High grade prostatic intraepithelial neoplasia does not display loss of heterozygosity at the mutation locus in BRCA2 mutation carriers with aggressive prostate cancer
Liam Kavanagh, kConFab, Research Department, Peter MacCallum Cancer Centre, Locked Bag 1, A'Beckett St, Vic. 3002, Australia. e-mail: email@example.com
What's known on the subject? and What does the study add?
The risk of developing aggressive prostate cancer is increased for men carrying a pathogenic germline mutation in BRCA2. An earlier study by the Kathleen Cuningham Consortium for Research into Familial Breast Cancer showed that BRCA2 mutation carriers displayed a loss of heterozygosity (LOH) within their prostate cancer tissue in the majority of cases, thus implying that the prostate cancer in these men occurred as a result of LOH for BRCA2. High grade prostatic intraepithelial neoplasia (HGPIN) has been considered a precursor to prostate adenocarcinoma in some, but not all, cases of prostate adenocarcinoma.
The study found that there was no LOH for BRCA2 in HGPIN. From this small cohort of BRCA2-positive men, we suggest HGPIN is not necessarily a precursor to their prostate cancer development. The presence of HGPIN in a TRUS biopsy in these men at risk of high risk disease is not an indication for prostatectomy.
- • To determine if high grade prostatic intraepithelial neoplasia (HGPIN), which is considered a precursor to the development of prostate adenocarcinoma, displays the same genetic hallmarks as adenocarcinoma.
- • To identify, using molecular genetic techniques, if HGPIN is a precursor of tumour development and progression in men carrying a pathogenic germline mutation in BRCA2.
PATIENTS AND METHODS
- • Ten participants from the Kathleen Cuningham Consortium for Research into Familial Breast Cancer cohort of high-risk breast cancer families were identified, with (i) a diagnosis of aggressive prostate cancer and presence of HGPIN, (ii) a pathogenic BRCA2 mutation, and (iii) access to archival prostate tissue specimens.
- • Loss of heterozygosity (LOH) at the BRCA2 gene was examined using mutation-specific PCR and sequencing of DNA from laser microdissected HGPIN.
- • Within this cohort of 10 pathogenic BRCA2 carriers, no patient displayed LOH at the mutation locus within HGPIN, irrespective of whether or not corresponding adenocarcinoma DNA displayed LOH.
- • Although HGPIN is considered a precursor to cancer, as no LOH was observed, this assay does not provide a genetic marker that may be considered a positive predictor of tumorigenesis in BRCA2 carriers.
- • In this group of high-risk men, early screening via prostate-specific antigen testing, rectal examination and prostate biopsy may be prudent to permit the detection and the optimum clinical management of prostate cancer.
prostatic intraepithelial neoplasia
high grade PIN
Kathleen Cuningham Consortium for Research into Familial Breast Cancer
loss of heterozygosity
ductal carcinoma in situ
haematoxylin and eosin
Royal College of Pathologists Australia
Prostate cancer is the most common, non-skin-related, malignancy in men in developed countries, and is a major cause of morbidity and mortality . Risk factors include family history  and germline genetic mutations . Other than the presence of a BRCA2 mutation, the search for high penetrance genes associated with an increased risk has been unsuccessful. It has become apparent that, perhaps owing to the genetic heterogeneity of prostate cancer, common genetic variants associated with a small increased risk [4,5] of prostate cancer may prove as informative as highly penetrant genes.
The breast cancer susceptibility genes, BRCA1 and BRCA2, identified in the 1990s as being a major cause of breast and ovarian cancer [6–9], have been implicated in a number of other malignancies. These include pancreatic cancer, melanoma, and prostate cancer [10,11]. The risk of developing these cancers when predisposed to a BRCA1 or BRCA2 pathogenic mutation increases substantially for members of these families [11,12]. For example, the relative risk of developing prostate cancer in BRCA1 carriers <65 years of age is 1.82 (1.0–3.3 95% CI, P= 0.05) , while the risk increases to 4.6 across all ages (3.5–6.2 95% CI) for BRCA2 carriers .
