Purpose: Human epidermal growth factor receptor-2 (HER2) and epidermal growth factor receptor (EGFR) expression have been associated with disease progression in patients with prostate cancer. We tested the hypothesis that plasma levels of HER2 and/or EGFR are associated with prostate cancer stage and prognosis in patients with clinically localized disease.
Experimental Design: We measured preoperative plasma HER2 and EGFR levels using commercially available ELISAs on banked plasma from 227 patients treated with radical prostatectomy and bilateral lymphadenectomy for clinically localized prostate adenocarcinoma.
Results: Median preoperative plasma EGFR and HER2 levels were 31.4 ng/mL (interquartile range, 19.2 ng/mL) and 10.0 ng/mL (interquartile range, 2.7 ng/mL), respectively. HER2 was elevated in patients with seminal vesicle invasion (P = 0.033). In separate multivariate analyses that adjusted for the effects of standard preoperative predictors, lower EGFR, higher HER2, and higher HER2/EGFR ratio were associated with prostate-specific antigen (PSA) progression (P = 0.003, P < 0.001, and P < 0.001, respectively). In separate multivariate analyses that adjusted for the effects of standard postoperative predictors, lower EGFR and higher HER2/EGFR ratio were associated with PSA progression (P = 0.027 and P < 0.001, respectively). Among the patients who experienced PSA progression, HER2 was significantly higher (P = 0.023) and EGFR was lower (P = 0.04) in those with features of aggressive disease (i.e., development of metastasis, PSA doubling time <10 months, and/or failure to respond to local salvage radiation therapy).
Conclusion: Preoperative plasma HER2 and EGFR were associated with prostate cancer progression after radical prostatectomy. Plasma HER2 and EGFR may provide a tool for predicting long-term recurrence-free survival and early metastasis.
- prostatic neoplasms
- radical prostatectomy
Prostate cancer is the most commonly diagnosed cancer in men and the second leading cause of cancer death in men in the United States (1). Whereas local therapies with curative intent such as radical prostatectomy and radiotherapy result in durable disease control, up to 40% of patients experience disease progression (2). Disease progression in these patients is often due to early dissemination of microscopic metastatic disease that remains undetected by standard pretreatment staging modalities. Identification of patients who are likely to fail definitive local therapy would be helpful for selecting patients best suited for local adjuvant radiation therapy or clinical trials of early systemic intervention (3). Therefore, there is a clear need for novel biomarkers that are specifically associated with biologically aggressive prostate cancer for improved staging and prognostication. In addition, the emergence of new therapeutic approaches for prostate cancer cannot flourish without a set of markers to serve as prognosticators and/or targets.
The ErbB family of receptor tyrosine kinases, which includes epidermal growth factor receptor (EGFR/ErbB1) and human epidermal growth factor receptor 2 (HER2/neu/ErbB2), plays a crucial role in growth, differentiation, and motility of normal and cancer cells. Receptor activation induces dimerization and initiates a signaling cascade that results in phosphorylation and activation of downstream targets. HER2 and EGFR expression have been associated with advanced disease stage, metastasis, shortened survival, poor response to chemotherapy, and even failure of endocrine therapy (4–7). In breast cancer, for example, the gene encoding the HER2 protein is amplified in 20% to 25% of breast cancer patients, and anti-HER2 antibodies have proved to be extremely useful in their treatment (8, 9). In prostate cancer, tissue levels of both HER2 and EGFR are overexpressed (10) with the highest levels in patients with hormone-refractory prostate cancer (5, 11, 12). In addition, HER2 and EGFR expression have been reported to be associated with disease progression after androgen deprivation in patients with metastatic prostate cancer (12, 13). The extracelullar domains of the HER2 and EGFR proteins are frequently cleaved and released into the circulation, where they can be detected by ELISA. Similar to tissue levels, circulating levels of HER2 and EGFR have been associated with progression to metastasis and response to hormonal and chemotherapy (14–16). Elevated circulating levels of HER2 have been associated with progression to hormone refractory prostate cancer (5). However, the predictive value of circulating levels of HER2 and EGFR for prognosis of patients with clinically localized prostate cancer has not yet been studied.
We hypothesized that men with apparently clinically localized prostate cancer harboring aggressive disease would have altered plasma levels of HER2 and EGFR that would be associated with a higher risk of biochemical progression despite effective local control of disease. Therefore, we assessed preoperative plasma levels of HER2 and EGFR in a consecutive cohort of patients with clinically localized prostate cancer treated with radical prostatectomy and bilateral lymphadenectomy.
