High Ferritin Levels Flag Deadly Risk in Tuberculosis Patients with Low Blood Counts

A simple blood test that costs pennies to perform may help clinicians identify which tuberculosis patients are quietly sliding toward one of the most lethal complications of infectious disease. A new retrospective cohort study from Wenzhou Central Hospital, affiliated with Wenzhou Medical University in China, suggests that among pulmonary tuberculosis patients whose blood counts have begun to collapse, a serum ferritin level above 500 micrograms per liter marks a dramatic spike in the risk of dying during hospitalization. The finding, published in BMC Infectious Diseases, is exploratory in nature and drawn from a small cohort, but it points to a readily available biomarker that could serve as an early warning signal in a patient group where minutes and hours matter.

The clinical problem the researchers set out to address is a deceptively difficult one. Tuberculosis, caused by Mycobacterium tuberculosis, is usually a chronic infection marked by cough, fever, night sweats and weight loss. But in a subset of patients, the disease triggers a runaway inflammatory response that can spiral into hemophagocytic lymphohistiocytosis, or HLH, a hyperinflammatory syndrome in which the immune system turns on the body’s own blood-forming and blood-filtering tissues. In HLH, activated macrophages and T cells flood the circulation with inflammatory cytokines and begin devouring red cells, platelets and white cells in the bone marrow and spleen. The result is cytopenia, a shortage of one or more blood cell lines, and, if untreated, multi-organ failure and death.

Early HLH is notoriously hard to spot in a tuberculosis ward. The fever, fatigue and wasting that herald the syndrome look almost identical to the toxemic symptoms of active tuberculosis itself, and the systemic inflammatory response that accompanies severe infection can mimic the syndrome’s laboratory signature. By the time the classic findings appear, such as very high triglycerides, low fibrinogen, low or absent natural killer cell activity and hemophagocytosis visible on a bone marrow biopsy, the patient may already be critically ill. Ferritin, an iron-storage protein that also behaves as an acute-phase reactant, sits at the center of the standard diagnostic scoring systems for HLH precisely because extreme hyperferritinemia is one of its most consistent features. What has remained unclear, however, is whether ferritin carries prognostic weight, not just diagnostic weight, in tuberculosis patients who already show falling blood counts.

To probe that question, the team led by Zhengxing Wu and colleagues assembled a retrospective cohort of pulmonary tuberculosis patients admitted to their hospital between 2018 and 2025 who had bicytopenia or pancytopenia, meaning a reduction of two or more of the three principal blood cell lines: red cells, white cells and platelets. From the hospital’s clinical records they identified 43 such patients and divided them into two groups according to serum ferritin. Twenty-four patients had hyperferritinemia, defined in this study as a ferritin concentration above 500 micrograms per liter, while 19 patients had ferritin levels at or below that threshold. The researchers then compared the two groups on clinical characteristics, a panel of laboratory indicators and, most importantly, in-hospital mortality.

The headline result was stark. In the hyperferritinemia group, 37.50 percent of patients died during their hospital stay, compared with just 5.26 percent in the lower-ferritin group, a difference the authors report as statistically significant with a P value of 0.026. In other words, patients with elevated ferritin were roughly seven times more likely to die in hospital than their counterparts with lower ferritin values. A bar chart comparing the two mortality rates, included in the paper, underlines how sharply the two groups diverged despite their shared diagnosis and shared degree of cytopenia.

Intriguingly, the conventional laboratory markers that clinicians often lean on to gauge severity did not separate the groups. Platelet counts and lactate dehydrogenase, or LDH, an enzyme released when tissues are damaged or when cells are being destroyed at high rate, showed no statistically significant differences between the hyperferritinemic and lower-ferritin patients, with P values above 0.05. That absence of difference is itself informative. It suggests that ferritin may capture a dimension of inflammatory dysregulation, plausibly the cytokine-driven macrophage activation that defines HLH, which routine markers of cell destruction and marrow suppression do not fully reflect. In a patient whose platelets look similarly depressed on paper, a ferritin above 500 micrograms per liter may be the clue that the disease process has crossed into a far more dangerous phase.