In the context of breast and ovarian cancers, BRCA1 and BRCA2 are well known classical tumour suppressor genes . Breast tumours arising in mutation carriers usually display inactivation of both copies of the corresponding gene [14–18], where one allele is inactivated as a result of the germline mutation and the second allele is somatically inactivated, often as a consequence of deletion [14,16,18] but sometimes by epigenetic changes [15,17,19,20]. Several groups have demonstrated in familial breast and prostate tumours that the loss of the wild-type allele in heterozygote carriers of a BRCA1 or BRCA2 mutation is the common mechanism of inactivation in tumours [18,21]. The observation that loss of heterozygosity (LOH) occurs in some BRCA carriers within areas of ductal carcinoma in situ (DCIS) , as well as other non-tumour tissue in the surrounding breast tumour microenvironment, suggests that there may be early genetic events that occur before the distinct morphological changes observed in the mammary gland . The microenvironment adjacent to adenocarcinoma of the prostate that contains prostatic intraepithelial neoplasia (PIN) and moderate to severe degrees of inflammation has also been identified as potentially useful in the diagnosis of clinically significant disease . A similar study carried out by Hu et al. , demonstrates LOH of the M6P/IGF2R gene in laser-dissected high grade PIN (HGPIN) tissue. This suggests that this mutation may be an early event in tumorigenesis, and supports the notion that M6P/IFR2R functions as a tumour suppressor gene in this disease.
The hypothesis that malignancy arises through progression from premalignant precursors has been established for a number of cancers, the best example being colorectal carcinogenesis . Recent advances in molecular genetics are changing our original understanding of tumorigenesis in many tumour streams, with the exact role of premalignant lesions in these pathways proving to be a complex model to map. In breast cancer, for example, there are low grade and high grade pathways of tumorigenesis. A recent breast cancer study has shown a significant proportion of high grade oestrogen-receptor-positive (ER+) invasive breast cancers contain genomic changes characteristic of low grade ER+ lesions; that is, high grade ER+ breast cancer may originate from low grade lesions [27,28].
Prostatic intraepithelial neoplasia was first described as a precursor to some prostatic carcinomas in the 1960s . It consists of architecturally benign prostatic acini lined by cytologically atypical cells and is categorized into low and high grade. HGPIN can be distinguished from low grade PIN by the presence of prominent nucleoli, and has other features such as increased chromatin content and positive cytoplasmic staining for α-methylacyl-CoA racemase (AMACR) [29,30]. Phenotypically, HGPIN and adenocarcinoma are similar, and the rates of cell proliferation and death in both forms are raised, compared with those in normal prostates . Furthermore, there are distinct genetic and molecular alterations common to both HGPIN and adenocarcinoma, such as frequent allelic loss of 8p12-21, LOH at 8p22, alterations in expression of P16, TP53, bcl-2, c-myc and AMACR [30–32]. The TMPRSS2-ERG fusion protein, which is present in 50% of prostate cancer, is also present in ∼20% of cases of HGPIN and is not identified in BPH or normal prostate tissue .
In 1997, Qian et al.  reported that 86% of whole-mount radical prostatectomy specimens with cancer contained HGPIN, usually within 2 mm of the primary site, while Bostwick et al.  have shown isolated HGPIN to be present in ∼30–50% of biopsies that later progressed to adenocarcinoma.
There is a need to provide conclusive evidence to support the predictive value of HGPIN in this well defined group of BRCA2 mutation carriers, as it has been shown that these men have highly aggressive disease and dramatically reduced survival compared with the general population . A definitive result on the role of HGPIN, under the assumption that LOH occurs only because tumorigenesis arose from the germ-line BRCA2 mutation, may have important clinical implications for early treatment intervention; therefore, the aim of the present study was to examine for LOH at the BRCA2 mutation locus in microdissected HGPIN from radical prostatectomy specimens to identify if HGPIN can be used as a potential marker for the development of aggressive prostate cancer.
MATERIALS AND METHODS
SUBJECTS AND ETHICS
All cases of prostate cancer were obtained from the Kathleen Cuningham Consortium for Research into Familial Breast Cancer (kConFab); see Mann et al. . In brief, kConFab families are recruited if they have a strong family history of breast and/or ovarian cancer, known to be carrying a mutation in BRCA1, BRCA2, ATM or TP53, or classified as high risk using the Australian National Breast and Ovarian Cancer Centre criteria and willing to donate a fresh tumour and blood specimen.
Eligibility for inclusion in the present study required each participant to: (ii) be a carrier of a germline pathogenic mutation in BRCA2; (ii) have a confirmed diagnosis of prostate cancer in which LOH was examined at the mutation locus in adenocarcinoma; and (iii) have access to archival tissue that has pathological evidence of HGPIN. Mutation status was confirmed by either a predictive clinical mutation test or by extended genotyping within the family. All prostate cancers were verified by a diagnostic pathology report. Written informed consent was obtained from all participants (or proxy consent from the next of kin where those affected with prostate cancer were deceased) and ethics approval for the project was obtained from the human research ethics committee at the Peter MacCallum Cancer Centre.