Materials and Methods
Patient population. All studies were undertaken with the approval and oversight of the Institutional Review Board. The study consisted of 227 patients treated with radical prostatectomy and bilateral lymphadenectomy for clinically localized prostate adenocarcinoma. All 671 patients with the intent to treat their clinically localized prostate cancer (cT1c-3a, NX, M0) with radical prostatectomy from 7/1994 to 11/1997 were potential candidates for this analysis. After consent was gained, preoperative plasma specimens were obtained for 587 of these men. Sixty-four men initially treated with hormonal therapy, 19 who were treated with definitive radiotherapy, and 7 who were treated with cryotherapy before surgery were excluded from the analysis. No disease follow-up information was available for 43 men and they were also excluded. The cohort was randomly divided into two using S-Plus software. This left 227 men for analysis.
No patients were treated preoperatively with either hormonal or radiation therapy and none had secondary cancer. Serum total prostate-specific antigen (PSA) was measured by the Hybritech Tandem-R assay (Hybritech, Inc.). Staff pathologists from our institutions, who were blinded to clinical outcome, examined all prostatectomy specimens. Evaluation of the radical prostatectomy specimen was done as previously described (17) in accordance with the guidelines of the College of American Pathologists (18).
Plasma HER2 and EGFR measurements. Plasma samples were collected after a preoperative overnight fast on the morning of the day of surgery, at least 4 weeks after transrectal ultrasound–guided needle biopsy of the prostate. Blood was collected into Vacutainer CPT 8-mL tubes containing 0.1 mL of molar sodium citrate (Becton Dickinson Vacutainer Systems) and centrifuged at room temperature for 20 min at 1,500 × g. The top layer corresponding to plasma was decanted using sterile transfer pipettes. The plasma was immediately frozen and stored at −80°C in polypropylene cryopreservation vials (Nalgene, Nalge Nunc). For quantitative measurements of HER2 and EGFR levels, we used commercially available microtiter ELISAs (Oncogene Science/Bayer Corp.). Every sample was run in duplicate and the mean was calculated for data analysis. Differences between the two measurements were minimal (intra-assay precision coefficient of variation, 6.4 ± 5.3% for HER2 and 5.9 ± 5.2% for EGFR). The capture antibodies specifically recognize the extracellular domain fragments of each HER2 and EGFR and do not cross-react.
Postoperative follow-up. Patients were generally scheduled to have a digital rectal examination and serum total PSA evaluation postoperatively approximately every 3 months for the first year, semiannually from the second through the fifth year, and annually thereafter. PSA progression was defined as a sustained elevation, on two or more occasions, of total PSA >0.2 ng/mL and was assigned to the date of the first value >0.2 ng/mL. Post-progression serum PSA doubling time (DT) was calculated using the formula DT = log(2) × T / [log(final PSA) − log(initial PSA)] (ref. 19). All patients had at least three PSA measurements available post-progression. Sixteen patients were treated with salvage radiation therapy. A complete response to salvage radiation therapy was defined as the achievement and maintenance of an undetectable serum total PSA level (0.2 ng/mL). Patients were categorized into those with features of aggressive progression and those without. Features of aggressive progression included evidence of metastases (positive bone scan), PSA doubling time <10 months, and/or failure to respond to local salvage radiation therapy as previously described (2, 20–23).
Statistical analysis. Differences in variables with a continuous distribution across categorical variables were assessed using the Mann-Whitney U test and the Kruskal-Wallis test. Logistic regression was used for multivariate analysis of binary outcome variables. Estimates of the probability of remaining free from biochemical recurrence were calculated using the Kaplan-Meier method. Multivariable analysis was done using the Cox proportional hazards regression model. The proportional hazards assumption was verified by tests of correlations with time and examination of residual plots. HER2, EGFR, and HER2/EGFR ratio had a skewed distribution. HER2, EGFR, and HER2/EGFR ratio were analyzed as categorical variables on the basis of their tertile distribution. The lowest category was used as the referent category when calculating the hazards ratios. The accuracy of each tested multivariate models was quantified with receiver operating characteristic–derived area under the curve (AUC; ref. 24). The Hanley and Hajian-Tilaki adaptation (25) of the DeLong et al. (26) method and the associated SAS macro were used.3 In Cox regression models, the AUC is substituted with Harrell's concordance index (24, 27). A value of 100% indicates perfect predictions, whereas 50% is equivalent to a toss of a coin. Internal validation of the accuracy estimates was done using 200 bootstrap resamples (28). The comparisons of accuracy estimates were done using the nonparametric method of DeLong et al. (26). Statistical significance in this study was set as P ≤ 0.050. All reported P values are two sided. Analyses were done in the SAS and S-Plus environments (PC Version 2000 Professional, Insightful Corp.). The S-Plus design library was used.
