Karanam, V.

Abstract Rationale Per the 2021 Sepsis Campaign, it is recommended that hospitals use a performance improvement program for sepsis screening, utilizing machine learning to improve specificity. Around 20% of sepsis at the Mayo Clinic, Rochester site is hospital acquired. As inpatient sepsis is usually detected by extreme physiologic derangement or clinical concern, we seek to leverage risk factors beyond traditional SIRS and sepsis criteria to create a machine and deep learning based risk stratification model to improve the precision of sepsis monitoring. Methods We designed a retrospective observational case control study, including patients diagnosed with sepsis as an inpatient at the Mayo Clinic, Rochester hospital and compared them to patients without a sepsis diagnosis over the last 7 years. Sepsis diagnosis was labeled by a combination of sepsis ICD codes and initial lactate elevation. We excluded transfers from outside hospitals, direct admissions for sepsis and patients diagnosed with sepsis in the emergency department. Data was extracted using Mayo Data Explorer, and data preparation and validation were performed using JMP Statistical Software and Python libraries. Feature selection was critical as we examined all data from previous contacts with healthcare, environmental factors, and dynamic trends of inpatient objective data; amongst 5,000 options of lab tests and vital measurements, we selected near 100 to reduce model complexity. We used XG Boost to train our model with a baseline cohort of 100 sepsis patients and 1000 patients without sepsis. Results We created a model that categorically classified patients into three sepsis risk level profiles (low-risk, medium-risk, high-risk) with performance metrics of a specificity of 97%, sensitivity of 65% and area under the receiver operating characteristic curve of 95%. Amongst laboratory data, lactate and creatinine had the highest feature importance (figure below). Our model also identified variables with the least importance to further refine our architecture. Conclusions Our model is one of the first sepsis risk stratification models whose algorithm is optimized for non-ICU inpatient wards, while the majority of similar studies were conducted on ICU and emergency department patients due to limited data availability. The model’s high specificity will allow for earlier sepsis recognition. By including our full cohort of 7,500 patients with sepsis and 35,000 patients without sepsis, we hope to improve class imbalance and increase sensitivity. By integrating transformer architecture, we seek to eliminate variables of low significance to create a model with more practical use, while maintaining high specificity. This abstract is funded by: None

Abstract Rationale Portopulmonary hypertension (POPH) and hepatopulmonary syndrome (HPS) are two distinct pulmonary vascular complications of liver disease. POPH is characterized by pulmonary hypertension and elevated pulmonary vascular resistance while HPS is characterized by intrapulmonary vasodilation and hypoxemia. POPH and HPS reflect opposing pulmonary vascular responses with markedly different therapeutic and transplant implications. Accurate distinction on imaging may aid clinical decision-making and risk stratification. We aimed to quantitatively characterize and compare chest CT imaging features of POPH and HPS compared to controls with isolated portal hypertension (PoH). Methods Three cohorts of HPS, POPH, and PoH were included (n = 5 per cohort). POPH patients with intrapulmonary vasodilation on contrast echocardiography were excluded. Contrast-enhanced chest CTs with slice thickness ≤ 1.5mm obtained closest to the time of diagnosis were selected (mean 2.7 years following diagnosis). CT images were analyzed using a deep learning algorithm developed in-house that segmented and computed volumes of central, interlobar, peripheral, upper peripheral, and lower peripheral pulmonary arteries (PAs); total, upper, and lower pulmonary veins (PVs); and the four cardiac chambers. Resulting volumes were indexed by body-surface area. One-way analysis of variance (ANOVA) was performed with a post-hoc pairwise Mann-Whitney U test for comparison. Results Patients with HPS were 80% male, had a mean age of 48.6±17.4, and mean MELD-Na of 15.4±3.4. The POPH cohort was 20% male, had a mean age of 56.8±11.9, and mean MELD-Na of 13.4±4.0. The PoH cohort was 100% male, had a mean age of 62.8±4.8, and mean MELD-Na of 12.8±8.8. HPS patients had smaller central PA volumes (mean 26.4±1.8 mm2/m2) compared to both PoH (mean 38.6±9.8 mm2/m2, p = 0.03) and POPH patients (44.8±10.8 mm2/m2, p = 0.008). HPS patients, however, had larger lower peripheral PA (mean 30.7±1.9 mm2/m2) and PV (mean 28.3±6.7 mm2/m2) volumes compared to PoH (PAs mean 20.1±4.8 mm2/m2, p = 0.008; PVs mean 18.2±2.8 mm2/m2, p = 0.02) and larger lower peripheral PA volumes compared to POPH patients (PAs mean 22.1±5.1 mm2/m2, p = 0.03). There was no significant difference in volumes between POPH and PoH for all segments. Conclusion Compared to PoH and POPH, HPS is associated with lower central PA volume and higher peripheral PA volume in the lower lungs. Volumes were similar between POPH and PoH; however, this may reflect limited statistical power given the small sample size. Larger studies are needed to further delineate imaging distinctions between POPH, HPS, and PoH. This abstract is funded by: NHLBI

