Clinical and Translational

2 downloads 0 Views 248KB Size Report
20 and CD15) in the differential diagnosis of renal tumors using a tissue microarray that included 23 CCRCC, 20 PRCC, 17. ChRCC, and 19 oncocytomas.
Pathology Research Clinical and Translational A Publication of the Division of Pathology and Laboratory Medicine at The Cleveland Clinic ◆ Issue 1 Winter 2005

In This Issue Page 2

Molecular Genetic Markers in the Diagnosis and Therapeutic Management of Breast Cancer

4

Immunohistochemical Markers in the Differential Diagnosis of Renal Neoplasms

6

Real-Time PCR for the Detection of Legionella Pneumophila

ON THE COVER: Simultaneous demonstration of HER2 (ERBB2) gene amplification and encoded HER2 protein expression in the breast cancer cell line SK-BR-3, an illustration of the translational research performed by the interdisciplinary Molecular Genetic Markers in Breast Cancer research group (MGMBC). The CODFISH (concomitant oncoprotein detection with fluorescence in situ hybridization) technique, developed in the Cleveland Clinic Division of Pathology and Laboratory Medicine, is illustrated.

CLINICAL AND TRANSLATIONAL

THE CLEVELAND CLINIC FOUNDATION

1

COVER STORY

The MGMBC group includes: (bottom row) Shannon Tarr, Karen Hawkins, Joanne Woletz, B.S.N., R.N., Catherine Vanderboom, M.S.N., R.N., Kelly Simmerman, and Marybeth Hartke; (second row) David Hicks, M.D., Bryan Williams, Ph.D., Joseph P. Crowe, Jr., M.D., Director Breast Center, Raymond Tubbs, D.O., David Weng, M.D., James Cook, M.D., and Halle Moore, M.D.; (top row) James Pettay, Joseph DiDonato, Ph.D., G.Thomas Budd, M.D., Graham Casey, Ph.D., and Eric Swain.

Molecular Genetic Markers in the Diagnosis and Therapeutic Management of Breast Cancer By David G. Hicks, M.D.

A

denocarcinoma of the breast is a leading cause of cancer mortality among women in the United States with an estimated 200,000 new cases and 40,000 deaths occurring in 2003. The successful treatment of breast cancer depends on a number of complex factors, including the discovery of the tumor early in its course of development, the biology underlying the disease, and appropriate treatment. The decision to stratify patients newly diagnosed with breast cancer as to the need for adjuvant therapy has traditionally been based on the tumor grade and the anatomic extent of disease (1). However, it is increasingly clear that current prognostic factors in clinical use do not 2

provide sufficient information to allow accurate individual risk assessment for node-negative breast cancer patients, emphasizing the need for new prognostic and therapeutic strategies for early-stage breast cancer. This is particularly true in light of the development of new, targeted cancer treatments that require companion diagnostic tests to predict their effectiveness. In this regard, a greater understanding of the molecular and cellular basis of breast cancer phenotypes and their relevance and how this pertains to clinical behavior and aggressiveness of disease are important first steps in identifying and evaluating potential new molecular targets for clinical practice, as well as the develop-

ment of new diagnostic biomarkers. The Molecular Genetic Markers in Breast Cancer working group (MGMBC) is an interdisciplinary group of clinicians, pathologists, and basic scientists from The Cleveland Clinic led by Joseph Crowe, M.D., director of the Cleveland Clinic Breast Center, who meet weekly to facilitate the development of an integrated translational research program in breast cancer. Through an innovative, multidisciplinary collaboration involving clinical and basic research programs, the group has sought to synergistically apply the latest research findings to discover better ways to successfully treat and eradicate this disease. When a diagnosis of breast cancer is PATHOLOGY RESEARCH

made, the immediate questions of importance to the patient and her doctor are the future implications of the disease and her treatment options. In current clinical practice, there is an unmet need for robust prognostic and predictive markers that can be applied at the time of initial diagnosis to help stratify an individual patient’s risk and her likelihood for recurrence. This information would be invaluable clinically to help optimize adjuvant systemic therapies for those patients with a high risk for recurrence, while optimizing local and hormonal therapies for patients with an excellent prognosis, thus avoiding unnecessary and potentially toxic treatment regimens that may be of little or no benefit. Fundamental advances in molecular biology, including expansion of our knowledge of DNA sequences and the development of powerful array technologies for analyzing gene expression, have made it possible to begin to explore the specific genetic alterations that underlie human malignancy. These discovery tools are being used to help identify clinically relevant biologic differences between tumors, which can be translated into new diagnostic molecular tests (2-4). In support of the MGMBC collaborative research efforts, the molecular pathology laboratories of the Division of Pathology and Laboratory Medicine are organizing the clinical tissue archives and developing the infrastructure to prepare and quantitatively analyze breast cancer tissue microarrays by immunohistochemistry and in situ hybridization. Pathologists are working with clinicians dedicated to the care of breast cancer patients to develop an IRB-approved breast cancer database with tumor samples linked to treatment and clinical follow up. Through this collaborative partnership, novel basic research findings can be rapidly and rigorously evaluated against large numbers of clinical samples. Synergistically, the direct interrogation of human tumor samples can refine the focus of the mechanistic studies being conducted in cell culture within basic science laboratories. The MGMBC group is uniquely wellpositioned to study these exciting questions given the state-of-the-art research CLINICAL AND TRANSLATIONAL

