Cord-blood disappointment

One of the supposed advantages of cord-blood transplantation is the suggestion that intensive matching is not required. Advocates of this view will be disappointed by this comment in Lancet Oncology

After years of research and debate, findings after umbilical-cord blood transplantation have mirrored those of other sources for haemopoietic stem-cell transplantation—ie, matching of the HLA loci between donor and recipient does matter. Thus cord-blood transplantation now faces the same hurdles as bone-marrow or peripheral-blood stem-cell transplantation. In The Lancet Oncology, in a clear and convincing analysis, Mary Eapen and colleagues1 show that additional matching of donor and recipient for HLA C improved the outcome of patients with a cord-blood transplant. This improvement was identified independently of other known risk factors for outcome after haemopoietic stem-cell transplantation, such as the patient's age or disease stage. Results were based on a homogeneous group of more than 800 patients with leukaemia or myelodysplastic syndrome and with sufficient information for the class I HLA antigens HLA A, B, and C and the class II antigen DRB1 in donors and recipients. Pairs with or without matching for HLA C could be compared with pairs with no, single, or multiple mismatches at HLA loci other than HLA C. Effects of matching for HLA C were greatest when no HLA antigen or only one other HLA antigen differed between the donor and recipient.

The importance of matching for HLA in general was underlined by two additional findings: results were better with full matching than with a single, double, or multiple mismatch at any of the class I or class II antigens; and, in the case of a single mismatch for one class I antigen, results were better when the class II antigens DRB1 were identical.

These findings accord with results from unrelated-donor transplants and are reassuring to immunologists (it was difficult to understand why cord blood should behave differently to other sources). These findings might be disappointing for some who thought minimal matching sufficient, but they have clear consequences. The degree of matching between donor and recipient, including HLA C, needs to be integrated into the algorithm used to select for or against transplantation. The risk of disease should be balanced against the transplant risk, so that overall a better outcome with regard to survival, quality of life, and cost, compared with a non-transplant strategy, is achieved. If cord-blood transplants are to be used in early disease, such as for patients with high-risk acute myeloid leukaemia in first complete remission, the search strategy should aim for the best match and should now include HLA C. The value of high-degree matching in unrelated-donor transplants was recently shown to be specifically important for patients with early disease and low-risk characteristics. Hence, an optimally matched transplant product might tip the balance in favour of immediate transplantation in first remission, whereas a suboptimal product would favour a watch and wait or non-transplant strategy. To allow sufficient time for the selection process, any donor search for an unrelated-living or cord-blood donor needs to be started at diagnosis.

The findings of Eapen and colleagues further support the role of standardised reporting of transplant data, as recently stipulated by the WHO's guiding principles on organ and cell transplantation. Only by such cooperation can results be yielded. Furthermore, other antigens such as HLA DQ, HLA DP, or minor histocompatibility antigens are likely to be important in cord-blood transplants because they are present in unrelated-volunteer donor transplants. Typing for HLA C should now become mandatory, accompanied with storage of samples for later analysis. Cord-blood banks will have to provide this service. The additional costs will be recouped through improved transplant outcomes. Quality management systems such as the Foundation for the Accreditation of Cellular Therapy, the Joint Accreditation Committee: European Group for Blood and Marrow Transplantation and the International Society for Cellular Therapy, or Netcord should implement full typing for class I and II antigens into their standards. These steps will help to further improve the procedure and define those selected patients who will clearly benefit from a cord-blood transplant.

Transplants - a conclusion

So, what is the conclusion? Who should have a transplant and when. As usual, the best results are achieved in people who don't need one, and transplanters delight in doing these because it makes their results look good. However, all transplants have a risk of killing the patients, and the procedure is not so safe that it should be performed if it is not necessary.

Perhaps the most obvious group to benefit are patients without co-morbidities who need treatment but have del 17p. Even in this group there is probably a need to debulk the disease before the transplant and how to do this is controversial.

The second group is for patients who have failed to respond to a fludarabine containing regimen or relapsed following treatment with one within 12 months.

The only other sensible advice is not to muck around with other regimes like PCR, cladribine or various experimental treatments. If a transplant is necessary it should be performed as soon as the criteria are met.

There is no information about which reduced conditioning regime should be used, but almost certainly a non-ablative treatment is preferable to an ablative one. T depletion is still in the melting pot. Almost certainly it is better to go to an experienced unit that has performed many previous transplants in CLL (not one that has specialised in other diseases).

What are the alternatives to transplantation?

