|
|
||
 |
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 13 (May 1), 2006: pp. 1982-1989 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.02.9439 Impact of IGF-I and CYP19 Gene Polymorphisms on the Survival of Patients With Metastatic Prostate Cancer
From the Department of Urology and the Department of Medical Information Science Akita University School of Medicine, Akita; Department of Urology, Graduate School of Medicine, Chiba University, Chiba; and the Department of Urology, Graduate School of Medicine, Kyoto University, Kyoto, Japan. Address reprint requests to: Tomonori Habuchi, MD, Department of Urology, Akita University School of Medicine, 1-1-1 Hondo Akita 010-8543, Japan; e-mail: thabuchi{at}doc.med.akita-u.ac.jp
PURPOSE: The prognosis of metastatic prostate cancer significantly differs among individuals. While various clinical and biochemical prognostic factors for survival have been suggested, the progression and response to treatment of those patients may also be defined by host genetic factors. In this study, we evaluated genetic polymorphisms as prognostic predictors of metastatic prostate cancer. PATIENTS AND METHODS: One hundred eleven prostate cancer patients with bone metastasis at the diagnosis were enrolled in this study. Thirteen genetic polymorphisms were genotyped using polymerase chain reaction-restriction fragment length polymorphism or an automated sequencer with a genotyping software. RESULTS: Among the polymorphisms, the long allele (over 18 [CA] repeats) of insulin-like growth factor-I (IGF-I) and the long allele (over seven [TTTA] repeats) of cytochrome P450 (CYP) 19 were significantly associated with a worse cancer-specific survival (P = .016 and .025 by logrank test, respectively). The presence of the long allele of either the IGF-I or CYP19 polymorphisms was an independent risk factor for death (P = .019 or .026, respectively). Furthermore, the presence of the long allele of both the IGF-I and CYP19 polymorphisms was a stronger predictor for survival (P = .001). CONCLUSION: The prognosis of metastatic prostate cancer patients is suggested to be influenced by intrinsic genetic factors. The IGF-I (CA) repeat and CYP19 (TTTA) repeat polymorphisms may be novel predictors in prostate cancer patients with bone metastasis at the diagnosis.
Prostate cancer is becoming one of the most common cancers in males in Japan.1 The introduction of prostate-specific antigen (PSA) screening has increased the identification of early-stage prostate cancer, and the establishment of surgical and radiation techniques may improve the outcome in these patients. Meanwhile, the prognosis of metastatic prostate cancer patients remains poor, although approximately 80% of untreated patients respond to androgen deprivation therapy. However, it has also been acknowledged that the prognosis of metastatic prostate cancer differs significantly among individuals.2 The prediction of prognosis and stratifying patients by their risk of progression are important for personalized treatments and follow-up strategies. To date, various clinical and biochemical parameters as well as tumor characteristics have been reported to predict the survival of patients with metastatic prostate cancer. As clinical and biochemical factors, lower performance status, pain score, extent of disease on bone scan, pretreatment serum testosterone, serum alkaline phosphatase (ALP) and acid phosphatase, PSA, and lower hemoglobin (HGB) were reportedly associated with treatment response or patient survival.3-7 Pathologic and immunohistochemical analyses also demonstrated that nuclear texture, oligosaccharide sialyl Lewis (x), c-erbB-2 (Her2/neu), and tissue factor could be predictors of survival.5,8,9 The factors indicating the characteristics of cancer and polymorphisms as host genetic factors possibly influence the prognosis of cancer. Recently, various genetic polymorphisms were reportedly associated with a risk of prostate cancer although the real influence of these polymorphisms remains controversial.10-16 It has also been suggested that these polymorphisms may further modify the progression of cancer and define the response to therapy as well as cancer susceptibility. Since most prostate cancers show androgen-dependent growth, many previous analyses were performed with an emphasis on the polymorphisms of genes involved in biosynthesis and the metabolism of steroids and androgens. The association with susceptibility to prostate cancer has so far been demonstrated in vitamin D receptor (VDR), androgen receptor (AR), steroid-5-alpha-reductase, alpha polypeptide 2 (SRD5A2), cytochrome P450 (CYP) 11A1, CYP17, and aromatase (CYP19).10-14,16 Other genetic polymorphisms such as transforming growth factor beta 1 (TGF-ß1), cyclin D1 (CCND1), epidermal growth factor (EGF), human epidermal growth factor receptor 2 (HER2/neu), insulin-like growth factor (IGF-I), insulin-like growth factor binding protein-3 (IGFBP-3), and PSA have also been evaluated on the basis of their functions as regulators of cell cycle, differentiation, and apoptosis.17-23 Most of these studies were designed to assess the association between polymorphisms and prostate cancer risk, tumor grade, or extension at the time of diagnosis using a case-control study model. To date, no definite host genetic factors influencing the prognosis of advanced prostate cancer patients have been reported. A long-term longitudinal study is required to evaluate the real impact of polymorphisms on disease progression, as well as treatment response or patient survival. In this study, we analyzed 13 gene polymorphisms to examine the hypothesis that polymorphisms as host genetic factors are associated with the survival of metastatic prostate cancer patients.
