Clinical Outcomes in COVID-19 Patients Treated with Immunotherapy

Simple Summary Patients with cancer who contract COVID-19 are very vulnerable to increased complications and illness while actively being treated with chemotherapy or immune checkpoint inhibitors (ICIs). The aim of this retrospective review was to describe the disease course and identify specific risk factors and overall outcomes in COVID-19-affected patients who are also diagnosed with cancer. We examined whether treatment (chemotherapy vs. ICIs) was associated with clinical outcomes, including hospitalization rates, ICU admissions, and any-cause mortality. A total of 121 patients were examined in this study, and 61 (50.4%) received immunotherapy treatment within 12 months. COVID-19-related ICI mortality was higher compared to patients receiving chemotherapy, but patients with better functional status and COVID-19 vaccination had reduced mortality. ICI cessation or delay is unwarranted as long there has been a risk–benefit assessment undertaken with the patient. However, further investigation still needs to be undertaken with a larger cohort, with an emphasis on timing and outcomes between ICI therapy and COVID-19 infection. Abstract Introduction: The full impact of COVID-19 infections on patients with cancer who are actively being treated with chemotherapy or immune checkpoint inhibitors (ICIs) has not been fully defined. Our goal was to track clinical outcomes in this specific patient population. Methods: We performed a retrospective chart review of 121 patients (age > 18 years) at the University of Alabama at Birmingham from January 2020 to December 2021 with an advanced solid malignancy that were eligible to be treated with ICIs or on current therapy within 12 months of their COVID-19 diagnosis. Results: A total of 121 patients were examined in this study, and 61 (50.4%) received immunotherapy treatment within 12 months. One quarter of the patients on ICIs passed away, compared to 13% of the post-chemotherapy cohort. Patients who were vaccinated for COVID-19 had lower mortality compared to unvaccinated patients (X2 = 15.19, p < 0.001), and patients with lower ECOG (0.98) were associated with lower mortality compared to patients with worse functional status (0.98 vs. 1.52; t = 3.20; p < 0.01). Conclusions: COVID-19-related ICI mortality was higher compared to patients receiving chemotherapy. However, ICI cessation or delay is unwarranted as long there has been a risk–benefit assessment undertaken with the patient.


Introduction
In December 2019, coronavirus disease 2019 , caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), emerged in China and subsequently led to a global pandemic in early 2020 [1]. As of November 2022, the COVID-19 pandemic ing protection against the development of severe COVID-19 by preventing lymphocyte exhaustion [43].

Figure 1. Mechanism of action of PD-1/PD-L1 and CTLA4 inhibitors.
Antigen presentation occurs on the surfaces of antigen-presenting cells (APCs) via T-cell receptors (TCR) and major histocompatibility complex (MHC). There are stimulatory and inhibitory signals, such as cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and programmed cell death-1 (PD-1). Some of the upregulators of T-cells and their cognate ligands are CD27-CD70 and CD28-B7. Inhibitory T-cell receptors and their ligands include LAG3-MHC, CTLA4-B7 and PD1-PDL1. CTLA-4 reduces signaling via its co-stimulatory receptor, CD28, by binding to CD80 and CD86 on APC. CTLA-4 sends an inhibitory signal to T-cells. A CTLA-4 inhibitor (ipilimumab) stops autoreactive T-cells. PD-1 binds to programmed death-ligand 1 (PD-L1), which results in cancer evasion from the immune system. Blockade of PD-1 (via pembrolizumab, nivolumab, cemiplimab or dostarlimab) or PD-L1 (via atezolizumab, durvalumab or avelumab) increases anti-tumor immune activity. The image used has been modified from a prior submission [11]. The potential long-term benefits or harmful effects of ICIs in patients with cancer and COVID-19 infection are not well known and still must be investigated. It is vital to understand the full scope and impact of ICI treatment in COVID-19 patients, to provide better quality of care. Thus, it is the aim of our retrospective chart review to describe the disease course and identify specific risk factors and overall outcomes in a very diverse patient population with an underlying malignancy and COVID-19 infection who received IMT or chemotherapy.

