = 0.50) in men with a history of AO exposure, nor was there an increased risk of seminal vesicle invasion (P = 0.20). Similar to previous investigations [7,8], we found that men with a history of AO exposure were not significantly more likely to have a Gleason sum of ≥4 + 3 (P = 0.19) when adjusting for age at RP, year of RP, race, BMI, preoperative stage, and biopsy Gleason sum.

With similar or more favourable clinical and pathological characteristics (Tables 1,2), we would expect that patients with a history of AO exposure would have similar or more favourable long-term outcomes after primary therapy with RP. However, we found an increased risk (RR 1.47, 95% CI 1.08–2.00, P = 0.02) of biochemical progression after RP in men with AO exposure, after adjusting for clinicopathological findings. Although biochemical progression or PSADT after recurrence, as markers of biological aggressiveness of prostate cancer in patients with AO exposure, have not been specifically addressed in previous studies, mortality has been investigated and a relatively higher mortality rate was found in chemical-plant workers and farmers with prostate cancer and exposure to dioxins than in unexposed individuals [3,5]. In the study with chemical-plant workers, there was a quantitative level of exposure and those with higher levels of exposure had a higher risk of all-cancer mortality [3]. The patients in these previous studies did not receive treatment, so a complete parallel cannot be drawn, but the similarities in the increased biological aggressiveness resulting from AO exposure in these studies generally support our finding of an increased risk of biochemical progression and shorter PSADT after recurrence (8.4 vs 18.2 months, P = 0.001), despite similar or more favourable clinical and pathological characteristics than in unexposed patients. By contrast, the study by Akhtar et al.[9] showed that AO exposure was not associated with increased all-cancer mortality in Operation Ranch Hand veterans compared with age-matched controls and veterans serving in the same region.

Although our results suggest some important associations between a history of AO exposure and the clinical and pathologic characteristics of these patients and their disease, there were relatively few patients with a history of AO exposure, despite the large initial sample. This might have limited the power to detect other modest but important and clinically relevant associations. Another limitation might be selection bias, as only men selecting RP were included and men with advanced disease were not. Another limitation of this study is that we did not quantify levels of AO or dioxin, and the status of exposure is purely subjective in patients with financial incentives (compensation for healthcare and disability from the federal government) to associate their diagnosis with a history of AO exposure. Tests for extremely low levels of dioxins in serum and adipose tissue are under development and when available will help to provide definitive evidence confirming or refuting our findings. Stratification based on increasing levels of exposure would help to clarify if a dose–response exists for biological aggressiveness, and potentially clarify if the level of exposure is related to the increased risk of biochemical progression and shorter PSADT after recurrence. However, at present, diligent screening for probable prostate cancer among patients with probable AO exposure is reasonable.

In conclusion, veterans with a history of AO exposure compared to unexposed counterparts were more likely to be black men, be treated in more recent years, and present with lower-risk clinical and pathological features before and after RP. Similar associations existed with AO exposure and clinical and pathological characteristics when examined by race. Despite low-risk features, AO exposure was associated with an increased risk of biochemical progression and shorter PSADT after RP for prostate cancer. When stratified by race, the association of AO exposure and poorer outcomes after treatment was equal between the races.

ACKNOWLEDGEMENTS

Funding Support: Department of Veterans Affairs, National Institute of Health R01CA100938 (W.J.A.), NIH Specialized Programs of Research Excellence Grant P50 CA92131–01A1 (W.J.A.), the Georgia Cancer Coalition (M.K.T.), the Department of Defense, Prostate Cancer Research Program (S.J.F.), and the American Urological Association Foundation/Astellas Rising Star in Urology Award (S.J.F.). Views and opinions of, and endorsements by the author(s) do not reflect those of the US Army or the Department of Defense.

CONFLICT OF INTEREST

None declared.

Ancillary