Validated machine learning tools to distinguish immune checkpoint inhibitor, radiotherapy, COVID-19 and other infective pneumonitis

BACKGROUND

Pneumonitis is a well-described, potentially disabling, or fatal adverse effect associated with both immune checkpoint inhibitors (ICI) and thoracic radiotherapy. Accurate differentiation between checkpoint inhibitor pneumonitis (CIP) radiation pneumonitis (RP), and infective pneumonitis (IP) is crucial for swift, appropriate, and tailored management to achieve optimal patient outcomes. However, correct diagnosis is often challenging, owing to overlapping clinical presentations and radiological patterns.

METHODS

In this multi-centre study of 455 patients, we used machine learning with radiomic features extracted from chest CT imaging to develop and validate five models to distinguish CIP and RP from COVID-19, non-COVID-19 infective pneumonitis, and each other. Model performance was compared to that of two radiologists.

RESULTS

Models to distinguish RP from COVID-19, CIP from COVID-19 and CIP from non-COVID-19 IP out-performed radiologists (test set AUCs of 0.92 vs 0.8 and 0.8; 0.68 vs 0.43 and 0.4; 0.71 vs 0.55 and 0.63 respectively). Models to distinguish RP from non-COVID-19 IP and CIP from RP were not superior to radiologists but demonstrated modest performance, with test set AUCs of 0.81 and 0.8 respectively. The CIP vs RP model performed less well on patients with prior exposure to both ICI and radiotherapy (AUC 0.54), though the radiologists also had difficulty distinguishing this test cohort (AUC values 0.6 and 0.6).

CONCLUSION

Our results demonstrate the potential utility of such tools as a second or concurrent reader to support oncologists, radiologists, and chest physicians in cases of diagnostic uncertainty. Further research is required for patients with exposure to both ICI and thoracic radiotherapy.

CYFRA 21-1 Predicts Progression in Idiopathic Pulmonary Fibrosis: A Prospective Longitudinal Analysis of the PROFILE Cohort

Rationale: Idiopathic pulmonary fibrosis (IPF) is a progressive and inevitably fatal condition for which there are a lack of effective biomarkers to guide therapeutic decision making. Objectives: To determine the relationship between serum concentrations of the cytokeratin fragment CYFRA 21-1 and disease progression and mortality in individuals with IPF enrolled in the Prospective Observation of Fibrosis in the Lung Clinical Endpoints (PROFILE) study. Methods: CYFRA 21-1 was identified by immunohistochemistry in samples of human lung obtained at surgery. Concentrations of CYFRA 21-1 were measured using an ELISA-based assay in serum samples collected at baseline, 1 month, and 3 months from 491 individuals with an incident diagnosis of IPF who were enrolled in the PROFILE study and from 100 control subjects at baseline. Study subjects were followed for a minimum of 3 years after their first blood draw. Measurements and Main Results: CYFRA 21-1 localizes to hyperplastic epithelium in IPF lung tissue. Peripheral CYFRA 21-1 concentrations were significantly higher in subjects with IPF than in healthy control subjects in both the discovery (n = 132) (control: 0.96 ± 0.81 ng/ml; vs. IPF: 2.34 ± 2.15 ng/ml; P < 0.0001) and validation (n = 359) (control: 2.21 ± 1.54 ng/ml; and IPF: 4.13 ± 2.77 ng/ml; P < 0.0001) cohorts. Baseline concentrations of CYFRA 21-1 were able to distinguish individuals at risk of 12-month disease progression (C-statistic, 0.70; 95% confidence interval, 0.61-0.79; P < 0.0001) and were predictive of overall mortality (hazard ratio, 1.12 [95% confidence interval, 1.06-1.19] per 1 ng/ml increase in CYFRA 21-1; P = 0.0001). Furthermore, 3-month change in concentrations of CYFRA 21-1 separately predicted 12-month and overall survival in both the discovery and validation cohorts. Conclusions: CYFRA 21-1, a marker of epithelial damage and turnover, has the potential to be an important prognostic and therapeutic biomarker in individuals with IPF.

The Transferrin Receptor CD71 Delineates Functionally Distinct Airway Macrophage Subsets during Idiopathic Pulmonary Fibrosis