Study Design and Population
We performed a retrospective chart review of 121 patients (age > 18 years) with a solid malignancy at the University of Alabama at Birmingham (UAB) between January 2020 and November 2021. Any patients with an established stage 3 or 4 solid malignancy diagnosis, apart from breast, prostate cancer and gastrointestinal stromal tumors (GIST), who tested positive for SARS-CoV-2 through nasopharyngeal swab nucleic acid amplification or serological testing, were included in this study. Asymptomatic individuals who did not require hospitalization were also included. Of the 121 patients in this review, 61 received IMT treatment within 12 months. Treatment response was assessed per iRECIST criteria [44]. Clinical and laboratory data were obtained in full accordance with UAB's Institutional Review Board (IRB) and the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule. The declassified data collected were stored on a UAB encrypted and password protected share-drive. UAB encrypted and password protected share-drive Any deviation from the protocol was documented and reported accordingly.
Laboratory blood work in the study included baseline white blood cells (WBC), WBC during initial COVID-19 diagnosis, lymphocyte count during initial COVID-19 diagnosis, c-reactive protein (CRP), creatinine, estimated glomerular filtration rate (eGFR), aspartate transaminase (AST), alanine transaminase (ALT), lactose dehydrogenase (LDH) and ddimer. Initial imaging was also documented with the most common finding (presence and location of opacities). Oxygen saturation by pulse oximeter (SpO2) was measured at admission. Admission (COVID-19 floor or intensive care unit (ICU)) and treatment (oxygen therapy, mechanical ventilation, use of vasopressors, continuous renal replacement therapy (CRRT), antiviral therapy, antibiotics and glucocorticoids) of COVID-19 were conducted according to UAB's institutional protocol and at the treating provider's discretion.
Clinical outcomes measured included oxygen requirement, hospitalization rate, ICU admission, use of mechanical ventilation, death of any cause and overall survival (OS). OS or "days from date of pathological diagnosis" was defined as the period from the date of diagnosis until the last office visit. "Days from end of treatment" measured the time from COVID-19 diagnosis until the last office visit.

Statistical Analysis
Frequency and proportion calculations were used to describe patient demographics and overall characteristics. Chi-square tests were utilized to assess differences among the groups for categorical variables, and t-tests analyzed the differences between groups for continuous measures (Tables 1-5).

Baseline Characteristics
Sample descriptive statistics can be found in Table 1. A total of 121 patients were examined in this study and 61 received IMT treatment within 12 months. Fifty-six percent of patients received pembrolizumab, 15.0% nivolumab, 18.0% atezolizumab, 7.0% durvalumab, 1.0% ipilimumab and 3.0% ipilimumab/nivolumab. The three most common irAEs in decreasing order were endocrine (6 cases; 4 adrenal insufficiency, 1 hypophysitis and 1 hypothyroid), gastrointestinal (5 cases; 5 colitis) and skin (4 cases; 3 rash and 1 psoriasis). The median age at diagnosis for IMT patients was 62 years, 62.0% were male, and 66.0% of the study population were white. For the 60 patients on chemotherapy, the median age at diagnosis was 65 years and 53.0% were male. Seventy-three percent of the study population were white and 62.0% were former smokers (>12 months), 25.0% never smoked and 13.0% were current smokers in the IMT group (Table S3). Meanwhile, in the chemotherapy cohort, 48.0% were former smokers (>12 months), 32.0% were never smokers and 20.0% were current smokers ( Table 2). Of the 61 patients treated with IMT, the most common cancer was lung (33.0%), followed by hepatocellular carcinoma (HCC, 13.0%) and renal cell carcinoma (RCC, 11.0%, Table 1 and Table S1). Hypertension (49.0%), diabetes mellitus (26.0%), hyperlipidemia (21.0%) and secondary malignancies (18.0%) were the most common comorbidities in decreasing order. Within the 60 patients on chemotherapy, the most common cancers were lung (33.0%), HCC (12.0%) and head and neck cancers (10.0%). The most common comorbidities in the same patients, in decreasing order, were hypertension (53.0%), diabetes mellitus (28.0%), hyperlipidemia (23.0%) and COPD (18.0%). The PD-L1 expression in the IMT group was 15, compared to 11 in the chemotherapy group. Moreover, the ECOG was 1.0, versus 1.1 in the IMT and chemotherapy groups, respectively. Additionally, 56.0% of the IMT cohort received a COVID-19 vaccine (8 people received a vaccine before their diagnosis), compared to 62.0% in the chemotherapy group (9 people received a vaccine before their diagnosis; Table 3 and Table S4).