Patients From July 1980 to July 2003, 111 native Japanese patients with prostate cancer with bone metastasis at diagnosis were enrolled in this study. The patients were diagnosed at Akita University Hospital (Akita, Japan) and its related community hospitals, Chiba University Hospital (Chiba, Japan) and Kyoto University Hospital (Kyoto, Japan), and only incident cases were included. This study was approved by the institutional review board (the ethical committee) of the Akita University School of Medicine, the Chiba University Graduate School of Medicine, and the Kyoto University Graduate School of Medicine. Written informed consent was obtained from all patients for the use of their DNA and clinical information. All patients enrolled in this study had metastatic prostate cancer with no previous treatments. Prostate needle biopsy specimens provided material for pathologic diagnosis, and metastasis was identified by x-rays, computed tomography scans, or bone scintigraphy. After diagnosis, all patients underwent surgical castration or luteinizing hormone-releasing hormone (LH-RH) analogs with or without antiandrogens as the initial hormone therapy. Other optional therapies, including estrogens, antiandrogen agents, steroids, palliative radiation, or a combination of these were added to or replaced by the preceding therapies when treatment failure was noted. Pathologic grading of a biopsied specimen was determined according to the General Rule for Clinical and Pathological Studies on Prostate Cancer by the Japanese Urological Association and the Japanese Society of Pathology,24 which is based on the WHO criteria and according to the Gleason score.25 All pathologic grading was based on needle biopsy specimens judged by local pathologists with no designated primary pathologist. Well-, moderately-, and poorly differentiated carcinomas generally correspond to Gleason scores of 2 to 4, 5 to 7, and 8 to 10, respectively.24,25 As both grading systems were used by local pathologists, the tumor grade system was newly categorized as follows: low-grade cancer included well-differentiated or Gleason 2 to 4 carcinomas; intermediate-grade cancer included moderately-differentiated or Gleason 5 to 7 carcinomas; and high-grade cancer included poorly-differentiated or Gleason 8 to 10 carcinomas. In 10 patients, the final pathologic grade was not determined because no grade information was described in the final report or a different grading system was applied by the local pathologists. In this study, performance status and extent of disease were excluded from the analysis because they are not objective and may not be suitable in an inter-institutional study. Pretreatment HGB, ALP, lactate dehydrogenase (LDH), and PSA levels were measured before the initial treatment of prostate cancer and PSA was measured every 3 months thereafter. An independent end point reviewer in each institution (H.S., T.S., and H.F.) determined the cause of death on the basis of standardized extractions from the patient files. For this determination, genotype data of patients were not revealed to the reviewers.
Genotyping Analysis
Statistical Analysis The end point of this study was cancer-specific death defined as death from prostate cancer or death from other causes strongly associated with progression of disease. Survival time was calculated from the date of prostate cancer diagnosis to the day of death due to cancer-specific death, with deaths from other causes censored. To compare survival rates, patients were dichotomized by the median value of age and PSA, by normal limits in HGB, ALP, LDH, by the tumor grade system (ie, high versus low/intermediate), and by criteria previously reported in each polymorphism. For repeat-length polymorphisms, the number of repeats was dichotomized and the long allele of the AR, CYP19, and IGF-I polymorphisms were defined as having more than 18 (CAG) repeats, more than seven (TTTA) repeats, and more than 18 (CA) repeats, respectively according to the previous reports.11,13,16,29 Similarly, the repeat number of the CYP11A1 polymorphism was dichotomized by the presence of four (TTTTA) repeats according to the previous study.13 Differences in survival were tested using the logrank test. Hazard ratios (HRs) and 95% CIs for cancer death were assessed by the Cox proportional hazard regression models. Prostate specific antigen, HGB, ALP, and polymorphisms of the IGF-I and CYP19 were employed as a variable set in the analysis, model A. Of the variable set, the IGF-I and CYP19 polymorphisms were replaced by a combination of the two polymorphisms as an alternative analysis, model B. Cancer-specific survival was estimated using the Kaplan-Meier method. Median follow-up time was computed among censored cases. All statistical analyses were performed using SPSS software version 12.0 (SPSS, Inc, Chicago, IL) and two-sided P values < .05 were considered to indicate statistical significance.
