We conducted a single-centre comparative observational study involving 490 hospitalized elderly and non-elderly adult patients with COVID-19 infection associated with CSS, who received baricitinib in addition to the standard-of-care treatment. A higher prevalence of female gender, essential hypertension, chronic cardiovascular disease, chronic renal disease, chronic cerebrovascular disease, diabetes mellitus, and active oncological malignancy was observed among elderly patients. A substantial 76.33% of the total cohort had not received vaccination at the time of inclusion. At baseline, elderly patients experienced more severe symptoms and required oxygen supportation with Venturi masks or non-invasive ventilation, with extended infiltrations on chest CT scans. Furthermore, a significantly lower absolute lymphocyte count, and elevated levels of interleukin-6 and CRP were recorded among this group. Regarding the time from hospitalization to baricitinib initiation and ICU admission rates, we did not detect any differences between subgroups. However, a longer hospitalization period was observed among the elderly. Furthermore, we observed a 19.80% severe secondary infection rate and a 10.82% non-severe secondary infection rate in the total cohort during the 1-year follow-up period. During the 90-day and 1-year follow-up, severe secondary infections were significantly more prevalent among elderly patients. The 30-day, 90-day, and 1-year all-cause mortality rates were also significantly higher among the elderly patients compared to the non-elderly. In conclusion, elderly patients demonstrated a propensity for long-term morbidity and mortality, irrespective of baricitinib treatment.

Age and specific comorbidities are well-established risk factors for COVID-19 disease progression [8]. Differences in morbidity and mortality between elderly and non-elderly COVID-19-infected patients arise from a complex interplay of age-related mechanisms [13]. First, the pathological alteration of the immune system during aging, defined as immunosenescence, affects both innate and adaptive immunity, which contribute to chronic illnesses and a reduced capacity to effectively combat infections [14, 15]. An important aspect of immunosenescence is “inflamm-aging”, characterized by the accumulation of proinflammatory cytokines due to damaged macromolecules or cells, resulting in a low-level sterile chronic inflammation [15]. Immune cells adapt to continuous signalling, leading to a decline in their function with age. Aging also influences haematopoietic stem cells, causing a shift towards myeloid cell production, resulting in lymphopenia, a well-established prognostic factor for poorer COVID-19 outcomes [14]. With increased myeloid cell production, their function and activation rate are enhanced, altering the proinflammatory cytokine profile [15]. Upregulation of interleukin-6 and downregulation of interferon-γ further also inhibit the cellular lymphocyte response [15]. In our study, lymphopenia was significantly higher among elderly patients, reflecting the suggested cellular and subcellular mechanisms in a real-world cohort.

Furthermore, in the course of aging, there is an augmentation of cellular senescence and the senescence-associated secretory phenotype [16]. The latter compromises cytokine production by senescent cells, thereby inducing extracellular matrix degradation, a pro-inflammatory state, pro-coagulation, and an augmented complement activation [16]. The secretory phenotype disseminates in a paracrine manner among cells, thereby hastening the aforementioned processes [16]. In addition to the aging processes, it is noteworthy that SARS-CoV-2 has the capacity to induce senescence and the proinflammatory phenotype [17]. In the geriatric population, both pre-existing and virus-induced senescence concomitantly contribute to and exacerbate proinflammatory cytokine production. This, in turn, contributes to the manifestation of cytokine storm syndrome, thrombosis, and tissue injury-induced fibrosis. Consequently, this cascade leads to multiple organ failure and an elevated mortality rate among the elderly [17].

Current literature also emphasizes the role of gastrointestinal tract dysbiosis in increased morbidity and mortality among the elderly [18]. The aging process of the gastrointestinal tract is characterized by alterations in the intestinal microbiota, with an abundance of Bacteroidetes and a decrease in anti-inflammatory elements, contributing to a proinflammatory environment similar to “inflamm-aging” and a weakened intestinal barrier [19]. Additionally, microbiome studies also suggest a SARS-CoV-2-induced dysbiosis and a colonization of opportunistic pathogens in the gastrointestinal tract, facilitating secondary infections [18].

