POSTTRANSPLANT LYMPHOMAS

Gerhard Opelz, Heidelberg, Germany

Chair: Alex M. Davison, Ancrum, United Kingdom

Gerhard Opelz, Heidelberg, Germany

Prof Gerhard Opelz Department of Transplantation Immunology University of Heidelber Heidelberg, Germany

---

Slide 1

Thank You Alex. I would like to thank the organisers for providing me with this opportunity to present our data on non-Hodgkin’s lymphoma to you. When I say our data I don’t mean my data, it’s data contributed by a great many people around the world, as I will tell you in a minute.

Slide 2

This will be an epidemiological talk on lymphoma derived from a set of data provided by some 300 transplant centres in 46 different countries. We started a transplant registry called the Collaborative Transplant Study Registry some 25 years ago and we initiated collecting data on post transplant tumours in 1984 and did that sort of prospectively. Once a year we asked each centre to confirm in writing that the tumour data were correct and accurate and complete, and a centre that did not comply and send in this confirmation, we did not count in the tumour analysis. So this is probably as accurate a set of tumour data as you can get in the transplant literature. It’s probably not 100% perfect since there’s probably still some under reporting but we’ve made all kinds of efforts to ensure the accuracy of the tumour data that we are reporting on.

Slide 3

Now, in the first slide on non-Hodgkin’s lymphoma I’d like to explain a little bit. This is the cumulative rate of non-Hodgkin’s lymphoma that we studied here in 223,000 kidney transplant recipients transplanted during the 20-year period from 1985 to 2004. This is a normal Kaplan-Meier curve and as you can see and importantly, and everyone in this room probably knows that already, the incidence of lymphoma is particularly high during the first post-transplant year. If you look at this over 2 years, there were 350 cases of lymphoma per hundred thousand patients, 0.35%, but in the first year the rate was already 0,25% and only 0.1 was added in the second year. That is why everybody speaks about lymphomas as an early post-transplant disease. That is the critical phase and once you have come over the first critical phase, the rate is very low and you might say it is negligible.

Slide 4

Well, if you take that very same curve, as I’ve told you we’ve collected data over a 20 year period, and you extend that curve from 2 years to 15 years of follow up, what you’d have expected ist hat the curve would level off and go more or less horizontally. But surprisingly what you see is a curve like this. This is the first year, alright the rate was very high, the increase was steep in the first year and you see some levelling off thereafter. But there’s a relentless increase in the incidence of lymphomas over the entire follow up period, so that after 15 years of follow up we have 3% of lymphoma in this population. Now we talk about a sizable fraction of the patients who develop this very dangerous disease. So this is something to bear in mind and very important to consider. It is not a disease where the risk is early and then there’s no risk anymore. Maintenance immunosuppression quite clearly contributes to the continuous increase in lymphoma.

Slide 5

Lymphomas occur in the normal population too. So how would this compare to the normal population? Because there you would also expect an occasional lymphoma. In order to do that you have standardise. We’ve gone to the large tumour registries around the world, we have collaborated with the International Cancer Institute in Lyon and we have calculated this curve, a background curve for the normal population, same size, 223,000 individuals with the same age, gender and geographical distribution. You arrive at a control group, at an expected rate of non-Hodgkin’s lymphomas in that population that looks like this. It is customary in the cancer literature that you calculate the so-called standard incidence ratio, in other words the relative risk that these patients have compared to an age, sex and geographically matched control population. As you can see, that risk is 11 times background. So after 15 years of follow up, 11 times more transplant recipients develop a lymphoma as compared to the normal population.

Slide 6

If you take that very same scale and add to it a number of curves, namely the lymphoma curves for transplant recipients other than kidney transplants, you get curves that look like this. They go off the scale. It cannot be shown, so we have to change the scale in order to make this interpretable. When we do that, we now find that of all these organs the lowest curve is the kidney transplant recipient curve. The others are shown for pancreas, liver, heart,  lung, heart-lung transplants. You look at the relative increases it is sometimes phenomenal, for example, with heart-lung transplants about 180 times background. The shapes of the curves are also quite different because in heart-lung transplant recipients you have this fantastic steep increase early on to 5% within a couple of years. So it is in all likelihood a question of strength of immunosuppression and some other factors and I will say a few words about this later on.

Slide 7

But certainly in these transplant situations the incidence of lymphoma is high. We talk about 6% incidence after 10 years of follow up. Also increasing with time and there’s no levelling off except for that one curve here all the others keep on increasing as we follow these patients.