facilities and expertise in cellular, cancer, and molecular biology existing within the Lerner Research Institute. These resources, combined with the Clinic’s active clinical programs serving a large patient population with breast cancer, enable both retrospective and prospective studies. The large pathology archives of both snapfrozen and paraffin-embedded breast cancer tissue samples and the molecular pathology infrastructure and expertise allow for the identification of distinct patient subsets and rapid analysis of molecular and genetic markers — molecular signatures that can be correlated with patient outcomes. An early success from these efforts has been a multidisciplinary collaboration among the laboratory of Lerner Research Institute cell biologist Josephine Adams, Ph.D., the Department of Hematology and Medical Oncology, and the molecular pathology laboratories of the Division of Pathology and Laboratory Medicine. Dr. Adams’ laboratory includes extensive expertise in the study of fascin, (an actinbundling cell motility protein), cell-ECM adhesion, and cellular motility. The group has begun to investigate whether fascin expression in breast cancer can be used to help identify a subset of patients who will experience a more aggressive clinical course via enhanced tumor cell motility and metastasis. In a preliminary study of 210 consecutive newly diagnosed breast cancer

Figure 1: Dual immunostaining for fascin (red) and estrogen receptor (brown) in infiltrating ductal carcinoma of the breast. Staining was performed on a Benchmark® automated stainer (Ventana Medical Systems, Tuscon, Ariz.). The expression of fascin, an actin-bundling cell motility protein, may be one of the mediators of aggressive disease in a subset of breast cancer patients through enhanced cell motility, invasion, and metastatic spread. The absence of fascin in ER- positive tumor cells is illustrated.

patients with clinical follow-up, fascin was markedly over-expressed in 16 percent of tumor samples as demonstrated by immunohistochemistry (Figure 1). The subset of breast cancer that demonstrated fascin expression were more commonly hormone receptor-negative, of higher nuclear grade, and clinically more aggressive with a significantly decreased diseasefree and overall survival (5-7). If fascin expression can be confirmed to correlate with a more aggressive clinical course for these tumors, this could offer a significant novel target for prognosis and therapy of breast cancer. This collaboration has lead to two presentations at national meetings and a manuscript currently in press with Clinical Cancer Research (5-7). In addition, a grant proposal is being prepared to support the continuation of these studies. Similar collaborations are currently under way with the laboratories of Lerner Research Institute investigators Joseph DiDonato, Ph.D., Graham Casey, Ph.D., and Bryan Williams, Ph.D. Dr. DiDonato has been awarded a grant to study TOLL-like receptor expression in breast cancer, and Dr. Casey has submitted a grant to study metastasis-suppressor gene expression pathways and signaling in breast cancer, both in collaboration with the MGMBC. Additionally, Thomas Budd, M.D., David Hicks, M.D., and Raymond Tubbs, D.O., have submitted grant applications to the Breast Intergroup for the study of TOP2A deletions and amplifications as breast cancer markers predicting clinical response to anthracycline chemotherapy. The most clinically relevant diagnostic testing to help determine the prognosis for breast cancer patients will continue to be an area of active investigation. Studies such as these continue to challenge traditional diagnostic pathology with opportunities to more closely link tumor diagnosis with therapeutic decision-making. The need for such testing will only increase with the development of new and novel targeted therapies. The future of breast cancer diagnosis and treatment will come from a better understanding of the molecular basis for a tumor’s biologic and clinical behavior. Continued on page 4 3

Immunohistochemical Markers in the Differential Diagnosis of Renal Neoplasms By Ming Zhou, M.D., Ph.D., and Cristina Magi-Galluzzi, M.D., Ph.D.

R

enal epithelial neoplasms include a wide array of malignant and benign tumors with vastly different biological behaviors and therapeutic options. Therefore, accurate classification of a renal tumor plays a critical role in patient management. Traditionally, the classification of renal neoplasms is based on hematoxylin and eosin (H & E) histology. Renal cell carcinoma (RCC) is the most common malignant renal tumor involving the adult kidney and is further sub-classified into clear cell (CCRCC), papillary (PRCC), chromophobe (ChRCC), collecting, and medullary carcinoma. However, many cases have overlapping morphological features and are difficult to classify accurately based on histomorphology alone. Different histological subtypes have distinct genetic changes. For example, CCRCC has alterations involving chromosome 3p, often resulting in inactivation

Ming Zhou, M.D., Ph.D., and Cristina Magi-Galluzzi, M.D., Ph.D., have determined that the application of a specific panel of immunomarkers can distinguish real neoplasm subclasses based on characteristic genetic changes.

