Citations

AbstractAn in-depth understanding of portal venous anatomy, particularly its variants, is essential for interventional radiologists performing complex hepatobiliary procedures such as transjugular intrahepatic portosystemic shunt (TIPS) creation. Anatomical variations in the portal venous system occur in approximately 35% of the population, and failure to identify these variants can result in significant procedural complications including vascular injury and postprocedural liver failure. This article discusses the most common (types I–V) portal vein variants and the importance of their preprocedural identification, their potential impact on procedural outcomes, and the role of advanced imaging techniques in mitigating risks associated with these anatomical variations.

Background: Monoclonal B-cell lymphocytosis (MBL) is a common premalignant condition that can be classified into chronic lymphocytic leukemia (CLL)-type (CD5, CD23, and CD20 co-expression), and non-CLL-type (CD5-negative) based on flow cytometry immunophenotypes. CLL-type MBL is relatively well characterized with increased risk of hematologic (especially lymphoid) malignancies, serious infections, and decreased response to vaccinations. However, the prevalence, genetic predispositions, and clinical consequences of non-CLL-type MBL remain to be fully elucidated. Herein, we report findings in non-CLL-type MBL in a large screening cohort. Methods: We screened 10,330 individuals (40 years or older with no prior history of hematologic cancers) for MBL with highly sensitive flow cytometry, in the Mayo Clinic Biobank, with patients recruited from primary care-based clinics. Individuals were genotyped by Regeneron Genetics. Logistic regression was used to estimate odds ratios (OR) and 95% confidence intervals (CI). Among those individuals residing in the 27 counties surrounding Rochester, Minnesota, we identified incident hematologic malignancies and melanomas using ICD codes and the Rochester Epidemiology Project, a population based medical records linkage system. All diagnoses were manually reviewed for confirmation. Cox regression was used to calculate hazard ratios (HR) and 95% CIs, adjusted for age and sex. Results: We identified 1,736 individuals with MBL and 8,341 individuals without MBL. Among those with MBL, 246 (14.2%) had non-CLL-type MBL as the major clone. The median age of non-CLL-type MBL was 77 years (range, 47-95). The prevalence of non-CLL-type MBL significantly increased with age, with 0.5% prevalence among those between 40 to 59 years of age, 2.3% prevalence among those between 60 and 79 years, and 7.8% prevalence in those 80 years of age or older. The prevalence of non-CLL-type MBL was significantly higher (P < .001) among males (3.1%) than females (2.0%). Using the genome-wide complex trait analysis tool, the heritability of non-CLL-type MBL was estimated to be ~26%, suggesting inherited genetics plays a role in etiology. To identify common inherited variants, we first evaluated the known CLL susceptibility variants, many of which also associated with CLL-type MBL. Among the 42 variants, none were statistically associated (all P > .05) with non-CLL-type MBL. We next conducted a genome-wide association study to identify novel variants, and the most significant variant resided in 10q26.3 (minor allele frequency [MAF]control = 0.7%, MAFMBL = 3.5%, OR = 6.4, 95% CI 3.6-11.3, P = 5.7 * 10-8). Among the 6,478 individuals (192 with non-CLL-type MBL and 6,286 without MBL) followed for incident hematologic malignancies, 36 developed lymphoid and 23 developed myeloid malignancies after a median of 4.2 years of follow up. Similar to CLL-type MBL, non-CLL-type MBL was associated with a 5.2-fold risk for lymphoid malignancies (95% CI 2.0-13.0, P < .001). Elevated but not significant risks were observed for myeloid malignancies (HR = 2.3, 95% CI 0.5-10.3, P = 0.27). The lymphoid malignancies developed by individuals with non-CLL-type MBL included diffuse large B-cell lymphoma, multiple myeloma, marginal zone lymphoma, and other low-grade lymphomas. We recently reported that individuals with CLL-type MBL are at higher risk for developing melanoma (HR = 1.9, 95% CI 1.3-2.8). Interestingly, non-CLL-type MBL had no evidence of an association with melanoma risk (HR = 0.4, 95% CI 0.1-1.8). We did not observe a significant difference in overall survival between individuals with non-CLL-type MBL vs those without MBL. Conclusions: In the largest screening cohort to date, we evaluated the etiology and clinical outcomes of non-CLL-type MBL, an understudied yet common condition that occurs in ~4 million adults in the United States. Non-CLL-type MBL is more common in men and older individuals, and evidence supports that inherited genetics plays an etiological role with 10q26.3 locus identified as a novel susceptibility locus. Individuals with non-CLL-type MBL had significantly increased risk of lymphoid malignancies, similar to CLL-type MBL, but, distinct from CLL-type MBL, no evidence of an increased risk of melanoma. Our study provides insights into the prevalence, genetic predispositions and clinical consequences of non-CLL-type MBL and how it differs from the classic CLL-type MBL.

