Dr Anne Vincent Salomon, Anatomopathologist, Head of Department at the Curie Institute
Interview in October 2018
Currently, we realize that the analysis of the tissue by the pathologists is really useful to the surgeon, to give him an indication on the quality of his operation. This is essential for the patient since the information that the pathologist will have given from his analysis will then determine the treatments that will be proposed to her: the more precise the level of information, the more we will be able to adjust the treatments proposed to the patients in order to be precise, to target only the tumor cells and to spare the normal cells.
Thus, this medicine of precision really starts with the pathological analysis of the tumor, it is enriched by molecular analysis and it leads to a range of treatment possibilities that we hope to see refined in the near future.
The anatomopathology report (anapath) is a checklist that will help the surgeon, the medical oncologist and the radiotherapist who will eventually complete the treatment.
What is a cancer cell?
The cells of the body, for a great number of them, divide, they renew themselves, they are very disciplined, very organized. The body is a perfectly oiled machine, which every day allows the cells that have to divide to do so. When they divide, they multiply their chromosomal material, their chromosomes, they double them, to be able to separate the chromosomal material in the two daughter cells, without there being any error and so that they really resemble the mother cell.
But the problem is that there are many “spelling mistakes” that are created, that are written, at each doubling of this chromosomal material. So, in the normal body, in the healthy body, we have all the enzymes – the enzymes are very specialized little workers – that will correct the spelling mistakes in the DNA of the chromosomes.
As we age, these spelling mistakes may tend to be less well repaired, or under other certain circumstances. I will take the example of the skin cancer that everyone knows well, melanoma. The sun will induce anomalies, spelling mistakes that we call “mutations” in our jargon, and it will cause so many of them in certain cells that they will be overwhelmed, that they will no longer be able to repair the spelling mistakes. The accumulation of these mistakes completely disrupts the cell division signal, and they will run amok and divide out of order. They tend to repeat themselves and multiply enormously.
In fact, cancer is really a runaway division of cells, which gives them the ability to move. So, dividing without respecting the normal orders and starting to move in the body, these are the two essential characteristics of cancer cells.
Cancer removal and safety margins
The best treatment for cancer, particularly breast cancer, is surgery: the tumor, the cancer, and the healthy, normal tissue around it are removed. And this safety margin must be at least a few millimeters, but we understand intuitively that if we leave more than five millimeters, ten millimeters, around the tumor, well that increases the chances that all the tumor tissue has been perfectly removed.
Therefore, when we look at this surgical specimen, we measure the size and the distance of the tumor from all the margins. But what are the really important margins? It’s the ones that are superior, inferior, external, internal, that is to say, on the right, on the left, on the top and on the bottom, in relation to the patient’s breast, and so it’s these margins that we examine.
So, we look at them macroscopically, like this, with the eye, we measure, and by palpating. Then, with the mapping of the tumor, which we do once we have fixed the surgical specimen in formalin, we will again verify, under the microscope, that the tumor is far from this safety margin, and that there is no cancer cell in situ that would be invisible and impalpable, that neither the surgeon nor the pathologist in the macroscopic state can spot but that we see under the microscope.
Size of the tumor
Afterwards, there is something extremely important, which is the size of the tumor. When the surgeon gives us the operating room, we examine it, we open it like a book, and there we measure the size of the tumor with the induration, that is to say that we really palpate it, as if we were examining the body, but we do it outside the body on this operating room and we palpate the size of the tumor. The size is also essential for the prognosis. The smaller the tumor, the better the prognosis; the larger it is, the more complicated it will be and the treatments will have to be adjusted. So why? Most probably because the smaller the tumor, the less tumor cells it has inside, and earlier I told you that the cells move and acquire the ability to move.
In fact, the whole problem with cancer is the spreading from the diseased breast to give these small micro-metastases which lurk for weeks, years, and which afterwards can give metastases elsewhere than in the breast, without us understanding yet exactly why. But all our efforts are to try to predict whether this small tumor has been able to metastasize. So, grade and size are really important building blocks in defining prognosis.
Differentiation and grades
A “well-differentiated” cancer is one that closely resembles a normal mammary gland: this is the first characteristic that we observe when we look at the biopsy or the surgical specimen, and then we look at the capacity of the tumor cells to divide. We can see this with the image of the mitoses, that is to say the chromosomes that are dividing, so we will count them. If the tumor is aggressive, it will have many mitoses, and we count them very precisely to establish a grade and this grade is also composed of the shape of the nuclei. The nucleus is the heart of the cell that contains the chromosomes, so it’s easy to understand that the more abnormal chromosomes there are, the more of a mess the chromosome counts are, the bigger and more atypical the nuclei will be, and the more different they will be from one cell to another, because this will reflect the growing anarchy that is possible in the tumor cells
This is the first level of classification: differentiation, mitosis and nuclear size. From this we will derive a grade that is extraordinarily robust, the Elston and Ellis grade, which is 1, 2, or 3, and which is already a very good indication of the classification of the cancer. So, grade 1 is very well differentiated, it’s not very serious. Grade 3 is undifferentiated and more serious. This is the solid base on which the definition of prognosis is based.
