The Impact of Diagnosis to Treatment Interval on the Outcome of Patients with Acute Myeloid Leukemia

Introduction:

Acute myeloid leukemia (AML) has historically been considered an oncologic emergency, requiring immediate intervention and initiation of therapy. Although urgency to initiate treatment has been debated, no guidelines exist to address the impact of diagnosis to treatment interval (DTI) in patients with AML. Recent FDA approvals have added treatment options using targeted therapy as monotherapy in older patients (e.g. - ivosidenib), or in combination with chemotherapy in younger patients (e.g.- midostaurin) in the newly diagnosed setting with improved outcomes. Unfortunately, results of molecular mutation testing often return several days after the initial diagnosis of AML has been made.

While immediate induction is important to control disease and to minimize disease-related morbidity and mortality (Sekkers et al), molecular testing is becoming increasingly important for treatment decisions, and there may be benefit in waiting for these results. In addition, in 'real world settings', several barriers (e.g., lack of resources, correct diagnostic testing) often delay time to diagnosis, and initiation of therapy. We conducted a retrospective study in two academic centers in a major metropolitan area in the United States to evaluate the impact of DTI on the outcomes of patients with AML.

Methods:

We collected data from 196 patients who presented to George Washington University Hospital (120 patients) between June 2010 and July 2018 and Georgetown University hospital (76 patients) between January 2014 and December 2018. All patients were older than 18 years. Patients with acute promyelocytic leukemia and those who decided to pursue treatment elsewhere were excluded. Patients characteristics including age, gender, race, smoking past medical history and disease characteristics (Cytogenetics, FISH and NGS when available) were collected using retrospective chart review. Type of treatment received (low intensity vs high intensity), and the time from diagnosis to initiation of therapy were collected. We recorded mortality at day 30 and day 90. Multivariable logistic regression models were used to evaluate the association of DTI with mortality independently from patient age. Chi-square test was used to examine the association of 30- and 90-day mortality with DTI.

Results:

A total of 140 patients were used in the analysis, 71% of whom were treated within 5 days of diagnosis (DTI <1-5 days), 16% in 6-10 days, and 12% in >10 days. Incidence of 30-day mortality was 15%, 14%, and 0% (p=.24), while 90-day mortality was 28% 18%, and 13% for <1-5, 6-10, and >10 days respectively (p=.33). Neither association was statistically significant, and adjusting for age and type of therapy did not change the findings.

Conclusion:

Our data does not support that shorter DTI is associated with improved mortality regardless of age or type of therapy. Although there was no statistical significance, there was a numerical trend of improved mortality in patients with delayed DTI (6-10 days) compared to those with a DTI of <1-5 days. Larger prospective studies need to be conducted to investigate DTI in AML patients based on patient's age, disease characteristics, and treatment type. In particular, given the growing importance of knowing molecular mutations at diagnosis, used for both prognosis and incorporation of potential targeted agents, it will be important to understand how long we can safely wait to treat patients without affecting outcomes.