What Are the Limitations of IPSS-M for MDS Prognosis?

The Molecular International Prognostic Scoring System (IPSS-M) is one of the most comprehensive validated risk tools for estimating how Myelodysplastic Syndromes (MDS) might behave. While a detailed risk score can create a sense of certainty, it is crucial to understand its limitations. The IPSS-M is a prognostic model—it estimates outcomes for large groups of people based on features of their disease at a given time. It cannot perfectly predict your individual outcome, and it is not a predictive tool that tells you exactly how you will respond to a specific treatment ^[1][2]. Furthermore, its reliability depends heavily on the quality of your medical tests, and it does not account for your overall health, frailty, or personal goals.

How Test Quality and Missing Information Affect Your Score

To calculate an IPSS-M score, your doctor inputs clinical measurements (like your hemoglobin and platelet blood counts), data from your bone marrow biopsy, chromosome analysis (cytogenetics), and genetic testing (Next-Generation Sequencing or NGS). The score is only as reliable as this input data:

• Blood Counts and Blast Counts: The score relies on accurate blood counts and the percentage of immature, abnormal cells (blasts) in your bone marrow. The primary measure for blasts in MDS scoring is typically done by a pathologist looking at cells under a microscope (morphology). Another method, flow cytometry, is supportive but can sometimes produce different estimates, especially if the marrow sample is diluted or of poor quality ^[3].
• Genetic Testing Quality and Missing Data: NGS panels vary significantly between hospitals. Some panels cover all the genes required for the IPSS-M, while others are more limited. If a gene is not tested, it does not mean the gene is normal. While the IPSS-M calculator can handle some missing data using statistical methods, missing or technically uncertain results increase the uncertainty of your final score ^[4][1].
• Complex Mutations: Simply knowing a gene is mutated isn’t always enough to predict risk. For example, humans usually have two copies of every gene. The TP53 gene behaves differently depending on whether one or both copies are mutated (its allelic state) and what other mutations are present alongside it ^[5][6].

Populations vs. Individuals

• Estimates for a Group: The IPSS-M groups patients into risk categories and provides statistical estimates for overall survival (how long patients live on average) and the risk of the disease transforming into acute myeloid leukemia (AML). These represent averages for a defined cohort, not a personal countdown or precise forecast ^[1][2].
• Population Differences: While the IPSS-M has been widely validated, its performance and calibration can vary in populations that were underrepresented in the original development studies. For example, some studies in Asian and South American groups have shown different performance levels, reflecting variations in patient selection, healthcare access, and regional treatment practices ^[7][8].
• Boundary Conditions: Disease classification matters. For patients whose disease sits on the boundary with a higher blast count (10% to 19% blasts, sometimes classified as MDS/AML or MDS with excess blasts depending on the specific medical classification system), standard MDS models like IPSS-M may be less effective at accurately separating risk groups ^[9].

What the Score Cannot Capture

Perhaps the most important limitations of the IPSS-M are the real-world factors it does not measure:

• Treatment Response: The IPSS-M is a baseline prognostic tool. It estimates risk from the data available at the time it is calculated. It does not reliably predict whether you will respond well to specific MDS therapies, such as hypomethylating agents (HMAs), nor is it an estimate of what would happen if you received no treatment at all ^[1][10].
• Dynamic Changes: The score is a snapshot in time and does not automatically update on its own. For example, developing a need for frequent blood transfusions is clinically important and provides additional prognostic information that a baseline score misses. A new or increasing transfusion need should prompt clinical reassessment rather than relying solely on your original score ^[11].
• Comorbidities and Frailty: The IPSS-M focuses entirely on the genetics and features of your MDS. It does not measure your age, other medical conditions (comorbidities), or your physical reserve and ability to recover from illness (frailty). These factors are crucial for determining overall survival and whether you are a candidate for intensive treatments like a stem cell transplant ^[12][13].
• Your Personal Goals: No mathematical model can measure your quality of life or what matters most to you. Treatment decisions should always be a shared decision-making process that balances your clinical risk score with your priorities, whether that involves independence, symptom relief, treatment burden, or longevity ^[14][15]. For instance, two patients with the exact same IPSS-M score might make completely different treatment decisions based on their age, other health conditions, and personal priorities.