Translational pathology of neoplasia

Abstract

With the increasing use of individualized medical care (personalized medicine) in treating and managing patients with cancer, the utilization of biomarkers in selecting and tailoring such medical approaches also is increasing and becoming more important. Specifically, many therapies are effective against only a subgroup of a specific type of tumors and exposing patients with different non-responsive subgroups of the same tumor to ineffective therapies, not only exposes these patients needlessly to acute and chronic side effects of the therapy, but also adds to the costs of medical care. For example, the Oncotype Dx test for estrogen receptor positive tumors that are node negative has been used to identify low risk tumors for which surgery alone is an adequate therapy. Biomarkers may be used to aid in multiple aspects of medical care related to cancer, including early detection, diagnosis, risk assessment, as well as in predicting the aggressiveness of cancers (i.e., prognosis) and predicting the therapeutic efficacy of treatments (i.e., prediction). Biomarkers may be also used as surrogate endpoints to aid in evaluating therapies and preventive approaches. Types of biomarkers vary greatly and include histopathologic appearance, stage of the lesion, quantitative morphologic features, size of the lesion, metastatic pattern and extent of metastasis, as well as imaging and molecular features. The types of measurements of biomarkers also vary; for example, molecular features can be measured at the DNA, mRNA or protein levels as well as at regulatory levels (e.g., microRNA). The usefulness of each biomarker is limited by its sensitivity and specificity in fulfilling its role (e.g., in early detection) and the requirements of sensitivity and specificity to accomplish specific tasks are affected by multiple variables. For example, both very high specificity and sensitivity of a test are required to screen a population with a low prevalence of a specific tumor. The goal of this manuscript is to introduce the reader to how biomarkers may be used and the limitations on the uses of biomarkers in translational research.

Fig. 1

Demonstrates the complex interaction between a tumor, its surround and the immune system. It shows how various biomarkers may be associated with the presence of a tumor without coming directly or uniquely from a tumor. For example, PSA is produced at higher levels in the cells of normal glandular epithelia of the prostate than in cells of prostate cancer.

Fig. 2

Demonstrates the molecular subtyping of breast carcinoma. The abbreviations used are defined subsequently. (GATABP-3) GATA binding protein 3; (X-BBP-1) box binding protein 1; (TRF-3) trefoil factor 3; (HNF3α) hepatocyte nuclear factor – 3α; (PTP4A2) protein tyrosine phosphatase type IVA member; (TRAF3) Tumor necrosis factor receptor associated factor 3; (BAP-1) BRCA associated protein 1; (KPT5, -6 or -17) keratin 5, 6 or 17; (MTIX) metaiothionein 1x; (FABP7) fatty acid binding protein 7; (SFRP1) frizzled related protein 1; (ATF3) activating transcription factor 3; (CAV1 or 2) caveolin 1 and 2; (HGF) hepatocyte growth factor; (TGFβR2) transforming growth factor β receptor II; (ABCB1) multidrug resistance protein 1; (S100P) S100 calcium binding protein P; (FBX09) fatty acid syntase; (RALB) GTP binding protein; (RAB6A) member of RAS oncogene family; (FN1) fibronectin 1; (SDC1) syndecan 1. In the future, the complex cancer such as lung or renal carcinomas also are likely to be subdivided molecularly.

Fig. 3

Demonstrates how molecular analysis aids in identifying poorly differentiated malignant tumors, especially metastatic cancers, whose primary cancer is unknown. Typically the pathologist may suspect which type of tumor is the primary cancer and only a few of these molecular markers are used to confirm the initial impression.

Fig. 4

In Fig. 4, note biomarkers x and y may change at site Z1 without the resolution of the area of pre-invasive neoplasia. This would correlate with an effect of the drug and not a pattern indicative of a surrogate endpoint. Similarly, areas of pre-invasive neoplasia may “move” spontaneously, with or without resolution of the extent of the PINN. In this case, there is partial resolution of the area of PINN. The movements of PINN lesions over a period of months without even partial resolution of the PINN is why the use of “controls” is so important in studies to identify responses in SEBs to drugs used in the prevention of neoplasia [51].