Abstract Introduction: Pulmonary nodules are not uncommonly found in patients with Sjogren's disease. Although they frequently can be benign, nodules in the context of systemic autoimmune disease such as Sjogren's disease require broadened differentials including lymphoid intersitial pneumonia, follicular bronchiolitis, nodular lymphoid hyperplasia (NLH), amyloidosis and lung malignancy. Although lung carcinoma can occur, the most frequent malignant etiology is non-Hodgkin's lymphoma, including pulmonary extranodal marginal zone lymphoma of mucosa associated lymphoid tissue (MALT lymphoma). On the other hand, many of these differentials are benign; however, some such as NLH can progress to lymphoma while others such as nodular amyloidosis can be associated with MALT lymphoma. As pulmonary MALT lymphoma often presents with asymptomatic nodules or consolidations, the index of suspicion by pulmonologists is crucial for diagnosis. Description: A 66 year old female nonsmoker presented with incidentally found bilateral solid and subsolid pulmonary nodules, the largest being 1.3 centimeters in the left lower lobe, in the context of calcified lymph nodes and splenic and liver calcified granulomas. As she was asymptomatic, possible differentials included chronic granulomatous disease, inflammatory/autoimmune lesions and less likely multifocal pulmonary malignancy. Fungal serologies were negative. Antinuclear antibodies, SS-A, and SS-B serologies were strongly positive. Erythrocyte sedimentation rate and rheumatoid factor were also positive. Further diagnostic workup led to a diagnosis of Sjogren's disease. Repeat 3 month CT chest showed stability of the lymphadenopathy and pulmonary nodules. However, despite stability, our pre-test probability for neoplastic pulmonary nodules had increased due to her autoimmune diagnosis. PET-CT scan was completed, showing metabolically active nodules bilaterally with a SUV range of 2-3.5, with the largest nodule having the highest uptake. Transbronchial biopsy of the largest nodule was consistent with MALT lymphoma. Discussion: The diagnosis of Sjogren's disease led to more aggressive efforts towards diagnostic certainty including a PET-CT scan and ultimately transbronchial biopsy. It is postulated that autoimmune activity in the pulmonary lymphoid tissue is what gives rise to MALT lymphoma. In one retrospective study on pulmonary nodules in patients with Sjogren's disease, greater than 70% of biopsied nodules were malignant, of which half represented MALT lymphoma. MALT lymphoma is diagnosed at the stage of disseminated disease in one third of cases, frequently recurs and has potential for aggressive transformation. Therefore, in the case of multiple pulmonary nodules in patients with Sjogren's disease, the evidence of underlying autoimmunity should increase the index of suspicion for pulmonary malignancy, including MALT lymphoma.

