Citations

AbstractBackgroundThere is a knowledge gap regarding lobar versus sublobar resection for atypical carcinoid (AC) of the lung. As such, the authors sought to understand and analyze the outcomes of sublobar resection versus lobectomy in this patient population.MethodsA retrospective analysis using the National Cancer Database was performed to compare overall survival (OS) between patients treated with lobectomy and patients treated with sublobar resection for AC of the lung between the years 2004 and 2016. Patient characteristics were compared with χ2 tests. The Kaplan–Meier method was used to estimate OS distributions, and the log‐rank test was used to compare distributions by treatment strategy. A multivariable Cox regression model was used to assess associations between the treatment strategy and OS. A propensity score matching method was also implemented to further eliminate treatment selection bias in the study sample.ResultsThe database identified 669 patients with T1–T4 and N0–N3 lung ACs that were surgically resected. Unadjusted Kaplan–Meier survival curves did not demonstrate an OS difference between lobectomy and sublobar resection (p = .094). After propensity score matching, curves demonstrated a numerical improvement in OS with lobectomy; however, it was not statistically significant (p = .5). In a subgroup analysis, lobectomy and node‐negative disease were associated with the best OS, whereas sublobar resection and node‐positive disease were associated with the worst OS (p < .0001). Nodal involvement was associated with worse survival, regardless of surgical treatment (p < .0001).ConclusionsIn patients with T1–T4 and N0–N3 ACs of the lung, lobectomy was not associated with an improvement in OS in comparison with sublobar resection.

Abstract The recent development of imaging and sequencing technologies enables systematic advances in the clinical study of lung cancer. Meanwhile, the human mind is limited in effectively handling and fully utilizing the accumulation of such enormous amounts of data. Machine learning-based approaches play a critical role in integrating and analyzing these large and complex datasets, which have extensively characterized lung cancer through the use of different perspectives from these accrued data. In this review, we provide an overview of machine learning-based approaches that strengthen the varying aspects of lung cancer diagnosis and therapy, including early detection, auxiliary diagnosis, prognosis prediction, and immunotherapy practice. Moreover, we highlight the challenges and opportunities for future applications of machine learning in lung cancer.

Objective Heart failure (HF) imposes a substantial burden and the prevalence of HF is high in patients with chronic kidney disease (CKD). HF results in multiple hospital admissions, but whether HF subtypes worsen long-term outcomes and renal function in patients with CKD remains inconclusive. Methods The study comprised 10 904 patients with CKD aged ≥20 years who underwent echocardiography between 1 January 2011 and 31 December 2018. The patients were stratified into four groups: non-HF, HF with reduced ejection fraction (HFrEF), HF with mildly reduced ejection fraction (HFmrEF) and HF with preserved ejection fraction (HFpEF). The primary end points were all-cause mortality, major adverse cardiovascular events (MACEs) and adverse renal outcomes. Results In inverse probability of treatment weighting-adjusted method, the risk of all-cause mortality and MACEs relative to the non-HF group was greatest in the HFrEF group (HR 3.18 (95% CI 2.57 to 3.93) and HR 3.83 (95% CI 3.20 to 4.59)), followed by the HFmrEF (HR 2.75 (95% CI 2.22 to 3.42) and HR 3.08 (95% CI 2.57 to 3.69)) and HFpEF (HR 1.85 (95% CI 1.59 to 2.15) and HR 2.43 (95% CI 2.16 to 2.73) groups. In addition, the HFrEF group had the greatest risks of end-stage renal disease (HR 2.58 (95% CI 1.94 to 3.44)) compared with other groups. Conclusions HF is associated with subsequent worse clinical outcomes, which may be more pronounced in patients with HFrEF, followed by those with HFmrEF and those with HFpEF relative to non-HF group.

PURPOSEThe advancement of natural language processing (NLP) has promoted the use of detailed textual data in electronic health records (EHRs) to support cancer research and to facilitate patient care. In this review, we aim to assess EHR for cancer research and patient care by using the Minimal Common Oncology Data Elements (mCODE), which is a community-driven effort to define a minimal set of data elements for cancer research and practice. Specifically, we aim to assess the alignment of NLP-extracted data elements with mCODE and review existing NLP methodologies for extracting said data elements.METHODSPublished literature studies were searched to retrieve cancer-related NLP articles that were written in English and published between January 2010 and September 2020 from main literature databases. After the retrieval, articles with EHRs as the data source were manually identified. A charting form was developed for relevant study analysis and used to categorize data including four main topics: metadata, EHR data and targeted cancer types, NLP methodology, and oncology data elements and standards.RESULTSA total of 123 publications were selected finally and included in our analysis. We found that cancer research and patient care require some data elements beyond mCODE as expected. Transparency and reproductivity are not sufficient in NLP methods, and inconsistency in NLP evaluation exists.CONCLUSIONWe conducted a comprehensive review of cancer NLP for research and patient care using EHRs data. Issues and barriers for wide adoption of cancer NLP were identified and discussed.