Risk Factors
There were several risk factors associated with hospitalization. Patients who identified as white were at a significantly increased risk for overall hospitalization in lower acuity units compared to black patients (62.9% vs. 37.1%, X 2 = 13.3, p < 0.01; Table 1). Higher admissions were also seen in patients with increased frailty, as evidenced by the higher ECOG (1.1 vs. 1.7, 1.0; t = 2.4; p = 0.05). Specific cancer subtypes excluding NSCLC, liver, RCC and H&N were associated with increased admission (X 2 = 13.5, p = 0.03). Several specific laboratory findings were associated with increased admission. Patients admitted to lower acuity units had a mean AST value of 40.9, compared to 99.1 for patients in critical care units (t = 2.8; p < 0.001). Patients admitted to the ICU were more likely to be unvaccinated compared to those admitted to the lower acuity units (1.4% vs. 15.7%; X 2 = 10.2, p < 0.001). Oxygen supplementation was also higher in critically ill patients in the ICU (88.9% vs. 27.7, X 2 = 13.5; p < 0.001) compared to the lower acuity units. Treatmentspecific risk factors associated with admission to the ICU compared to a regular floor unit included antibiotics (77.8% vs. 30.0, X 2 = 8.0, p < 0.001), antivirals (66.7% vs. 17.1, X 2 = 11.1, p < 0.001) and steroids (77.8% vs. 34.3, X 2 = 6.3, p < 0.01). Smoking status was not associated with treatment outcomes or hospitalization.

Survival Outcomes
Survival outcomes differed between the ICI cohort and chemotherapy cohort (Table 4). In the ICI cohort, 15 patients died, but only five deaths were directly related to COVID-19 ( Figure S1). In comparison, eight patients in the non-immunotherapy group died, and four deaths were directly attributed to COVID-19 ( Figure S1). The OS in the ICI cohort was 780 days, compared to 706.0 days in the chemotherapy cohort (t = −0.6, p = 0.6). The average time since COVID-19 diagnosis in non-smokers was 159 daysversus 220.0 days smokers. There was an association between reduced mortality in patients with a lower ECOG (<1) versus those patients that were deceased (1.5; t = 3.2; p < 0.01). Other factors associated with worse mortality were higher oxygen needs (X 2 = 13.5, p < 0.001), being unvaccinated for COVID-19 (X 2 = 15.2, p < 0.001), the use of antiviral therapy (X 2 = 5.4, p < 0.05) and the use of steroids (X 2 = 4.3, p < 0.1). No mortality differences were observed according to smoking history, types of comorbidities or cancer origin, including "hot" vs. "cold" tumors (Table S2).