Abstract Introduction EGPA is diagnosed clinically in patients with asthma, rhinosinusitis, and eosinophilia with end-organ involvement including neuropathy, lung infiltrates, cardiomyopathy, or kidney, gastrointestinal, or skin involvement. The five-factor score (FFS) differentiates non-severe (FFS = 0) and severe (FFS > 0) EGPA. Forme fruste, or partially treated, EGPA occurs when presumed severe eosinophilic asthma is treated with corticosteroids that can mask end organ damage typical of EGPA. We present a case of forme fruste non-severe EGPA that failed established treatments for non-severe and severe EGPA. Case Presentation This is a 38-year-old female presenting to the pulmonology clinic with three years of cough, dyspnea, wheezing, rhinorrhea, and chest heaviness. She was initially given prednisone for adult-onset asthma with improvement but was unable to taper due to symptom recrudescence on doses lower than 20 milligrams. Allergic bronchopulmonary aspergillosis testing was negative, fractional exhaled nitric oxide was elevated, pulmonary function testing showed moderate airway obstruction, and absolute eosinophil count was 11x10^3/microliter despite prednisone use. Chest computed tomography (CT) showed mild bronchial wall thickening. She was given benralizumab followed by mepolizumab, inhaled corticosteroid with a long-acting beta agonist, cetirizine, nasal budesonide, and two cycles of rituximab for possible EGPA at an outside hospital. At the current evaluation, Chest CT showed multiple nodules, patchy ground glass opacities, and bronchial wall thickening. CT sinuses showed paranasal sinus mucosal thickening with sinus outlet obstruction and nasal endoscopy showed polyps. Anti-neutrophil cytoplasmic antibody was negative, urinalysis showed mild hematuria, and bronchoalveolar lavage was negative for infection. Bone marrow did not suggest neoplasm and hematology workup did not identify other causes of hypereosinophilia. EGPA was considered highest on the differential. She was offered sinus surgery and lung biopsy for enrollment in refractory EGPA trials, but no further medical therapy was identified. Discussion This case shows the importance of recognizing potential forme fruste EGPA rather than asthma refractory to systemic glucocorticoids. Early suspicion is crucial to initiation of the proper regimen. Non-severe EGPA warrants treatment with corticosteroids and mepolizumab. Azathioprine and cyclophosphamide have demonstrated a comparable benefit in relapsing non-severe EGPA after steroid treatment, but in a single low power study. In our patient, who was refractory to glucocorticoids with mepolizumab and rituximab, there is no strong data suggesting further treatment not involving high dose steroids. In her case, chronic steroid use led to morbid obesity and obstructive sleep apnea, highlighting the importance of further research in non-severe EGPA treatment.

Abstract Rationale Immune checkpoint inhibitors (ICI) have become standard care for many cancers. Their mechanism of action mediates immune related adverse events including ICI pneumonitis. Diagnosis of this is difficult to differentiate from infectious and radiation pneumonitis as it is a diagnosis of exclusion often requiring bronchoscopy and imaging interpreted by expert radiologists. This study investigates the use of CALIPER, a computer-aided lung informatics program, in identifying differences between Computed Tomography (CT) patterns in ICI, radiation, and infectious pneumonitis that may aid in diagnosis noninvasively. Methods CT images of ten patients with ICI pneumonitis, ten patients with infectious pneumonitis, and nine patients with radiation pneumonitis, all diagnosed by clinical expertise and evaluation with bronchoalveolar lavage, were analyzed with CALIPER. CALIPER classified each CT voxel into either vessel-related structures (VRS), normal, mild low-attenuation area (LAA), moderate LAA, severe LAA, reticular (R), honeycomb (HC) or ground-glass (GG) features. Each lung was divided into upper, middle, and lower lobes. Exploratory analyses explored the relationship between CALIPER components and diagnosis, comparing ICI and radiation pneumonitis with infectious pneumonitis. Diagnosis was analyzed using multinomial logistic regression adjusted for presence of pleural effusion. Additionally, diagnosis was assessed as infectious versus a pooled group of radiation and ICI pneumonitis and analyzed using logistic regression adjusted for pleural effusion. Due to the exploratory nature of analysis, no correction was applied for multiplicity of testing. Results Twenty-nine patients were analyzed. ICI pneumonitis patients had less moderate low attenuation in the left upper (p = 0.016) and middle peripheral lung (p = 0.038), and radiation pneumonitis patients had less moderate low attenuation in the left upper peripheral lung (p = 0.031). When comparing the pooled group of ICI and radiation pneumonitis images with infectious pneumonitis, there was more moderate LAA in left upper (0.012) and middle (0.034) peripheral lung in the pooled group compared to infectious pneumonitis. There was no association with VRS, GG, HC, R, mild, or severe LAA patterns. Conclusions Although areas of moderate low attenuation were more prevalent for infectious pneumonitis, these results taken in the context of low sample size, exploratory analysis, and lack of clinical reasoning to support these findings make them likely insignificant. Based on this limited exploratory analysis, it is unlikely that we can solely use CALIPER as noninvasive evaluation to differentiate radiation and ICI pneumonitis from infectious pneumonitis. This highlights the need for a more robust program to aid in noninvasive diagnosis.