Besides the cellular and molecular mechanisms, frailty is also an age-related condition marked by bedridden states, dietary deficiencies, and polypharmacy, and is associated with higher rates of hospitalization and mortality, regardless of COVID-19 [13, 20]. At the physiological level, it can be defined as a decline in organ function and reserves, which, at a systemic level, leads to a reduction in daily activities. Beyond the age of 65, frailty affects approximately 15% of the elderly population, with a higher prevalence among female patients [13]. Geriatric hospitals, nursing homes, and long-term care facilities where frail elderly patients accumulate have been recognized as epidemiological hotspots for COVID-19 transmission, exposing them to a higher overall risk of contracting the disease [21].

Among the elderly, decreased drug bioavailability, lower levels of transport proteins due to malnutrition, hypoperfusion in peripheral blood, alterations in body fat and water balance, drug-drug interactions, and decreased liver or kidney function are expected regardless of SARS-CoV-2 infection [22]. Additionally, baricitinib is indicated for the treatment of COVID-19 infection with a severe clinical course and an associated cytokine storm [23]. This comprises an uncontrolled and hyperresponsive systemic inflammatory milieu, with an excess of proinflammatory cytokines resulting in multiorgan failure through vascular and immune-mediated pathways [23]. Therefore, baricitinib pharmacokinetics modulate both due to age-related and immune-mediated processes [22, 23]. Dose modification or the termination of the drug is indicated in cases of severely decreased kidney function (eGFR < 15 ml/min/1.73m2), suspicion of drug-induced liver injury (aspartate aminotransferase or alanine aminotransferase levels at least five times above the ULN, with a suspected drug-induced liver injury), severe lymphopenia (absolute lymphocyte count < 200 cells/µl), and severe neutropenia (absolute neutrophil count < 500 cells/µl) [3]. However, in our study, there were no cases that required baricitinib dose modification or necessitated drug termination due to a severe adverse event.

Previous studies in the literature have suggested that infected and hospitalized elderly patients experience significantly higher rates of secondary infections. In a single-centre observational cohort study that assessed pulmonary secondary infection rates and clinical outcomes among older adults testing positive for SARS-CoV-2, approximately 43% were diagnosed with suspected superimposed infections through positive bacterial cultures or radiological findings. The study reported increased mortality rates and extended hospitalizations specifically among elderly patients with secondary infections [9]. In comparison to our study, this report documented higher secondary infection rates, potentially attributable to the fact that included patients were likely to be bedridden, heightening the risk of pulmonary infections. An Indian retrospective observational study focused on the mortality rates of elderly patients hospitalized for COVID-19. Their findings indicated notably higher mortality rates among patients with at least three comorbidities, severe COVID-19 disease courses, and those experiencing multi-organ failure, elevated creatinine, or acute liver injury complications [24]. Similarly, a case–control observational study including 179 elderly patients hospitalized for non-COVID-19-related diseases found that 89.9% of patients contracted the disease in geriatric hospital settings. The study identified a mortality rate of 14.4% among elderly patients with a negative SARS-CoV-2 RT-PCR test, and a 29.2% mortality rate among COVID-19 patients [25]. Similar mortality rates were observed during our 1-year follow-up period among elderly patients. Furthermore, a retrospective observational study compared previously SARS-CoV-2-infected and hospitalized elderly patients with historical controls suffering from seasonal influenza between March 2020 and August 2022. The study reported a heightened long-term risk of hospital readmission or all-cause death at 30 days, 90 days, and 180 days post-discharge. The authors also noted a significant decline in all-cause mortality over the study period during the pandemic [26]. At the 1-year follow-up, we were also able to observe higher rates of hospital readmission due to severe secondary infections and mortality among the elderly.