Rationale: Idiopathic pulmonary fibrosis (IPF) is a devastating progressive disease with limited therapeutic options. Airway macrophages (AMs) are key components of the defense of the airways and are implicated in the pathogenesis of IPF. Alterations in iron metabolism have been described during fibrotic lung disease and in murine models of lung fibrosis. However, the role of transferrin receptor 1 (CD71)-expressing AMs in IPF is not known. Objectives: To assess the role of CD71-expressing AMs in the IPF lung. Methods: We used multiparametric flow cytometry, gene expression analysis, and phagocytosis/transferrin uptake assays to delineate the role of AMs expressing or lacking CD71 in the BAL of patients with IPF and of healthy control subjects. Measurements and Main Results: There was a distinct increase in proportions of AMs lacking CD71 in patients with IPF compared with healthy control subjects. Concentrations of BAL transferrin were enhanced in IPF-BAL, and furthermore, CD71- AMs had an impaired ability to sequester transferrin. CD71+ and CD71- AMs were phenotypically, functionally, and transcriptionally distinct, with CD71- AMs characterized by reduced expression of markers of macrophage maturity, impaired phagocytosis, and enhanced expression of profibrotic genes. Importantly, proportions of AMs lacking CD71 were independently associated with worse survival, underlining the importance of this population in IPF and as a potential therapeutic target. Conclusions: Taken together, these data highlight how CD71 delineates AM subsets that play distinct roles in IPF and furthermore show that CD71- AMs may be an important pathogenic component of fibrotic lung disease.

A randomised, placebo-controlled study of omipalisib (PI3K/mTOR) in idiopathic pulmonary fibrosis

Phosphatidylinositol 3-kinases (PI3Ks) and mammalian target of rapamycin (mTOR) play a role in the pathogenesis of idiopathic pulmonary fibrosis (IPF). Omipalisib (GSK2126458) is a potent inhibitor of PI3K/mTOR.

A randomised, placebo-controlled, double-blind, repeat dose escalation, experimental medicine study of omipalisib in subjects with IPF was conducted (NCT01725139) to test safety, tolerability, pharmacokinetics and pharmacodynamics. Omipalisib was dosed at 0.25 mg, 1 mg and 2 mg twice daily for 8 days in four cohorts of four subjects randomised 3:1 to receive omipalisib or placebo (two cohorts received 2 mg twice daily).

17 subjects with IPF were enrolled. The most common adverse event was diarrhoea, which was reported by four participants. Dose-related increases in insulin and glucose were observed. Pharmacokinetic analysis demonstrated that exposure in the blood predicts lung exposure. Exposure-dependent inhibition of phosphatidylinositol 3,4,5 trisphosphate and pAKT confirmed target engagement in blood and lungs. 18F-2-fluoro-2-deoxy-d-glucose(FDG)-positron emission tomography/computed tomography scans revealed an exposure-dependent reduction in 18F-FDG uptake in fibrotic areas of the lung, as measured by target-to-background, ratio thus confirming pharmacodynamic activity.

This experimental medicine study demonstrates acceptable tolerability of omipalisib in subjects with IPF at exposures for which target engagement was confirmed both systemically and in the lungs.

Preparation for a first-in-man lentivirus trial in patients with cystic fibrosis

We have recently shown that non-viral gene therapy can stabilise the decline of lung function in patients with cystic fibrosis (CF). However, the effect was modest, and more potent gene transfer agents are still required. Fuson protein (F)/Hemagglutinin/Neuraminidase protein (HN)-pseudotyped lentiviral vectors are more efficient for lung gene transfer than non-viral vectors in preclinical models. In preparation for a first-in-man CF trial using the lentiviral vector, we have undertaken key translational preclinical studies. Regulatory-compliant vectors carrying a range of promoter/enhancer elements were assessed in mice and human air–liquid interface (ALI) cultures to select the lead candidate; cystic fibrosis transmembrane conductance receptor (CFTR) expression and function were assessed in CF models using this lead candidate vector. Toxicity was assessed and ‘benchmarked’ against the leading non-viral formulation recently used in a Phase IIb clinical trial. Integration site profiles were mapped and transduction efficiency determined to inform clinical trial dose-ranging. The impact of pre-existing and acquired immunity against the vector and vector stability in several clinically relevant delivery devices was assessed. A hybrid promoter hybrid cytosine guanine dinucleotide (CpG)- free CMV enhancer/elongation factor 1 alpha promoter (hCEF) consisting of the elongation factor 1α promoter and the cytomegalovirus enhancer was most efficacious in both murine lungs and human ALI cultures (both at least 2-log orders above background). The efficacy (at least 14% of airway cells transduced), toxicity and integration site profile supports further progression towards clinical trial and pre-existing and acquired immune responses do not interfere with vector efficacy. The lead rSIV.F/HN candidate expresses functional CFTR and the vector retains 90–100% transduction efficiency in clinically relevant delivery devices. The data support the progression of the F/HN-pseudotyped lentiviral vector into a first-in-man CF trial in 2017.

Linezolid-induced hypoglycaemia in a patient with type 1 diabetes

Linezolid is a relatively novel oxazolidinone and its use is currently limited to secondary care in the UK. It is a reversible, non-selective monoamine oxidase inhibitor, but at the doses used for antibacterial therapy it does not exert an anti-depressive effect. It has a potentially serious side effect profile and is monitored intensively by the Commission on Human Medicines and Medicines and Healthcare products Regulatory Agency. Herein we report the rarely recognised complication of linezolid-induced hypoglycaemia in a patient with type 1 diabetes.