SARS-CoV-2 infection and the resulting COVID-19 pandemic have had profound effects on global morbidity and mortality, leading to widespread illness, long-term health complications, and millions of deaths, while placing unprecedented strain on healthcare systems and societies worldwide. The intensive care units faced extraordinary challenges. The emergence and spread of the Delta variant (SARS-CoV-2 virus Delta B.1.617.2) resulted in a surge of infections and record numbers of hospitalizations, particularly among the unvaccinated population. Approximately 5–7% of patients with COVID-19 caused by the Delta variant have developed critical lung injury and acute respiratory distress syndrome (ARDS). In severe ARDS, when mechanical ventilation is unable to maintain adequate oxygenation and/or CO₂ elimination, venovenous extracorporeal membrane oxygenation (V-V ECMO) may be considered to support gas exchange and minimize ventilator-induced lung injury (1, 2). At the peak of the pandemic, V-V ECMO support was implemented by tertiary cardiovascular centers to treat critically ill COVID-19 patients (3).

The determinants of in-hospital mortality and the selection criteria were reported in our previous study (4). However, limited data are available on the long-term outcomes after discharge from the hospital following ECMO support (5–8). During our therapeutic efforts, the most frequently raised question was whether survivors who had received ECMO support would regain an acceptable quality of life. During the follow-up period, the most common clinical complaints of our patients were tachycardia, effort dyspnea, fatigue and reduced physical capacity. We hypothesized that these symptoms could result from organ damage of cardiac or pulmonary origin caused by severe COVID-19. Our secondary hypothesis was that survivors of critical illness remain at increased risk of morbidity even after hospital discharge. These long-term consequences, widely referred to as post-intensive care syndrome (PICS), include mental, physical/organic, and neurocognitive deficits that negatively affect recovery, rehabilitation, and further quality of life (9). The aim of our present study was to investigate the long-term outcome of patients receiving ECMO support for severe COVID-19 disease-causing ARDS. We investigated in detail the cardiac, pulmonary, immunological, and mental status and quality of life of the patients during a follow-up period of at least 24 months.

We have found no significant organ damage behind the clinical symptoms frequently reported by our patients. In the recent literature, several studies have been conducted with the objective of evaluating the physical and mental status of patients who are affected by the acute phase of SARS-CoV-2 infection. Impaired pulmonary function and pulmonary-related symptom burden are frequently associated with postacute sequelae of coronavirus disease 2019 (PASC), which many patients with this disease experience after acute infection (23). This may have an impact on patients’ quality of life and general well-being. Pulmonary functionality can be evaluated objectively through PFTs, diffusion measurements, and 6MWTs. In accordance with the findings of other studies (24, 25), the PFTs of our patient population were almost normal, with only slight reductions in results and a restrictive pattern. It has been previously reported that CO diffusion capacity is impaired in patients with underlying residual fibrotic-like changes in the lungs, with some studies indicating that CO diffusion capacity does not return to normal at follow-up visits (24, 26, 27). In our study, patients who were able to undergo longitudinal measurements showed a notable steady improvement in DLCO and KLCO. DLCO (diffusion capacity) and 6MWD or exercise testing are often more sensitive to gas exchange impairment. Although standard PFT parameters did not change significantly, analysis of DLCO/6MWD values (where available) may provide important additional information. Based on this consideration, we recommend regular DLCO recording in long-term follow-up, as DLCO decline has been previously associated with radiological residuals, even when spirometry appeared “normal.”

During the COVID-19 pandemic, pulmonary radiomorphological features on chest CTs have been thoroughly studied; therefore, a typical progression pattern was observed during the course of the disease (28, 29). Chest CT imaging and assessment of the lung parenchyma played crucial roles in the diagnosis and follow-up of our study cohort. Considering that the patients included in this study suffered from particularly severe COVID-19 pneumonia, the findings of the baseline scans performed on admission showed severe and widespread pneumonia with predominantly peripheral ground-glass opacities (GGOs) and patchy consolidation patterns, corresponding to the known signs of the early active phase of the infection. It should be noted that in some patient cases, different stages of the disease could be observed on the same scan even at the 24-month evaluation. As a sign of disease progression, GGOs and consolidation patterns sometimes coexisted with varying degrees of residual subpleural banding on the baseline scan. A trend toward significant regression was observed at the 6- and 24-month follow-up CT scans. Nonetheless, residual fibrotic lesions were present on the majority of scans. In two patients, residual GGOs persisted on follow-up examinations (30). The observed presence of residual lung changes aligns with international results on this topic. However, although our cohort of patients was considerably smaller than the populations studied in more extensive international investigations, the prevalence of residual fibrosis appears to be relatively greater in the patients evaluated in our study (31). This may be attributed to the extreme severity of the illness in our study population. Most localized fibrotic bands (focal scarring) are clearly visible on radiological images, but do not necessarily cause measurable changes in standard global respiratory function tests (spirometry, TLC). Compensatory mechanisms (average performance of other, normal parts of the lung) maintain overall lung function. This is CT–PFT discordance, which has been described in several follow-up studies, and no close correlation has been found between quantitative CT parameters and PFT results (32, 33).

