Leaders in Pharmaceutical Business Intelligence Group, LLC, Doing Business As LPBI Group, Newton, MA

Personalized Medicine is not yet here

Writer and Curator:  Larry H Bernstein, MD, FCAP

LPBI

3.4 Personalized Medicine is not yet here

ESMO Personalized Medicine
Written by Dr Marina Garassino for ESMO
https://www.esmo.org/content/download/20122/337223/file/ESMO-Patient-Guide-Personalised-Cancer-Medicine.pdf

The aim of personalised medicine is clearly to make therapy more efficient for patients. A very, very small step in the process is to try to identify for every patient the main molecular driver of their tumour. We have to understand that patients differ between each other, although they may have the same cancer type; for example, every patient with breast cancer or bowel cancer will have a unique tumor. This is entirely new knowledge, so what we are trying to do now in the medical community is to identify for each patient his/ her type of disease and then to give the drug that will work best. We are moving forward with an incredible amount of new data and innovative knowledge on genetic characteristics and subsequent proteomic changes* in the tumor. The challenge is now about how to exploit this information in order to offer targeted treatment and generally improve patient care.

For a number of years we have classified tumors according to their site of origin and using a classification system called “TNM”. Researchers and clinicians once thought that all cancers that derived from the same site were biologically similar and they differed perhaps only in their pathohistological* grading. This grading is a score which classifies tumors from 1 to 3, where 1 is the least aggressive tumor and 3 is the most undifferentiated tumor. Other clinical differences were distinguished based on the presence of regional node metastases or distant metastases. Most of the tumors were therefore classified within the “TNM” system, where T corresponds to the diameter of the primary tumor, N to the presence of regional nodes, and M to distant metastases. For at least three decades, personalization of oncology was based only on these parameters and on the patient’s physical condition, and even now these represent the fundamental elements for treatment decisions. Chemotherapy, surgery and radiation therapy were once the only treatment options for cancer. Although these treatments are still used, oncologists know that some patients respond better to certain drugs than to others and that a surgical approach is not always indicated. In recent years, researchers have studied thousands upon thousands of samples from all types of tumors. They have discovered that tumors derived from the same body site can differ in very important ways.

Firstly, there is histology*. The pathologist is able to distinguish different subtypes of cancer with the microscope. When a patient is diagnosed with a cancer, he/she will undergo a biopsy or a fine-needle aspiration. In some tumor types, debulking or removal of the primary tumor also allows sampling for tissue examination. Some cells of the tumor which have been removed will be taken and analyzed. This examination allows the pathologist to confirm a cancer diagnosis, but, through particular colorations of the tissue sample, the pathologist is also able to provide clinicians with a lot of additional information, such as the tumor’s histological characterization, its hormone sensitivity, and its grade of differentiation*.

For example, in the treatment of lung cancer the histology provides very useful tools to decide the best drug for the treatment of the patient. Clinical studies have shown that for a patient with lung adenocarcinoma* there might be more chance of a response if the drugs pemetrexed or bevacizumab are added to the chemotherapy, while for a patient with lung cancer of squamous* histology, it would be more beneficial to add gemcitabine or vinorelbine. A similar example may be observed Personalization of Oncological Treatments: The Story 12 for other cancers. For the treatment of esophageal cancer it is mandatory to know if the tumor is squamous or not, because although deriving from the same organ, the treatment approach is completely different.

This information is a useful tool in the first step of the personalization process. For example, lung cancer can be divided as a first step into non-small cell lung cancer and small cell lung cancer, which are two completely different neoplasms*. Within the non-small cell lung cancer category, there are again several different tumor types. Breast cancer can also be divided into two major categories: the hormone-sensitive neoplasms and the HER2-positive diseases. Lung and breast cancers are only two examples, because it is possible to recognize several entities within the same tumor type for many other cancers.

Molecular subsets of lung adenocarcinoma Lung cancer subtypes
Figure 2. Lung Cancer – Not One Disease: Histological (Tissue) and Molecular Subtypes of Lung Cancer (not shown) On the left side, four histological subtypes of lung cancer. On the right side, a pie chart showing the percentage distribution of molecular subsets of lung adenocarcinoma. Adapted from Petersen I. Dtsch Arztebl Int 2011; 108(31-32):525-531 (left) and Pao W & Hutchinson KE. Nature Med 2012; 18(3): 349-351.

Personalization depends on a multidisciplinary approach; we need a range of experts, because we need the medical oncologist, the surgeon and the expertise of the molecular pathologist, who should be part of the team in a more effective, integrated way than before. We don’t need the pathology report alone; we need to interact with all professionals, including nurses, who are dealing with the patient. This, to me, will create a lot of problems in terms of organization of care and in terms of cost, but it is the only way to bring together knowledge on the biology and pathology of tumors for effective treatment in every single patient. Our effort at ESMO is to bring this broad knowledge to the general public, to medical oncologists and to the community of doctors involved in cancer.