To account for the possibility that age, a well-known determinant of survival in severe infection, was confounding the association, the researchers performed a logistic regression analysis adjusted for age. Ferritin above 500 micrograms per liter remained independently associated with in-hospital mortality, with an odds ratio of 10.011 and a 95 percent confidence interval stretching from 1.041 to 96.272, yielding a P value of 0.046. The authors are appropriately candid about what that confidence interval means. Spanning nearly two orders of magnitude, it signals considerable statistical uncertainty, a direct consequence of the small sample size. The true effect could plausibly be modest, or it could be enormous; the data cannot yet distinguish between those possibilities with precision.

That caveat threads through the entire study, and the authors frame their conclusions as exploratory and hypothesis-generating rather than definitive. Retrospective cohort designs of this kind are vulnerable to selection effects, missing data and the possibility that ferritin elevation in some patients reflected causes other than tuberculosis-driven hemophagocytosis, including liver disease, transfusion history or other inflammatory conditions. The single-center design and the modest number of 43 patients further limit generalizability. The 500 micrograms per liter threshold, while a recognized cutoff in HLH diagnostic criteria, was not derived from an optimized receiver operating characteristic analysis in this cohort, and the study did not establish a dose-response relationship between ferritin concentration and mortality risk. Prospective validation in larger, multi-center cohorts, ideally with serial ferritin measurements and standardized workups for secondary HLH, will be needed before any clinical algorithm can be built on these findings.

Even so, the practical implications are worth taking seriously. Ferritin is already measured routinely in febrile, severely ill patients in most hospitals, and the assay is fast, cheap and widely available even in resource-limited settings where the burden of tuberculosis is heaviest. If the association holds up, a ferritin above 500 micrograms per liter in a tuberculosis patient with two or more depressed blood cell lines could trigger closer evaluation for secondary HLH, earlier consideration of immunomodulatory treatment, transfer to higher-acuity care or simply more vigilant monitoring, decisions that currently rest on clinical gestalt in a syndrome that hides behind tuberculosis’s own face. The study, funded by the Wenzhou Science and Technology Plan Project and approved by the ethics committee of Wenzhou Central Hospital, was conducted in accordance with the Declaration of Helsinki, and because of its retrospective design the requirement for written informed consent was waived. The authors declare no competing interests, and the article is published open access under a Creative Commons Attribution 4.0 license, making the full dataset of findings freely available to clinicians and researchers worldwide who may wish to test the hypothesis in their own populations.

For a disease that kills well over a million people each year, tuberculosis has no shortage of severity scores, imaging findings and laboratory panels. What it has lacked is a simple, early flag for the hyperinflammatory catastrophe that can erupt when the infection and the immune response collide. This small study does not prove that ferritin is that flag, and its authors are the first to say so. But it offers a concrete, testable signal: among tuberculosis patients whose blood counts are falling, the iron stores circulating in their plasma may be quietly announcing which of them is fighting for their life. The next step, larger and prospective, will determine whether that announcement comes soon enough to change the outcome.

Subject of Research: Serum ferritin as a prognostic marker for in-hospital mortality in pulmonary tuberculosis patients with cytopenia

Article Title: The impact of hyperferritinemia on prognosis in pulmonary tuberculosis patients with bicytopenia/pancytopenia: an exploratory retrospective cohort study

Article References: Wu, Z., Shi, J., Zhou, J., Li, S., Zhou, Y., & Jiang, X. (2026). The impact of hyperferritinemia on prognosis in pulmonary tuberculosis patients with bicytopenia/pancytopenia: an exploratory retrospective cohort study. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14613-8

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14613-8

Keywords: tuberculosis, ferritin, hyperferritinemia, cytopenia, hemophagocytic lymphohistiocytosis, in-hospital mortality, prognosis, retrospective cohort study, biomarker, LDH, platelet count, BMC Infectious Diseases