Patients and Treatment Characteristics The mean age (± standard deviation) of the patients was 70.6 ± 8.7 years (range, 45 to 89; median, 71 years). The mean follow-up period was 46.3 ± 33.4 months (range, 1 to 153; median, 37 months). Pretreatment PSA, HGB, ALP, and LDH levels are shown in Table 2. Of 111 patients, bone metastasis alone, additional lymph node metastasis, and other visceral metastasis were seen in 60 patients (54.1%), 37 patients (33.3%), and 14 patients (12.6%), respectively. As the initial hormone therapy, surgical castration was selected in 57 patients (51.4%) and LH-RH analog was administrated in 54 patients (48.6%), while 42 patients (37.8%) and 69 patients (62.2%) were treated with monotherapy of surgical castration or LH-RH analog and combined androgen blockade, respectively. None of the patients were initially treated with anticancer drugs. Seventy patients (63.1%) had a decreased PSA level < 4 ng/mL, 36 patients (32.4%) did not achieve normalized PSA, and there was no information regarding the lowest PSA level in five patients (4.5%).
Survival Analysis The 5- and 10-year cancer-specific survival of all patients was 40.1% and 22.2%, respectively, with a median survival time of 51 months. The survival was compared between two groups divided by the median value of age and PSA at diagnosis, by the normal limit for HGB, ALP, and LDH, and by tumor grade (high versus intermediate/low grade). Univariate analysis using the logrank test showed that a PSA level above 260 ng/mL (P = .015), HGB below (P < .001) and ALP above the normal limit (P < .001) were associated with poor survival (Table 2). Age (P = .365), LDH level at diagnosis (P = .272), and tumor grade (P = .084) were not influential factors of survival. The genotype frequency of each SNP was indicated in Table 3. The repeat number of the IGF-I (CA) repeat polymorphism ranged from 15 to 21 and 18 genotypes were observed. The most frequent genotype was 17/19 (21 patients), followed by 16/17 (13 patients), 18/19 (11 patients), 17/18 (nine patients), 16/19 (nine patients), 18/18 (eight patients), 19/19 (eight patients), 16/18 (seven patients), 17/17 (seven patients), 15/19 (five patients), 19/20 (four patients), 16/16 (three patients), and others (one patient for each). The repeat number of the CYP19 (TTTA) repeat polymorphism ranged from 7 to 13 and seven genotypes were observed. The most frequent genotype was 7/7 (46 patients), followed by 7/11 (41 patients), 7/12 (12 patients), 11/11 (five patients), 11/12 (four patients), 7/13 (two patients), and 12/12 (one patient). The number of the CYP11 (TTTTA) repeat polymorphism ranged from 3 to 9 and 9 genotypes were observed. The genotype frequency was as follows: 6/6 (39 patients), 4/6 (32 patients), 4/4 (19 patients), 6/8 (13 patients), 8/8 (three patients), 3/4 (two patients), and others (one patient for each). The repeat number of the AR (CAG) repeat polymorphism ranged from 18 to 37 and 18 genotypes were observed. The genotype frequency was as follows: 25 repeats (22 patients), 24 repeats (17 patients), 26 repeats (14 patients), 23 repeats (12 patients), 28 repeats (eight patients), 27 repeats (seven patients), 22 repeats (seven patients), 21 repeats (six patients), 19 repeats (four patients), 30 repeats (four patients), 20 repeats (two patients), 32 repeats (two patients), and others (one patient for each).
Of 13 polymorphisms, two polymorphisms, the IGF-I (CA) repeat polymorphism and the CYP19 (TTTA) repeat polymorphism, were associated with survival. Patients with at least one copy of the long allele for the IGF-I polymorphism showed worse survival compared with those without the long allele (P = .016; Table 3, Fig 1). The rates of 10-year cancer-specific survival were 16.8% and 28.8%, and the median survival time was 41 months and 61 months for patients with and without the long allele of the IGF-I polymorphism, respectively. Similarly, the survival of patients with at least one copy of the long allele for the CYP19 polymorphism were worse than without the long allele (P = .025; Table 1, Fig 2). The rates of 10-year cancer-specific survival were 16.7% and 29.7%, and the median survival times were 41 months and 61 months for patients with and without the long allele of the CYP19 polymorphism, respectively.