We have to deeply analyze each tumor of every patient in order to identify those genetic characteristics that make the tumor able to survive. As a result, we can choose the appropriate drugs to target the specific alterations. The clearest examples of this process are in melanoma, lung cancer and breast cancer. For instance, in lung cancer, the presence of mutations in the epidermal growth factor receptor (EGFR) renders the tumor highly sensitive to EGFR tyrosine kinase inhibitors. When oncologists identify these mutations in a patient’s tumor, they may observe that the lesion disappears a few weeks after treatment. A similar response may be observed after treatment with BRAF inhibitors in patients with melanoma or with gastrointestinal stromal tumors (GIST) that express the c-kit gene. Unfortunately, oncogene addiction is not the only process underlying carcinogenesis* and tumor growth. The tumor environment and so-called “epigenetic” alterations* play an important role in rendering the fight against cancer more and more challenging. Despite the enormous recent advances, a specific alteration has not been identified in all cancers. The hope is that the possibility of sequencing the full genome – which means every gene – will give us new insights and therefore new drugs for our patients.

In the DNA of some individuals a “germline” mutation* may be present. This means that a particular mutation is conferring susceptibility to that person to develop a particular type of cancer during his/her life. For instance, BRCA is an alteration for which there is a particular predisposition to have a breast cancer or ovarian cancer in one’s life. A woman with a BRCA gene mutation can transmit this alteration to her female descendants, so her daughters and following generations of female family members can therefore inherit this predisposition.

Mutations that are not germline are called somatic mutations*, which are acquired mutations and are found generally only in the tumor. Distinct from germline mutations, somatic mutations are not inherited.

The move from blockbuster or empirical medicine* towards personalized medicine is a stepwise process. We are currently on the second step of stratified medicine and moving up the stairs towards personalized medicine.

Will molecular pathology evolve from pathology? You need to give a name to a tumor, and a pathologist is the professional who gives a name to tumors. The variety of cancers is broad; when we say “sarcoma”, “carcinoma”, or “lymphoma”, we actually say nothing, because we have hundreds and hundreds of diseases within these categories that need to be recognized. And the reason for recognizing them is exactly related to personalization. The biology of cancer is very complex, and admittedly we have been very naive in the past. We always thought that the problem was how genes become altered in the cancer cell, but actually it is even more complex than that and also involves the way genes direct how they are read; it is the flow of information that comes from genes to the making of their proteins which is as important as the aberration of the genome.

We are facing obstacles currently because the whole issue of tissue sampling has been regulated under the umbrella of privacy, which is of course important. Defending your rights as a human being is a key issue, but we should also try to focus a little bit on the necessity to use that tissue. Of course, we need to have rules, but the approach we are currently facing is basically preventing clinical research and translational research under the excuse of protecting our privacy as human beings, and this is an increasing obstacle. We as researchers, as molecular geneticists, as pathologists, are really looking into a future in which it is becoming increasingly difficult to try to answer the basic question of cancer genomics. Why? Because it is becoming increasingly difficult to use tissue for these purposes.

With the new therapeutic approach and the use of targeted therapy, molecular testing is gaining a very relevant role. It is very important for us, as advocates, to educate patients in these issues. So patients have to receive very clear and transparent information. It should be the doctor who explains to the patient the reason why molecular testing is performed; the doctor has to explain that molecular testing will find whether there is some tumor characteristic which can be targeted with one of these therapies, in order to determine if maybe the patient is the right candidate to receive targeted therapy and perhaps to benefit from it. The communication between the doctor and patient must be very accurate and must educate, meaning that the patient has to understand the precise situation. This can be important also to empower the patient in treatment decisions, but it is important that he/she knows that not every patient may be a candidate for receiving targeted therapy and to understand why this is the case.

• Different tumour types are increasingly divided into very small subgroups carrying a rare molecular alteration.
• Most new drugs are targeting these infrequent events.
• Clinical trials are testing the use of high throughput molecular technologies* in the context of personalized cancer medicine.
• There are a growing number of newer techniques to optimize genomic testing, including the virtual cell program, which foresees testing of a piece of patient’s tumor tissue in the laboratory in order to mimic what would happen in the human body (e.g. drug sensitivity).
• Clinical research is today focusing on target identification at the patient level.

