A new computer modeling programme has been devised to help predict the likelihood of complete remission or early death in patients aged 60 years or older undergoing intensive chemotherapy for acute myeloid leukaemia (AML). The findings are reported in an Article Online First and in an upcoming Lancet, and are being presented at the American Society of Hematology meeting in Orlando, Florida. The Article is by Dr Utz Krug, Professor Carsten Müller-Tidow and Professor Thomas Büchner, University of Münster, Germany, and colleagues.

Acute myeloid leukaemia is a cancer of the blood, in which white blood cells accumulate in bone marrow and interfere with regular blood cell production. It is the most common acute leukaemia affecting adults, and its incidence increases with age. Its prominence as a cause of death will rise as populations age worldwide.

About 50% of patients (age ≥60 years) who have acute myeloid leukaemia and are otherwise medically healthy (ie, able to undergo intensive chemotherapy) achieve a complete remission (CR) after intensive chemotherapy, but with a substantially increased risk of early death (ED) compared with younger patients. The decision whether or not to treat acute myeloid leukaemia with intensive chemotherapy can be very difficult. Haematologists mostly rely on personal experience and preference in making recommendations to individual patients.

Risk factors known to affect the chances of achieving CR risk include: age at diagnosis; serum concentration of lactate dehydrogenase* at diagnosis; leukaemia secondary to a previous haematological disease including various bone marrow-related conditions; leukaemia secondary to previous treatment with cytotoxic drugs or radiation for another disease (eg, breast cancer). Genetic changes, such as chromosomal abnormalities and gene mutations, in leukemic cells also affect survival and remission prospects for AML patients, both older and younger than 60 years.

Since no treatment algorithms are in clinical use that take these different variables into account, the authors obtained and verified the scores and developed a web-based application for the prediction of the probability of CR and risk of ED after intensive induction chemotherapy in patients who are 60 years or older, have acute myeloid leukaemia, and are otherwise medically healthy.

The study included 1406 patients (aged ≥60 years) with AML, but otherwise medically healthy, who were treated with one or two courses of intensive induction chemotherapy in the German Acute Myeloid Leukaemia Cooperative Group 1999 study. Risk prediction was validated in an independent cohort of 801 patients (aged >60 years) with acute myeloid leukaemia who were given two courses of cytarabine and daunorubicin in the Acute Myeloid Leukaemia 1996 study.

The researchers found that body temperature, age, de-novo leukaemia versus leukaemia secondary to cytotoxic treatment or an antecedent haematological disease, haemoglobin, platelet count, fibrinogen levels, and serum concentration of lactate dehydrogenase were significantly associated with CR or ED. The probability of CR with knowledge of cytogenetic and molecular risk (score 1) was from 12% to 91%, and without knowledge (score 2) was from 21% to 80%. The predicted risk of ED was from 6% to 69% for score 1 and from 7% to 63% for score 2. The predictive power of the risk scores was confirmed in the independent patient cohort from 1995 (CR score 1, from 10% to 91%; CR score 2, from 16% to 80%; ED score 1, from 6% to 69%; and ED score 2, from 7% to 61%).

Importantly, the authors noted that in their analysis, 25% of patients with high cytogenetic risk, who would often not be offered intensive treatment, were predicted to still have more than 40% probability of achieving CR. This subgroup might potentially benefit from an intensive approach, since a stem cell transplantation in remission becomes increasingly available even in this age group. The authors explain that without stem cell transplantation, the relapse rate in these high-risk patients is so high that an intensive approach is still considered to be only palliative, ie, in the vast majority of patients the disease will return within two years. Only a stem cell transplantation seems to be a potential curative approach in those high-risk patients. In the past, stem cell transplantation was associated with an excess toxicity and mortality in those elderly patients, and only recently improved stem cell transplantation protocols are available which allow this approach also in AML patients aged over 60 years.

By contrast, only 14% of patients with either low cytogenetic or molecular risk were not in the quarter of patients with the highest prediction for CR, and these patients still had a greater than 50% probability of CR, justifying an intensive approach.

The authors say: "Use of the score predictor might be most helpful for patients with acute myeloid leukaemia who achieve a very low or very high predicted rate of CR. For patients predicted to have about 50% probability of CR, not much might be gained by obtaining the score for CR... However, the scores for acute myeloid leukaemia have to be used with caution, and the responsibility for the final treatment recommendation remains with each patient's haematologist."

They conclude: "The scores for acute myeloid leukaemia might aid in the decision of whether intensive induction chemotherapy is the preferred option in an older, otherwise medically healthy patient who is diagnosed with acute myeloid leukaemia, and thus they could change the clinical practice for a large number of patients with this leukaemia."

In a linked Comment, Dr Felicetto Ferrara, Cardarelli Hospital, Naples, Italy, says: "The acute myeloid leukaemia score predictor could be very helpful, and would probably be limited to patients with a very low predicted rate of complete remission (ie