of von Hippel-Lindau gene. PRCC, in contrast, has trisomy 7 and 17. ChRCC is hypodiploid with extensive chromosomal loss. Although potentially more precise, the classification based on genetic changes

has yet to be vigorously validated in clinical trials to become a routine diagnostic tool for RCC. Proteomics, on the other hand, offers a cross-sectional representation of the

Molecular Genetic Markers in the Diagnosis and Therapeutic Management of Breast Cancer Continued from page 3 This understanding will most certainly suggest new and novel targeted therapeutic strategies and will help us reach the goal of tailoring individual treatment regimens for newly diagnosed breast cancer patients to maximize the benefit from therapy. The pathology laboratory has a critical role to play in these new developments and must continue to work closely with clinicians and basic scientists to help translate relevant biologic differences among tumors into new companion diagnostic testing, which can be reliably and reproducibly utilized in clinical diagnosis. ◆ 4

References 1. Jemal A, Tiwari RC, Murray T et al. Cancer Statistics 2004. Ca Cancer J Clin 2004; 54:8-29. 2. Van de Vijver MJ, He YD, van t Veer LJ et al. A gene expression signature as a predictor of survival in breast cancer. N Engl J Med 2002; 347:1999-2009. 3. Perou CM, Sorlie T, Eisen MB et al. Molecular portraits of human breast tumors. Nature 2000; 406:747-752. 4. Sorlie T, Perou CM, Tibshirani R et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA 2001;98:10869-10874. 5. Yoder BJ, Tso E, Skacel M, Choueiri T, Tarr S, Budd T, Tubbs RR, Adams J, Dawson AE, Hicks DG. The expression of fascin, an actin bundling

motility-associated protein in breast cancer. Mod Pathol 2004; 17:215A. 6. Yoder BJ, Pettay J, Tarr SM, Ruddy T, Tso E, Budd T, Adams J, Tubbs R, Crowe J, Chouieri T, Hicks D. Fascin expression correlates with hormone receptor negative breast cancer and a more aggressive clinical course: an immunohistochemical and image analysis study. (accepted for the 27th annual San Antonio Breast Cancer Symposium). 7. Yoder BJ, Tso E, Skacel M, Pettay J, Tarr S, Budd T, Tubbs R, Adams J, Hicks DG. The expression of fascin, an actin-bundling motility protein, correlates with hormone receptor negative breast cancer and a more aggressive clinical course. Clin Cancer Res In Press.

PATHOLOGY RESEARCH

H&E

CK7

CK20

CD15

CCRCC Figure 1: Expression of CK7, CK20, and CD15 in clear cell RCC (CCRCC), chromophobe RCC (ChRCC), and oncocytoma.

ChRCC

Oncocytoma

Table 1: Expression of CK7, CK20 and CD15 in Renal Neoplasms CK7 CCRCC PRCC ChRCC Oncocytoma

% + cases

% focal +

26.1% 100% 100% 78.9%

21.7% 5% 35.3% 78.9%

CK20 % diffuse + % + cases % + cases

4.3% 95% 64.7% 0%

complex gene expression in a tumor. Each tumor often has a unique, or signature protein expression pattern. We recently studied the utility of a panel of immunohistochemical markers (cytokeratins 7 and 20 and CD15) in the differential diagnosis of renal tumors using a tissue microarray that included 23 CCRCC, 20 PRCC, 17 ChRCC, and 19 oncocytomas. Different histological subtypes of RCC showed distinct expression patterns of cytokeratin 7 (CK7), cytokeratin 20 (CK20), and CD15 (Table 1 and Figure 1). The majority of CCRCC (74%) were negative for CK7. All CCRCC were CK20 negative, but CD15 expression was detected in 65% of CCRCC. Nineteen of 20 (95%) PRCC were strongly positive for CK7, and only one (5%) was positive for CK20. Eighteen PRCC (90%) were positive for CD15. All ChRCC were positive for CK7, with focal positive staining in 35% and diffuse positive staining in 65% of cases. CK20 was detected in 6% of ChRCC. CD15 expression was seen in 17% of ChRCC, with only 1.5% of cells staining. CLINICAL AND TRANSLATIONAL

CD15

0% 5% 5.9% 52.6%

65.2% 90% 17.6% 73.7%

% + cells

53.9% (10-100) 47% (1.7-100) 1.5% (1-1.7%) 39.8% (3.3-100)

Oncocytomas, on the other hand, showed a very different pattern, with weak and focal CK7 staining pattern present in 79% of cases. Most oncocytomas were positive for both CK20 and CD15 (53% and 74%, respectively). Three different patterns of CK20 expression were observed in oncocytomas: paranuclear dotlike, perinuclear, and membranous. Based on these results, we designed several subtype-specific classifiers, or immunomarker panels, to maximize the diagnostic utility of these markers in the differential diagnosis of renal tumors. We tested the CCRCC classifier “CD15+ (>5%)” or “CK7-” and the ChRCC classifier “CD15- (