Brain imaging in lung cancer staging: A real-world, multi-centre study of prevalence of brain metastases, impact on treatment and re-modelling of the NICE health economic analysis.

INTRODUCTION: In 2019, the National Institute for Health and Care Excellence (NICE) updated their recommendations with respect to brain imaging in the staging of non-small cell lung cancer (NSCLC) based on an analytic cost-effectiveness model using published data and modelling assumptions from committee experts. In this study, we aimed to re-run this model using real-world multi-centre UK data. MATERIALS AND METHODS: Retrospective data was collected on consecutive patients with radically treatable clinical stage II and III lung cancer from eleven acute NHS Trusts during the calendar year 01/01/2018 to 31/12/2018. Following a written application to the NICE lung cancer guideline committee, we were granted access to the NG122 brain imaging economic model for the purpose of updating the input parameters in line with the real-world findings from this study. RESULTS: A total of 444 patients had data for analysis. The combined prevalence of occult brain metastases was 6.2% (10/165) in stage II and 6% (17/283) in stage III, compared to 9.5% and 9.3% used in the NICE economic model. 30% of patients with clinical stage III NSCLC and occult BMs on pre-treatment imaging went onto complete the planned curative intent treatment of extracranial disease, 60% completed SRS to the brain and 30% completed WBRT. This compares to 0%, 10% and 0% in the NICE assumptions. The health economic analysis concluded that brain imaging was no longer cost-effective in stage II disease (ICERs £50,023-£115,785) whilst brain imaging remained cost-effective for stage III patients (ICERs 17,000-£22,173), with MRI being the most cost-effective strategy. CONCLUSION: This re-running of the NICE health economic model with real-world data strongly supports the NICE guideline recommendation for brain imaging prior to curative-intent treatment in stage III lung cancer but questions the cost-effectiveness of CT brain imaging prior to curative-intent treatment in stage II lung cancer.

Acquired Hemophilia of Unknown Etiology in an Elderly Man: Case Report.

BACKGROUND Acquired hemophilia is a rare but potentially dangerous bleeding disorder caused by autoantibodies against coagulation factors. It affects 1 to 1.5 per 1 million people each year. While 50% of cases could be idiopathic, other causes include malignancies, diabetes, pregnancy, infection, and autoimmune disorders. CASE REPORT We report a case of a 90-year-old male who developed a spontaneous hematoma on the dorsum of his right hand, with no prior history of trauma or any other mucosal bleeding. His activated partial thromboplastin time (aPTT) was found to be prolonged (>180 seconds) with a very low level of factor VIII (0.1%). CONCLUSIONS As workups did not identify the source, including malignancy and autoimmune diseases, of his acquired hemophilia, it is believed to be idiopathic. He was started on intravenous recombinant factor VIIa (NovoSeven) to control the bleeding in combination with an immunosuppressive therapy of cyclophosphamide and prednisolone. In approximately 10% of patients with acquired hemophilia, underlying malignancy, such as squamous cell cancer, chronic lymphocytic leukemia, non-Hodgkin lymphoma, and multiple myeloma can present and commonly develop in elderly patients. Therefore, patients diagnosed with idiopathic acquired hemophilia should be given long-term follow up.

Is Digital Tomosynthesis on Par With Computed Tomography for the Detection and Measurement of Pulmonary Nodules?

Chest digital tomosynthesis (DT) has potential advantages compared to computed tomography (CT) such as radiation dose reduction. However, the role of DT in pulmonary nodule management remains investigative. We compared DT against CT for pulmonary nodule detection and size measurement. A clinical population comprising 54 nodules from 30 patients and a screening population comprising 42 nodules from 52 patients were included. Scans were independently read by two radiologists. Agreement in nodule measurements between readers and between modalities was assessed by Bland-Altman analysis using a 95% level of significance. The DT true positive fraction for the two readers was 0.44 and 0.39 in the clinical population, and 0.10 and 0.05 in the screening population. No significant inter-modality bias was observed between DT and CT measurements of nodule size, but the range of variation between modalities was approximately 30%. Inter-reader DT measurements also showed no significant bias, with a range of variation of approximately 15%. We conclude that DT has poor nodule detection sensitivity compared to CT. However, DT showed good measurement reproducibility and may be useful for monitoring growth of existing pulmonary nodules.