Flanking 5-µm sections of archival prostate specimens (nine radical prostatectomy and one TURP specimen) were stained with haematoxylin and eosin (H&E) and reviewed by a uropathologist (D. C.) according to the International Society of Urological Pathology guidelines 2005 and structured reporting guidelines of the Royal College of Pathologists Australia (RCPA) [36,37]. Areas of HGPIN and adenocarcinoma were identified and outlined with a marker pen on the H&E slides for each study participant.
LASER CAPTURE AND NUCLEIC ACID ISOLATION
Tissue laser capture microdissection
After cresyl violet staining  of five serial 5-µm formalin-fixed paraffin-embedded archival prostate specimen sections, HGPIN and adenocarcinoma were dissected using laser capture microdissection on the Arcturus VeritusTM Microdissection system (Applied Biosystems, Foster City, CA, USA). The laser capture microdissection was carried out for specific regions that had previously been scored on the corresponding flanking H&E sections. To ensure a sufficient amount of DNA was extracted, several foci of HGPIN were captured and dissected from all of the HGPIN cases.
Nucleic acid isolation
DNA was isolated from laser microdissected HGPIN and adenocarcinoma cells using the DNeasy blood & tissue kit (Qiagen, Hilden, Germany) and a modification of the protocol by Wu et al. , which involves a 3-day incubation at 56°C with the addition of 240 µg proteinase K at 24-h intervals.
LOSS OF HETEROZYGOSITY TESTING
To determine LOH at the specific family mutation locus of BRCA2 within HGPIN and corresponding adenocarcinoma, DNA analysis was performed via mutation-specific PCR and Sanger sequencing.
Nine archival radical prostatectomy tissue blocks and one TURP specimen were obtained from BRCA2 pathogenic mutation carriers with clinically localized prostate adenocarcinoma. The median age of diagnosis of this group was 61.5 (53–71) years; and 9/10 of the prostate adenocarcinomas had a Gleason score ≥7 (Table 1). One of the adenocarcinomas was staged as pT1b (TURP specimen), two as pT2a, three as pT2c, and the remaining four as pT3a or pT3b (Table 1). All of the participants in the study had a family history of breast cancer. In addition, a number of participants also had a family history of ovarian cancer and/or bowel cancer, while five had a family history of prostate cancer in 1st or 2nd or 3rd degree.
Table 1. Clinical characteristics of study participants with HGPIN
| 1||BRCA2c.538_539dupAT (p.Ser181PhefsX5)||71||T2a||3+4 (7)||No|
| 2||BRCA2c.5946delT (p.Ser1982ArgfsX22)||64||T2c||4+5 (9)||Yes|
| 3||BRCA2c.8904delC (p.Val2969CysfsX7)||65||T3b||5+4 (9)||Yes|
| 4||BRCA2c.5286T > A (p.Tyr1762X)||66||T1b†||4+5 (9)||Yes|
| 5||BRCA2c.1813delA (p.Ile605TyrfsX9)||62||T2c||4+4 (8)||Yes|
| 6||BRCA2c.9117G > A (p.Val2985fs)||58||T3a||4+3 (7)||No|
| 7||BRCA2c.7757G > A (p.Trp2586X)||54||T3a||3+3 (6)||No|
| 8||BRCA2c.3847_3848delGT (p.Val1283LysfsX2)||61||T3a||4+5 (9)||INC|
| 9||BRCA2c.8297delC (p.Thr2766AsnfsX11)||53||T2c||3+4 (7)||No|
|10||BRCA2c.3847_3848delGT (p.Val1283LysfsX2)||53||T2a||3+4 (7)||No|
Nine of the 10 cases were examined for LOH at the BRCA2 locus in the laser capture microdissected adenocarcinoma, with 4/9 displaying LOH. The final sample did not produce a conclusive LOH result. By contrast, none of the 10 cases of laser capture microdissected HGPIN displayed LOH at the BRCA2 mutation locus.
The present study is the first to show that identified areas of HGPIN do not show LOH at the BRCA2 mutation locus in men with aggressive prostate cancer. Furthermore, this study confirms our previous findings that LOH is observed in associated adenocarcinoma. To date, no genetic tests have been developed to identify hallmarks of early prostate disease, nor have any genetic markers been elucidated for the clinical management for HGPIN or prostate cancer.