Introduction Intravenous Immunoglobulin G (IVIg) is a pooled antibody Immunoglobulin G product that is used for a large and ever-growing number of medical indications. It is also known to be associated with an increased risk of thromboembolic complications such that the Food and drug administration (FDA) has mandated a “black box” warning alerting physicians to this risk. Particular lots of IVIg preparations have been recalled due to a higher-than-expected rate of thromboembolic adverse events (“Evaluation of the Risk of Thromboembolic Events After Immunoglobulin Administration, 2006-2012”. Sentinel Product Assessment, US FDA, 7/13/2017). While a number of mechanisms have been proposed to explain the thromboembolic risk, platelet activation induced by small amounts of aggregated IgGs in IVIg preparations remains the leading hypothesis. Currently, testing of IVIg lots specifically looking at their ability to induce platelet activation is not routinely performed due to challenges with acquisition of fresh platelets for testing, and lack of availability of technically simple platelet activation endpoints. Recent breakthroughs in platelet storage technology have demonstrated that platelet viability can be preserved for long periods (months-years) by controlled-rate cryopreservation which also allows platelets to retain FcγRIIa-mediated signaling (FcγRIIa is the IgG receptor on platelets). This technology has been successfully leveraged for the detection of platelet-activating heparin-induced thrombocytopenia (HIT) and Vaccine-induced immune thrombotic thrombocytopenia (VITT) antibodies (Blood, 140 (25) 2722-2729, 2022). Goals of the study We hypothesized that long-term stored cryopreserved platelets could be used to detect platelet-activating IVIg preparations using a technically simple ELISA endpoint that measures levels of thrombospondin-1 (TSP-1), a platelet granule protein, released upon platelet activation. If successful, this tool could help de-risk IVIg from the standpoint of thrombotic adverse reactions. Methods Aggregated IgG was prepared by heat treatment of IVIg (Pel-Freez- Rogers, AR) as previously described (Immunol. Lett. (15) 311-316, 1987). Platelets were cryopreserved as previously described (Blood, 140 (25) 2722-2729, 2022) with minor modifications. Cryopreserved platelets were thawed and treated with heat-aggregated or untreated IVIg at concentrations ranging from 19 to 75mcg/mL. After incubation, supernatants were collected by centrifugation and were assayed for TSP-1 by an antigen capture ELISA which has been shown previously to correlate with platelet activation. Results 10-month-old “off-the-shelf” cryopreserved platelets treated with buffer control displayed minimal platelet activation as measured by the TSP1 ELISA (Optical density, OD: 0.21). Similarly, platelets treated with IVIg (non-aggregated IgG) at concentrations ranging from 19 to 75mcg/mL displayed little-to-no platelet activation in the TSP1 ELISA (OD: 0.22-0.29). In contrast, cryopreserved platelets treated with heat aggregated IVIg demonstrated marked platelet activation in the TSP1 ELISA with ODs in the 0.60-0.82 range, correlating with increasing concentrations of heat-aggregated IVIg in the reaction. Conclusions De-risking IVIg from a prothrombotic standpoint has been challenging due to the need for fresh platelets for testing and the lack of technically simple means to assess platelet activation. Here, we leveraged breakthrough platelet storage technology coupled to a technically simple ELISA endpoint that allowed for rapid and easy identification of platelet activating forms of IVIg. This easy-to-use self-contained test kit could be implemented as a quality control/patient safety tool to de-risk IVIg preparations from the standpoint of platelet activation-mediated thrombosis. Future studies aim to test a large number of IVIg lots, including “problem” lots associated with thromboembolic events to assess the incidence of platelet activating IgG molecules in IVIg preparations. Acknowledgments This work was supported in part by a grant from the National Institutes of Health (C.G.J.: HL147734).