Following the severe COVID infection dominantly GGO was seen over a longer time, consistent with the slow resolution of alveolar-level injury. In regions where the insult was likely more pronounced, limited scar formation can be observed as a residual structural change, although pulmonary function remains preserved. The general improvement of the lung parenchyma was associated with significant improvement in the functional status, partly attributable to recovery of the respiratory pump, including improved diaphragmatic strength. The clinical significance of morphological radiological residuals (subjective symptoms, quality of life, progression) requires further, longer-term follow-up.

Cardiovascular diseases are known risk factors for severe COVID-19, and SARS-CoV-2 infection can also cause cardiovascular complications (e.g., heart failure, myocardial infarction, thromboembolism, myocardial injury, arrhythmias) (34). Initial data revealed an alarmingly high rate of cardiac MR abnormalities suggestive of myocarditis even after mild COVID-19 (35), but this high rate was not confirmed by subsequent studies (36–38). In the group of patients with severe COVID-19 detected in the present study, no patient had an increase in hs-troponin T levels suggestive of myocarditis, and their left and right ventricular functions were good both during acute infection and during follow-up. On echocardiography, patients had mildly dilated right ventricles in the acute period as a consequence of severe pulmonary injury (39, 40). Cardiac MR examinations during follow-up did not reveal contrast enhancement or mapping abnormalities suggestive of myocardial damage. A comparison of the cardiac MR parameters of COVID-19 patients and healthy controls revealed that COVID-19 patients had lower ventricular volumes, which based on our test results can be explained by the physical inactivity of the patients after ECMO treatment.

Our immunological findings demonstrated that patients infected with SARS-CoV-2 infection, whose condition had become sufficiently critical to require ECMO support were not only able to mount an immune response, but that this response was also sustained over the long term. This finding is also supported by the fact that the effect of the humoral and cellular immune response in the ECMO group was more pronounced than that in the control group in the long term. In the control group, the humoral and cellular immune responses declined after 12 months, whereas in the ECMO group, cellular immunity was consistently elevated, which was only partly explained by the vaccination. The immunological background of this requires further research.

Age has a significant impact on the dynamics of innate and adaptive immune responses (41). In older individuals, immunosenescence leads to a reduced pool of naive T cells, accumulation of senescent or terminally differentiated T cell subsets, impaired antigen presentation, and diminished cytotoxic responses (42). Such changes may impair effective viral clearance and the ability to control inflammation. Simultaneously, inflammaging may prime the immune system to mount a dysregulated, over-exuberant inflammatory response (“cytokine storm”) upon infection or other insults (43). In contrast, younger, immunocompetent individuals may mount a more balanced but robust immune response, capable of producing a more explosive pro-inflammatory response under certain circumstances. For example, studies in COVID-19 patients have shown that age influences immune cell subset distributions and cytokine profiles: younger patients may have different patterns of T cell activation, exhaustion, and cytokine production compared to older patients (44). While many existing data emphasize that older age is a risk factor for severe disease, the immunological picture is complex, and in some younger patients, hyperinflammatory responses may also contribute to more severe tissue injury (45, 46). Several autopsy studies report that certain clinical conditions (e.g., severe COVID-19) younger patients show different, sometimes more pronounced organ/histological changes, which supports the idea that there may be age-related differences in tissue damage (47). Our study population was relatively young (mean age ∼34 years), which may have influenced both clinical presentation and long-term outcomes. One potential immunological explanation for our observations may be related to age-specific differences in immune response dynamics. In our cohort, the relatively young age might have allowed for a sufficiently strong and rapid immune response that cleared the insult efficiently, but at the same time induced localized tissue changes (e.g., fibrotic remodeling) through inflammatory processes, without resulting in sustained systemic immune exhaustion. Nevertheless, we acknowledge that mechanistic confirmation of this hypothesis requires further dedicated studies. While we primarily focused on clinical, radiological, and functional outcomes, future directions should include standardized immunological assays, such as flow cytometry for T and B lymphocyte subsets (naïve, memory, exhausted), activation and co-stimulatory markers, and serum cytokine profiling (e.g., IL-6, IL-1β, TNF-α, IL-10). These would allow correlation of immunophenotype with clinical and morphological outcomes and help to elucidate whether localized fibrotic changes relate more to direct viral injury or to immune-mediated damage.