Slide 8

Are these really dangerous diseases? Everyone knows, if a patient develops a lymphoma, you reduce the immunosuppression and sometimes the lymphoma disappears. Well, what we can say is that nearly half of the patients die within the first year after diagnosis. This is patient survival after diagnosis during the first year. A 40% death rate and after 5 years it is about a 50% death rate and a 50% survival of patients who have a lymphoma. Compared here to patients who develop other types of tumours, you get an idea how the lymphoma compares to other tumours. It behaves very similar to a colon cancer and the only tumour that clearly is much worse is a lung cancer.

Slide 9

One can express this in terms of numbers by calculating the background survival of non-cancer patients, non-transplant and non-cancer patients of the same age category and you get again a standardised incidence ratio of death. Now, it’s the reverse. That is very interesting because it considers age. Many of the lymphoma patients are young. Therefore, the death rate is 34 times increased to the background population and the only tumour that supersedes lymphomas is lung cancer with a 47 times increase, whereas these others have all lower relative increases of death.

Slide 10

So it is clearly a vicious tumour and one of the worst tumours you can get. To show you that the death rates after diagnosis are similar, shown here are tumours that developed in North America or in Europe and they are very similar. So that there’s no secret as to how one treats these lymphomas.

Slide 11

I am making a point of this because the incidence of lymphoma is dramatically different in Europe and in North America, it’s much higher in North America, nearly twice as high, and there’s all kinds of evidence that we have published that I’m not going into detail here, but suffice to say that the indirect evidence points to more aggressive, stronger, higher dose immunosuppression in North America compared to Europe being the reason for this much higher rate of lymphoma in America than in Europe.

Slide 12

The disappointing thing is that we have not really seen a striking increase in survival after the diagnosis of lymphoma. Because now everybody knows, of course, that you reduce immunosuppression and that you give an anti-CD20 antibody. But even in this most recent period we do not see even the slightest hint of a significant improvement in survival and I have to say I have just seen two papers that impressed me and in small analyses also point out that the results with the anti-CD20 antibody are not as good as initially was claimed. There are some cases where lymphoma regresses quickly, but overall and on the whole the effect is not dramatic and one does not see much of a real improvement, so this is of continuing concern to us because it is a serious problem.

Slide 13

It does not matter at what time post transplant lymphoma occurs. I have seen papers in the literature claiming that late occurring lymphomas have a poor prognosis, I’ve seen papers claiming that early occurring lymphomas have a poor prognosis but as you can see, this is days after transplantation occurrence, days after transplantation. It’s more or less all the same within a 10% range or so.

Slide 14

And because this has been discussed a lot and discussed especially in the Heart Transplant literature, here are the heart and lung transplant recipients. Again, the time of occurrence after transplantation does not significantly influence the outcome of the patient.

Slide 15

What does influence the outcome is the location of the lymphoma. In the kidney patient, if the lymphoma is diagnosed in the kidney and limited to the kidney, the transplanted organ, the prognosis is relatively good similar to a patient who has a lymph node localisation of the lymphoma. The gastric localisation has an intermediate prognosis and then all the others are similarly poor. I should say that there is a preference of localisation and this would also take too much time, but kidney patients tend to have an over representation of lymphomas in the kidney. Heart and lung patients have the lymphoma over represented in the lung, and the liver transplant recipients have the lymphoma over represented in the liver. So that may have something to do with the prognosis because perhaps the lymphomas occurring in the transplant are diagnosed quicker and treated earlier, I do not know.

Slide 16

What clearly has an influence is the age of the recipient. A patient who is younger than 30 has about a 60% 5-year patient survival after diagnosis, whereas a patient over the age of 50 has only a 30% 5-year patient survival. So age is very important. Age is a fascinating variable when one studies lymphomas.

Slide 17

Here I’m just showing you a summary slide for all tumours except the skin and lymphoma. This is what you see regularly and you could show one tumour after the other looking similarly. The older the patient gets, just like in the normal population, you get more and more tumours. So the older the patients get, the higher the incidence of cancer, no surprise to anyone.

Slide 18

But the lymphomas are a very important exception. This is the only exception we know where the old patients, the older than 60 patients have the same incidence of lymphoma as the less than 10 year old patients. If you look at the shape of the curves, these young patients, the children have a very steep increase early on. The explanation, of course, is, well most people think that that is because the children have a primary infection with EBV. Most of the lymphomas are B-cell lymphomas that are driven by EBV infection. The T-cell control is eliminated or impaired because of immunosuppressive drug treatment and that is why these B cells proliferate and form lymphomas, so that the children are at particularly high risk of getting lymphomas because they have the highest rate, of course, of primary infection for EBV.