Background: Chronic lymphocytic leukemia (CLL) is one of the most common leukemias in the United States (US), with an estimated incidence of 20,700 cases in 2024. Population-based studies conducted in Olmsted County (OC), Minnesota, reported an increase in CLL incidence over a 55-year period (1935 through 1989), which was attributed to increased detection of early-stage cases in individuals over age 50. In an updated study from 2000 to 2010 in OC, we reported no further increase in the incidence of CLL using the 2008 International Workshop on CLL (IWCLL 2008) guidelines. Additionally, when we compared the 1996 National Cancer Institute Working Group (NCI-WG 96) criteria with the IWCLL 2008 guidelines incorporating the monoclonal B-cell lymphocytosis designation, we found that the use of the IWCLL 2008 guidelines reduced the overt CLL incidence, modified the distribution of Rai Stages at diagnosis, and shortened the median time to first treatment (TTFT). Over the last two decades, 5-year survival for CLL has increased due to new therapies and improved supportive care. Herein we provide updated data on the incidence, prevalence, and natural progression of CLL in OC and studied similar features in an expanded 6-county region around Mayo Clinic, Rochester. Methods: We utilized the Rochester Epidemiology Project (REP), a population-based medical records linkage system, to identify CLL cases from 2000 to 2020 using International Classification of Diseases codes. We included individuals residing in OC and the expanded 7-county region (REP-7, including OC), for which data became available starting in 2010. CLL cases were included if they met the IWCLL 2008 criteria. All CLL diagnoses were confirmed by manual review of medical records and data on CLL-related factors were collected, including treatment status, prior history of other cancers, and overall survival (OS). Age- and sex-specific incidence rates were derived from REP census data and adjusted to the 2020 US white population. OS analyses were conducted using the Kaplan-Meier method. TTFT was calculated with death as a competing risk. Cox regression was used to evaluate the effect of decade on TTFT and OS among patients in OC. Results: From 2010-2020, 242 patients were diagnosed with CLL in REP-7, and between 2000-2020, 174 patients were diagnosed in OC. Among the 242 CLL patients in REP-7, the median age at diagnosis was 71 years (range 30-96),158 (65%) were male, 236 (99%) were white, and 74 (31%) had a history of other cancer, with non-melanoma skin cancer being the most common. After age and sex adjustment to the 2020 US white population, the CLL incidence in REP-7 and OC from 2010 to 2020 was 7.8 and 8.1 per 100,000 individuals, respectively. In OC, the age- and sex-adjusted CLL incidence from 2000 to 2020 was 7.5 per 100,000 individuals, showing no significant trend changes over time. Among the 242 patients diagnosed with CLL in REP-7 between 2010 and 2020, 102 received treatment with a median TTFT of 9.8 years. In OC from 2000 to 2020, 79 out of 174 patients received treatment with a median TTFT of 9.4 years. When comparing across two decades in OC, the median TTFT was 12.8 years for the period from 2000 to 2009, decreasing to 5.9 years between 2010 and 2020, but was not significantly different after accounting for age and sex. The median OS for patients diagnosed with CLL in REP-7 from 2010 to 2020 was 11 years. The median OS for patients in OC between 2000 to 2020 was 12.6 years. When comparing across decades in OC, the median OS was 8.9 years for the period from 2000 to 2009, increasing to 13.6 years for those diagnosed between 2010 and 2020, but this trend was not significantly different after accounting for age and sex. Prevalence data for REP-7 and OC were analyzed as of 01/01/2020. Additionally, prevalence data for OC were available for 01/01/2010. In REP-7, the prevalence of CLL was 97 per 100,000 individuals on 01/01/2020. In OC, the CLL prevalence was 86 per 100,000 individuals on 01/01/2010, and 100 per 100,000 individuals on 01/01/2020. Conclusions: We observed no substantial increase in the incidence rates of CLL in REP-7 from 2010 to 2020, nor in the 20 years of available data from OC. However, the prevalence of CLL in OC increased by 16% from 2010 to 2020. Additionally, evidence indicated a longer OS among newly diagnosed CLL cases in OC compared to previous decades, suggesting a potential impact of improved therapeutic approaches on OS.