There is a level of diversity that we will call histological, to say that these tissues take different forms. Under the microscope, the tumor cells will organize themselves, weave a tumor tissue that will have different shapes, and to these different shapes, sometimes different prognoses are linked, and therefore it is a third level of classification. If I give an example, infiltrating lobular breast cancer is the second major type of breast cancer, it represents 15% of our patients, often older women. These lobular cells are like that, round, small, isolated, and they will infiltrate the breast in the form of small cells detached from each other, it does not make a solid tissue, they are small cells isolated from each other.
In situ or invasive cancer
Under the microscope, the tumor will be characterized as infiltrating or in situ.
The tumor in situ remains in the lactiferous duct (which conducts milk to the nipple), so it remains protected from contact with blood vessels by the basal membrane and the myoepithelial cells (these are somewhat complex terms but they are the bark of the lactiferous duct), which really confine the tumor cells to the interior of this network of channels. So, there can be no metastasis. Thus, we inform if it is infiltrating or in situ. If it is in situ, we will give indications of the dynamism and potential aggressiveness of this in situ, by giving its nuclear grade, its size, whether or not this in situ is associated with micro-calcifications, and whether the surgical excision margins are distant from these in situ lesions.
For infiltrating cancer, more information is given, by isolating the count of mitoses since it is really the characteristic of the aggressive tumor cell to produce many of them. We will also specify if there are already tumor cells present in the vessels in the tissue that has been removed.
You have to imagine that there are vessels everywhere in the body and around the cancers of course, there were previously existing vessels and moreover, the tumors have the capacity to make vessels to feed themselves, and unfortunately that makes a circulation route for the tumor cells, so that they can go elsewhere. So, we give this information whether or not there are tumor cells in the vessels.
It will also show whether or not there are lymphocytes in the tumor. Lymphocytes are white blood cells that can, under certain circumstances, help breast cancer cells to die. They are very good cells in our body that, when present in large numbers, predict a better prognosis. So, we quantify them, we say if there are a lot or not, and then we go for the infiltrating cancer again, to specify its distance from the margins.
Molecular characteristics of cancers
Then there is another level, the molecular level, where we go down to the level of the molecular characteristics of tumors. There, breast cancers are reclassified in a different way and this has a direct impact on the treatment.
There are the so-called “luminal” cancers, which are tumors that express hormone receptors, estrogen and progesterone receptors, which provide a treatment weapon because we can block these hormone receptors.
Then there are cancers that are called HER2. What is HER2? It is a class of tumor where during the transformation, one of the mutations is a kind of error. It happens, when the chromosomes replicate to make two cells identical to the cell that gives birth to them, there is a kind of repetition, an error, where on one of the chromosomes, there is a gene called HER2, discovered thirty years ago now.
We have understood that in breast cancer, in certain ovarian cancers, in certain stomach cancers, this error gets replicated in a large number of copies. We also understood that this gene, like the protein it produces, gives a very strong order of mitosis, when it is in large numbers, then it is the total runaway of the cells. American researchers, such as Dennis Slamon in particular, have understood that this protein could be blocked with an antibody, as a kind of immunotherapy – it was one of the first immunotherapies -: therefore, this mitosis signal will be blocked and this therapy is extremely effective, because there is the anomaly in question.
So that’s the second class, and then there’s the third class, which is called “triple negative”, where there’s no expression of estrogen and progesterone receptors, and there’s no large numbers of HER2. These are tumor cells that really have a strong ability to divide, to multiply.
We understood that these three classes were associated with very different prognoses, depending precisely on their ability to divide.
Every time there are new technological advances in science, it takes a little time, but it always ends up having positive repercussions for a better definition of this tumor identity card. At the moment, we have a tool called “molecular signatures” that we will use in very specific situations when tumors express estrogen receptors, when they are larger than 20 mm, or between 10 and 20 mm, and when there is hesitation: should chemotherapy be given or not, will it benefit the patient, will it improve her chances of survival without recurrence?
What’s a little difficult to understand is that there are several types of signatures. But even if they don’t all share exactly the same lists of genes, they all give an indication of the tumors’ capacity to divide and to be potentially aggressive. So whatever signature we take, we will refine the information that was proposed by the pathologist on his report for estrogen receptor positive tumors.
A targeted therapy is a therapy that has the ability of only killing the tumor cells and to spare the rest. So, the most spectacular targeted therapy we’ve seen in the last 15 years is anti-HER2. In this case, there is too much HER-2 on the cell, and the antibodies that are the anti-HER2 drug bind to the tumour cell and kills it.
In the past, having a tumor that had a lot of HER2, too much HER2, was a serious tumor. Patients would relapse in three years, four years, five years, they would metastasize and we had no tools to treat them, no drugs to treat them. The discovery of HER2 and anti-HER2 therapies has completely transformed the prognosis. Now they have an excellent prognosis, which is similar to women who have very well-differentiated tumors that express estrogen receptors and are small. The patients who were offered this anti-HER2 treatment after surgery are doing very well.
