Prognostic Risk Assessment in MDS: Incorporating the IPSS-R into Practice

Dr. Rami S. Komrokji
Rami S. Komrokji, MD
Professor of Medicine and Oncologic Sciences
University of South Florida College of Medicine
Vice Chair, Malignant Hematology Department
Moffitt Cancer Center
Tampa, Florida

Managing MDS recently interviewed Rami S. Komrokji, MD, an expert in malignant hematology at Moffitt Cancer Center and the University of South Florida, to learn more about prognostic risk assessment in myelodysplastic syndromes (MDS). In this interview, Dr. Komrokji discusses the importance of using prognostic risk assessment tools, such as the IPSS-R, to guide treatment decisions with patients and to individualize therapy.

Why is it important to perform a prognostic risk assessment and what specific information can be ascertained?

When managing MDS, clinicians should perform a prognostic risk assessment when a patient is diagnosed. In addition to giving us an idea about prognosis, these assessments also provide important information to guide treatment discussions with patients and their caregivers so that we can tailor therapy according to the disease risk. Prognostic risk assessments offer valuable information on overall survival and the likelihood of MDS evolving to acute myeloid leukemia (AML).

Prognostic risk models have evolved significantly over time. The first model, the French-American-British (FAB) Morphology Group classification, was released in 1982.1 Since that time, several additional risk classification systems have been developed to classify MDS prognosis and assess the potential for survival and evolution to AML. Years after the FAB model was introduced, the World Health Organization (WHO) issued a new classification system.2 In addition, several other risk models were developed, including the:

  • IPSS (International Prognostic Scoring System)3
  • IPSS-R (Revised IPSS)4
  • WHO Classification-based Prognostic Scoring System (WPSS)5
  • MD Anderson Comprehensive Scoring System (MDA-CSS)6
  • Lower-Risk MDS Prognostic Scoring System (LR-PSS) from MDA7

With so many different models to choose from, it can be confusing for community oncologists to decide on which prognostic risk assessment to use. The most widely accepted model is the IPSS-R, but guidelines from the National Comprehensive Cancer Network recommend using any prognostic risk assessment when patients are first diagnosed with MDS.8 Physicians are urged to become familiar with one of the models, apply it, and use it in clinical practice. In many cases, community oncologists will perform prognostic risk assessments without objectively examining variables in those models. Many of the assessments have the same variables (eg, cytopenias, blast percentages, and cytogenetics), but newer models like the IPSS-R and MDA-CSS add to or refine prognostic information, which can influence treatment decisions.

What limitations should community oncologists be aware of with the IPSS-R?

First developed and published in 1997, the IPSS was intended to be calculated at the time of an MDS diagnosis in untreated patients. This model was the gold standard for prognostic risk stratification for many years. Both the FAB and the IPSS served as the basis of clinical trials leading to the FDA approval of several MDS therapies, including lenalidomide, azacitidine, and decitabine. An important shortcoming of the IPSS model is that it is not dynamic. It fails to provide enough detail on cytogenetic classification and doesn’t account for depth of cytopenias. For example, patients with low platelet counts can receive the same point value in the IPSS as those with higher platelet levels. In real-world settings, patients with low platelet counts often require transfusions more frequently, making them inherently different from those with higher platelets levels.

To improve on shortcomings with the IPSS, developers of the IPSS-R based their modifications on a larger sample of MDS patients. The IPSS-R built on its predecessor by dividing patients into five cytogenetic prognostic subgroups (very good, good, intermediate, poor, and very poor) instead of three (good, intermediate, and poor). The IPSS-R also addresses specific cytogenetic abnormalities within each subgroup, such as deletion 7q and deletion 11q, among others. This allows for more accurate prognostication with the IPSS-R than the IPSS.9

Other efforts were made to overcome IPSS shortcomings with the IPSS-R, including altering the cutoff for blast cells in bone marrow. The IPSS-R cutoff is now divided into four groups: 1) ≤2%, 2) >2% to <5%, 3) 5% to 10%, and 4) >10%. In addition, IPSS-R accounts for the depth of cytopenias for anemia, thrombocytopenia, and absolute neutrophil count. With IPSS-R, patients receive a score based on the sum of blasts, cytopenias, and cytogenetics. Patients are then classified into one of five risk groups: 1) very low, 2) low, 3) intermediate, 4) high, and 5) very high. This is different from the original IPSS, which has just four risk groups.

Several research groups have validated the IPSS-R, and these studies suggest that this prognostic model adds value to the IPSS. It upstages patients who were previously classified as lower-risk by the IPSS. With more variables under consideration, the IPSS-R offers clinicians a better tool to predict prognosis and risks for transformation to AML. With very low-risk MDS, the median survival is almost 8.8 years and the time to AML transformation is not reached. With very high-risk MDS, the median survival is 0.8 years and about 25% would develop AML in less than a year.4

Despite progress seen with the IPSS-R, the model has limitations. It is still not intended to be a dynamic model used at the time of diagnosis specifically for MDS and it does not include MDS/MPN (myeloproliferative neoplasms) or therapy-related MDS. Another important limitation with the IPSS-R is that the blast percentage cutoff is sometimes not reported in bone marrow results, especially in the community setting. Blasts may register as less than 5%, but more detail is needed to optimally classify the blast percentage in the cutoffs utilized in the IPSS-R.