Monoclonal gammopathy of thrombotic significance (MGTS) is a recently recognized thrombophilia characterized by the production of a monoclonal antibody that binds to platelet factor 4 (PF4) and stimulates platelet activation (Blood. 2022 Jul 7;140(1):73-77). Typical signs of MGTS include a history of recurrent thrombosis, persistent thrombocytopenia, and positive laboratory results in PF4/polyanion enzyme-linked immunosorbent assay (ELISA) and serotonin release assay (SRA) testing, as in the related thrombotic disorder heparin-induced thrombocytopenia (HIT). Here, we present a case of MGTS, presenting with transient thrombocytopenia coincident with thrombotic episodes and negative test results in SRA and PF4/polyanion-based testing platforms (ELISAs and automated assays). A 56-year-old male patient initially presented with acute left lower extremity deep vein thrombosis (DVT) and a pulmonary embolism. Following the administration of warfarin, the patient experienced a second DVT four months later and was transitioned to a factor Xa inhibitor. Fifty-two months later, the patient developed occlusion of the splenic and superior mesenteric veins despite additional anticoagulation therapy. Further testing led to a diagnosis of an IgG2 kappa monoclonal gammopathy of undetermined significance (MGUS). Eleven months later, the patient again suffered breakthrough thrombosis, specifically radial artery thrombosis, and a thrombotic cerebrovascular accident, leading to the addition of aspirin to the patient's treatment regimen. Notably, where detailed blood counts were available in the patient's clinical history, the patient experienced only transient thrombocytopenia coinciding with his two most recent thrombotic episodes and has had platelet counts in the normal range in the four-year period since his last thrombotic event. After initiating dual anticoagulant/antiplatelet therapy, the patient has not developed recurrent thrombocytopenia or new thrombotic events. Antigen-based and functional testing for HIT antibodies (ELISA, SRA, and automated immunoassays) were negative in this patient, unlike in the previously described index case of MGTS, where serological testing mimicked classical HIT serology (i.e., HIT ELISA and SRA positivity). The sample was also non-reactive in a vaccine-induced immune thrombotic thrombocytopenia (VITT)-specific ELISA employing uncomplexed PF4 targets, as previously described (Am J Hematol. 2022 May;97(5):519-526). Conversely, the patient's antibody consistently stimulated platelet activation in a PF4- and IgG-dependent manner in functional testing using PF4-treated platelets. This included PF4-treated platelets that were freshly isolated and used in the PF4-dependent P-selectin expression assay (PEA) and cryopreserved platelets used in the thrombospondin-1 release assay. Interestingly, consistently positive PEA results have been documented from four samples over the last four years, during which the patient had normal platelet counts. To further characterize the patient's IgG2 MGUS antibody, the patient's monoclonal was immuno-enriched. Patient sera and the immuno-enriched MGUS were then compared by mass spectrometry, which generated identical mass profiles. After confirming the effective enrichment of the patient's MGUS by mass spectrometry, the antibody was de novo sequenced or “reverse-engineered” to obtain the antibody's amino acid sequence. The antibody was then expressed and tested in ELISA-based and functional platelet assays. Importantly, the recombinant antibody replicated the PF4-dependent platelet activation observed with patient serum and did not bind PF4/polyanion targets in ELISA testing. Together, these data strongly support the identification of a new class of ELISA- and SRA-negative anti-PF4 monoclonal antibodies that cause MGTS. This case illustrates the diagnostic challenges in MGTS. MGTS patients may not exhibit persistent thrombocytopenia, and diagnostic tests for detecting HIT antibodies may be ineffective in accurately diagnosing this new thrombophilia. This case emphasizes the importance of reviewing platelet counts in the peri-thrombotic period to recognize transient thrombocytopenia and establishes the utility of PF4-enhanced functional assays to detect pathogenic anti-PF4 antibodies, even in settings where classical HIT testing (i.e., ELISA/SRA) is negative.

