Although many advances have been made in the treatment of primary myelofibrosis (PM), all agree that transplantation is the only cure. Although many transplant physicians, commonly known as “transplanters,” have helped contribute to these results none more so than Professor Nicolaus Kroeger. His knowledge regarding this subject is readily apparent from this wonderfully lucid and informative interview with Zhen.
In addition to my professional familiarity with his published papers, I have come to know Professor Kroeger on a personal basis through my unique membership in the European LeukemiaNet. I’ve interacted with him on many occasions, and have found him to be a warm, thoughtful, sensitive individual emphasizing the hazards as well as the successes of transplantation. He notes that although the various risk scores are indeed important, (and Zhen’s new computer-based score is such a new and useful tool for our patients) there are many other items regarding patient selection for transplantation including for example, social factors, family support, and patient personality. Although he is recognized worldwide as a “transplanter” par excellence, I think he takes even more pride when his patients and his colleagues describe him as a “very good physician.”
-Richard T. Silver, MD
Stem cell transplantation: Expected outcomes.
Zhenya: I’m concerned about outcomes. We don’t know what the stem cell transplant (SCT) outcomes are going to be. One problem I see is that the data we use might be skewed because by the time people come into SCT they might be at high risk or even worse, at end stage, so overall survival figures are poor. What do you think about that?
Kröger: Yeah so the question still is, to speak simply, the intention with transplant is always to cure the patient of the underlying disease, in this case myelofibrosis. But beside cure there are other possibilities of outcome, thus actually there are three outcome possibilities, The first is cure, the disease has disappeared and will never come back. The second possibility is relapse, means the transplant procedure went well, the disease disappeared but came back after a certain time period. And the third outcome possibility is mortality which is not due to myelofibrosis but to complication which are mainly related to the transplant procedure such as severe Graft Versus Host disease (GVHD), severe infection or organ failure, which we call therapy-related mortality
Three possibilities. Three different risk factors.
And as there are only three possibilities, thus at the end for every patient you should always end up with 100%. If we take all myelofibrosis patients independent risk factors about 60% will be cured, about 20 to 25% will experienced relapse and 15 to 20% will die from therapy- related complications after allogeneic stem cell transplantation. However these percentages can differ from patient to patient depending on different risk factors. It’s important to know that the percentages will also differ in every patient over time because one can easily imagine with increasing age the complications are less tolerated and the risk of therapy-related mortality is increasing. Or, if the disease is progressing to a more advance stage and becomes more resistant to therapy, the risk of relapse is increasing. And, logically, if risk of therapy-related mortality or the risk of relapse is increasing then the possibility of cure will become less. In other words, if you do the transplant very early, if the patient is relatively young and in good condition and the disease is not very advanced, then the rate of cure is higher, the risk of relapse and dying from therapy-related mortality is lower.
To calculate the risk of cure, of relapse and of therapy-related mortality you have to take three different areas of risk factor into account
- Disease-specific risk factors are related to the disease, such as transfusion dependency, blasts in peripheral blood, cytogenetic and others. Usually disease-specific risk factor are included in disease-specific risk scores such as DIPSS (Dynamic International Prognostic Scoring System) or DIPSS plus
- Patient-specific risk factors are related to the patients performance status , comorbidities or age and others
- Transplant-specific risk factors are, for instance, number of transplanted stem cells , HLA compatibility of the donor , female donor for a male recipient and others
All risk factors can influence probability of survival, but disease-specific risk factors have major risk of relapse; patient-specific and transplant -specific risk factors influence mainly risk of therapy-related mortality. Thus we can have patients with the same disease risk (e.g DIPSS Intermediate-2) but with different outcomes because of different patient conditioning and different stem cell donors.
So you always have a lot of variables that influence the outcome that need to be taken into account.
A comprehensive risk score
Z: So how do we put those all together into an algorithm that can can come up with a reasonable sense of what the outcome might be?
Kröger: Exactly. So what I just mentioned is just the outcome after stem cell transplantation. Counselling patients includes also calculating the probability of survival without transplantation. The question here is how long is the life expectancy without transplant? Here the disease-specific risk factor such as DIPSS is very useful to determine the median survival according disease-specific risk .
Median survival means at what time point are 50% of the patients with a certain risk status already dead and 50% still alive. This may help in treatment decision for transplantation by balancing the risk and the benefit. However for an optimized balance of risk and benefit we would need a transplant-specific risk score model which would include disease-specific, patient-specific and transplant-specific risk factors. Such a comprehensive transplant specific risk algorithm does not exist so far, but we in Europe at EBMT (European Society for Blood and Marrow Transplant) and CIBMTR (Center for International Blood and Marrow Transplant Research) in US are are currently working on such a transplant-specific risk score. We are both collecting a lot of patients who receive transplant, collecting all this information and then trying to create a score that can predict the survival but also can predict risk of therapy -related mortality (TRM), and predict the risk of relapse according the different disease-, patient- and transplant-specific risk factors.
Z: How far along is that work?
Kröger: I hope both groups can present first results at the next meeting of the Amercan Society of Hematology (ASH) in December this year. We work closely together and share our date to validate the EBMT and the CIBMTR cohorts. And once we have this score we can at least predict the outcome as to NRM and survival outcome, but this is just for the timing. But then we can really counsel the patient properly regarding the risk of dying of therapyrelated causes, the risk of relapse and the probability of being cured..
So many variables
This is actually what we need and what we don’t have for Myelofibrosis so far. For this project we need a lot of patients, It is not possible to create a valid score with only 100 patients. There are so many variables so you need at least four to six hundred patients. On the other hand what is also important for the decision to go to transplant is the outcome without transplant. I mentioned this earlier. Patients usually want to know ”If I don’t go for transplant what is my life expectancy?” I try to counsel the patient best as I can explaining the risk and benefits , the chance of cure and the outcome without transplant.
Beside DIPSS or DIPSS+, new disease specific scores such as MIPSS (Mutation-enhanced International Prognostic Scoring System) includes also molecular abnormalities and may help further for a better determination of outcome
I’d like to mention also that the decision for transplant is always an individual decision. Different personalities, life plannings and social factors have to be taken into account as well. Usually a patient needs a lot of social support after transplantation which should be organized in time .
Closer to the truth
Z: On your paper Impact of allogeneic stem cell transplantation on survival of patients less than 65 years of age with primary myelofibrosis, your retrospective decision model seems roughly based on DIPSS score not on disease-specific or patient-specific factors…
Kröger The idea was just to make the community aware that if you have at least Intermediate-2 or High Risk disease you should consider transplant and not wait too long. And if you have a Low Risk disease there’s no need to go for transplant immediately. This was the intention. This disease is a continuum. If you have Low Risk today doesn’t mean you have Low Risk tomorrow. In this abstract, but also in another paper we used a statistical model to estimate outcome after transplantation vs a non transplant approach, because no prospective comparative study exist. This model helps us today to recommend stem cell transplantation , but this model has also some weakness because it is not a prospective study and it is only based on DIPSS score and does not include important patient specific and transplant specific risk factors, as mentioned earlier. Let me say these studies are estimations to get closer to the truth.
The SSTT and missing data
Z: The SSTT (SCT Spectrum Transplant Timing tool) concerns me. It’s useful to start the conversation but it lacks all mutational and genotypic inputs. More importantly, the conversation it stimulates will be with a hematologist who almost certainly is not going to be an MPN specialist and works in an environment of uncertain outcomes. It is good to warn the patient of risk levels but are we providing a pathway for improved outcomes?
Kröger : The SSTT is a wonderful and important project. We could suggest the patient ask for the molecular genetics especially for the driver mutations, which are JAK2V617F, MPL or CALR. I still think it’s pretty early to recommend transplantation according molecular genetics. What comes out if you look at the MIPSS score, the CALR mutation is a bit better, the triple negative, EZH2, SRSF2, IDH1/ 2, and ASLX-1 a bit worse. Currently, to be honest, we can’t make any decision based on molecular genetics alone. It may help sometimes but it’s only another disease- specific factor. We still have to confirm all the data concerning molecular genetics in the transplant setting
The major problem is to create awareness of allogeneic stem cell transplantation. Awareness to the hematologist and maybe also to the patients. Patients with myelofibrosis should be referred early to an experienced transplant center and the decision for timing should be made with stem cell transplanter ,hematologist and patients including their family.
Z: What causes the risk of relapse. Is it incomplete conditioning?
Kröger: No, the purpose of the conditioning is to suppress the immune system of the patient to enable engraftment of the stem cell of the donor. if this immunosuppression is too low the stem cell will be rejected which results in “graft failure.” The other aim of the conditioning is to destroy the malignant cells, however the major “killer” of the malignant cells is not the conditioning but rather the immune cells (T cells) from the donor which causes a “graft versus myelofibrosis effect.” Retrospective comparison in myelofibrosis between different intensities of the conditioning regimen did not show a significant higher risk of relapse if reduced intensity is used. More important is to harness the mentioned graft versus myelofibrosis effect after stem cell transplantation to prevent relapse. The risk of relapse depends on the status of the disease, as more advanced, the higher is the risk, For example, if the disease is already transforming to acute myeloid leukemia the risk of relapse is significantly higher in comparison to patients who received transplantation in non advanced phase (e.g DIPSS Intermediate-2)
The double edged sword – Cure, relapse and GVHD
GVHD and graft versus myelofibrosis is most likely only the other side of the same coin. Graft Versus Host Disease can be dangerous but it can be very helpful. Patients who experience some degree of GVHD have a lower risk of relapse than those without any GVHD. Thus, we need a lot of experience to balance the risk of GVHD and of harnessing the graft versus myelofibrosis effect.