In patients who are fludarabine refractory or have a p53 dysfunction the available treatments are limited. The first drug that comes to mind is alemtuzumab (Campath). Although Campath does not require intact p53 to kill CLL cells, it is very poor at eliminating bulky disease. Why this should be is unclear, since in the first |Lancet paper on the treatment of two patients with CLL with Campath, bulky disease was not a problem. However, Campath is made differently now and it may be that the new system is not as good at attaching sugars to the molecule. Hence the new alemtuzumab may be ineffectively glycosylated. A new version of Campath is being developed.

Another drug that deals with p53 deficient CLL cells is high dose steroids (either methylprednisolone or dexamethasone). These drugs also deal with bulky disease, but they are likely to produce severe steroid side effects, such as high blood pressure, fluid retention, alteration in body shape (a melon on cocktail sticks), thin skin, thin bones, diabetes and most important in this context, immunodeficiency.

This is why many doctors are reluctant to combine high dose steroids with alemtuzumab. Fungal, viral and bacterial infections are all likely and must be guarded against. Prophylactic cotrimoxazole (either Bactrim or Septrin) is needed to prevent pneumocystis, aciclovir (or similar) to prevent herpes simplex and zoster, and one of the many antifungals to prevent candida and aspergillus. Twice weekly screening for reactivated CMV is also necessary.

However, in my experience, when used in specialized centers this is the most effective treatment available.

There are advocates for high dose steroids with rituximab and some evidence that this combination may also be effective in bulky fludarabine resistant disease. We await a head to head comparison.

Lenalidomide (Revlimid) is reported to be effective in drug resistant CLL though the numbers of drug resistant and p53 defective patients treated in clinical trials is still very few. The other agent with this property is Flavopiridol, but this is notoriously difficult to administer and available in very few centers.

Clinical trials of other drugs that may be useful, like acadesine, are continuing, but as yet treatment for this difficult problem remains uncertain.

Transplants: ablative or non-ablative?

Continuing from yesterday. Should one go for an ablative or non-ablative transplant? Ablative means using sufficient chemotherapy to destroy as much of the leukemia as possible, while not worrying about how much bone marrow is destroyed because it is going to be replaced by the graft anyway; non-ablative means using immunosuppressive drugs to enable the marrow to engraft and relying on graft-versus-leukemia to destroy the tumor.

The paper in Blood has looked at the results on ablative transplants published by the European Bone Marrow Transplant Registry (EBMTR) and the International BMTR. The long term overall survival was 46% and another 46% died of the treatment (the remaining 8% died of relapsed leukemia). The best results came from the Dana Farber at Boston with only 24% treatment related mortality and 55% overall survival at 6 years. However, none of the Boston patients had fludarabine refractory disease. The median age at transplant ranged between 39 and 48 at different centers.

In the total experience from Europe and America, if an unrelated donor was used, the results were less good, with only 33% overall survival at 5 years.

Non-ablative transplants can be used in older patients - up to 72 years of age, with a median age of between 50 and 57. This type of transplantation is fairly recent, but we do have 7 reports of studies from different units (Seattle, Houston, Boston, Germany, Spain and the UK) on both sides of the Atlantic involving 346 patients.

The degree of myeloablation and immunosuppression varied between centers with Seattle being least for both, Spain being the most myelosuppressive and the UK the most immunosuppressive. Only Seattle has 5-year follow up and they report 23% treatment-related mortality and 50% overall survival similar to Boston's figures on ablative transplants, but 87% of the Seattle patients were refractory to fludarabine and the patients were older. The other studies are less mature, but treatment-related mortality ranged from 16% to 34% and overall survival from 48% to 72%. One suspects that these figures will deteriorate as time passes.

I have interpreted these figures a little. Generally they don't talk about treatment-related mortality, but non-relapse mortality to take account of the fact that some patients commit suicide, others have heart attacks and others are run-down by Greyhound buses. However, most non-relapse deaths are caused by the treatment - even suicides.

One of the reasons that patients commit suicide is that they can't live with chronic graft-versus-host disease. This was least common in those groups treated with more immunosuppressive drugs like alemtuzumab (in the UK) and ATG (in Germany). The worry is that more immunosuppression means more relapses. The data are not mature enough to be sure of this.

So far my conclusions are that about half of CLL patients can expect to be cured by a transplant. Reduced intensity conditioning extends our ability to transplant patients who are older - up to the early seventies - and have co-morbidities. The exact form of conditioning that should be used is still being used, but even among survivors there is a high risk of both acute and chronic graft-versus-host disease which may be very unpleasant.

Transplantation in CLL may justifiably be referred to as a work in progress.

Who needs a transplant?

I think another article about bone marrow transplants is due. A paper has appeared in Blood detailing the European experience in CLL, so it is timely to review this and to reacquaint readers with the issues.

