Suspecting Hyperferritinemic Sepsis in Iron-Deficient Population: Do We Need a Lower Plasma Ferritin Threshold?

Objectives: Hyperferritinemia is being suggested to identify patients with sepsis-induced macrophage activation syndrome for early intervention. However, data among iron-deficient children are scarce. This study was planned to explore the biological behavior of plasma ferritin in children from communities with a high frequency of iron deficiency with septic shock and its association with the outcome.

Design: Prospective observational study.

Setting: Tertiary care teaching hospital in a low-middle income economy of South Asia.

Patients Or Subjects: Patients (6 mo to 12 yr) (n = 42) with septic shock and their healthy siblings as controls (n = 36). Patients/controls with blood transfusion/iron supplement during last 6 months or with any chronic disease were excluded.

Interventions: None.

Measurements And Main Results: Ferritin was measured in patients at enrollment and then at 1 month of hospital discharge while they were not on iron supplementation and in controls as indicative of baseline level. Patients' median age was 30 months (13.5-87 mo), 31% were malnourished, majority (86%) had anemia, and two thirds had microcytic hypochromic red cells. Ferritin at admission was 763 ng/mL (480-1,820 ng/mL) in nonsurvivors, whereas 415 ng/mL (262-852 ng/mL) in survivors (p = 0.11). Pediatric Logistic Organ Dysfunction score and C-reactive protein correlated positively with plasma ferritin (p = 0.03 and p = 0.01, respectively) at enrollment. Elevated ferritin of greater than 500 ng/mL (relative risk, 2.48; 95% CI, 0.95-6.43) and greater than 1,000 ng/mL (relative risk, 1.94; 95% CI, 0.94-4.02) were associated with higher mortality but not independently. Among survivors, the 1-month follow-up ferritin fell significantly to 97 ng/mL (16-118 ng/mL) (p = 0.001). However, it was still significantly higher than that in sibling controls (19 ng/mL [10-54 ng/mL]) (p = 0.003).

Conclusions: Ferritin rises significantly in septic shock patients despite iron deficiency and seems to correlate with the severity of inflammation and organ dysfunction. Even a lower threshold (of 500 or 1,000 ng/mL) could predict higher mortality. It may suggest the need for redefining the plasma ferritin threshold for suspecting hyperferritinemic sepsis and sepsis-induced macrophage activation syndrome in these patients. Larger studies with frequent ferritin measurements are desirable to validate these initial observations.