What drives COVID-19 mortality in cancer?

Roshini Claire Anthony
06 Aug 2020

Among patients with cancer who develop COVID-19, progressive disease and a higher ECOG* performance status (PS) are tied to an increased risk of mortality, according to early results ascertained from the COVID-19 and Cancer Consortium (CCC19) registry.

“Older age, poor performance status, and progressing cancer were strongly associated with increased mortality, especially in subsets admitted to the intensive care unit (ICU) and/or intubated,” said lead author Associate Professor Jeremy L. Warner from Vanderbilt University Medical Center, Nashville, Tennessee, US, who presented the findings at ASCO20.

The study cohort comprised 928 adults (median age 66 years [30 percent aged 75 years], 50 percent male) with prior or active haematologic or invasive solid malignancies from 106 participating institutions (primarily in North America) in the first 30 days of the registry initiation (March 17–April 16, 2020). The most common malignancies were breast and prostate cancer (21 and 16 percent, respectively). Forty-three percent of patients had active cancer, 39 percent were on treatment, and 45 percent in remission.

After a median 21 days of follow-up, 121 patients had died (13 percent). Fifty percent of patients (n=466) required hospitalization upon diagnosis of COVID-19, 23 percent of whom died. Fourteen percent of patients required ICU admission (38 percent died), 12 percent required mechanical ventilation (43 percent died), and 44 percent required supplemental oxygen. The composite outcome of death, severe illness requiring hospitalization, ICU admission, and mechanical ventilation occurred in 26 percent of the patients (n=242). Among patients admitted to the ICU or who required mechanical ventilation, mortality risk was elevated in those aged 75 years or with ECOG PS of 2.

Factors predicting all-cause mortality at 30 days included age (partially adjusted** odds ratio [padjOR], 1.84 for every decade increase, 95 percent confidence interval [CI], 1.53–2.21), male sex (padjOR, 1.63, 95 percent CI, 1.07–2.48 vs female), former smoking (padjOR, 1.60, 95 percent CI, 1.03–2.47 vs never smokers), and number of comorbidities (padjOR, 4.50, 05 percent CI, 1.33–15.28 for 2 vs 0). [ASCO20, abstract LBA110; Lancet 2020;395:1907-1918]

Cancer-related factors predicting mortality included ECOG PS (2 vs 0/1: padjOR, 3.89, 95 percent CI, 2.11–7.18) and active malignancy vs remission/no disease (stable or responding disease: padjOR, 1.79, 95 percent CI, 1.09–2.95; progressing disease: padjOR, 5.20, 95 percent CI, 2.77–9.77).

Patients who received the combined therapy of azithromycin and hydroxychloroquine were also at an increased mortality risk compared with use of neither drug (padjOR, 2.93, 95 percent CI, 1.79–4.79), though use of either drug alone had no bearing on mortality.

In a post hoc analysis of patients who received azithromycin and hydroxychloroquine as combined or individual treatments, factors that increased their risk of mortality at 30 days were ECOG PS 1 (univariate OR, 4.1), Rhesus-positive blood type (OR, 3.2), non-Hispanic ethnicity (OR, 3.1), receipt of cancer treatment <2 weeks prior to COVID-19 diagnosis (OR, 4.0; p<0.001 for all), and statin use at baseline (OR, 2.6; p=0.003).

“While these findings are provocative, we believe that there is significant confounding by indication [higher hospitalization risk or disease severity] and that carefully planned prospective studies are needed to truly demonstrate the risk or benefit of these drugs,” cautioned Warner. The drug combination may not have led to increased mortality, but rather, been prescribed to patients with more severe COVID-19 illness, the authors noted.

Conversely, tumour type (haematologic vs solid tumour; padjOR, 1.40), obesity (padjOR, 0.99), and type of active cancer therapy (cytotoxic chemotherapy vs none: padjOR, 1.47; non-cytotoxic therapy*** vs none in the 4 weeks pre-COVID diagnosis: padjOR, 1.04) were not significantly associated with an increased risk of 30-day mortality, nor were race/ethnicity or recent surgery.

“[This] suggests that curative surgical resections, adjuvant chemotherapy, and maintenance chemotherapy could continue during the SARS-CoV-2 pandemic with extreme caution,” the authors said, highlighting that this is not a recommendation.


Real world implications

“[P]atients with cancer appear to be at increased risk of mortality and severe illness due to SARS-CoV-2 infection, regardless of whether they have active cancer, are on anticancer treatment, or both, [and several] important subgroups [such as ECOG PS 2 and active cancer] … appear to be at increased risk for adverse outcomes,” noted Warner and co-authors.

“These findings have implications for patients and healthcare providers who will be confronted with difficult decisions during the SARS-CoV-2 pandemic, such as whether to withhold or continue anticancer treatments, and whether to accelerate end-of-life planning under some circumstances,” said Warner and co-authors.

“[In addition,] these findings have important policy implications including, but not limited to, the need for increased surveillance and testing for SARS-CoV-2 [and] minimizing healthcare system exposure.”

“The main lesson that we might deduce … is that standard oncological care should be offered if feasible, including chemotherapy administration,” said Professor Philip Poortmans from Iridium Kankernetwerk, Wilrijk-Antwerp, Belgium, Associate Professor Valentina Guarneri from the University of Padova, Italy, and Dr Maria-João Cardoso from the Champalimaud Clinical Centre, Lisbon, Portugal, in a commentary. [Lancet 2020;395:1884-1885]

“Will COVID-19 negatively affect active oncological treatments or, on the contrary, might anticancer therapy be protective against the cytokine storm caused by SARS-CoV-2? Are disease stage and status important for these interactions?” they questioned.


