A Real-World Analysis of the Outcomes of Septic Shock in Hematologic Patients with Acute Myeloid Leukemia

Background: Acute Myeloblastic Leukemia (AML) is an aggressive blood cancer characterized by bone marrow failure, with resulting neutropenia. Myelosuppressive treatment aimed at achieving complete remission is also complicated by neutropenia, leading to an increased risk of infection, febrile neutropenia (FN), sepsis and septic shock (SS). The Surviving Sepsis Campaign (SSC), developed within the context of Critical Care, provides guidelines for the approach to non-hematologic patients (pts) with sepsis and SS. We aim to look at real-world data on the outcomes of AML pts with SS treated according to SSC guidelines.

Methods: We studied all AML pts admitted to a single Center Hematology ward over a 3 year period, selecting the sub-cohort of non-palliative pts with an intercurrent diagnosis of septic shock. Patients with treatment-limiting decisions precluding the use of amines were excluded. Disease (AML and SS) characteristics, treatment approaches and outcomes were analyzed.

Results: Over the study period, 37 AML pts (48.6% male, median age: 62.2 years old) were diagnosed with shock. A third (35.1%) had AML with recurrent genetic abnormalities; 18.9% had AML with myelodysplasia (MDS)-related gene mutations; 24.3% had AML, NOS and a further 8.1% had AML, NOS progressed from MDS. There were one pt each with AML with mutated TP53; AML with MDS-related cytogenetic abnormalities; AML, NOS progressed from MDS/NPM; therapy-related AML, NOS; and myeloid sarcoma.

The median interval from the diagnosis of AML to the diagnosis of shock was 1.6 months (0-56.7), and in 35.1% of pts shock complicated the inaugural admission for AML.

FN was present in 73.0% of pts, starting at a median of 14 days (0-94) after admission; in 22.2% of these pts, FN was diagnosed upon presentation to the emergency room (ER); shock was identified a median of 1 day (0-39) after the diagnosis of FN; in 27.8% of pts, the two diagnoses were synchronous; in 13.5%, criteria for shock were present in the ER.

In 97.3% of pts, the etiology of shock was septic (one pt had pure cardiogenic shock), and empiric antibiotics were started on all; the most common first-line empiric antibiotics used were meropenem (n=10), cefepime (n=9) and piperacillin/tazobactam (n=8). Ultimately, 75.7% of pts had at least one germ isolated in microbiological cultures (with coinfection with 2 agents in 18.9% of pts, and with 3 in 5.4%). The most common agents were Klebsiella spp (with 17 isolates), E. coli (5 isolates), P. aeruginosa (4 isolates) and E. faecium and S. epidermidis with 3 isolates each. Useful Antibiotic Sensitivity Test (AST) results were obtained in 96.4% of pts with isolates, and in 55.6% these results led to an AST-guided change in antibiotics.

By definition, all pts had cardiovascular dysfunction, and in 13.5% of pts this was the sole dysfunction; 32.4% had 2 organ dysfunctions, 18.9% had 3, 2.7% had 5 and 8.1% had 6. Respiratory dysfunction was the second most common, in 67.6% of the cohort, followed by neurologic (45.9%), renal (35.1%), hematologic (29.7%) and hepatic (10.8%).

Aminergic support was provided with noradrenaline (NA) in 81.1% of pts (2.7% had dopamine, DA; and 13.5% had either a synchronous or sequential association of NA and DA). Only 18.9% of pts were accepted into an intermediate (IMCU) or intensive care (ICU) unit. Amines were continued to exitus in 43.0% of pts; in 57.0% cardiovascular disfunction was resolved and amines stopped, although 25.6% eventually died of complications of the septic episode. At last follow-up, only 10.8% of pts were alive; overall, 67.6% of pts died as a consequence of shock, 12.1% later died due to progressive disease, 6.1% due to a subsequent infectious complication, and 5.4% died outside our Centre, and the cause of death could not be established. Median overall survival (OS) after the diagnosis of shock was 9 days, with a median overall survival after AML of 123 days.

Discussion: In our cohort, we found that septic shock was a rapidly progressive complication of AML, occurring a median of 1 day after a diagnosis of FN; in a third of pts, shock complicated their first admission for AML (being present in the ER in an eighth of pts).

Multiorgan dysfunction was the norm; in over a quarter of patients, although cardiovascular dysfunction was resolved, the vital outcome was still unfavourable. Overall, nearly 70% of the cohort died as a direct consequence of the episode of sepsis and SS, with a median OS of just over a week after shock was diagnosed.

No relevant conflicts of interest to declare.