Importantly, physicians will need to take extra steps to individualize treatments for patients who fall into the intermediate-risk group using the IPSS-R. These patients have many variables that must be taken into consideration when deciding on therapy, including age, comorbidities, molecular abnormalities, transfusion dependency, and functional status. Other important factors may include LDH and the presence of bone marrow fibrosis. Ultimately, individualized treatment is more complex for patients in this category and it is challenging to determine if patients should be treated as having high- or low-risk disease.

How can community oncologists tailor therapies based on prognostic risk, and what role does genetic testing play to support information collected in risk assessments?

The only curative procedure for MDS is allogeneic stem cell transplantation (SCT), and data collected in prognostic risk assessments can be used to determine if the disease risk will exceed transplant-related mortality risk. Based on prognostic data, clinicians can tailor therapy accordingly. For example, when managing patients with higher-risk MDS, allogeneic SCT is a possible treatment option, but an alternative approach may be to use hypomethylating agents (eg, azacitidine) to improve overall survival. For lower-risk MDS, clinicians can take a stepwise approach. The goal of treatment for these patients is to alleviate cytopenia and manage anemia. Treatments may include erythropoiesis-stimulating agents (ESAs), hypomethylating agents, lenalidomide, anti-thymocyte globulin (ATG), and cyclosporine. Risks with SCT outweigh benefits in lower-risk MDS. For intermediate-risk MDS, clinicians must weigh the risks and benefits of SCT and decide on the best timing to administer this treatment. Younger patients gain the greatest survival benefit with SCT when receiving transplants early in the disease course. Older patients may be candidates for SCT if they have a good performance status and no comorbidities.

Efforts are being made to incorporate data from genetic testing into risk assessment models. Research has shown that some somatic mutations are predictive of outcome. A recent study found that mutations in TP53, EZH2, ETV6, RUNX1, and ASXL1 are important predictors of poor overall survival in MDS, independent of established risk factors,10 and other somatic mutations can also be used to complement risk assessment models. Of note, the mutations in TP53 are especially important in treatment decisions for MDS. If TP53 mutations are detected at the time of SCT, the benefits from transplant are short lived. Patients with this mutation may have better outcomes with hypomethylating agents than intensive chemotherapy.

With recent advances in MDS treatments, we now have several targeted therapies that are directed toward specific genetic mutations, including IDH1, IDH2, and FLT3. These therapies should be considered for those at risk for progressing to AML. Testing MDS patients for these genes and others can help us tailor therapy accordingly. With genetic testing becoming more widely available, it may eventually enhance our ability to diagnose and classify patients, and perhaps predict therapeutic responses in the way it has with AML and MPN.11

What are the key points for community oncologists regarding diagnosing and risk stratifying patients with MDS?

Many MDS patients are referred to specialists without having received a formal prognostic risk assessment. Community oncologists need to recognize the importance of these tools and perform assessments when patients are diagnosed with MDS. Prognostic information collected in risk assessments should be conveyed to patients and caregivers and used to tailor therapy. Several risk assessment models are available, each with strengths and limitations, but the key is to become familiar with one model and use it objectively. The IPSS-R is one of the most validated prognostic risk tools available. At the time of diagnosis, higher-risk patients should be considered for SCT whereas lower-risk patients can be managed using a stepwise approach. Intermediate-risk patients will require individualized treatment. Data from risk assessments can be complemented by using genetic testing for somatic mutations using next-generation sequencing tools.


  1. Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the classification of the myelodysplastic syndromes. Br J Haematol. 1982;51:189.
  2. Vardiman JW, Thiele J, Arber DA, et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood. 2009;114(5):937-951.
  3. Greenberg P, Cox C, LeBeau MM, et al. International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood. 1997;89(6):2079-2088.
  4. Greenberg PL, Tuechler H, Schanz J, et al. Revised international prognostic scoring system for myelodysplastic syndromes. Blood. 2012;120(12):2454-2465.
  5. Malcovati L, Della Porta MG, Strupp C, et al. Impact of the degree of anemia on the outcome of patients with myelodysplastic syndrome and its integration into the WHO classification-based Prognostic Scoring System (WPSS). Haematologica. 2011;96(10):1433-1440.
  6. Kantarjian H, O’Brien S, Ravandi F, et al. Proposal for a new risk model in myelodysplastic syndrome that accounts for events not considered in the original International Prognostic Scoring System. Cancer. 2008;113(6):1351-1361.
  7. Garcia-Manero G, Shan J, Faderl S, et al. A prognostic score for patients with lower risk myelodysplastic syndrome. Leukemia. 2008;22(3):538-543.
  8. National Comprehensive Cancer Network. Myelodysplastic syndromes. Version 2.2018—February 15, 2018. Available at Accessed June 13, 2018. Login required.
  9. Deeg HJ, Scott BL, Fang M, et al. Five-group cytogenetic risk classification, monosomal karyotype, and outcome after hematopoietic cell transplantation for MDS or acute leukemia evolving from MDS. Blood. 2012;120(7):1398-1408.
  10. Bejar R, Stevenson K, Abdel-Wahab O, et al. Clinical effect of point mutations in myelodysplastic syndromes. N Engl J Med. 2011;364(26):2496-2506.
  11. Bejar R. Prognostic models in myelodysplastic syndromes. Hematology Am Soc Hematol Educ Program. 2013;2013:504-510.