Targeted therapy drugs work differently to standard chemotherapeutic drugs. They attack cancer cells and, in particular, the targets which are strategic points for cell survival, cell replication and metastases. They generally create little damage to normal cells. In fact, these drugs tend to have different side effects to traditional chemotherapeutic drugs. Targeted therapies are used to treat many kinds of tumors: certain types of lung, pancreatic, head and neck, liver, colorectal, breast, melanoma and kidney cancers. Targeted therapies are a major focus of cancer research today

Many future advances in cancer treatment will probably come from this area. There are many different targeted therapies in use and new forms are appearing all the time. Depending on the type of cancer and the way it spreads, targeted therapy can be used to cure the cancer, to slow the cancer’s growth, to kill cancer cells that may have spread to other parts of the body or to relieve symptoms caused by the cancer.

We can divide targeted therapies into two main categories: antibody drugs and small molecules. Antibody drugs are man-made versions of immune system proteins that have been designed to attack the external part of cells at certain targets, generally called receptors. Receptors can be considered the antennas of the cells. They transmit signals from the surrounding environment to the nucleus of the cell. Some receptors are fundamental to the vital processes of the cell. Targeting certain receptors means preventing the transmission of some survival signals to the tumor cells.

Trastuzumab (Herceptin®) is, after tamoxifen, the second targeted therapy drug ever used to treat cancer and it is a monoclonal antibody directed at a receptor called HER2. This targeted therapy greatly improves the survival rate of women with breast cancer expressing the HER2 receptor. Therefore, the determination on tissue blocks of the presence of expression of HER2 is one of the best examples of personalization of treatment.

A knowledge of the cancer characteristics and a determination of the tissue characteristics of each patient allows the doctor to select patients for the best treatment.

Other examples of monoclonal antibodies are cetuximab and panitumumab, which have been developed to treat colon cancer. At first it seemed as if these drugs were a failure, because they did not work in many patients. Then it was discovered that if a cancer cell has a specific genetic mutation, known as KRAS, these drugs will not work.

This is another excellent example of using individual tumor genetics to predict whether or not a treatment will work. In the past, the oncologist would have had to try each therapy on every patient and then change the therapy if the cancer continued to grow.

The other type of targeted therapy drugs are not antibodies. Since antibodies are large molecules, this other type is called “small-molecule” targeted therapy drugs. The small molecules attack cancer cells from the inner vital processes. Also, in this case, the small molecules prevent the broadcast of vital signals that regulate the survival of the tumor. There are several examples of targeted drugs that changed the natural history of some cancers.

One example is imatinib mesylate (Gleevec®), which is used in GIST, a rare cancer of the gastrointestinal tract, and in certain kinds of leukemia. Imatinib targets abnormal proteins, or enzymes, that form on and inside cancer cells and promote uncontrolled tumor growth. Blocking these enzymes inhibits cancer cell growth. Gefitinib (Iressa®) is used to treat advanced non-small cell lung cancer. This drug hits the internal part of the EGFR. These receptors are found on the surface of many normal cells, but certain cancer cells have many more of them. EGFR take in the signal that tells the cell to grow and divide. When gefitinib blocks this signal, it can slow or stop cell growth. However, gefitinib does not work in all patients when trying to treat lung cancer, but only

Personalization of Treatment in a particular subtype. About 10% of patients show genetic alterations called “EGFR mutations” in their tumors at diagnosis. These particular mutations mean that the EGFR is always turned on and therefore there is a continuous signal to the cell to grow and divide. Gefitinib is able to switch off this signal and to stop cell growth in this subtype of patients. After a few weeks, the tumor disappears. Unfortunately, these mutations are rare and they are mainly present in never-smokers, who are the minority of patients.

Another, similar example in lung cancer is provided by crizotinib (Xalkori®). Patients with ALK translocations, which is another rare type of alteration present mainly in never smokers, experience a rapid shrinkage in their tumors when treated with this drug.

Another example of small molecules is represented by sunitinib (Sutent®). This drug is used to treat advanced kidney cancer and some GIST. Sunitinib is considered a multitarget agent because it blocks the vascular endothelial growth factor (VEGF) receptor and other enzymes. By doing all of this, sunitinib slows cancer growth and stops tumors from creating their own blood vessels to help them grow and metastasize. In this case, no biomarkers have been identified to help select patients who are responders from patients who are nonresponders.

Exploring the clinical utility of comprehensive genomic testing. After the patient’s informed consent, tumor and normal DNA is extracted in a certified laboratory. After targeted somatic mutation testing, more extended testing is performed in a research environment. Test results are shared with the treating oncologists, and validation of research findings is pursued if any clinically relevant research findings are found. Therapeutic decisions are based only on validated test results.

We really have to strengthen and reinforce in the future all the collaborative ways to work, without any – or minimal, at least – competitive ways of thinking. We have to work together to make the science evolve and forget about the national or regional representation of research that we have had in the past. I think the priority now is to have really good networks of institutions in order to make new treatments rapidly reach our patients.