Prevalence and Predictive Value of Anemia and Dysregulated Iron Homeostasis in Patients with COVID-19 Infection

Infections with SARS-CoV-2 can result in severe clinical manifestations. As such patients present with systemic inflammation, we studied the prevalence and predictive value of anemia of inflammation (AI) or functional iron deficiency (FID), originating from immune-mediated alterations of iron homeostasis. Within this retrospective analysis of 259 hospitalized patients with COVID-19, we found that, upon admission, 24.7% were anemic, with the majority suffering from AI (68.8%). Anemia was associated with a significantly higher in-hospital mortality (OR 3.729 (95%CI 1.739–7.995), p = 0.001) but not an increased frequency of intensive care unit (ICU) admission or need for mechanical ventilation. FID was present in 80.0% of patients upon admission, linked to more advanced inflammation and associated with significantly longer hospital stay. Notably, a ferritin/transferrin ratio > 10 predicted a five-fold higher risk of ICU admission and an eight-fold higher risk of the need for mechanical ventilation. Anemia and alterations of iron homeostasis are highly prevalent in hospitalized COVID-19 patients. Iron metabolism biomarkers and hemoglobin can contribute to risk stratification of patients, as initial anemia is associated with increased mortality, whereas alterations of iron homeostasis with a higher ferritin/transferrin ratio reflect more advanced inflammation and predicts subsequent insufficient pulmonary oxygenation with the need for ICU admission and mechanical ventilation.


Introduction
Infections with the severe acute respiratory syndrome-related coronavirus 2 (SARS-CoV-2) emerged as a worldwide pandemic, termed as the coronavirus disease 2019 (COVID-19) affecting millions of people [1,2]. While the majority of infections have a mild clinical course, up to 20% of infected patients need to be hospitalized, mainly because of pneumonia, with eventual admission to intensive care units (ICU) and a need for mechanical ventilation [3,4]. Such severe COVID-19 infections are characterized by a hyper-inflammatory state being associated with increased concentrations of inflammation markers such as C-reactive protein (CRP), interleukin-6 (IL-6) but also of ferritin [5,6]. The latter is the major iron storage protein, the expression of which is induced by iron loading or inflammation. In COVID-19 patients, elevated ferritin levels were related to disease severity, 2 of 11 development of acute respiratory distress syndrome (ARDS) and death [5,7,8]. Inflammation leads to typical alterations of iron homeostasis hallmarked by increased iron acquisition and retention within macrophages along with reduced intestinal iron absorption [9]. This results in a reduction of circulating iron levels and a reduced availability of the metal for erythropoiesis where it is needed for the production of hemoglobin. Together with cytokine-mediated inhibition of erythropoiesis, shortened erythrocyte half-life and reduced biological activity of the red cell hormone erythropoietin, this results in the development of anemia of inflammation (AI) [10]. This type of anemia is characterized by reduced circulating levels of iron and transferrin, a reduced saturation of transferrin with iron (TfS) while ferritin levels are normal or increased, which contrasts AI with iron deficiency anemia (IDA) where ferritin levels are low and transferrin levels are usually elevated [10,11]. The presence of anemia has been linked to an unfavorable clinical course of many diseases, including infections or inflammatory disorders [12][13][14][15][16]. Thus, we investigated the prevalence and causes of anemia and alterations of iron homeostasis in a cohort of hospitalized patients with COVID-19 upon admission and linked it to the clinical course and outcome from the disease.

Study Population
We retrospectively analyzed the data of 259 patients with PCR-confirmed SARS-CoV-2 infections who needed hospitalization either at the Innsbruck University Hospital (n =129) or Hospital St. Vinzenz Zams (n = 130) between 25 February and 20 May, 2020. The study conformed to the principles of the Declaration of Helsinki and was approved by the ethics committee of the Medical University of Innsbruck (ethical vote: 1167/2020, approved 24 July 2020).

Outcome Analysis and Measurements
Fatal events, ICU admission and need/duration of invasive ventilation were recorded during the patients' hospital stay. The event-free survival was defined as the period of time between the hospitalization date and patients' in-hospital death.
Blood samples were taken from patients at baseline (day 1 ± 1) and were analyzed by fully automated tests in the laboratories of the hospitals, which undergo regular quality control. Laboratory findings were then extracted from the clinical information system.

Statistical Analysis
We used the Shapiro-Wilk test to analyze for Gaussian distribution. Variables are depicted as n (%) or medians (25th, 75th percentile), since most biomarkers were not normally distributed. Mann-Whitney-U test, Kruskal-Wallis test or Pearson chi-square tests were performed to test for significant differences between groups. Logistic regression analysis was performed to analyze the effects of risk factors on the probability of complications and death. Risk factors that were significant in univariate logistic regression analysis were considered for multivariate logistic regression analysis.
Variables that showed no Gaussian distribution were logarithmized with the natural logarithm for the logistic regression analysis. Spearman rank correlation test was used to correlate continuous variables.
All tests were two-tailed and p-values < 0.05 were regarded as statistically significant. Statistical analysis was performed using SPSS Statistics Version 25.0 for Macintosh (IBM Corporation, Armonk, NY, USA).

