Assessing the Elderly MDS Patient for Comorbidities, Frailty and Treatment Fitness


Increasing age is an important risk factor for the development of myelodysplastic syndromes (MDS), with a median age at diagnosis of 71 years. (1) In the MDS population, therefore, comorbidities are frequent and can complicate management.

Absolute age, however, is not the driving factor in MDS treatment. The “elderly” tennis player arriving to the clinic is far different than the same-age nursing home resident arriving in a wheelchair. Aging is a multidimensional and highly individualized process, and differences in physical and cognitive functional status can impact treatment selection and outcomes.

While MDS management usually begins with a determination of risk, the presence of comorbidities and frailty in the elderly can make the calculation of a “risk score” more complex. The International Prognostic Scoring System (IPSS), (2) the revised IPSS (IPSS-R), (3) and the WHO 2008 guidelines (4) can be useful in assessing risk of transformation to leukemia and can inform decision-making, but they do not take into account age-related health, functional, cognitive, and social factors that can affect outcomes as well as treatment feasibility. It may be unclear, therefore, which patients will benefit from intensive, potentially curative treatment and who will fare well with observation and best supportive care or low-intensity treatments.

Impact of Comorbidities
At least 50% of MDS patients have some comorbid condition. (5-8) In one study, this proportion reached 75%, with fully 55% of patients having a cardiovascular condition and 16% an endocrine abnormality. (6) Presence of comorbidities affected survival, which dropped from a median overall survival of 3.5 years without comorbidities to less than 1 year when three such conditions were identified.

The association between comorbidities and survival, independent of IPSS risk, has been shown in studies that have measured comorbidity according to the Hematopoietic Stem Cell Transplantation-Specific Comorbidity Index (HCT-CI), (9) the Adult Comorbidity Evaluation-27 instrument (ACE-27), (6) and the MDS-Specific Comorbidity Index (MDS-CI) developed by Italian investigators. (10)

These are among the tools that can be used to assess comorbidities, but no single tool explicitly focuses on the elderly MDS patient. The National Comprehensive Cancer Network (NCCN) MDS Panel, therefore, makes no specific recommendation for which comorbidity index to use, but does recommend thorough evaluation of the presence and extent of comorbid conditions in the management of MDS. (11)

Clinicians can benefit from knowing which conditions are the most worrisome. The ACE-27 includes numerous conditions related to the cardiovascular, respiratory, gastrointestinal, renal, endocrine, neurological, and immunological systems, as well as a number of psychiatric and rheumatologic illnesses, malignancies, substance abuse and body weight. (6) The following conditions generate the highest comorbidities scores on ACE-27 (12):

  • Cardiovascular: recent (≤6 months) myocardial infarction, hospitalization for congestive heart failure, ventricular arrhythmia, pulmonary embolism or bypass or amputation for peripheral arterial disease; ejection fraction <20%; diastolic blood pressure ≥130 mmHg; untreated thoracic or abdominal aneurysm
  • Respiratory: marked pulmonary insufficiency, restrictive lung disease or COPD, chronic supplemental oxygen, CO2 retention, baseline pO2 >50 torr, FEV1 <50%
  • Gastrointestinal: recent (≤6 months) ulcers requiring transfusion, portal hypertension or esophageal bleeding; acute or chronic pancreatitis with major complications
  • Neurological: acute stroke, severe dementia, paraplegia or hemiplegia, or chronic neuromuscular disorder
  • Others: end-stage renal disease, diabetes causing hospitalization and/or end-organ failure, recent suicide attempt, connective tissue disorder with secondary end-organ failure, cancer, withdrawal from alcohol or illicit drugs

Using the WPSS/MDS-CI
Italian researchers determined that five conditions impact survival in MDS patients and they included them in the MDS-CI (Table 1). (13) The WHO Classification-Based Prognostic Scoring System (WPSS) is an MDS prognostic scoring system that is used primarily in Europe. To determine whether the MDS-CI assessment could improve the accuracy of the WPSS, the researchers integrated these two instruments and found that the MDS-CI had a significant effect on overall survival in the very low/low and intermediate WPSS risk subgroups (P <.001). (13) The analysis confirmed the prognostic value of comorbidity in MDS.

