Search

The significance of extreme hyperferritinemia and its association with certain diagnoses and prognoses are not well characterized. We performed a retrospective analysis of adult patients with at least one total serum ferritin (TSF) measurement ≥ 5000 µg/L over 2 years, in three university hospitals. Conditions associated with hyperferritinemia were collected, and patients were classified into 10 etiological groups. Intensive care unit (ICU) transfer and mortality rates were recorded. A total of 495 patients were identified, of which 56% had a TSF level between 5000 and 10,000 µg/L. There were multiple underlying causes in 81% of the patients. The most common causes were infections (38%), hemophagocytic lymphohistiocytosis (HLH, 18%), and acute hepatitis (14%). For TSF levels > 10,000 µg/L, there were no solid cancer or hematological malignancy without another cause of hyperferritinemia. Isolated iron-overload syndromes never exceeded TSF levels > 15,000 µg/L. Extreme hyperferritinemia (TSF levels > 25,000 µg/L) was associated with only four causes: HLH, infections, acute hepatitis and cytokine release syndromes. A total of 32% of patients were transferred to an ICU, and 28% died. Both ICU transfer rate and mortality were statistically associated with ferritin levels. An optimized threshold of 13,405 μg/L was the best predictor for the diagnosis of HLH, with a sensitivity of 76.4% and a specificity of 79.3%. Hyperferritinemia reflects a variety of conditions, but only four causes are associated with extreme hyperferritinemia, in which HLH and acute hepatitis are the most common. Extreme hyperferritinemia has a poor prognosis with increased mortality.

Glycosylated ferritin (GF) has been reported as a good diagnostic biomarker for adult-onset Still’s disease (AOSD), but only a few studies have validated its performance. We performed a retrospective study of all adult patients with at least one GF measurement over a 2-year period in one hospital laboratory. The diagnosis of AOSD was based on the expert opinion of the treating physician and validated by two independent investigators. Patients’ characteristics, disease activity, and outcome were recorded and compared. Twenty-eight AOSD and 203 controls were identified. Compared to controls, the mean GF was significantly lower (22.3% vs. 39.3, p < 0.001) in AOSD patients. GF had a high diagnostic accuracy for AOSD, independent of disease activity or total serum ferritin (AUC: 0.674 to 0.915). The GF optimal cut-off value for AOSD diagnosis was 16%, yielding a specificity of 89% and a sensitivity of 63%. We propose a modified diagnostic score for AOSD, based on Fautrel’s criteria but with a GF threshold of 16% that provides greater specificity and increases the positive predictive value by nearly 5 points. GF is useful for ruling out differential diagnoses and as an appropriate classification criterion for use in AOSD clinical trials.