Patients Characteristics
We retrospectively analyzed 259 hospitalized patients with COVID-19 infection with a median age of 68 years (53−80 years): 157 men and 102 women. Women were significantly older compared to men (74 years vs. 64 years, p = 0.015). During a median hospitalization time of 9 days (5-17 days), 32 patients died and 53 patients were transferred to the ICU (of whom 35 patients needed mechanical ventilation). Of the patients who died, 40.6% died in the ICU and 21.9% died while on mechanical ventilation. While death rates did not significantly differ between men and women (14.6% vs. 8.8%, p = 0.164), men were more often transferred to ICU (24.8% vs. 13.7%, p = 0.030) and needed mechanical ventilation more frequently (18.5% vs. 5.9%, p = 0.004) compared to women.

Disturbances of Iron Homeostasis in Patients with COVID-19 Infection
Iron metabolism biomarkers were available from 222 patients. Upon admission, 194 patients (88.2%) had an abnormal iron homeostasis, and the majority of those had functional iron deficiency (FID, n = 176, 79.3%; Table S2). The presence of FID was associated with significantly poorer clinical conditions and a longer hospital stay, but was not linked to an increased risk of in-hospital death, ICU admission or the need for mechanical ventilation (Tables 2 and 3 and Table S1). When analyzing variables of iron homeostasis in terms of patient outcomes, no significant relations were found between mortality rate and elevated ferritin levels, low Tf levels, or low TfS (Table 2). However, elevated ferritin levels were associated with longer hospital stays (rs = 0.251, p < 0.001), an increased risk for ICU admission (OR 2.780 (95%CI 1.874−4.124), p < 0.001, Table 3) and the need for mechanical ventilation (OR 3.497 (95%CI 2.124−5.757), p < 0.001, Table S1). In addition, low Tf levels upon admission but not a low TfS were associated with a longer hospital stay (rs = −0.286, p < 0.001), an increased risk for ICU admission (OR 0.059 (95%CI 0.015−0.226), p < 0.001, Table 3) and the need for mechanical ventilation (OR 0.079 (95%CI 0.018-0.337), p = 0.001, Table S1). Based on the findings of a significant association of either ferritin or transferrin with the clinical course and previous observations indicating that a ferritin/transferrin ratio can discriminate well between true and FID [18] and identify IDA in non-inflammatory subjects with borderline ferritin levels [19], we studied whether a ferritin/transferrin ratio would be an even more powerful predictor of the patients' clinical course. A ferritin/transferrin ratio > 10 (n = 37) well discriminated patients with a higher risk of ICU admission (OR 5.702 (95%CI 2.625-12.388), p < 0.001, Figure 1c and for need of mechanical ventilation (OR 8.054 (95%CI 3.369−19.249), p < 0.001, Table S1, Figure 1d compared to patients with a ferritin/transferrin ratio ≤ 10 (n = 185). An elevated ferritin/transferrin ratio was also associated with longer hospital stays (rs = 0.259, p < 0.001).
conditions and a longer hospital stay, but was not linked to an increased risk of in-hospital death, ICU admission or the need for mechanical ventilation (Tables 2, 3 and S1). When analyzing variables of iron homeostasis in terms of patient outcomes, no significant relations were found between mortality rate and elevated ferritin levels, low Tf levels, or low TfS (Table 2). However, elevated ferritin levels were associated with longer hospital stays (rs = 0.251, p < 0.001), an increased risk for ICU admission (OR 2.780 (95%CI 1.874−4.124), p < 0.001, Table 3) and the need for mechanical ventilation (OR 3.497 (95%CI 2.124−5.757), p < 0.001, Table S1). In addition, low Tf levels upon admission but not a low TfS were associated with a longer hospital stay (rs = −0.286, p < 0.001), an increased risk for ICU admission (OR 0.059 (95%CI 0.015−0.226), p < 0.001, Table 3) and the need for mechanical ventilation (OR 0.079 (95%CI 0.018-0.337), p = 0.001, Table S1). Based on the findings of a significant association of either ferritin or transferrin with the clinical course and previous observations indicating that a ferritin/transferrin ratio can discriminate well between true and FID [18] and identify IDA in non-inflammatory subjects with borderline ferritin levels [19], we studied whether a ferritin/transferrin ratio would be an even more powerful predictor of the patients' clinical course. A ferritin/transferrin ratio > 10 (n = 37) well discriminated patients with a higher risk of ICU admission (OR 5.702 (95%CI 2.625-12.388), p < 0.001, Figure 1c and for need of mechanical ventilation (OR 8.054 (95%CI 3.369−19.249), p < 0.001, Table S1, Figure 1d compared to patients with a ferritin/transferrin ratio ≤ 10 (n = 185). An elevated ferritin/transferrin ratio was also associated with longer hospital stays (rs = 0.259, p < 0.001).

