Citations

Background/Objective: To externally validate our previously published machine learning model for identifying hospitalized patients with community-acquired pneumonia (CAP) at low risk of respiratory deterioration or death using the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Methods: This is a retrospective cohort study of adult patients (≥18 years) who were admitted with CAP and acute hypoxemic respiratory failure using the publicly available MIMIC-IV critical care dataset (Beth Israel Deaconess Medical Center, Boston, MA). We conducted an external validation study of a previously developed gradient boosting machine (GBM) model without recalibration using data available within the first 6 h of hospital admission. Results: For the primary composite outcome (need for advanced respiratory support [high flow nasal cannula (HFNC), non-invasive mechanical ventilation (NIMV), invasive mechanical ventilation (IMV)] or in-hospital death), the gradient boosting model demonstrated comparable performance in the derivation and external validation cohorts. The area under the receiver operating characteristic curve (AUC) was 0.713 in the Mayo cohort (n = 4379) and 0.689 in the MIMIC-IV cohort. Accuracy was 0.612 (95% confidence interval [CI], 0.595–0.628) versus 0.606 (95% CI 0.600–0.611), specificity 0.574 versus 0.523, sensitivity 0.723 versus 0.754, NPV 0.860 versus 0.842, and PPV 0.364 versus 0.401, respectively. For secondary outcomes, model discrimination was comparable between cohorts. The AUC for in-hospital mortality was 0.727 in the Mayo Cohort versus 0.733 in MIMIC-IV; for IMV, 0.736 versus 0.724; and for NIMV, 0.732 versus 0.708. Conclusions: Our machine learning algorithm demonstrated good discrimination and a high negative predictive value for identifying low-risk hospitalized CAP patients for respiratory deterioration or death in the external MIMIC-IV dataset, supporting its potential utility for prognostic enrichment in pneumonia clinical trials.

Abstract IntroductionFungal infections, particularly in the Midwest, can often be asymptomatic and may mimic primary lung cancer. This case illustrates pulmonary histoplasmosis in an 80-year-old woman with a history of lung cancer highlighting the diagnostic challenges presented when both conditions coexist.Case SummaryAn 80-year-old woman (prior smoking history of 47-pack-years) with synchronous lung cancers (poorly differentiated non-small cell carcinoma with squamous differentiation in the left lower lobe and squamous cell carcinoma in the right lower lobe) previously underwent radiation therapy to both lesions. She was admitted for shortness of breath and cough and was initially diagnosed with pneumonia based on CT findings. Despite multiple courses of antibiotics, her symptoms persisted over 2 months, improving only temporarily with a short course of prednisone. A follow-up CT revealed increasing confluent thick-walled cavitation in the posterior right upper lobe and superior segment of the right lower lobe concerning for cavitating/necrotizing pneumonia versus possible cancer progression. A PET-CT scan (Fig.1A) showed significant FDG uptake in the right upper lobe lesion and mild uptake in mediastinal lymph nodes complicating the differentiation between cancer and infection. Bronchoscopy with biopsy of an endobronchial lesion in the posterior segment of right lower lobe revealed necrotizing granulomatous inflammation (Fig.1B) with GMS stain showing the presence of fungal organisms suggestive of histoplasmosis. Histoplasma PCR from bronchoalveolar lavage was also positive. She was subsequently started on antifungal therapy for a planned one-year treatment course. DiscussionHigh positron emission tomography (PET) avidity in chest imaging is often linked to aggressive tumors but can also indicate benign conditions such as fungal infections. A “flip-flop sign” on PET imaging, where nodal activity surpasses that of a lung nodule, is typically seen in benign lung lesions, particularly fungal infections. However, infections and inflammatory conditions may also show high PET avidity mimicking malignancy. Endobronchial lesions, common in lung cancer, are less frequently associated with infections. In patients with a history of cancer, active lung infection, and PET-avid lesions without the flip-flop sign, invasive testing is usually required to differentiate between infection and malignancy. In this case, our patient with a history of lung cancer and a PET-avid lung lesion was found to have pulmonary histoplasmosis after invasive testing highlighting the need for comprehensive evaluation.ConclusionThis case emphasizes the importance of considering fungal infections in the differential diagnosis of PET-avid pulmonary lesions, particularly in patients with a history of malignancy.