Patients were classified into four groups according to the presence of the long allele of each polymorphism. Patients with long alleles of both the IGF-I and CYP19 polymorphisms showed the worst pattern in Kaplan-Meier plots. The P values of all three possible pair-wise logrank tests controlled by the worst pattern were less than .05 (Fig 3). There were no significant differences in survival between three groups except for the worst pattern. The rate of 10-year cancer-specific survival for patients with long alleles of both polymorphisms and other combinations was 8.8% and 29.4%, respectively, and the median survival times were 28 months and 61 months, respectively. The remaining 11 polymorphisms were not associated with patient survival.
Using five prognostic factors significantly determined by logrank test of dichotomized groups on serum markers as well as gene polymorphisms, the Cox proportional hazard regression analyses were performed against two models, A and B (Table 4). In model A, the presence of at least one copy of either the long allele of the IGF-I or CYP19 polymorphism was also an independent risk factor predicting survival with HRs of 2.012 (95% CI, 1.120 to 3.623; P = .019) or 1.976 (95% CI, 1.086 to 3.595; P = .026) compared with no long allele, respectively. Furthermore, in model B, the presence of one or more copies of the long allele of both the IGF-I and CYP19 polymorphisms was a stronger predictor for survival (HR, 2.570; 95% CI, 1.436 to 4.597; P = .001).
In metastatic prostate cancer, resistance to hormone therapy and the androgen-independent growth of cancer cells are considered key factors of patient survival because hormone therapy is the standard and main treatment in most of these patients.6,34 In this study, we evaluated the polymorphisms of 13 genes previously reported to be associated with susceptibility to or progression of prostate cancer and found that the polymorphisms of IGF-I and CYP19 as well as conventional biochemical markers (ie, pretreatment HGB, ALP, and PSA) were independent predictors of survival. IGF-I regulates cell proliferation, differentiation, apoptosis and transformation through the IGF-I receptor with both paracrine and autocrine mechanisms, is required for development of the prostate gland and is involved in prostate cancer.21 IGF-I also activates the AR directly in the absence of androgens, suggesting that IGF-I increases prostate cancer cell growth in both an androgen-dependent and androgen-independent manner.35 A prospective case-controlled study demonstrated that men with high levels of serum IGF-I were at an increased risk of developing clinically evident prostate cancer within the next 5 to 10 years.36 Furthermore, a strong association between the plasma level of IGF-I and extraprostatic and distant metastatic prostate cancer was recently indicated in a relatively large case-controlled study.37 The IGF-I contains a polymorphic region composed of (CA) repeats approximately 1 kilo basepair upstream from the transcription initiation site, which is suggested to influence the gene transcription level.38 Although the exact correlation between the polymorphism and the circulating IGF-I level has not been fully determined, several studies suggest that serum IGF-I levels increase with the number of 19-alleles.39-41 A recent large cohort study demonstrated that the serum IGF-I level was significantly higher in carriers of the 192-bp allele, which corresponds to the 19 (CA) repeats allele, than noncarriers of the allele.41 The same group further reported that the significantly higher serum IGF-I level was observed in carriers of 192-bp or 194-bp which is identical to the long allele in this study.42 Thus, the cut off point employed in this study is considered to be appropriate. Meanwhile, some conflicting results in regard to the association between the IGF-I polymorphism and serum IGF-I level.43,44 Additional studies are needed to clarify the effect of the polymorphism on not only the serum IGF-I level, but also on the tissue IGF-I levels in the prostate. It is known that the 19 (CA) repeat allele of the IGF-I polymorphism predicted a susceptibility to several types of cancer.45 Recently, we demonstrated that the 19 (CA) repeat allele carriers had a significantly increased risk for prostate cancer or BPH compared with noncarriers of the allele.29 Although no association between the aggressiveness of prostate cancer and the IGF-I polymorphism were observed in our previous case-control study,29 this study suggested that the 19 or more (CA) repeat allele is associated with the aggressive phenotype or resistance to hormone therapies in advanced prostate cancer. In an animal study, transgenic adenocarcinoma of a mouse prostate (TRAMP) model showed an increased expression level of IGF-I mRNA in prostate tissue with cancer progression and metastasis.46 Since IGF-I stimulates the growth of both metastatic and androgen-independent prostate cancer cells, these results indicate that a higher level of circulating and/or intra tumor IGF-I as influenced by the IGF-I polymorphism might have a significant impact on the survival of metastatic prostate cancer patients. If the results are confirmed in larger studies, modification of the IGF-I level or IGF-I signaling may be a promising therapeutic target for the treatment of advanced prostate cancer. Aromatase