First off, we don’t call it ‘bone marrow transplantation’ anymore, because we quite often take the graft from the blood. We tend to talk about stem cell transplants, though this is confusing now that infusions of embryonic stem cells are being used for some neurological diseases. The authors entitle their paper “Allogeneic hematopoietic cell transplantation for chronic lymphocytic leukemia: ready for prime time?”

Note the ‘allogeneic’. 25 years ago when I started in transplantation, we used to think that autografting; taking the donated marrow from the patient him (or her) self; would avoid the nasty complications of rejection and graft-versus-host disease. However, nobody is cured by an autograft. Tom McCune, who for a long time was thought to be the one patient to be cured in this way, has relapsed. I had a patient who went 13 years after an autograft and died of something else without relapsing, but I suspect he would have relapsed over time if he hadn’t got lung cancer.

So, let’s be clear, the purpose of allografting is to cure. How does it cure? Well, it’s not by allowing us to give more chemotherapy and then rescuing the damaged marrow with a graft. We used to think that was what we were doing, and the superiority over autografting was because the autograft would always be contaminated by residual disease; hence the fashion for laundering the marrow with monoclonal antibodies. However, allografts are also superior to transplants from identical twins whose marrow was completely clean.

No, it turns out that the efficiency of an allograft depends on its ability to mount an immune attack on the leukemia – the so-called ‘graft-versus-leukemia’ effect. This being so there was no longer any need to give the whopping doses of chemotherapy or radiotherapy, just enough immunosuppression was needed to ensure the graft would take. These days most transplants use ‘reduced intensity conditioning’ and because of this we can do them in increasingly older people. I used to joke that my ambition was to get so old that nobody would think of doing a transplant on me. I have a way to go as they are done routinely up to the age of 70 and exceptionally in people who are older. This is good news for patients with CLL since the median age of presentation is 72, and though some patients present at an earlier age (my youngest is 21) and older patients can often be managed to the end of their natural lifespan with palliative chemotherapy, nearly 50% of patients fall within the age when a transplant is feasible.

So who should have a transplant? I guess the answer is that bit is a judgement issue, since there are no randomized clinical trials to give us an answer. Even reduced intensity conditioning allografts are hazardous. So like matrimony, transplantation is not to be enterprized, nor taken in hand, unadvisedly, lightly, or wantonly, to satisfy man’s carnal lusts or appetites, like brute beasts that have no understanding; but reverently, discreetly, advisedly, soberly and in the fear of God, duly considering the causes for which transplants were ordained.

Transplantation should be reserved for patients with poor-risk disease, among them, those who are fludarabine refractory. The outlook for these patients is dismal. Before the availability of monoclonal antibodies, the overall response rate for the first salvage therapy after relapse from a fludarabine-induced remission was only 22% with a median overall survival (OS) of 10-12 months, and even though alemtuzumab is licensed for fludarabine-refractory patients, the response rate is only 33% and the median OS is only 16 months. FCR will produce a higher response rate in fludarabine refractory patients (58%), but the length of remission is probably not very long. I say probably because the paper does not tell us, though the more intense regimen of OFAR gives an overall survival of less than a year in this situation.

Define ‘fludarabine refractory’. This means that there was no PR or CR following treatment with standard doses of fludarabine or that if there was a response, progression occurred within 6 months of stopping treatment.

The major reason for failing to respond adequately to fludarabine is loss of p53 activity. p53 is a protein discovered more than a quarter of a century ago by Richard Lane at Dundee in Scotland. It is called the ‘guardian of the genome’. Its function is to detect damage to DNA and assess whether the damage is reparable. If it is, then damage repair is initiated, if it is not, then p53 starts up a program that results in cell death. Most anti-cancer drugs (including fludarabine, chlorambucil and bendamustine) use p53 induced apoptosis (programmed cell death) to kill tumor cells. We can detect that p53 has gone missing with FISH, since the p53 gene resides on the short arm of chromosome 17. The technical term for those who have lost it is del 17p.

Don’t we have two chromosome 17s? Isn’t the gene for p53 on the other chromosome 17 sufficient? In some cases it may be. We have certainly seen some del 17p patients who have a benign disease; these usually have mutated IGHV genes and do not require treatment according to the IWCLL guidelines. However, whenever a patient with del 17p requires treatment according to the IWCLL guidelines, the other chromosome 17 carries a p53 gene that codes for a disabled p53 molecule. Such patients are refractory to fludarabine and almost always refractory to FCR.

In clinical trials around 5-7% of patients presenting for treatment for the first time have del 17p, but in trials of patients presenting for salvage treatment following relapse, the incidence of del 17p is about 30%.

So who needs a transplant? Anyone aged 70 or under without significant co-morbidities who needs treatment and had del 17p or who is refractory to fludarabine.