At the beginning of the HER2 story, it was quite spectacular. We had never seen anything like it, because pathologists don’t prescribe drugs, so we never had contact with the pharmaceutical industry. All of a sudden, we saw them arrive. There was one doctor in particular who, in one of these pharmaceutical companies, to be smart enough to say to himself: we must help pathologists to standardize their working methods even better if we want the right patients to receive the right treatment. And the success of the anti-HER2 treatment comes from the intelligence of this collaboration between the pharmaceutical industry doctors and the pathologists.
The digital revolution
There is a revolution coming thanks to the digitization of our little HES slides, which will no longer be observed under the microscope but virtualized, that is to say that they will go from being a glass slide to a virtual image thanks to a slide scanner that will capture the image and transform it into a digital image. And we will no longer look through a microscope but on a computer, on a screen. This accumulation of fantastic data represented by the images gives rise to interactions with those who work in artificial intelligence. Deep learning” and “machine learning” are really helping to design new tools to compensate for possible fatigue or laborious, repetitive tasks such as counting mitoses. Help by an artificial intelligence algorithm will be possible soon and will allow us to have more time to look for new or more precise parameters. This will be a real revolution, which is in the process of happening and which is quite exciting.
For triple negatives, I haven’t talked about it so far, there’s a revolution brewing, which is immunotherapy and PARP inhibitors. We’ll talk about both topics.
Immunotherapy is the easiest to understand. Earlier, we were talking about lymphocytes, the white blood cells that are in tumors and which, when there are many of them, indicate a good prognosis. Immunotherapy uses one of the functions of the immune system, which is to constantly recognize that our cells are ours and that we must not destroy them. In the normal body, there are white blood cells whose role is to always say “no, no, it’s okay, everything is fine, we must not destroy that cell, it is part of the body…”.
When tumors grow, unfortunately, these immune system cells, like idiots, keep saying “no, no, everything is fine, it’s the body’s cells”. But it turns out that they are cancer cells. So, the Nobel Prize in Medicine, which has just been awarded to those who have understood immunotherapy, is in fact to block these immune cells which say “these are tumour cells, but everything is fine, they are cells of the body, don’t destroy them”, so it is to go and block them by saying to them: “shut up, you absolutely must destroy these cells, they are abnormal cells”. This allows the other cells of the immune system to go and destroy the tumour cells. So that’s what checkpoint inhibitors are, this immunotherapy is exactly that, it’s to mute the cells of the immune system which must not make the body believe that cancer cells are body cells and that they must therefore be left there.
Breast cancer, as we know, in 5 to 10% of cases, occurs in the context of a hereditary mutation of either BRCA 1 or BRCA 2, or P53 or PALB 2… Finally, we are beginning to know more about the genes responsible for a hereditary family transmission… These genes, BRCA1, BRCA2, PALB2, P53, are genes that are normally used to repair spelling mistakes in DNA. So, if they themselves have a mutation, that is to say that they have a huge spelling error on their sequence, they will no longer be able to function, they will no longer correct spelling errors. So, it is easy to understand that at that point, women will have an increased risk of having an accumulation of spelling mistakes in their breast cells and of getting cancer.
Because the body is a marvelous machine, there are backup systems, so there is not only one way to repair these spelling mistakes, there is a slightly less elegant one, because it is a little less perfect, it uses enzymes whose abbreviation is PARP… So now, if we say to ourselves: BRCA1, BRCA2 do not work in tumor cells, there are always PARPs that come to repair little spelling mistakes. And the tumor cells don’t care about having spelling mistakes because they don’t obey anyone anymore and on the contrary, it gives them the ability to divide and move, these spelling mistakes. But that’s not too much to ask, because they are still cells.
So, the PARPs repair a little bit what is needed for them to continue to live and to do their dirty role of dirty tumor cell. Now, if we block the PARPs when BRCA1 or BRCA2 are not working, the tumor cells will be suffocated by the overaccumulation of mutations and they will eventually die. So, it only works for the moment when there is a BRCA1 or BRCA2 mutation, but it is a great hope…
Tumor or somatic genomics, we sometimes say it is the analysis, the identity card of the spelling mistakes in the genes of tumor cells, only. That is to say that we, pathologists, will take a small piece of this tumor, extract the DNA, either do it ourselves or with the geneticists, and sequence the DNA to look for the presence or absence of these mutations that interest us. That’s in the tumor. It’s not hereditary. It came with the history of the tumor, it’s what made the tumor malignant, that it’s a cancer.
Search for constitutional genetic mutations
Genetics, which we call constitutional, means taking the person’s normal DNA, either by making a small smear on the cheek and recovering cells inside the cheek, or by taking a blood sample, and we will look at the person’s normal white blood cells, their DNA, that is to say, what is present in all of these cells, which has been transmitted to them by their father and mother. So, we will look for mutations that are constitutional because they are carried by all the cells in the person’s body and that he or she is able to transmit to his or her children.