Slide 19

Some factors that increase the risk of lymphoma have been clearly identified. The most important one I believe is induction treatment with antibodies, with ATG or OKT3 where we have this type of lymphoma rate, whereas patients without antibody induction rate have a much lower rate of lymphoma. Statistically highly significantly different. So antibody induction is one important thing and secondly use of antibodies for rejection treatment in addition, because this one curve represents patients who had no rejection at all.

Slide 20

These were patients who were treated for rejection with steroids only and these were patients who, in addition to steroids, received ATG or OKT3. Again doubling the rate of lymphoma, so antibodies very clearly promote the occurrence or the development of lymphomas in these patients. We had a paper in the journal Transplantation last month which, well we thought was very interesting because it shows that different preparations of ATG have different lymphomagenic capacity or properties.

Slide 21

These are different manufacturers and you’ll find that OKT3, ATGAM and thymoglobulin are associated with high rates of lymphoma, whereas the Fresenius ATG has a lymphoma rate that is almost identical to patients who received no antibody at all. IL-2 receptor antibodies, monoclonal antibodies against the IL-2 receptor, do not produce lymphomas over baseline.

Slide 22

The assumption was always that an ATG that is effective, that improves graft survival, has the side effect of promoting lymphomas. The publication was exactly about that point, that that is not true, because what you find is that in terms of efficacy there are 2 curves that clearly are different, that are higher than the others, and those are the thymoglobulin and the IL-2 receptor antibody group. This is statistically highly significantly different from the other curves. But if you recall, these two curves did not go together in terms of lymphoma incidence because the thymoglobulin group developed a lot of lymphomas, whereas the IL-2 receptor curve was more or less like background. So here we have a dissociation of the probability of developing lymphoma and efficacy. So, of course, everything speaks then for use of the IL-2 receptor antibody rather than thymoglobulin.

Slide 23

Another important thing to know is that FK 506 or tacrolimus also increases the rate of lymphoma. We are now very sure about this. We have already published that a couple of years ago. It’s now very clear that the rate is higher compared to cyclosporin with or without MMF or azathioprine. As shown here, the increase is attributable to tacrolimus. It’s not prohibitive, we talk about 0.4% of lymphomas over a one-year period. But if you consider this fact, then it’s something to think about when you, for example, treat patients at a high risk, the elderly or children, I would say, or heart transplant recipients. I would certainly keep that in mind when I treat patients, and the dosage of course also.

It is probably the total load of immunosuppression that is responsible for the development of lymphomas. The lower you can go, the better. But as we know it’s a two edged sword. If you go too low, you get rejection. If one goes overboard, one increases the risk of inducing lymphomas. Lymphomas continue to be a very serious tumour that leads to death in the majority of patients who develop it. Thank you very much.

 

Slide 24

Chairman: Professor Opelz thank you for that excellent summary. I think it would be wise to have some questions after each presentation and then there will be the chance for presentations or questions at the end of the session. So are there any? Yes please, could you say who you are and where you’re from?

Questions: My question is in a stable transplant patient say more than 5 years out what order of immunosuppression would you consider in between calcinurine inhibitors, cellcept or azathioprine and prednisone?

Prof Opelz: Well, I don’t think I want to enter that type of discussion because you find lymphomas also in patients who do not get calcineurine inhibitors, in fact, we have always been unable to confirm early claims that the rate of lymphoma was higher with cyclosporin than during the pre-cyclosporin times. Steroids and azathioprine probably have more or less the same effect as calcinurine inhibitors and whether you exchange one for the other and then really can reach a reduction in risk of lymphoma is questionable. There is probably a dose relationship, it’s not very solid. We have not published. The problem is we cannot distinguish which dose it is because usually a patient that gets a higher dose say of cyclosporin also gets a higher dose of steroids and gets a higher dose of azathioprine or MMF, so it’s difficult to say it’s one or the other. There is probably a dose relationship and maybe it’s not so critical which type of agent you give.

Question: If you say that the incidence of lymphomas is so much increased by T-cell directed antibody treatments what does that say about the pathogenesis of these lymphomas?

Prof Opelz: Well, this is not my work but what people say is that the T cell has this equilibrium. There is T cell control that prevents B cell proliferation. That is equilibrium is disturbed especially if you have antibodies against T cells, you disturb the equilibrium and you let the B cells proliferate.

Chairman: Well, thank you very much I think we will move on to the second presentation today.