encoded by CYP19 is a key enzyme that converts testosterone into estrogen in males, and is suggested to play an important role in the development of benign prostate hyperplasia and prostate cancer.47 The CYP19 has a tetranucleotide (TTTA) repeat polymorphism in intron 4 and the polymorphism is reportedly associated with a risk of breast cancer, prostate cancer, and postmenopausal bone metabolism.16,48,49 The biologic function of the CYP19 polymorphism in enzyme activity or the expression level of aromatase has not been clarified. Because the polymorphism is located in the midportion of the intron, it is not likely to be associated with regulation of the transcript, splicing, or alteration of enzyme activity, but it may influence mRNA stability or in linkage disequilibrium with unidentified functional polymorphisms. Recently, Haiman et al48 demonstrated that the presence of at least one allele more than 7 (TTTA) repeats of the CYP19 polymorphism was associated with a lower level of plasma androstendion, a higher level of estrogens, and a higher estrone to androstendione ratio. Therefore, we used the 7 (TTTA) repeats as a cut off in this study, and other studies also used a similar cut off.16,50 Although we did not measure pretreatment testosterone levels, our results are in line with previous reports which indicated that a lower pretreatment testosterone level predicted a lower response to hormone therapy and poor survival for patients with metastatic prostate cancer.51 Advanced prostate cancers are adapted for a lower androgen environment and may achieve hormone-independent cell growth earlier in the clinical course. Those cancers are suggested to resist various hormone therapies.8 However, as the circulating testosterone level has a circadian variation and is influenced by aging, it seems difficult to prove an association between the testosterone level and the polymorphisms.52 A recent study observed no association between the CYP19 polymorphism and susceptibility or aggressiveness of prostate cancer, suggesting that the racial or environmental difference also influence the significance of the polymorphisms in a development or progression of the cancer.50 In Japanese population, the prognosis of metastatic prostate cancer is possibly affected by the CYP19 polymorphism affecting the testosterone metabolism. Androgen ablation therapy has been the standard treatment for metastatic prostate cancer since Huggins reported for the first time in 1941.53 However, once cancer cells achieve resistance to hormone therapy, the mean survival is only approximately 12 months. Recently, two studies demonstrated a moderate survival benefit treated with chemotherapy using a taxane analog in hormone-refractory advanced prostate cancer.54,55 However, the timing and target of this chemotherapy remain a controversial issue. Such chemotherapy at an earlier stage after diagnosis may further improve the prognosis of patients with poor survival, which may be defined by genetic polymorphisms as well as conventional risk factors. These results may provide a basic concept that this therapeutic modality, including the administration of early chemotherapy, may be modified according to the genetic polymorphism. Furthermore, the modification of the activity of the CYP19 enzyme or the IGF-I signaling pathway may be a promising therapeutic target for the treatment of metastatic prostate cancer. Interestingly, although recent programs on large-scale SNPs instituted by public and private initiatives are expected to reveal disease-associated genes,56,57 the significant results with polymorphisms obtained in this study were from both simple dinucleotide repeat polymorphisms. Although we do not know whether there are any other significant SNPs in the vicinity of or linked with dinucleotide repeat polymorphisms, our results may underscore the importance of simple repeat polymorphisms in the genesis and progression of common diseases. In conclusion, to our knowledge, this is the first study demonstrating the possible association of two polymorphisms, the CYP19 (CA) repeat polymorphism and the IGF-I (TTTA) repeat polymorphism, with the survival of prostate cancer patients with distant metastasis. It is also suggested that intrinsic factors of cancer hosts as well as the cancer status and characteristics affect the prognosis of prostate cancer. Understanding the impact of genetic polymorphisms on the prognosis of prostate cancer could lead to the diversification and individualization of prostate cancer treatments, to follow-up strategies, and to novel therapeutic modalities.
The authors indicated no potential conflicts of interest.
Supported by grants-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology of Japan Grants No. 16591579, 00293861, 14207061, and 16591582, Princess Takamatsu Cancer Research Fund (2003), Public Trust Haraguchi Memorial Cancer Research Fund (2003), The Japanese Foundation for Prostate Research, and the grant-in-aid from the Japanese Urological Association (2003). This study was presented at the Prostate Cancer: Epidemiology and Natural History (I) moderated poster session at the 100th Annual Meeting of the American Urological Association, San Antonio, TX, May 21-26, 2005. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
Related Editorial
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