Molecular genetics is advancing our understanding about premalignant lesions and their role in tumorigenesis, and is assisting in the development of new clinical tools for cellular description, diagnosis and drug targets. There have been advances in the clinical management and therapy of premalignant lesions to prevent the development of invasive disease, for example, DCIS of the breast , high grade cervical intraepithelial neoplasia  and the removal of multiple polyps in colorectal tissue . HGPIN has been considered a precursor lesion to the development of adenocarcinoma evidenced by a number of biochemical, morphological and genetic changes usually associated with cancer [29,30,33,43,44]. In addition, normal and tumour tissue dissected from radical prostatectomy specimens have provided evidence that the human adult prostate epithelium is maintained by a population of multi-potent stem cells as single clonal units of both PIN and adenocarcinoma .
As a result of the findings in 2008 that 71% prostate adenocarcinomas displayed LOH at the mutation locus in BRCA2 mutation carriers , this study was performed under the assumption that mutations in BRCA2 drive prostate cancer development. The ensuing expectation was that the LOH event would also occur in the precursor HGPIN, thereby providing a genetic marker in the same manner as for DCIS in the breast . There are two important observations from the expansion of our original work. Firstly, in doubling the size of the original study with an additional 13 prostate adenocarcinoma cases, we have seen the percentage of cases displaying LOH fall from 71% to 50%, leading us to query if the presence of a BRCA2 mutation alone is the main driver of tumorigenesis in this group of men.
The need for a greater understanding about the presence and role of a BRCA2 mutation in the development of prostate cancer in this group of high risk men was highlighted in our 2011 study, in which we performed a retrospective study of 136 men with a diagnosis of prostate cancer from breast cancer-rich families. BRCA2 mutation carriers who developed prostate cancer had a median survival of 4.5 years. Interestingly, those men who did not carry the family-specific BRCA2 mutation, yet had a strong family history of breast cancer, also had a poor median survival of 4.9 years . As LOH at the BRCA2 mutation locus was not identified in the HGPIN cases of the present study, an alternative mechanism of gene inactivation, including epigenetic changes or a later mutational event along the BRCA2 pathway, not specific to HGPIN tissue, may be at play.
Ribeiro et al.  have identified two distinct initiating events that lead to prostate tumour development, one involving HGPIN and loss of chromosomal arm 8q, the other involving 13q loss which is independent of HGPIN. Ribeiro et al. suggest that these two pathways converge as the tumour progresses, resulting in genetically complex and heterogeneous cancers . As the BRCA2 gene is located on chromosome 13q  and as loss of 13q is rarely seen in HGPIN  it is perhaps not unexpected that the present study did not observe LOH at the BRCA2 mutation locus. There are many reports where loss within a region of a gene is not accompanied by aberrations of the opposing allele, suggesting that the initial event, be it a mutation or a genomic deletion, may not directly relate to carcinogenesis at all, but rather be a ‘passenger’ event . Alternatively, our results may indicate the phenomenon of haploinsufficiency, whereby the presence of the first hit (BRCA2 mutation) may impact upon tumorigenesis by promoting loss of an alternative tumour suppressor gene or by predisposing the mutant cells to alternative tumour subtypes [47,48].
Using this well characterized group of men, our future research will examine the whole genome for genetic abnormalities observed in the adenocarcinoma and HGPIN of BRCA2 mutation carriers. By contrasting these changes to the changes seen in patients from BRCA2 carrier families who do not carry the mutation, we hope to identify and map alternative genetic pathways in the development of aggressive disease, with the aim of identifying new genetic markers that are common to both groups. With this information, we may be better equipped to propose an alternate genetic mechanism for prostate tumour development.
We would like to thank Associate Professor Ian Campbell, Dr Ella Thompson and Professor Joe Sambrook for their invaluable assistance, advice and guidance throughout the development of this project and preparation of the manuscript. We would also like to thank Dr Sarah Ellis (Head of Microscopy Core, Peter MacCallum Cancer Centre) for her expert assistance and instruction on laser capture microdissection. We also wish to thank the kConFab data manager, Eveline Niedermayr; the kConFab research nurses and staff; the heads and staff of the Family Cancer Clinics; and the Clinical Follow-Up Study (funded by NHMRC grants 145684, 288704 and 454508) for their contributions to this resource. Special thanks go to the many families who contribute to kConFab. kConFab is supported by grants from the National Breast Cancer Foundation, the National Health and Medical Research Council (NHMRC) and by the Queensland Cancer Fund, the Cancer Councils of New South Wales, Victoria, Tasmania and South Australia, and the Cancer Foundation of Western Australia.
CONFLICT OF INTEREST