The artificial intelligence-based model ANORAK improves histopathological grading of lung adenocarcinoma

The introduction of the International Association for the Study of Lung Cancer grading system has furthered interest in histopathological grading for risk stratification in lung adenocarcinoma. Complex morphology and high intratumoral heterogeneity present challenges to pathologists, prompting the development of artificial intelligence (AI) methods. Here we developed ANORAK (pyrAmid pooliNg crOss stReam Attention networK), encoding multiresolution inputs with an attention mechanism, to delineate growth patterns from hematoxylin and eosin-stained slides. In 1,372 lung adenocarcinomas across four independent cohorts, AI-based grading was prognostic of disease-free survival, and further assisted pathologists by consistently improving prognostication in stage I tumors. Tumors with discrepant patterns between AI and pathologists had notably higher intratumoral heterogeneity. Furthermore, ANORAK facilitates the morphological and spatial assessment of the acinar pattern, capturing acinus variations with pattern transition. Collectively, our AI method enabled the precision quantification and morphology investigation of growth patterns, reflecting intratumoral histological transitions in lung adenocarcinoma.

Modification of the GRACE Risk Score for Risk Prediction in Patients With Acute Coronary Syndromes

Importance

The Global Registry of Acute Coronary Events (GRACE) risk score, a guideline-recommended risk stratification tool for patients presenting with acute coronary syndromes (ACS), does not consider the extent of myocardial injury.

Objective

To assess the incremental predictive value of a modified GRACE score incorporating high-sensitivity cardiac troponin (hs-cTn) T at presentation, a surrogate of the extent of myocardial injury.

Design, Setting, and Participants

This retrospectively designed longitudinal cohort study examined 3 independent cohorts of 9803 patients with ACS enrolled from September 2009 to December 2017; 2 ACS derivation cohorts (Heidelberg ACS cohort and Newcastle STEMI cohort) and an ACS validation cohort (SPUM-ACS study). The Heidelberg ACS cohort included 2535 and the SPUM-ACS study 4288 consecutive patients presenting with a working diagnosis of ACS. The Newcastle STEMI cohort included 2980 consecutive patients with ST-elevation myocardial infarction treated with primary percutaneous coronary intervention. Data were analyzed from March to June 2023.

Exposures

In-hospital, 30-day, and 1-year mortality risk estimates derived from an updated risk score that incorporates continuous hs-cTn T at presentation (modified GRACE).

Main Outcomes and Measures

The predictive value of continuous hs-cTn T and modified GRACE risk score compared with the original GRACE risk score. Study end points were all-cause mortality during hospitalization and at 30 days and 1 year after the index event.

Results

Of 9450 included patients, 7313 (77.4%) were male, and the mean (SD) age at presentation was 64.2 (12.6) years. Using continuous rather than binary hs-cTn T conferred improved discrimination and reclassification compared with the original GRACE score (in-hospital mortality: area under the receiver operating characteristic curve [AUC], 0.835 vs 0.741; continuous net reclassification improvement [NRI], 0.208; 30-day mortality: AUC, 0.828 vs 0.740; NRI, 0.312; 1-year mortality: AUC, 0.785 vs 0.778; NRI, 0.078) in the derivation cohort. These findings were confirmed in the validation cohort. In the pooled population of 9450 patients, modified GRACE risk score showed superior performance compared with the original GRACE risk score in terms of reclassification and discrimination for in-hospital mortality end point (AUC, 0.878 vs 0.780; NRI, 0.097), 30-day mortality end point (AUC, 0.858 vs 0.771; NRI, 0.08), and 1-year mortality end point (AUC, 0.813 vs 0.797; NRI, 0.056).

Conclusions and Relevance

In this study, using continuous rather than binary hs-cTn T at presentation, a proxy of the extent of myocardial injury, in the GRACE risk score improved the mortality risk prediction in patients with ACS.

Clear cell variant of atypical fibroxanthoma and pleomorphic dermal sarcoma: molecular characterization and review of the literature

Atypical fibroxanthoma (AFX) and pleomorphic dermal sarcoma (PDS) are unusual cutaneous tumors that typically arise in sun-damaged skin of elderly individuals. Several histopathologic variants have been described, but the clear cell variant is particularly rare with only 18 cases of AFX and 1 case of PDS reported. Here, we present two cases of clear cell AFX and PDS highlighting key histopathologic findings and molecular alterations assessed by next generation sequencing (NGS).