Discussion
Our data indicate that anemia, and specifically AI, is prevalent in patients with severe SARS-CoV-2 infection and that anemia is associated with longer hospital stays, poor clinical conditions and poor survival. This may be linked to reduced tissue oxygenation along with anemia being a reflection of co-morbidities such as impaired renal function or older age [20] or advanced inflammation [10]. It is also possible that multi-morbid patients were more likely to be anemic and less likely to be transferred to ICU when their physical condition deteriorated, a notion which needs to be systemically studied. This would be in line with the observation that only 40.6% of all fatalities occurred on ICU and only 21.9% of the patients who died had been on mechanical ventilation. Moreover, anemic patients had a higher prevalence of comorbidities such as arterial hypertension, cardiovascular disease or chronic kidney disease, all of which are known risk factors for COVID-19-associated death [5]. Moreover, it will be important to perform a future analysis on whether or not pre-existing anemia is already a risk factor for COVID-19 infection and/or a fatal clinical course [21]. Of interest, high ferritin and low transferrin levels were associated with an increased risk for ICU admission and the need for mechanical ventilation. This is in accordance with the association of altered iron homeostasis with more advanced inflammation, the latter promoting lung injury and respiratory failure [22,23]. We introduced a ferritin/transferrin ratio which seems to be a robust and easily available marker for risk stratification at initial presentation of patients in the hospital in terms of ICU admission and the need for mechanical ventilation in patients with COVID-19. FID per se did not result in higher mortality rates, risk for ICU admission, or a need for mechanic ventilation which would be in line with the fact that infection triggered macrophage iron retention as a host defense mechanism to withhold the nutrient iron from invading microbes [24][25][26], thereby possibly reducing the risk of secondary bacterial and fungal infection in those patients, a notion which needs to be prospectively studied. The finding of a close association of ferritin levels with markers of inflammation is in a line with ferritin as an acute phase protein which is induced by several cytokines such as interferon-gamma or IL-6, two major drivers of macrophage activation and lung injury [22]. However, in contrast to other studies, higher ferritin levels were not associated with an increased risk of death [6][7][8]. The association of low transferrin levels with a more severe clinical course also warrants further investigation in order to see whether they may contribute to a more severe course of the disease by preventing detoxification of pro-oxidative free iron in the circulation, thereby contributing to vascular and/or lung tissue injury [27][28][29] or by the inability to limit the access of SARS-CoV-2 to its nutrient iron [30]. Alterations of iron homeostasis may contribute to the pathogenesis of severe COVID-19 infection by mechanisms which need to be disentangled by future research. This also harbors the potential for therapeutic intervention by treatments which modulate iron availability such as application of transferrin, hepcidin-modifying agents or iron chelation. If increased levels of non-transferrin bound iron are found to be associated with tissue damage and a worse outcome, either iron chelators, transferrin or hepcidin agonist may help to reduce iron catalyzed radical formation and cellular damage in organs such as the lung or the vasculature. Furthermore, once the impact of alterations of iron availability on adaptive and innate immune functions as well as on viral pathogenicity and replication have been clarified, therapeutic spatiotemporal alterations of iron availability for either the virus or immune cells may favorably affect the course of the infection and associated pathologic inflammation [24][25][26][27]31]. Nonetheless, both anemia and inflammation-induced disturbances of iron homeostasis are important clinical predictors for risk stratification of SARS-CoV-2-infected patients, thereby improving the clinical management specifically of those at the highest risk [32].

Limitations
This was a retrospective observational analysis of COVID-19 patients with the need for in-hospital treatment in the region of Tyrol. As it is in the nature of a retrospective analysis, these results do not prove any causality. Additionally, we did not have iron metabolism variables available for all patients included in the study, depicting a probable selection bias and resulting in a smaller sample size in the multivariate analysis. Hemoglobin as a continuous variable was not a significant predictor in multivariate regression analysis, depicting the possibility of residual confounding for anemia classification. Finally, regional economic/social conditions must be taken into account when interpreting the results, which is why the findings do not allow unrestricted generalization for all COVID-19 patients.

Conclusions
The results of our study indicate that anemia, and specifically AI, is prevalent in patients with severe COVID-19 disease and is associated with an unfavorable outcome. We also introduced a ferritin/transferrin ratio which seems to be a robust and easily available marker for risk stratification at hospital admission of patients with SARS-CoV-2 infection. A higher ferritin/transferrin ratio reflects more advanced inflammation and predicts subsequent insufficient pulmonary oxygenation with the need for ICU admission and mechanical ventilation. Further studies are needed to investigate probable (direct) effects of changed iron homeostasis on the pathogenesis and severity of COVID-19 infections, as well as the potential for therapeutic interventions by modulating iron availability.