Association of preoperative plasma EGFR and HER2 with clinical and pathologic characteristics. Median patient age was 61.7 years (range, 39.4-75.5 years). Median preoperative serum total PSA was 6.1 ng/mL (range, 0.5-49.0 ng/mL). The clinical and pathologic characteristics of the 227 consecutive prostatectomy patients and their association with preoperative plasma EGFR and HER2 levels as well as HER2/EGFR ratio are shown in Table 1 . Preoperative plasma EGFR was significantly lower in patients with higher biopsy and pathologic Gleason sum (P = 0.014 and P < 0.01, respectively). Preoperative plasma HER2 was significantly elevated in patients with seminal vesicle invasion and higher pathologic Gleason sum (P = 0.033 and P = 0.028, respectively). Ratio of preoperative HER2/EGFR was significantly higher in patients with higher biopsy and pathologic Gleason sum (P = 0.006 and P < 0.001, respectively).
Association of preoperative plasma EGFR and HER2 with PSA progression after surgery. Of the 227 patients, 42 (18.7%) experienced PSA progression after radical prostatectomy. Median follow-up for patients alive at the time of analysis was 42.7 months (interquartile range, 43.4 months). PSA progression-free survival estimates were 81.3 ± 2.9% (± SE) at 3 years, 77.2 ± 3.5% at 5 years, and 68.6 ± 5.6% at 7 years. The distribution of HER2 and HER2/EGFR ratio in patients who experienced biochemical progression was skewed toward the highest third compared with patients who did not experience biochemical progression (Fig. 1A and C ). The distribution of EGFR in patients who experienced biochemical progression was skewed toward the lowest third compared with patients who did not experience biochemical progression (Fig. 1B).
In separate preoperative multivariable analyses that adjusted for the effects of standard preoperative predictors (Table 2 ), preoperative plasma EGFR, HER2, and HER2/EGFR ratio were associated with PSA progression (P = 0.003, P < 0.001, and P < 0.001, respectively). The standard preoperative multivariable model based on preoperative PSA, biopsy Gleason sum, and clinical stage achieved an AUC of 71.1%. Addition of either EGFR or HER2 improved the predictive accuracy of the standard preoperative multivariable model by 2.6% (P < 0.001) and 5.3% (P < 0.001), respectively. Conversely, the addition of the HER2/EGFR ratio did not improve the predictive accuracy (decrease in accuracy of 0.4%).
In separate postoperative multivariable analyses that adjusted for the effects of standard postoperative predictors (Table 3 ), preoperative plasma EGFR and HER2/EGFR ratio (P = 0.025 and P = 0.007, respectively), but not HER2 (P = 0.176), were associated with PSA progression. The standard postoperative multivariable model based on preoperative PSA, extracapsular extension, seminal vesicle invasion, lymph node invasion, surgical margin status, and pathologic Gleason sum achieved an AUC of 82.3%. Addition of either HER2 or EGFR improved the predictive accuracy of the standard postoperative multivariable model by 1.2% (P < 0.001) and 3.7% (P < 0.001), respectively. Addition of the HER2/EGFR ratio marginally improved predictive accuracy (0.1%; P = 0.3).
Association of preoperative plasma EGFR and HER2 with features of aggressive progression. Of the 42 patients who experienced PSA-progression, 18 (42.9%) patients were categorized as having features of nonaggressive progression because their PSA doubling times were ≥10 months (n = 16; median, 25; range, 14-242) and/or because they achieved a complete response to local salvage radiation therapy (n = 11). Twenty-four (57.1%) patients were categorized as having features of aggressive progression because of presence of metastases (positive bone scan; n = 4), their PSA doubling times <10 months (n = 17; median, 8; range, 3-9), and/or they failed to respond to local salvage radiation therapy (n = 5). Preoperative plasma HER2 levels were significantly higher in patients with features of aggressive disease progression (median, 11.7 ng/mL; range, 6.8-21.1 ng/mL) compared with those with features of nonaggressive disease progression (median, 9.3 ng/mL; range, 5.0-13.3 ng/mL; P = 0.023). Preoperative plasma EGFR levels were significantly lower in patients with features of aggressive disease progression (median, 21.5 ng/mL; range, 12.2-94.2 ng/mL) compared with those with features of nonaggressive disease progression (median, 32.2 ng/mL; range, 19.1-55.1 ng/mL; P = 0.040). In multivariable logistic regression analyses that adjusted for the effects of clinical stage, biopsy Gleason sum, and preoperative serum PSA, HER2 and HER2/EGFR ratio were positively associated with features of aggressive disease progression (test for trend: P = 0.037 and P = 0.011, respectively). In multivariable logistic regression analysis that adjusted for the effects of clinical stage, biopsy Gleason sum, and preoperative serum PSA, EGFR was inversely associated with features of aggressive disease progression (test for trend: P = 0.042).