Experiencing critical illness and undergoing ECMO therapy can impose a substantial psychological burden. A Canadian population-level cohort study examined the incidence of new mental health diagnoses among ECMO survivors compared to that among ICU survivors without ECMO treatment (48). The non-ECMO survivors had a 14.5 per 100 person-years incidence of new psychiatric diagnosis, while the ECMO survivors had 22.1 per year, which means that they are at increased risk (24%) of developing mental health problems. The most commonly reported mental health conditions (in both groups) were mood disorders, anxiety disorders and posttraumatic stress disorder (PTSD). A review in 2020 examined the cognitive, psychiatric and quality-of-life outcomes of ECMO treatment among survivors, with a median follow-up of 36 months (49). The most common psychiatric outcomes were depression, anxiety and PTSD. The prevalence of depression among the included studies reached 43%, the prevalence of anxiety ranged from 15 to 50%, and the prevalence of PTSD reached 41%. Regarding quality of life, the physical quality of life of ECMO survivors was worse than that of the standard population, although it seemed to improve over time. The mental quality of life did not differ significantly from the population norms. Knudson et al. (7) analyzed 31 studies examining the health-related quality of life and psychiatric symptoms of ECMO survivors. The authors found that physical quality of life was worse than mental quality of life, and a significant portion of the included patients reported anxiety, depression and PTSD symptoms (7). In our present study, we found significant associations between the duration of ECMO treatment and symptoms of depression, anxiety, and PTSD, indicating that longer periods of ECMO were linked to greater levels of psychological distress. This aligns with the findings of the previous studies discussed above. Overall, these findings suggest that ECMO survivors may face ongoing mental health challenges after treatment. This highlights the need for continued monitoring and support for ECMO patients, as well as interventions aimed at addressing their psychological needs. Additionally, the study underscores the importance of further research to better understand the long-term psychological effects of ECMO support.

Patients’ quality of life declined temporarily following ECMO support; however, by the second year of follow-up, most had returned to work, though with some limitations in their daily routine. Furthermore, no definite permanent organ or functional impairment was identified to explain the higher rates of subjective complaints (fatigue, palpitation, effort dyspnea). Our findings suggest that these symptoms are not directly related to the viral infection itself but rather represent sequelae of ICU treatment. This appears to be partly of psychological origin (depression, anxiety, posttraumatic stress), which led to lifestyle changes. During follow-up, the BMI of ECMO treated patients—already elevated at the onset of COVID-19 infection—increased significantly and by the end of the second year of follow-up, 88.9% were classified as morbidly obese. The poorer 6MWT results were primarily attributable to obesity, lack of stamina and physical inactivity rather than pulmonary or cardiac limitations. Recognizing this, we emphasize the importance of initiating exercise rehabilitation and psychological support early in the ICU, and maintaining them over the long term (50). At the same time, the relatively young age of our cohort (mean ∼34 years) may represent a potential source of selection bias, as younger individuals generally have fewer chronic comorbidities, better baseline cardiopulmonary reserve, and greater regenerative capacity. This may have contributed to the faster and more complete cardio-pulmonary recovery observed in our study compared with older or multimorbid populations. Larger studies across broader age ranges are required to confirm whether our findings apply more generally.

Our findings are subject to several limitations. First, limited sample size: this study was conducted at a single-center and involved a small sample size. The follow-up included only 9 ECMO-treated patients, which limited statistical power and generalizability. Second, lack of pretreatment baseline conditions: baseline (pre-treatment) data were unavailable, preventing us from attributing all observed changes solely to COVID-19 or ECMO treatment. Third, due to the heavy burden of the COVID-19 pandemic on the healthcare system, some data were missing at discharge (e.g., 6-month pulmonary function test results). Fourth, this cohort represents an unique group of ECMO survivors, making it impossible to select one appropriately matched control group; therefore, separate control groups were used for cardiological and immunological comparisons. A further limitation of our study is the relatively young age of the cohort (mean ∼34 years). This demographic may not reflect the typical patient population with chronic comorbidities or age-related immunological changes (immunosenescence, inflammaging), and therefore our findings may not be directly generalizable to older or multimorbid populations. Moreover, we did not perform komplex immunophenotyping (e.g., T-cell subsets, cytokine panels) which limits our ability to draw mechanistic conclusions regarding the role of age-dependent immune responses.

In conclusion, all our patients who survived the acute phase of COVID-19 disease with ARDS requiring ECMO treatment survived at least 2 years of follow-up. Remarkably, at the end of this period, all our patients still demonstrated a sustained and robust immune response. ECMO support in patients with ARDS from appropriately selected severe COVID-19 disease may result in favorable functional outcomes. Two years after recovery from infection we observed no lasting heart or lung damage. However, a significant proportion of patients still have psychological and musculoskeletal problems that affect their quality of life. Some of these problems are the result of COVID-19 disease and the associated ARDS, and some resulting of long ICU stays.