Abstract LB153: Clinical relevance of spatial intermixing of growth patterns and immune infiltration in lung adenocarcinoma-from TRACERx to LATTICe-A

Adenocarcinoma, the most common histologic variant of lung cancer, is morphologically diverse. The International Association for the Study of Lung Cancer (IASLC) grading system, based on the percentages of growth patterns within the tumour, is highly prognostic (Moreira et al. 2020). However, the clinicopathological significance of transitions between growth patterns, and the combinatorial effects of growth pattern and inflammatory cell infiltration are not yet known. We used a deep learning model to delineate six growth patterns (lepidic, acinar, papillary, micropapillary, solid, and cribriform) at pixel level on hematoxylin and eosin diagnostic whole slide images. The model was trained on 49 slides from the TRACERx cohort (AbdulJabbar et al. 2020), and subsequently applied to 4324 slides from 970 adenocarcinoma cases from the Leicester Archival Thoracic Tumor Investigatory Cohort (Moore et al. 2019). To examine how tumor growth patterns are spatially intermixed, we created a graph network of growth patterns. A linking criterion based on effective cell-cell communication distance was established, whereby adjacent compact tumor islands were linked together. Frequencies of 15 types of pairwise links were further evaluated. A higher intermixing score, measured as the Shannon diversity of link percentages, was associated with adverse relapse free survival (RFS) (p<0.001, Hazard Ratio (HR)=1.5, 95% Confidence Interval (CI)=1.3-1.8, n=966), independently of automated IASLC grading (p=0.001, HR=1.4, 95% CI=1.1-1.7). The clinical relevance of intermixing profiles was investigated by clustering patients into 3 groups, based on the similarity between link percentages. The group dominated by links involving high grade patterns (solid, micropapillary, cribriform), showed the highest risk of relapse (p<0.001, HR=1.7, 95% CI=1.4-2.2), followed by the group enriched with papillary-acinar links (p=0.006, HR=1.4, 95% CI=1.1-1.7). Although micropapillary subtype per se confers an unfavorable prognosis (Cao et al. 2016), its association with papillary morphology increased the risk of relapse (p=0.002, HR=4.2, 95% CI=1.6-11.0), independently of micropapillary burden. To investigate the immune microenvironment surrounding growth patterns, we quantified immune cells at the interface between growth patterns. We observed significantly reduced immune infiltration between micropapillary and papillary than between micropapillary and acinar, solid, and cribriform patterns (p<0.001). In conclusion, we showed that tumor growth pattern spatial intermixing is associated with adverse prognosis and immune infiltration. These findings offer novel insights into the spatial interplay of histological phenotypes and its clinical relevance, which may have an impact on immune escape.

Citation Format: Anca-Ioana Grapa, Hanyun Zhang, Xiaoxi Pan, Khalid AbdulJabbar, Jose Coelho-Lima, Ho Kwan Alvin Cheung, Sarah J. Aitken, David A. Moore, Charles Swanton, John Le Quesne, Yinyin Yuan. Clinical relevance of spatial intermixing of growth patterns and immune infiltration in lung adenocarcinoma-from TRACERx to LATTICe-A [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB153.

Abstract LB504: Automated grading of growth patterns in lung adenocarcinoma-from TRACERx to LATTICe-A

Lung adenocarcinoma exhibits distinct growth patterns (WHO, 2021) and the International Association for the Study of Lung Cancer (IASLC) grading system, based on the nature and proportion of histologic subtypes, is highly prognostic (Moreira et al. 2020). However, both recognition and quantification of growth patterns suffer from high interobserver variability among pathologists. Here, we present a large-scale application to automate the segmentation of histologic patterns and reproduce the IASLC grading system to stratify patients and predict prognosis.

A deep learning model was trained to recognize and segment 6 histologic patterns (lepidic, acinar, papillary, micropapillary, solid, and cribriform) on 49 whole-slide images from TRACERx (AbdulJabbar et al. 2020). This model was directly applied to an independent cohort, consisting of 4324 hematoxylin and eosin-stained sections from 970 patients (the Leicester Archival Thoracic Tumor Investigatory Cohort, Moore et al. 2019).

Growth pattern segmentation performance was first evaluated against 2433 hand annotations covering 6 patterns from 9 images at a pixel level using Dice coefficient. The average Dice was 0.502. Predicted predominant pattern per tumor was then compared to a subspecialty pathologist, achieving an overall agreement (51%), comparable to the interobserver rate among pathologists (52%, Boland et al. 2017). Discordant cases were more heterogeneous (p=4.5e-10, Shannon diversity based on pathological scores), underscoring the challenges posed by intratumor heterogeneity.