The next step?

“This is early and evolving data, and more time and analysis will be needed to confirm and expand on these findings,” said Warner. “Right now, we’re working to quickly get information about why some patients with cancer become infected with the SARS-CoV-2 virus and identify the factors that affect disease severity and death. We're also interested in the effects of treatments that are being used to treat patients with cancer who have COVID-19,” he said.   

In addition, ASCO President Howard A. Burris III who was not affiliated with the study, welcomed the use of registries to assess the impact of COVID-19 in patients with cancer.

“The cancer care community urgently needs data on the effects of COVID-19, specifically in patients with cancer. How we improve the care we provide these patients and reduce the number of deaths and severe consequences associated with this disease are among the top questions. The CCC19 registry is a great example of the community quickly coming together to identify and collect the data we need on a large scale,” he said.



COVID-19 mortality risk in thoracic cancer: Age, chemotherapy contributing factors

Analysis of the TERAVOLT# registry comprising patients with thoracic malignancies## and COVID-19 showed that older age, ECOG PS, and receipt of chemotherapy were tied to an increased risk of COVID-19 mortality.

“[B]aseline risk factors for [COVID-19] mortality [in patients with thoracic cancers] included age, PS, and the presence of comorbidities,” said study lead author Associate Professor Leora Horn, director of the Thoracic Oncology Program at Vanderbilt University Medical Center, Nashville, Tennessee, US.

In contrast, gender, body mass index (BMI), smoking status, and cancer type or stage had no effect on mortality risk, she added.

The TERAVOLT registry was set up to identify the factors – including cancer characteristics and treatments – that predispose patients with thoracic malignancies to elevated COVID-19 hospitalization and mortality risk. It also aimed to identify the clinical course of patients with thoracic cancer infected by SARS-CoV-2 and assess long-term outcomes of treatment adjustments and delays in this population, said Horn. Patients included had thoracic cancer plus RT-PCR confirmed or radiologically/clinically suspected COVID-19.

Initial analysis of the first 200 patients (median age 68 years, primarily from Europe) after a median 15-day follow-up showed a 33.3 percent COVID-19 mortality rate among patients with thoracic cancers, while 76 percent were hospitalized, and 9 percent admitted to the ICU. [American Academy of Cancer Research (AACR) 2020, session VCTPL09; Lancet Oncol 2020;doi:10.1016/S1470-2045(20)30314-4]

Univariate analysis suggested age >65 years, smoking status, chemotherapy, and comorbidities as factors increasing mortality risk; multivariate analysis deemed only age >65 years as a contributing factor.


Updated analysis

The present findings are the result of analysis of the first 400 patients (median age 67–70 years; BMI 24–25 kg/m2), conducted after a median 33 days since COVID-19 diagnosis and including a more international population. Patients were mostly male and were current/former smokers. At this analysis, 169 patients had recovered, 141 patients had died, and 118 patients had ongoing infection.

The deaths were primarily due to COVID-19 (79.4 percent), while 10.6 percent were due to cancer. About 78 percent of patients were hospitalized (median duration 10 days) and 8.3 percent were admitted to the ICU. Five percent required mechanical ventilation.

Patients who recovered were more likely to have no comorbidities (18.3 percent), while about 31 percent of patients who died had 1 comorbidity. The most common comorbidities were hypertension, chronic obstructive pulmonary disease, vascular disease, diabetes, and renal insufficiency.

Among patients who died, 46.8 percent were on chemotherapy, 22 percent on immunotherapy, 12.8 percent on targeted therapy, and 9.2 percent on radiotherapy.

Multivariate analysis showed that age >65 years was associated with an increased risk of COVID-19 mortality (hazard ratio [HR], 1.70, 95 percent confidence interval [CI], 1.09–2.63; p=0.018 vs <65 years), as was ECOG PS 1 (HR, 2.14, 95 percent CI, 1.11–4.11; p<0.001 vs ECOG PS 0). [ASCO20, abstract LBA111]

Receipt of steroids prior to COVID-19 diagnosis showed a trend toward a higher mortality risk (HR, 1.49, 95 percent CI, 1.00–2.23; p=0.052), as did treatment with anticoagulants.

“Therapy administered to treat COVID-19 is not significantly associated with outcome,” said Horn, noting the comparable proportions of patients who died or recovered after receiving each type of drug (eg, 24 percent vs 23 percent for anticoagulants, 14 percent vs 12 percent for antivirals).

In terms of cancer-specific treatment, receipt of chemotherapy (with or without other treatments) was tied to an increased COVID-19 mortality risk vs no treatment (HR, 1.71, 95 percent CI, 1.12–2.63) or vs immunotherapy or targeted therapy (HR, 1.64, 95 percent CI, 0.77–3.48; p=0.025 for both).

“Prior administration of chemotherapy, as a unique modality or in combination with immune checkpoint inhibitors, is associated with increased risk of death, while immunotherapy and tyrosine kinase inhibitors are not,” noted Horn. 

“A number of factors – pre-existing lung damage, smoking status, advanced age, and comorbidities – make patients with thoracic cancers especially vulnerable to COVID-19,” said Burris III. “These findings give us some insights into outcomes for patients with cancer who develop COVID-19,” he said.


Editor's Recommendations