Biological markers that can identify patients who will fail primary local prostate cancer treatment would be helpful for consideration of neoadjuvant and adjuvant therapy as well as for patient counseling. EGFR and HER2 are tyrosine kinase receptor transmembrane proteins that are activated in various cancers. Preclinical and clinical data show that the activation of EGFR and HER2 may be important for the progression of prostate cancer to androgen-independent disease (29–31). Whereas blood levels of EGFR and HER2 have been studied in metastatic prostate cancer, to our knowledge, no previous study has evaluated the predictive value of blood levels of HER2 and EGFR in patients with clinically localized prostate cancer.
Preoperative plasma levels of HER2 were elevated in patients with features of biologically aggressive prostate cancer such as seminal vesicle involvement and, more importantly, PSA progression after controlling for the effects of standard preoperative predictors. Preoperative HER2 was not associated with PSA progression in standard postoperative multivariate analysis but it still improved the predictive accuracy by 1.2%. Only 19% of our patient cohort experienced PSA progression after a median follow-up of ∼43 months. In addition, most of the patients in our study had favorable clinicopathologic features. PSA progression may result from local failure related to residual disease present after radical prostatectomy, occult nodal or distant metastatic disease present at the time of surgery, or a combination of these (2, 20–23). These forms of PSA progression have variable progression rates with regard to metastases and eventual death. Thus, the lack of independent association of preoperative plasma HER2 with PSA progression in standard postoperative analysis may be due to a lack of association with local and/or biologically more “indolent” disease. In support of this hypothesis, we found that patients with features of aggressive disease progression had higher levels of HER2 than those with features of nonaggressive disease progression.
Three previous studies have shown that serum HER2 is elevated in patients with metastatic prostate cancer compared with nonmetastatic disease (5, 32, 33). Furthermore, Osman et al. (5) found that higher serum HER2 was associated with prostate cancer–related death in patients with metastatic, castrate disease. At the tissue level, HER2 is overexpressed in many nonmetastatic prostate cancers, but HER2 amplification is a rare event (34). In contrast, the HER2 gene has been shown to be amplified and its protein to be overexpressed in the androgen-independent phase of the disease (11, 30, 31). Overexpression of HER2 is shown to not only stimulate proliferation but also enhance androgen receptor signaling by modulating its transcriptional activity and degradation in the presence of low androgen levels (35–37). Moreover, activation of the HER2 kinase signaling cascade can constitutively activate the androgen receptor and render prostate cancer cells refractory to androgen receptor inhibition in different model systems (29, 34, 37, 38). Circulating levels of HER2 have been correlated with gene amplification and tissue overexpression (39, 40). Taken together, these data suggest that preoperative plasma HER2 is a marker of early, clinically occult prostate cancer metastasis.
Lower preoperative plasma level of EGFR was an independent predictor of PSA progression after controlling for the effects of standard preoperative and postoperative predictors. Activation of EGFR generates complex signal transduction pathways that modulate many normal cellular processes such as proliferation, survival, adhesion, migration, and differentiation. Aberrant activation of EGFR has been shown to be critical for the maintenance of malignancy in a number of solid tumors including prostate cancer (7, 41, 42). To our knowledge, this is the first study to evaluate the prognostic role of circulating levels of EGFR patients with clinically localized prostate cancer. Circulating levels of EGFR have been reported to be decreased in patients with pancreatic, gastric, colon, lung, renal, prostate, and breast cancer (6, 43). Souder et al. (6) reported that decreased serum EGFR was associated with significantly reduced overall survival in patients with metastatic breast cancer.
The finding that circulating EGFR levels are decreased in aggressive cancer is not in contradiction with the fact that tissue levels are elevated. EGFR is released in the circulation either by proteolytic cleavage of the extracellular domain or alternate transcription of primary mRNA (44). EGFR can be activated as a homodimer or as a heterodimer with one of the other three members of the ErbB family. Increased circulating levels of its ligands, such as EGF and transforming growth factor α, may form complexes with circulating EGFR and, subsequently, may be cleared more rapidly from the circulation. Another explanation is that internal autocrine mechanisms activate EGFR, resulting in less available EGFR for proteolysis at the tumor cell surface. With either of these mechanisms, activation of the EGFR pathway may explain the decreased serum EGFR levels observed in patients with biologically and clinically aggressive prostate cancer.