Table 1 (10): Five Comorbidities Independently Associated with Risk

Comorbidity HR obtained through a multivariate Cox’s survival analysis with NLD as an outcome Variable weighted score (to be taken into account if the specific comorbidity is present)
Cardiac disease 3.57 (P<0.001) 2
Moderate-to-severe hepatic disease 2.55 (P = 0.01) 1
Severe pulmonary disease 2.44 (P = 0.005) 1
Renal disease 1.97 (P = 0.04) 1

Solid tumor

2.61 (P<0.001)


MDS-CI risk Sum of individual variable scores Proportion of patients in the learning cohort belonging to the risk group (%)
Low risk 0 546/840 (65%)
Intermediate risk 1 – 2 244/840 (29%)

High risk


50/840 (6%)

  NLD: non-leukemic death

Calculation of the MDS-specific comorbidity index (MDS-CI). The five comorbidities listed were found to be independently associated with the risk of non-leukemic death (NLD) in multivariable analysis, and each of them was assigned a score proportional to the regression coefficient of the multivariable Cox’s proportional hazards model. This score is taken into account if the specific comorbidity is present, and the MDS-CI is obtained as the sum of individual variable scores.

Implications for treatment planning can be drawn from risk stratification in the WPSS/MDS-CI model (Figure 1).

Figure 1 (10): MDS-CI Plus WPSS, Potential Combined Use in Decision-Making

Schematic representation of the potential of combined use of WPSS, a disease-related prognostic scoring system, and MDS-CI in clinical decision-making in MDS.

For patients with very low and low WPSS risk, assessment of comorbidity can help tailor monitoring schedules and can optimize supportive care. Among lower-risk persons aged ≥70, those with isolated erythroid lineage dysplasia have a life expectancy comparable to the general population. (14) In selected patients, treatment of anemia with erythropoiesis stimulating agents (ESAs) can positively impact outcome, (15,16) and there may be a survival benefit in preventing or adequately treating symptomatic anemia in MDS. (13) Especially in patients with clinically relevant comorbidity, especially cardiac disease, it is also important to optimally treat symptomatic anemia to limit the negative interaction between anemia and cardiac disease. (13)

When observation or a delayed treatment strategy is chosen for lower-risk patients, clinicians should not overlook the complications conveyed by comorbidities. (17,18) Patients requiring regular transfusions have been shown to be at risk for developing comorbidities, and these can be future obstacles to allogeneic transplantation. (19) Transfusion-dependency, in fact, is a major factor in deciding when to intervene in lower-risk MDS. (15)

Using ACE-27 Comorbidity Score Plus IPSS-R Risk
Using the ACE-27 comorbidity score, researchers from the University of Texas MD Anderson Cancer Center showed that assessment of comorbidity may enhance the prognostic ability of the IPSS-R. (20) The individual conditions of the ACE-27 are graded, and an overall comorbidity score (0-9) is generated. (12)

Adult Comorbidity Evaluation-27 PDF

The researchers assessed the severity of comorbidities using the ACE-27 on 600 MDS patients (median age 67) who were stratified by risk according to the IPSS-R. Putting these two risk estimates together, they created a prognostic model that incorporates comorbidities. The ACE-27 comorbidity score correlated well with survival, showing that patients with no comorbidities (score of 1) lived the longest, while those with severe comorbidities had the worst survival. The final prognostic score incorporated comorbidities, age, and IPSS-R, demonstrating significantly different survival probabilities (Table 2). Median overall survival was 53.3 months for patients in the lowest risk group, versus only 6.6 months for those in the highest (P ≤ .001).

Table 2 (20): Final Prognostic Score Incorporating Comorbidities, Age Group, and IPSS-R

Risk group

N = 576 (%)

Death N (%)

Median survival (mo)

95% CI

5 yr (%)

Low (values 0–4)

168 (29)

83 (49)




(values 5–6)

168 (29)

117 (70)




(values 7–8)

137 (24)

98 (72)




High >8

103 (18)

78 (76)




The Problem of Frailty in MDS
While comorbidity is important to consider, it is not the same as functional status and therefore may not completely separate the fit from the frail patient. Clinicians should also consider frailty, which is a state of decreased physiologic reserves that renders a person vulnerable to stressors. In MDS, the presence of anemia, bleeding, transfusion dependence, infection, and other occurrences related to disease can be considered stressors.

A thorough geriatric assessment, therefore, is universally recommended for determining the health status, including frailty, of older cancer patients. A 2013 meta-analysis of 13 studies of elderly patients with solid tumors showed that for about 50% of patients, a geriatric assessment revealed previously unknown problems. (21) Furthermore, up to 64% of patients suffered from severe treatment-related toxicity; nutritional status, functionality and comorbidity were associated with worse outcomes; and based on the geriatric assessment, clinicians changed the chemotherapy regimen 21% to 53% of the time.