Abstract Aims The increasing need for diagnostic echocardiography tests presents challenges in preserving the quality and promptness of reports. While Large Language Models (LLMs) have proven effective in summarizing clinical texts, their application in echo remains underexplored. Methods and results Adult echocardiography studies, conducted at the Mayo Clinic from 1 January 2017 to 31 December 2017, were categorized into two groups: development (all Mayo locations except Arizona) and Arizona validation sets. We adapted open-source LLMs (Llama-2, MedAlpaca, Zephyr, and Flan-T5) using In-Context Learning and Quantized Low-Rank Adaptation fine-tuning (FT) for echo report summarization from ‘Findings’ to ‘Impressions.’ Against cardiologist-generated Impressions, the models’ performance was assessed both quantitatively with automatic metrics and qualitatively by cardiologists. The development dataset included 97 506 reports from 71 717 unique patients, predominantly male (55.4%), with an average age of 64.3 ± 15.8 years. EchoGPT, a fine-tuned Llama-2 model, outperformed other models with win rates ranging from 87% to 99% in various automatic metrics, and produced reports comparable to cardiologists in qualitative review (significantly preferred in conciseness (P < 0.001), with no significant preference in completeness, correctness, and clinical utility). Correlations between automatic and human metrics were fair to modest, with the best being RadGraph F1 scores vs. clinical utility (r = 0.42) and automatic metrics showed insensitivity (0–5% drop) to changes in measurement numbers. Conclusion EchoGPT can generate draft reports for human review and approval, helping to streamline the workflow. However, scalable evaluation approaches dedicated to echo reports remains necessary.

Abstract Rationale Among tobacco-exposed persons with preserved spirometry (TEPSs), we previously demonstrated that different lung volume indices—specifically, elevated total lung capacity (TLC) versus elevated ratio of functional residual capacity to TLC (FRC/TLC)—identify different lung disease characteristics in the COPDGene cohort. Objective We sought to determine differential disease characteristics and trajectories associated with lung volume indices among TEPSs in the SPIROMICS cohort. Methods We categorized TEPSs (n = 814) by tertiles (low, intermediate, and high) of TLC or residual volume–to-TLC ratio (RV/TLC) derived from baseline computed tomography images and then examined clinical and spirometric disease trajectories in mutually exclusive categories of participants with high TLC without high RV/TLC ([TLC]high) versus high RV/TLC without high TLC ([RV/TLC]high). We examined differences in computed tomography–measured emphysema (Hounsfield units [HU] ⩽−950; parametric response mapping [PRM] of emphysema), air trapping (HU⩽−856; PRM of functional small airway disease; a disease probability measure for non-emphysematous gas trapping), airway geometry (the mean square root of wall area of a hypothetical airway with 10 mm internal perimeter), respiratory symptoms (on the modified Medical Research Council Dyspnea Scale; COPD Assessment Test [CAT]; St. George’s Respiratory Questionnaire [SGRQ]; and Short Form-12 [SF12]), and outcomes (annualized exacerbation rate) between the two categories at baseline and over follow-up time up to 8.5 years, using regression modeling adjusted for age, sex, height, weight, and smoking status (current vs. former smoker) and burden (pack-years). Results In TEPSs, the pattern of spirometric disease progression differed between participants with [TLC]high and those with [RV/TLC]high: There was increased forced vital capacity with stable forced expiratory volume in 1 second in participants with [TLC]high, versus unchanged forced vital