When activated, EGFR can pair with another EGFR, forming a homodimer, or with another member of the ErbB receptor family, such as HER2, to form a heterodimer. In patients with metastatic breast cancer, Souder et al. (6) found that the combination of serum EGFR and HER2 helped identify a subgroup of patients with decreased EGFR and normal HER2 who had significantly reduced survival. Another report of 807 patients with primary breast cancer showed that those with overexpression of both HER2 and EGFR by immunohistochemistry had the shortest survival (45). In colon cancer cell lines, inhibition of EGFR has been shown to increase activation of HER2 (46). These data suggest that HER2 and EGFR may have a cooperative/synergistic effect and that combining them into a ratio may improve prostate cancer prognostication. However, combination of HER2 and EGFR into a ratio did not improve the predictive accuracy for PSA progression after radical prostatectomy.
The findings of our study could lead to several applications. In addition to improving prognostication in patients with apparent clinically localized disease, plasma levels of HER2 and/or EGFR may help select patients for clinical trials of anti-HER2 and/or anti-EGFR therapies. Both HER2 and EGFR have been validated as therapeutic targets in breast, lung, and colon cancer. However, whereas HER2 and EGFR monoclonal antibodies and kinase inhibitors have shown promising activity in patients with these malignancies, they have failed to show significant activity in prostate cancer (47–50). Solit and Rosen (51) have given various explanations for the lack of efficacy of anti-HER2 and anti-EGFR therapies in prostate cancer. One explanation is that these therapies have been given to either all patients or those with HER2 or EGFR overexpression. HER2 and EGFR amplification are, however, rare events in prostate cancer, and thus the target may not be as important in this disease. In addition, the therapies were given in advanced cancer and metastatic setting, but the HER2/EGFR status was determined from the original tumor, which had been removed many years earlier. Furthermore, the tissue tests for protein overexpression by immunohistochemistry and gene amplification (i.e., fluorescence in situ hybridization) are both subject to technical problems. For immunohistochemistry, these include, but are not limited to, differences in the methods between laboratories and between operators and variation in reagents. Although the fluorescence in situ hybridization technology seems to be reproducible, the standardization of fixative solutions, fixation times, and digestion conditions remains a technical issue. Moreover, the fluorescence in situ hybridization instrumentation is expensive and not widely available in diagnostic pathology laboratories. The conversion of these assays from tests that require tumor tissue specimens to assays that can be done on blood would significantly increase their clinical value. Lipton et al. (15) reported that patients with metastatic breast cancer who had elevated pretreatment serum HER2 levels had a decreased response to first-line hormone therapy and decreased survival. Moreover, they found that serum conversion to HER2-positive status at the time of disease progression in patients with breast carcinoma on hormone therapy was an independent risk factor for decreased survival (52). This, together with the findings of the current study, supports the evaluation of anti-HER2 and/or anti-EGFR therapies in multidrug combinations for the treatment of prostate cancer patients with altered plasma HER2 and/or EGFR levels. Moreover, these markers could be followed to evaluate response to treatment/progression of disease.
This study has several limitations. First and foremost are the limitations inherent to sample size, relatively short follow-up, and small number of patients with PSA progression and aggressive disease progression, which may have limited our ability to detect small differences attributed to other variables. Although addition of preoperative plasma HER2 or EGFR improved the predictive accuracy of the standard preoperative and postoperative models by statistical significant margin, the clinical significance of a 1% to 5% improvement remains subtle. Therefore, prospective well-done studies are necessary to further delineate the clinical benefit of these biomarkers before clinical application. Moreover, correlation with tissue levels would help understand the biological and clinical relevance of plasma levels. Another potential limitation is the large variability of HER2 and EGFR levels among prostate cancer patients. It is unclear what mechanisms account for such a wide variation in shedding and what is the biological or clinical significance of the shedding.
Preoperative plasma levels of HER2 and EGFR each improved the accuracy of standard preoperative and postoperative models for prediction of PSA progression in patients with clinically localized prostate cancer treated with radical prostatectomy. Together with established clinical and pathologic features, HER2 and EGFR may help identify patients who may benefit from adjuvant targeted therapy and monitor treatment response. Larger studies with more events and longer follow-up are required for a more definitive statement about the association of preoperative circulating levels of HER2 and EGFR with prostate cancer progression and, more importantly, profound clinical end points such as metastases and survival.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Received February 8, 2007.
- Revision received June 25, 2007.
- Accepted July 5, 2007.