Discussion
Patients with cancer who were diagnosed with COVID-19 and were treated with ICIs had increased mortality compared to the same type of patients treated with chemotherapy. However, in the ICI cohort, the COVID-19-associated mortality accounted for one third of the cases, while, in the chemotherapy group, it accounted for one half of the cases. The overall mortality was 19% in our patient cohort, which is much higher than the case fatality rate of 3% in the general community [45]. In several other studies of patients with cancer and COVID-19, the 30-day mortality was reported as an average range of 10-40%, which is similar to our results [46][47][48].
A systemic review and meta-analysis of COVID-19 patients with cancer who were treated with chemotherapy found that there was an increased risk of death (OR: 1.9; 95% confidence interval: 1.3-2.7), after adjusting for confounding variables, compared to patients that received targeted therapies, IMT, surgery or radiotherapy [49]. Another meta-analysis of 6042 patients with COVID-19 showed that ICIs received within 30 days before diagnosis did not increase mortality or the severity of disease in affected patients [50]. In our review, there was one study that found a tendency of an increasing risk of mortality and severity of critical disease in COVID-19 patients treated with ICIs [51]. This could be related to the results being unadjusted and confounded by several variables (e.g., age, ECOG, comorbidities, smoking, presence of metastasis) in patients diagnosed with COVID-19. It should be noted that smoking increases the risk for intubation and mortality, but, due to our smaller sample size, this was not seen in our review [52].
We found that there were several risk factors associated with increased mortality. The factors that lead to worse disease severity and thus reduced OS were increased oxygen needs, being unvaccinated for COVID-19 and the use of antiviral therapy and steroids. However, one very important risk factor that led to worse mortality was an ECOG >1. It is known that a decline in functional status in patients with malignancies can result in an adverse outcome, but an ECOG of 2 is not a contraindication to treatment [53]. Patients with COVID-19 and an ECOG of 2 might require the postponement of therapy until there is an improvement in functional status, as it could potentially worsen outcomes.
One very important question that providers face is related to the benefit of postponing or temporarily discontinuing ICI therapy in patients with a solid malignancy. It is known that multiple office visits for ICI infusions raises the possibility of COVID-19 exposure and possible infection [54]. However, treatment delays in advanced malignancies can also be very detrimental, especially when considering the current improvements that have been made in treating COVID-19 infections. Another important consideration of ICI therapy in advanced or metastatic disease is the possibility of irAEs such as ICI pneumonitis, which can be masked by COVID-19 symptoms. In particular, 39-54% of patients with COVID-19 are symptomatic and require hospitalization [55,56]. In our own study, 53% of COVID-19affected patients treated with ICIs required admission to either the hospital or ICU. The most common symptom was dyspnea, which can also be seen in ICI pneumonitis, which accounts for a third of cancer-related deaths [57]. Nevertheless, this would appear to be less of a possibility in patients with COVID-19 as the standard-of-care treatment includes steroids, which are also used to treat ICI pneumonitis [58]. This retrospective chart review was able to provide valuable insights, but it also had several limitations. One specific constraint was related to the lack of a comparator group for outcomes (hospitalizations and mortality) in patients with an underlying malignancy who did not have COVID-19. Secondly, as with other retrospective chart reviews, selection bias can occur and it can be difficult to distinguish symptoms from COVID-19 and IMT pneumonitis. Detailed exploration of follow-up questions can be difficult to achieve due to limited information that was not obtained during the original chart abstraction. Additionally, with rarer malignancies, the full impact of COVID-19 will be more challenging to identify unless multicenter, long-term studies are conducted so that subgroup analysis can be performed in a more meaningful manner, with a larger pooled cohort (e.g., antibody/IL-6 therapy in thyroid cancer). Our study also had several strengths. One very important aspect of this retrospective chart review was the diverse patient population, with at least 25% of study cohort identifying as black, and there was a large number of ICI-treated patients for a single institutional study. We also examined several important variables, such as smoking history, COVID-19 vaccination status and irAEs.

Conclusions/Future Perspectives
ICI mortality was higher compared to patients receiving chemotherapy post-COVID-19. Risk factors for hospitalization and increased disease severity/mortality for COVID-19affected patients have been identified to aid in future risk stratification. There are still many unknown factors regarding COVID-19 survivors who are treated for an underlying malignancy. "Long COVID" refers to symptoms of COVID-19 that can last for months after the initial diagnosis and can have multi-organ effects. The specific triggers and management in patients with cancer are unknown and should be studied prospectively in larger cohorts.
In our opinion, ICI or chemotherapy cessation or delay is unwarranted if there is a risk-benefit assessment with the patient and there is an understanding that disease progression can be more detrimental. However, further investigation still needs to be undertaken to understand whether the PD-L1 pathway with the subsequent inflammatory cascade post-COVID-19 can impact overall survival and whether the timing between ICI and COVID-19 infection affects long-term symptoms/outcomes.

Data Availability Statement:
The de-identified patient data used in this study can be requested from Dr. Jessy Deshane at the University of Alabama at Birmingham.

Acknowledgments:
We would like to acknowledge and thank Jenny B Jung and Russell Griffin for their assistance with the data collection.

Conflicts of Interest:
The authors declare no conflict of interest.