capacity but nominally decreased forced expiratory volume in 1 second in those with [RV/TLC]high. Compared with participants with [TLC]high, TEPSs with [RV/TLC]high had less emphysema (by HU ⩽−950) but more airway disease (by HU ⩽−856; PRM of functional small airway disease; disease probability measure for gas trapping, and mean square root of wall area of a hypothetical airway with 10 mm internal perimeter), more respiratory symptoms (on the modified Medical Research Council Dyspnea Scale, CAT, SGRQ, and SF12), and more severe exacerbations at baseline. Over an average follow-up of 4.1 ± 2.4 years (range = 0.5–8.5 yr), TEPSs with [RV/TLC]high also had a higher likelihood of developing more severe spirometric disease (preserved ratio impaired spirometry or Global Initiative for Chronic Obstructive Lung Disease Classification 2) and worsening of their respiratory symptoms (on the CAT and SGRQ). Although the incidence rates of respiratory exacerbations, hospitalizations, and mortality were not significantly different between the two categories over the follow-up period, TEPSs with [RV/TLC]high were more likely to have been prescribed a respiratory inhaler at their last follow-up visit. Conclusions In these TEPSs from the SPIROMICS cohort, lung volume stratification by TLC versus RV/TLC identifies two pre-COPD phenotypes with distinct respiratory symptoms, radiographic features, and clinical trajectories. The characteristics of these pre-COPD phenotypes match those previously described in the COPDGene cohort using TLC versus FRC/TLC stratification.

Introduction: Among patients with cardiogenic shock (CS), high right atrial pressure (RAP) and low mean arterial pressure (MAP) are associated with in-hospital mortality. Pressure-adjusted heart rate (PAHR), defined as heart rate * RAP-to-MAP ratio (HR * [RAP/MAP]), integrates these parameters and has been used to identify cardiovascular dysfunction in critical illness. Higher PAHR values associate with higher surgical ICU mortality, but the prognostic significance of PAHR has not been assessed in cardiac intensive care unit (CICU) patients with CS. Hypothesis: We hypothesized that higher PAHR values are associated with higher risk of in-hospital mortality in patients with CS. Methods: The Critical Care Cardiology Trials Network (CCCTN) is a multinational registry of CICUs coordinated by the TIMI Study Group. Among CS admissions (2018-2023) undergoing invasive hemodynamic assessment within 24h of CICU admission, we calculated PAHR (in bpm) and assessed its relationship with in-hospital mortality. Patients on temporary mechanical circulatory support were excluded. Odds ratios were adjusted for age, sex, vasoactive-inotropic score (VIS), SCAI stage, and preceding cardiac arrest. Subgroup analyses stratifying by CS subtype were performed. Results: Among the 1411 CS admissions in the analysis cohort (mean age 61 years, 31% women), 18% had AMI-CS, 75% had HF-CS, and 7% had secondary CS. At the time of assessment (median 1.4h after CICU admission), 75% were on vasoactive support (median VIS 3.8). Median HR was 92 (Q1-Q3 77-109) bpm, RAP 15 (Q1-Q3 10-19) mmHg, MAP 75 (Q1-Q3 67-87) mmHg, and PAHR 17 (Q1-Q3 11-24) bpm. A stepwise gradient of higher in-hospital mortality with increasing PAHR values (p-trend <0.001; Fig ) was observed. In adjusted models, a higher PAHR was associated with higher in-hospital mortality (aOR per 10-bpm, 1.35 [95% CI, 1.15-1.58]), and had stronger prognostic associations with mortality than its individual hemodynamic components (Fig) . Conclusion: Higher PAHR is strongly associated with in-hospital mortality in CS patients. PAHR, a simple parameter which is readily calculable without needing a pulmonary artery catheter, may aid in identifying CS patients at higher risk of death.