New Model Predicts Survival in Treatment-Related MDS

Roxanne Nelson

December 16, 2011

December 16, 2011 — About 20% to 30% of cancer patients who have undergone chemotherapy and/or radiation treatment develop myelodysplastic syndrome (MDS); this represents a distinct form of the disease. A new prognostic model might help predict overall and leukemia-free survival among these patients with treatment-related MDS (t-MDS). In addition, the model might facilitate the development of risk-adapted therapeutic strategies in this population.

"Treatment-related MDS is a distinct subtype of myelodysplasia, and we have developed the first prognostic model specific for these patients," said first author Alfonso Quintás-Cardama, MD, who presented the data during an oral scientific session here at the American Society of Hematology 53rd Annual Meeting.

"It is a simple model, readily applicable to the clinical setting, and can be used to build adaptive therapeutic approaches for patients with t-MDS," said Dr. Quintás-Cardama, assistant professor in the Department of Leukemia at the University of Texas M.D. Anderson Cancer Center in Houston.

Need for a Prognostic Model

MDS related to previous treatment with radiation and/or chemotherapy differs from de novo MDS, in that it has a high frequency of chromosomal abnormalities (typically in the context of complex karyotype), a high rate of transformation to acute myeloid leukemia, and substantial resistance to the standard therapy used to treat MDS.

The most widely used prognostic model in MDS is the International Prognostic Scoring System, which was developed by the Myelodysplastic Syndrome Working Group to predict survival and risk for transformation to acute myeloid leukemia. It estimates risk on the basis of the percentage of blasts, karyotype, and number of cytopenias, but its development was based primarily on patients with de novo disease, explained Dr. Quintás-Cardama.

Other prognostic models have been developed, but they are also largely based on de novo disease; there is a need for a prognostic scoring system dedicated to t-MDS, Dr. Quintás-Cardama and colleagues explain. They developed a specific t-MDS prognostic model after evaluating the characteristics of a large cohort of patients with t-MDS.

Survival Factors Identified

Using the leukemia database at their facility, the authors identified 1950 patients with MDS who received treatment from 1998 to 2007. Of this group, 438 had 1 or more previous cancers that were treated prior to their MDS diagnosis. Of the 438 patients, 279 had received chemotherapy, radiotherapy, or both.

The median patient age was 67 years, and median time from previous treatment for a malignancy to the onset of MDS was 64.6 months (range, 2.6 to 532.0). Follow-up was a median of 6.0 months, and 178 patients in the cohort died after a median of 10.5 months.

On univariate analysis, the authors identified more than 10 significant factors that were associated with overall survival. Further analysis with a multivariate model revealed 7 factors that independently predicted survival: age (≥65 vs <65 years; hazard ratio [HR], 1.63), ECOG performance status (2 to 4 vs 0 to 1; HR, 1.86), cytogenetics (–7 and/or complex vs others; HR, 2.47), World Health Organization MDS subtype (refractory anemia with ringed sideroblasts, refractory anemia with excess blasts-1/2 vs others; HR, 1.92), hemoglobin (<11.0 vs ≥11.0 g/dL; HR, 2.24), platelets (<50 vs ≥50; HR, 2.01), and transfusion dependency (yes vs no; HR, 1.59).

Using these data from the multivariate analysis, the authors developed a model that divided patients into 3 prognostic groups.

Prognostic Groups

Prognostic Group Patients, n Risk Factors, n Median Survival, months
Good 57 0 to 2 34
Intermediate 154 3 to 4 12
Poor 61 5 to 7 5

The model also predicted a 1-year leukemia-free survival of 96% for patients with good prognosis, 84% for those with intermediate prognosis, and 72% for those with poor prognosis.

This model was subsequently validated in a test group of 189 patients with t-MDS diagnosed from 2008 to 2010. The median survival rates for low-, intermediate-, and poor-prognosis patients in this group were 26, 13, and 7 months, respectively (P < .001)

Need for Validation

This model is extremely predictive of outcome in t-MDS patients, which is a growing population of MDS patients, and uses variables similar to those in the larger MDS model, said Mikkael Sekeres, MD, MS, director of the leukemia program at the Cleveland Clinic Taussig Cancer Institute in Ohio. Dr. Sekeres was not involved in the study.

"It would be useful to validate it with data from an independent institution," he told Medscape Medical News. "With the panoply of prognostic tools at our disposal within the MDS field, as a group we need to determine the utility of ideal tools directed at specific conditions or the adaptation of a unifying instrument that could be applied in all MDS cases."

The authors have disclosed no relevant financial relationships.

American Society of Hematology (ASH) 53rd Annual Meeting: Abstract 967. Presented December 13, 2011.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: