Are Clinical Trial Exclusion Criteria for Patients With AML Too Strict?

Watch our interview with Abby Statler, MD, MPH.

In a study presented at the 2019 ASH Annual Meeting, researchers found that certain comorbidities in patients with acute myeloid leukemia (AML) do not affect treatment response or survival, even though these comorbidities are often used as exclusion criteria in clinical trials of AML treatments.

“While intended to protect patients, trial eligibility criteria, when excessive, exclude patients irrespective of expected toxicities,” explained lead investigator Abby Statler, PhD, MPH, from Cleveland Clinic. “These results suggest future AML trials may liberalize comorbidity and organ function eligibility criteria, which will likely improve recruitment rates and provide equitable access to investigational products.”

To determine which commonly used criteria affect patients’ outcomes, Dr. Statler and researchers evaluated 1,082 adults with AML (median age = 60.6 years; range = 50-69.1 years) who received induction chemotherapy at Cleveland Clinic between 2003 and 2019. They then compared outcomes among those with and without baseline comorbidities/
organ dysfunction that have been used in clinical trial eligibility criteria.

More than half of enrolled patients (53.1%; n=574) were male, 86.8% were white (n=939), and 45.7% were Medicare beneficiaries (n=494). Most patients had intermediate-risk disease (55.3%; n=598) and the majority received intensive cytarabine-based treatment (n=901; 83.3%).

Among the group, the most common patient comorbidities were vascular (n=540; 49.9%), endocrine (n=272; 25.1%), and neurological (n=220; 20.3%) in nature.

After a median follow-up of 12.9 months (range = 4.7-43), median overall survival (OS) among the entire population was 15.1 months (range = 0.03-189). After researchers controlled for treatment, age, sex, insurance, number of comorbidities, AML etiology, and European LeukemiaNet risk, liver disease was the only comorbidity associated with a difference in OS (hazard ratio [HR] = 1.90; p=0.004).

The following baseline characteristics were not associated with worsening OS:

  • elevated aspartate aminotransferase (AST)
    • 3 times the upper limit of normal (ULN) vs. normal (HR=1.02; p=0.94)
    • 3-5 times ULN vs. normal (HR=1.42; p=0.30)
  • elevated alanine transaminase (ALT)
    • 3 times ULN vs. normal (HR=0.76; p=0.45)
    • 3-5 times ULN vs. normal (HR=1.58; p=0.23)
  • elevated bilirubin
    • 5 times ULN vs. normal (HR=1.34; p=0.08)

Minor renal dysfunction also was not associated with OS when measured by elevated creatinine levels (HR for creatinine 1.5 times ULN vs. normal = 1.06; p=0.52) or creatinine clearance (HR for 84-60 mL/min = 0.95; p=0.62).

The investigators identified two conditions associated with increased mortality: baseline left ventricular ejection fraction (LVEF) abnormalities and prolonged corrected QT interval (QTc). As severity of these abnormalities increased, Dr. Statler reported, the effect size increased in a dose-response fashion:

  • LVEF:
    • 50% − 40% (HR=1.38 compared with LVEF >50%; p=0.04)
    • <40% (HR=1.66; p=0.009)
  • QTc:
    • >480 ms (HR=1.36 compared with QTc ≤480 ms; p=0.04)
    • ≥501 ms (HR=1.72; p=0.001)

In addition, the only comorbidities associated with response were liver comorbidities (odds ratio = 0.26; p=0.03), and no comorbidities or organ dysfunction were associated with dose modifications during treatment.

Although these findings are limited by the use of data from a single treatment center, they suggest that many eligible patients with AML are excluded from trials for minor laboratory abnormalities that would not affect outcomes, Dr. Statler concluded. She acknowledged that severe liver disease is an appropriate exclusion criterion, but that AST, ALT, and mild bilirubin abnormalities and mild renal dysfunction could be safely removed.

Study authors report no relevant conflicts of interest.


Statler A, Hobbs BP, Radivoyevitch T, et al. The impact of comorbidities and organ dysfunction commonly used for clinical trial eligibility criteria on outcome in acute myeloid leukemia (AML) patients receiving induction chemotherapy. Abstract #16. Presented at the 2019 American Society of Hematology Annual Meeting, December 7, 2019; Orlando, FL.