Patients with a high proportion of micropapillary in the tumor, identified by deep learning, had significantly worse relapse-free survival (RFS) in multivariate analyses including clinical parameters (p=0.00127, Hazard Ratio (HR)=6.4, 95% confidence interval (CI)=2.07-19.8, n=827), consistent with previous publication (Cha et al. 2014). The Kaplan-Meier curve for RFS was significantly differentiated with both automated and pathological IASLC grading (p<0.0001). Moreover, patients with predominantly high-grade patterns (solid, micropapillary, cribriform) identified by deep learning had significantly reduced RFS (p=6.22e-4, HR=1.5, 95% CI=1.18-1.8, n=970). The prognostic effect was stronger using IASLC grading (cutoff: 20%, p=5.56e-6, HR=1.7, 95% CI=1.36-2.2), comparable to the pathological score at IASLC grade 3 (cutoff: 20%, p=2.41e-5, HR=1.8, 95% CI=1.35-2.3). A similar performance was observed for overall survival regarding above analyses.

To the best of our knowledge, this study represents the largest application of deep learning to recognize tumor growth patterns in lung adenocarcinoma. Histologically heterogeneous growth patterns can be automatically identified using a method trained on an independent cohort. Automated tumor grading is significantly associated with patient outcomes, supporting its potential clinical utility.

Citation Format: Xiaoxi Pan, Khalid AbdulJabbar, Jose Coelho-Lima, Anca-Ioana Grapa, Hanyun Zhang, Ho Kwan Alvin Cheung, Sarah J. Aitken, David A. Moore, Charles Swanton, John Le Quesne, Yinyin Yuan. Automated grading of growth patterns in lung adenocarcinoma-from TRACERx to LATTICe-A [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB504.

1293 Hibernoma: Rare Case of An Intense 18F-FDG Uptake Mediastinal Lesion Mimicking Thymic Carcinoma

Hibernoma is a rare benign neoplasia derived from remnants of brown adipose tissue. Here, we describe an unusual case of mediastinal hibernoma which radiologically mimicked a thymic malignancy. The patients had presented with chest pain, associated with palpitations and dizziness suggestive of cardiac pathology. Investigation of coronary anatomy by CT scan, incompletely revealed the presence of the lesion situated in the region of the thymus gland, 48 mm x 94 mm. The mass was smooth walled with evidence of invasion of local structures. Whole-body positron-emission tomography (PET) demonstrated the mass had a very high uptake, SUV 55.2.

The mass was excised by total thymectomy through a sternotomy. Histopathological analysis revealed it to be fibro-fatty in nature. Light microscopy demonstrated the presence of large multi-vacuolated cells, with appearances typical of brown fat, a hibernoma. This is one of only 7 reported cases. Its findings highlight the importance of careful interpretation of PET studies in hibernoma cases. We propose hibernoma should be included as a differential diagnosis of growing mediastinal masses.

Prognostic value of admission high-sensitivity troponin in patients with ST-elevation myocardial infarction

BACKGROUND AND AIM

Although the diagnostic usefulness of high-sensitivity cardiac troponin T (hs-cTnT) is well established in ST-segment elevation myocardial infarction (STEMI), its prognostic relevance in risk stratification of patients with STEMI remains obscure. This study sought to determine the prognostic value of pre-reperfusion (admission) and post-reperfusion (12-hour) hs-cTnT in patients with STEMI treated with primary percutaneous coronary intervention (PPCI).

METHODS

Retrospective observational longitudinal study including consecutive patients with STEMI treated with PPCI at a university hospital in the northeast of England. hs-cTnT was measured at admission to the catheterisation laboratory and 12 hours after PPCI. Clinical, procedural and laboratory data were prospectively collected during patient hospitalisation (June 2010-December 2014). Mortality data were obtained from the UK Office of National Statistics. The study endpoints were in-hospital and overall mortality.

RESULTS

A total of 3113 patients were included. Median follow-up was 53 months. Admission hs-cTnT >515 ng/L (fourth quartile) was independently associated with in-hospital mortality (HR=2.53 per highest to lower quartiles; 95% CI: 1.32 to 4.85; p=0.005) after multivariable adjustment for a clinical model of mortality prediction. Likewise, admission hs-cTnT >515 ng/L independently predicted overall mortality (HR=1.27 per highest to lower quartiles; 95% CI: 1.02 to 1.59; p=0.029). Admission hs-cTnT correctly reclassified risk for in-hospital death (net reclassification index (NRI)=0.588, p<0.001) and overall mortality (NRI=0.178, p=0.001). Conversely, 12-hour hs-cTnT was not independently associated with mortality.

CONCLUSION

Admission, but not 12-hour post-reperfusion, hs-cTnT predicts mortality and improves risk stratification in the PPCI era. These results support a prognostic role for admission hs-cTnT while challenge the cost-effectiveness of routine 12-hour hs-cTnT measurements in patients with STEMI.