Baricitinib received its initial approval in November 2020, demonstrating its superior efficacy in combination with remdesivir over remdesivir monotherapy among hospitalized patients receiving high-flow oxygen or non-invasive ventilation [27, 28]. Historically, in rheumatoid arthritis therapy with 2 or 4 mg baricitinib doses, there were elevated incidences of mild to moderate upper respiratory tract infections, urinary tract infections, gastroenteritis, and reactivation of herpes simplex virus [29]. Additionally, severe infectious complications, such as herpes zoster or cellulitis, were reported [29]. Subsequently, early international clinical data indicated heightened rates of secondary infections also among hospitalized SARS-CoV-2-infected patients undergoing baricitinib therapy [3, 4]. Consequently, the Hungarian National Guideline recommended prophylactic antimicrobial measures upon completing baricitinib therapy [5]. However, novel real-world clinical data did not suggest significantly elevated secondary infection rates [6, 7, 28]. A systematic review from China assessed the safety and efficacy of baricitinib in hospitalized SARS-CoV-2-infected patients, comparing outcomes with those receiving a placebo or alternative treatments. This investigation revealed reduced rates of secondary infections in the baricitinib treatment arm [28]. Moreover, a retrospective cohort study compared secondary infection rates and 28-day mortality rates among patients receiving baricitinib plus remdesivir and dexamethasone (n = 43) and patients receiving only remdesivir and dexamethasone (n = 53) [7]. Baricitinib therapy was 2 or 4 mg depending on the decision of the physicians for a maximum of 14 days [7]. According to the study result, there were no significant differences among the two study arms in the secondary infection rate or 28-day mortality rate [7]. Also, a retrospective cohort study among transplant patients (89% solid organ recipients) compared secondary infection rates between baricitinib plus SOC-treated (n = 77) and only SOC-treated (n = 114) patients [6]. Similarly, secondary infection rates did not differ among the two subgroups [6]. Fundamentally, the aforementioned evidence suggests that the variance in secondary infection rates between patients treated for rheumatoid arthritis or severe COVID-19 infection may be attributed to the duration of treatment.

Previous studies have also investigated the impact of the immunomodulatory baricitinib in COVID-19-associated CSS among the elderly. A retrospective cohort study using propensity score matching evaluated elderly patients, defined as 70 years or older, and non-elderly patients with moderate-to-severe COVID-19 pneumonitis who received baricitinib compared to those who did not. The study observed an overall 48% reduction in all-cause mortality rates and an 18.5% absolute mortality risk reduction at 30 days among elderly patients who received immunomodulatory treatment [30]. Our study differs from this research in terms of study design. Our definitions are based on different age thresholds, and we compared elderly populations to non-elderly ones, both of which received baricitinib. In contrast, the present study involved a control population that did not receive immunomodulation beyond SOC.

In conclusion, our findings align with current literature data suggesting that elderly patients experience more severe microbiological and clinical outcomes, regardless of immunomodulatory therapy.

Our study had limitations. Firstly, the evolution of general knowledge influenced by emerging COVID-19 evidence led to adjustments in treatment protocols. Despite dedication, occasional delays in protocol updates might have occurred due to treatment availability issues. Secondly, our retrospective study included a long patient inclusion period, so multiple SARS-CoV-2 variants were included in the study, and routine variant typing is not available at our centre. Moreover, comorbidities were more frequent among elderly patients. Age and certain comorbidities are both risk factors for COVID-19 disease progression; thus, higher mortality rates were inherently expected among the elderly group [8]. Furthermore, the male gender was more common in the non-elderly subgroup, potentially influencing the insignificant results regarding the need for mechanical ventilation or admission to the intensive care unit. Finally, residual confounding might have biased our results during the 1-year follow-up period.

In our study, we found that elderly patients exhibited a tendency towards enduring morbidity and mortality, independent of baricitinib treatment. Certain national guidelines, such as the one in Hungary, recommended prophylactic antimicrobial measures and suggested a 3- to 6-month regimen of co-trimoxazole and acyclovir following the completion of baricitinib [5]. For future research, a redefinition of prophylactic requirements among both elderly and non-elderly patient subpopulations should probably be an intriguing question.