In another recent analysis of 15 studies of elderly patients with hematologic malignancies, geriatric assessment detected multiple health issues in patients with good performance status. Impairments in geriatric domains had predictive value for mortality and, in some studies, for treatment-related toxicity. (22)

Assessing the Elderly Patient
In the elderly cancer patient, clinicians are advised to look for disabilities and age-related conditions that can contribute to frailty. The comprehensive geriatric assessment (CGA) is recommended in the NCCN Clinical Practice Guidelines for older adults with cancer. (23) The CGA is a systematic process that applies a set of validated tools to evaluate somatic, functional and psychological domains. (24)

As this can be a task, clinicians can conduct a simpler screening first, to identify patients with troublesome deficits or challenging geriatric problems who should be further explored with the CGA. A number of simpler screening tools have been developed for this, such as the Clinical Frailty Scale (CFS). The CFS, which was derived from the Canadian Study of Health and Aging (CSHA), is based on clinical judgment and has a predictive validity indistinguishable from that of larger 70-item index the 70-item CSHA Frailty Index. (25)

Among the nine frailty categories on the CFS, each 1-category increment significantly increases the risk of dying or entering institutionalized care within approximately 5 years:

Clinical Frailty Scale PDF

  • Very fit: robust, active, energetic, well-motivated and fit; patients commonly exercise regularly and are in the fittest fit group for their age.
  • Well: without active disease symptoms, but less fit than people in category 1; these patients often exercise or are occasionally very active.
  • Managing Well: medical conditions are well controlled, but patients are not regularly active beyond routine walking.
  • Vulnerable: while not dependent on others for daily help, symptoms may limit activities; patients may complain of being “slowed up” or tired.
  • Mildly frail: “slowing down” is more evident, and help is required for high-order instrumental activities of daily living.
  • Moderately frail: help is needed with both instrumental and non-instrumental activities of daily living; patients often have problems with stairs and bathing.
  • Severely frail: completely dependent for personal care, though seemingly stable and not at high risk of dying within about 6 months.
  • Very severely frail: completely dependent, approaching end of life.
  • Terminally ill: approaching end of life; life expectancy <6 months, but not otherwise evidently frail.

Personalizing Therapy
Another tool to help individualize treatment of elderly MDS patients is the Multidimensional Geriatric Assessment (MGA), a fairly simple classification for treatment tolerance that is incorporated into a decision tree. (26) MDS-related risk is identified using the IPSS, then patients are assigned to one of three risk categories of the MGA: go-go/fit, slow-go/vulnerable, and no-go/frail. Specific treatment recommendations for each group were developed. (26)

Table 3 (26): Individualized Rx decision in elderly with high-risk MDS

Category Therapy recommendation Therapeutic target
Go-go/fit Best supportive care Hematologic improvement
Allo-HSCT Curation, prolonged OS and PFS
Azanucleosides Prolonged OS and PFS, hematologic improvement, relief of symptoms, improved QOL
Investigational agents Therapeutic target according to aim of the investigational study
Slow-go/vulnerable Best supportive care Hematologic improvement
Azanucleosides Prolonged OS and PFS, hematologic improvement, relief of symptoms, improved QOL, curation
Investigational agents Therapeutic target according to aim of the investigational study
No-go/frail Best supportive care Hematologic improvement, QOL
(Azanucleosides) Improved QOL, hematologic improvement, relief of symptoms
Investigational agents Therapeutic target according to aim of the investigational study

Under this algorithm, go-go patients may be proposed for therapies comparable to those given to younger patients, ie, attempts to cure disease or prolong life, as well as relieve symptoms and improve quality of life. For a small proportion of elderly patients in this group, allogeneic hematologic stem cell transplant may be possible. A recent prospective study of 1,280 subjects found that patients ≥65 had 2-year survival outcomes comparable to those <65. (27)

No-go patients, on the other hand, may be better served by palliative and best supportive care. For slow-go patients, clinicians consider age-related life expectancy and the patients’ MDS-related life expectancy, taking into account their expectations and prediction for treatment tolerance.

MDS management can also be complicated when the elderly patient has comorbidities and frailty, lacks access to care or a good support system, suffers from impaired cognitive or functional status, or experiences other age-related impediments. These factors form a multidimensional scenario that often leads to treatment de-escalation and the potential for inadequate treatment. The decision as to what constitutes adequate treatment, with acceptable toxicity, can be challenging for the clinician.

The elderly MDS patient’s overall health must be considered when prognosticating survival and treatment tolerance. While one simple scoring system is not currently available for this patient population, clinicians should assess for frailty and comorbidities using any of the validated instruments to do so. This will help guide treatment decisions and their counseling of patients and families. Where possible, interventions to treat comorbid conditions and improve frailty can enhance patients’ ability to tolerate treatments, and will give them the best chance of prolonged survival.

Managing MDS would like to recognize and thank Celgene Corporation for their educational support of

©2017 MediCom Worldwide, Inc. All rights reserved