Kinetics Analysis of Circulating MicroRNAs Unveils Markers of Failed Myocardial Reperfusion

BACKGROUND

Failed myocardial reperfusion occurs in approximately 50% of patients with ST-elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention (PPCI). It manifests as microvascular obstruction (MVO) on cardiac magnetic resonance (CMR) imaging. Although prognostically important, MVO is not routinely screened for. Our aim was to investigate the kinetics of circulating short noncoding ribonucleic acids [microRNAs (miRNAs)] following PPCI and their association with MVO in STEMI patients.

METHODS

Screening of 2083 miRNAs in plasma from STEMI patients with (n = 6) and without (n = 6) MVO was performed by next-generation sequencing. Two candidate miRNAs were selected and quantified at 13 time points within 3 h postreperfusion in 20 STEMI patients by reverse transcription and quantitative PCR. Subsequently, these 2 miRNAs were measured in a "validation" STEMI cohort (n = 50) that had CMR imaging performed at baseline and 3 months post-PPCI to evaluate their association with MVO.

RESULTS

miR-1 and miR-133b were rapidly released following PPCI in a monophasic or biphasic pattern. Both miRNAs were enriched in circulating microparticles. A second miR-1 peak (90-180 min postreperfusion) seemed to be associated with a higher index of microvascular resistance. In addition, miR-1 and miR-133b levels at 90 min post-PPCI were approximately 3-fold (P = 0.001) and 4.4-fold (P = 0.008) higher in patients with MVO, respectively. Finally, miR-1 was significantly increased in a subgroup of patients with worse left ventricular (LV) functional recovery 3 months post-PPCI.

CONCLUSIONS

miR-1 and miR-133b levels increase within 3 h of PPCI. They are positively associated with MVO and worse LV functional recovery post-PPCI.

Pathological changes secondary to pacing leads within the coronary veins

Changes within the Coronary veins secondary to pacing leads have not been described, this study assessed these changes in explanted hearts. Macroscopically fibrous sheaths formed around longstanding leads, leading to slit like channels for venous return in smaller veins. Histologically changes included bland fibrosis, a foreign body response to the lead, a chronic inflammatory response and inflammatory destruction of the media. Individuals responded in different ways with no clear relationship of changes to duration of the lead.

P1732The additive value of pre- and post-reperfusion cardiac troponin T levels in risk stratification of patients with ST-segment elevation myocardial infarction

Background

Despite therapeutic advances, mortality rates following ST-segment elevation myocardial infarction (STEMI) are still alarmingly high. There is conflicting evidence regarding the prognostic role of high-sensitivity cardiac troponin T (hs-cTnT) measurements before and after primary percutaneous coronary intervention (PPCI) over traditional predictors of mortality in STEMI patients.

Purpose

To determine the additive prognostic value of pre- and 12h post-PPCI hs-cTnT levels in STEMI patients

Methods

Retrospective longitudinal study including 3,113 consecutive STEMI patients treated with PPCI at a university hospital covering a population of 1.6 million in the North East of England. Clinical, procedural, and laboratory data were prospectively collected during patient hospitalization while hs-cTnT measurements were performed at admission to the catheterization laboratory and at 12h post-PPCI. Median follow-up was 53 months. The study endpoints were in-hospital and overall mortality. Mortality data were obtained from the UK Office of National Statistics.

Results

Admission hs-cTnT >515ng/L (4th quartile) was independently associated with in-hospital mortality [HR=2.39 per highest to lower quartiles; 95% CI: 1.44 to 3.97; p=0.001] after multivariate adjustment for a core clinical model of in-hospital mortality prediction. Likewise, admission hs-cTnT >515ng/L independently predicted overall mortality (HR=1.25 per highest to lower quartiles; 95% CI: 1.00 to 1.57; p=0.044) after adjustment for covariates significantly associated with this endpoint. Admission hs-cTnT correctly reclassified risk for in-hospital death [net reclassification index (NRI)=44.1%, p<0.001) and overall mortality (NRI=60.4%, p<0.001). Conversely, 12h hs-cTnT was not independently associated with mortality.

Conclusion

Admission, but not 12h post-reperfusion, hs-cTnT predicts mortality and improves risk stratification in the PPCI era. These results support the role of hs-cTnT in risk stratification of post-STEMI patients and challenge the cost-effectiveness of routine 12h hs-cTnT measurements.