The stem transplantation procedure lasts about four to six weeks During this period the new hematopoietic system starts working in the patients but it is still in a very vulnerable phase. Once the patient is discharged the work on balancing between GVHD risk and harnessing graft versus myelofibrosis starts. If GVHD is suppressed too strongly it is likely that also the graft versus myelofibrosis effect is suppressed too much, which would increase the risk of relapse. Monitoring the patient by molecular marker after transplantation (e.g JAK2, MPL or CALR) is helpful, because it shows residual disease or molecular remission, which is helpful for the physician to guide and adapt the drugs for immunosuppression. Thus, post transplant monitoring and treatment is very important to prevent relapse.
At the end if you have a lower risk of disease you have a lower risk of relapse. If you have a higher risk of disease you have a higher risk of relapse If you have a higher risk of disease and are also in bad condition or poorer condition then the risk of dying from therapy-related causes is also higher
Z: That would relate to GVHD as well?
Kroger: Yah. The immune effect of the donor T cells are crucial, as mentioned earlier. The immune cells are responsible for GVHD but also for the needed graft versus myelofibrosis effect.. GVHD is not always deterimental but can be also helpful , especially if it is not a severe GVHD and the risk of relapse is high . We should acknowledge that having GVHD is also preventing relapse. This “fine tuning” of immunosuppression after transplantation to balance between GVHD and graft versus myelofibrosis is crucial and requires some experience. One should be very careful to get this effect therefore the patient needs to be seen in the beginning once weekly or at least very frequently.
Awareness and decision making
The most important thing is creating awareness. In the end it’s always an individual decision and the patient should be very well informed and should have the best option for their life and for what they wish. And we should not say this is bad for you or you should do this. We’re trying to include the patient, to give them as much information as possible but also to give them some help because for the patient it’s a very difficult decision and they need some guidance.
Z: Earlier, we talked about a gain of greater than six months survival balanced against the quality of life (QOL) detriments imposed by SCT.
Kroger: First of all the QOL is an issue but really it’s not bad and it’s really becoming better. We did a retrospective study which shows that two years after stem cell transplantation the Quality of Life was identical to that of the general population. More recently we participated in a prospective study from the Mayo Clinic which showed continuous improvement in Quality of Life in the first year after stem cell transplantation.
The Median is not the message.
Second is median survival. If you want compare transplant results to non transplant results you cannot compare the median survival, because the survival curves have different slopes as you can see on both figure using DIPPS score for transplant patients and for non transplant patients. If you look at the score, the DIPSS score in MF, you start at 100 and then the curve goes slowly down. In High Risk it goes down fast. But if you follow the curves very carefully you can see that it goes down. At the end you are at zero, If you say median survival you just check to see what happens where 50% of the patients are dying and you look and can say OK this is the median survival.
In transplant if you see the curves go down in the beginning the curve is quite steep because this is therapy-related mortality But then, after one or two years, the curve turns become a plateau and there is no further death or relapse. Median survival is not the best thing to compare results after transplant to a non-transplant cohort .
As you can see in transplant figures at five years, there is no further relapse, it is significantly better. If you look only after one year than the transplant is worse because you have this therapy-related mortalityy which doesn’t exist in the non-transplant cohort
This is the reason why it is so difficult to compare by median survival.
Kroger: Here is the chart. So if you open the first line it is just a Risk Scoring system according to DIPSS.
If you look at the right curve the Kaplan-Meier curve.. For the High Risk group you can see the left side at 0.5 (means 50% of the patients) At this time point 50% of the patients are alive and 50% of the patients are dead. Thus you can go from this time point to the corresponding curves of e.g high risk or intermediate II risk If you go from the crossing point down to the line with the years , you can calculate the median survival for the corresponding risk group. Here for high risk it is about two years. That means at 0.5 all that is up have died already, what is down is still alive. If you look at the Intermediate 2 curve you end up about 5 years. You can see 50% is then dead,. 50% still alive. And you can see the curve is still going down.
Going down and 10 or 15 years the median survival is nearly zero for the non –transplant DIPSS. For this patient, just for argument sake, you always have two years and you have this five years. And if you go to the next line. According to DIPPS the High Risk in the beginning, the blue one, the curve goes steeply down. If you look at 50% at 0.5 you end up at 2 years. Thus if you just look at median survial you can say the median survival between transplant and non transplant is the same, only about two years. In contrast to the non transplant group you can see that the curve in the transplant cohort reached a “plateau” over time at about 50%. Thus no further patient died over time.. Therefore if you compare transplant and non-transplant you should not take the median survival.
Z: To what do you attribute early death, to the therapy or to the condition of the patient primarily?
Kroger: It’s both. We call it Non-Relapsed Mortality (NRM)… which means actually all causes of death which are not caused by the disease. It can be due by the treatment itself but if you are older, if you get the transplant at 60 and die at 70 from a heart attack, it’s also NRM. In older patients NRM is always a bit higher because over the years the likelihood of dying is higher in older than in younger patients.
Professor Nicolaus Kröger, MD, is Medical Director of the Department of Stem Cell Transplantation at the University Hospital Hamburg-Eppendorf, Germany. He is the current president of the European Society for Bone and Marrow Transplant (EBMT) and Past-Chair of its Scientific Council.
© 2018, MPNforum. All rights reserved under Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License