Neuronal Autoantibodies Linked to Lung Cancer Cognitive Impairment

Author: By Lynda Williams, Senior medwireNews Reporter 

medwireNews: The presence of neuronal autoantibodies is associated with a significantly increased risk of cognitive impairment, indicates a study of patients with non-small-cell lung cancer (NSCLC) and small-cell lung cancer (SCLC). 

“These autoantibodies might represent a potentially treatable mechanism of immune-mediated cognitive impairment among patients with lung cancer”, say study author Carsten Finke, from Charité-Universitätsmedizin Berlin in Germany, and co-workers in JAMA Oncology.

The team tested 127 NSCLC and 40 SCLC patients for anti-intracellular and neuronal surface autoantibodies previously linked to mild cognitive impairment, and also brain-directed autoantibodies with unidentified neuronal antigens. 

Overall, 36.5% of the 167 patients tested positive for a neuronal autoantibody, including 33.9% and 45.0% of the NSCLC and SCLC groups, respectively.  

Cell-based assays for autoantibodies against known antigens were recorded in 19.8% of the patients, with 9.6% having autoantibodies against neuronal surface antigens, 9.0% against intracellular (AIC) antigens and 1.2% against both. 

The cell-based assay findings showed that autoantibodies were more common in SCLC than NSCLC patients (35.0 vs 15.0%), and SCLC patients were more likely to have only intracellular antigens than their NSCLC counterparts (25.0 vs 3.9%) but had a similar likelihood of having only neuronal surface antigens (5.0 vs 11.0%)  

In addition, brain tissue immunohistochemistry detected autoantibodies that were not identifiable using cell-based assays in 16.8% of the lung cancer patients; these were more common in NSCLC than SCLC patients (18.9 vs 10.0%), although SCLC patients had greater binding of serum autoantibodies to hippocampus tissue (27.5 vs 7.9%). 

Men were more likely to have autoantibodies than women (41.0 vs 29.0%) and this difference was driven by their higher rate of neuronal surface autoantibodies (13.3 vs 6.5%), say Carsten Finke et al. 

Ninety-seven of the patients who were free from both brain metastases and a history of severe neurological or psychiatric disorder underwent a battery of neuropsychological assessments.  

Two thirds (67.0%) of these patients had cognitive impairment based on the International Cognition and Cancer Task Force criteria of a test score 2 standard deviations or more below the reference cohort on one or more cognitive domain. Specifically, 45.8% of patients had deficits in executive function and 43.3% in attention. 

Of note, the rates of cognitive impairment were comparable in NSCLC and SCLC patients (69.0 vs 61.5%) and men and women (69.6 vs 63.4%), with no differences detected by tumour stage or by type of treatment. 

But the team found that SCLC patients with neuronal autoantibodies were 11.0 times more likely to have cognitive impairment than SCLC patients without. And the 30.0% of SCLC patients with AIC autoantibodies were significantly more likely than those without to have cognitive impairment (odds ratio [OR]=8.3), verbal memory deficits (OR=44.0) and attention deficits (OR=36.8). 

Among NSCLC patients, the most common autoantibodies targeted the N-methyl-D-aspartate receptor (NMDAR), and patients with immunoglobulin A autoantibodies against NMDAR were 182 times more likely to have a verbal deficit than those without, as well as “substantial deficits” in verbal learning memory tasks.  

However, NSCLC patients with NMDAR autoantibodies did not have an increased risk of cognitive impairment or verbal memory deficit, the researchers observe. 

And patients with unidentified brain-directed autoantibodies, regardless of lung cancer histology, were 2.8 times more likely than patients without to have cognitive impairment, especially attention deficit impairments.  

The researchers note that eight SCLC patients were diagnosed with paraneoplastic syndrome before or during the study, and neurological deficits were more common in patients with neuronal autoantibodies than those without (75.7 vs 55.9%). 

The team highlights that 81.8% of the 11 patients with AIC autoantibodies had neurological deficits. This included all seven patients with AIC autoantibodies and cognitive impairment, whereas just 67.7% of the 31 cognitively impaired patients without AIC autoantibodies had neurological deficits. 

“The exact pathophysiological mechanisms underlying the observed association between neuronal autoantibodies and cognitive impairment warrant further investigation”, say Carsten Finke and co-workers.

They also urge research into the autoantibodies against unknown central nervous system antigens “to identify the target epitopes and better characterize associated clinical symptoms.”

The authors of an accompanying commentary believe that large-scale longitudinal lung cancer studies should be conducted to determine “whether the severity and rate of cognitive impairments mirror the burden of disease or whether it might be governed by separate processes”, as well as to better understand the impact of lung cancer treatment on cognition, including immunotherapies.

“As the underlying mechanisms of cancer-related and cancer therapy-related cognitive impairments become clearer, this study lays the foundation for forthcoming clinical and mechanistic studies and, ultimately, therapeutic intervention”, write Shawn Hervey-Jumper, from the University of California, San Francisco, USA, and Michelle Monje, from Stanford University in California, USA. 

References

Bartels F, Wandrey M-M, Aigner A, et al. Association between neuronal autoantibodies and cognitive impairment in patients with lung cancer. JAMA Oncol; Advance online publication 1 July 2021. doi:10.1001/jamaoncol.2021.2049

Hervey-Jumper SL, Monje M. Unravelling the mechanisms of cancer-related cognitive dysfunction in non-central nervous system cancer. JAMA Oncol; Advance online publication 1 July 2021. doi:10.1001/jamaoncol.2021.1900