Overcoming Heparin-Associated RT-qPCR Inhibition and Normalization Issues for microRNA Quantification in Patients with Acute Myocardial Infarction

Background Cardiac-enriched micro ribonucleic acids (miRNAs) are released into the circulation following ST-elevation myocardial infarction (STEMI). Lack of standardized approaches for reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) data normalization and presence of RT-qPCR inhibitors (e.g. heparin) in patient blood samples have prevented reproducible miRNA quantification in this cohort and subsequent translation of these biomarkers to clinical practice.

Materials and Methods Using a RT-qPCR miRNA screening platform, we identified and validated an endogenous circulating miRNA as a normalization control. In addition, we assessed the effects of in vivo and in vitro anticoagulant drugs administration (heparin and bivalirudin) on three RT-qPCR normalization strategies (global miRNA mean, exogenous spike-in control [cel-miR-39] and endogenous miRNA control). Finally, we evaluated the effect of heparin and its in vitro inhibition with heparinase on the quantification of cardiac-enriched miRNAs in STEMI patients.

Results miR-425–5p was validated as an endogenous miRNA control. Heparin administration in vitro and in vivo inhibited all RT-qPCR normalization strategies. In contrast, bivalirudin had no effects on cel-miR-39 or miR-425–5p quantification. In vitro RNA sample treatment with 0.3 U of heparinase overcame heparin-induced over-estimation of cardiac-enriched miRNA levels and improved their correlation with high-sensitivity troponin T.

Conclusion miRNA quantification in STEMI patients receiving heparin is jeopardized by its effect on all RT-qPCR normalization approaches. Use of samples from bivalirudin-treated patients or in vitro treatment of heparin-contaminated samples with heparinase are suitable alternatives for miRNA quantification in this cohort. Finally, we reinforce the evidence that cardiac-enriched miRNAs early after myocardial reperfusion reflect the severity of cardiac injury.

Telomerase Mediates Lymphocyte Proliferation but Not the Atherosclerosis-Suppressive Potential of Regulatory T-Cells

Supplemental Digital Content is available in the text.

Objective—

Atherosclerosis is an age-related disease characterized by systemic oxidative stress and low-grade inflammation. The role of telomerase and telomere length in atherogenesis remains contentious. Short telomeres of peripheral leukocytes are predictive for coronary artery disease. Conversely, attenuated telomerase has been demonstrated to be protective for atherosclerosis. Hence, a potential causative role of telomerase in atherogenesis is critically debated.

Approach and Results—

In this study, we used multiple mouse models to investigate the regulation of telomerase under oxidative stress as well as its impact on atherogenesis in vitro and in vivo. Using primary lymphocytes and myeloid cell cultures, we demonstrate that cultivation under hyperoxic conditions induced oxidative stress resulting in chronic activation of CD4⁺ cells and significantly reduced CD4⁺ T-cell proliferation. The latter was telomerase dependent because oxidative stress had no effect on the proliferation of primary lymphocytes isolated from telomerase knockout mice. In contrast, myeloid cell proliferation was unaffected by oxidative stress nor reliant on telomerase. Telomerase reverse transcriptase deficiency had no effect on regulatory T-cell (T_reg) numbers in vivo or suppressive function ex vivo. Adoptive transfer of telomerase reverse transcriptase^–/– T_regs into Rag2^–/– ApoE^–/– (recombination activating gene 2/apolipoprotein E) double knockout mice demonstrated that telomerase function was not required for the ability of T_regs to protect against atherosclerosis. However, telomere length was critical for T_reg function.

Conclusions—

Telomerase contributes to lymphocyte proliferation but plays no major role in T_reg function, provided that telomere length is not critically short. We suggest that oxidative stress may contribute to atherosclerosis via suppression of telomerase and acceleration of telomere attrition in T_regs.

Non-coding RNA regulation of T cell biology: Implications for age-associated cardiovascular diseases

Prevalence of age-associated cardiovascular diseases (CVD) has dramatically increased as a result of improvements in life expectancy. Chronic inflammation is a shared pathophysiological feature of age-associated CVDs, indicating a role for the immune system in the onset and development of CVDs. Indeed, ageing elicits profound changes in both the cardiovascular and immune system, especially in the T cell compartment. Although such changes have been well described at the cellular level, the molecular mechanisms underlying immune-mediated cardiovascular ageing remain largely unexplored. Non-coding RNAs (ncRNAs) comprise a heterogeneous family of RNA transcripts that regulate gene expression at the epigenetic, transcriptional, post-transcriptional, and post-translational levels. Non-coding RNAs have recently emerged as master modulators of T cell immunity. In this review, the state-of-the-art knowledge on ncRNA regulatory effects over T cell differentiation, function, and ageing in the context of age-associated CVDs, such as atherosclerosis, acute coronary syndromes, and heart failure, is discussed.