Abstract Introduction: Radiotherapy has long been examined in association with second malignancies, but the impact of radiotherapy for initial primary lung cancer (IPLC) on the risk of second primary lung cancer (SPLC) remains controversial. We evaluated associations between IPLC radiotherapy and the risks of SPLC and overall mortality among lung cancer survivors. Methods: We identified 689,785 patients diagnosed with IPLC in 1988-2013 from SEER. We applied cause-specific Cox (CSC) regression to evaluate association between IPLC radiotherapy and SPLC risk, adjusting for age at IPLC diagnosis, IPLC stage and histology, and prior history of cancer. To allow a latency period of exposure to IPLC radiotherapy on SPLC risk, we performed subgroup analyses restricted to patients who survived ≥5 and ≥10 years after IPLC diagnosis. Using Cox regression, we evaluated association between IPLC radiotherapy and overall mortality, adjusting for the same covariates used in the CSC analysis. Mediation analysis quantified the indirect effect of IPLC radiotherapy on overall mortality mediated through radiotherapy-driven SPLC diagnosis. Validation utilized data from 7,758 IPLC patients (496 SPLCs) at Mayo Clinic. Results: Of 689,785 IPLC patients, 18,818 (2.7%) developed SPLC in SEER. IPLC patients treated with radiotherapy had a reduced SPLC risk versus non-irradiated patients when evaluated at the time of initial diagnosis (cause-specific hazard ratio [csHR]=0.90, P&lt;1x10-6). However, the association between IPLC radiotherapy and SPLC risk was reversed when analyses were restricted to those who survived at least 5 years (csHR=1.08, P=0.01) and 10 years (csHR=1.16, P=0.009), thus long-term survivors treated with IPLC radiotherapy having a significantly increased risk of SPLC. Mayo Clinic data showed consistent results, with the effect of IPLC radiotherapy increasing from csHR of 2.41 (P&lt;1x10-6) to 3.38 (P=0.001) among 5-year and 10-year survivors, respectively. Similarly, compared to non-irradiated patients, those treated with IPLC radiotherapy showed a significantly increased risk of overall mortality among 10-year survivors (hazard ratio [HR]=1.46, P&lt;1x10-6 in SEER and HR=1.54, P=5.1X10-5 in Mayo Clinic). Mediation analysis showed that a substantial proportion of the total effect of IPLC radiotherapy on overall mortality—ranging from 32% to 59%—was due to radiotherapy-driven SPLC among 10-year survivors in SEER and Mayo Clinic. Conclusions: IPLC radiotherapy is associated with increased risks of SPLC and overall mortality, with its effects more pronounced among long-term survivors. Citation Format: Eunji Choi, Vicky T. Lam, Jacqueline A. Aredo, Ashok V. Kumar, Jason Wampfler, Julie T. Wu, David C. Christiani, Iona Cheng, Christopher I. Amos, Rayjean J. Hung, Leah M. Backhus, Joel W. Neal, Heather A. Wakelee, Ping Yang, Summer S. Han. Long term effect of radiotherapy on risk of second primary lung cancer and overall mortality among lung cancer patients [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 3445.

Abstract Lung cancer (LC) is the most common cause of cancer mortality and the third most common cancer by incidence in the United States. While environmental factors play an important role in its development, LC risk also exhibits a high degree of heritability. Linkage analyses and genome-wide association studies have identified multiple loci associated with increased LC risk; however, much of the heritability has yet to be explained by these loci. Structural mutations, including copy number variations (CNVs), contribute to phenotypic diversity through dosage and/or cis-regulatory effects. Unbiased CNV mapping is only possible with the use of high-depth, short-read sequencing and remains much more complicated and prone to false positives than SNP detection. As the tools necessary for CNV detection have improved, both somatic and germline CNVs have been shown to play an important role in disease, especially cancer, and often they account for a significant proportion of pathogenic variants. Little has been done to understand the role of germline CNVs in the biological pathways of hereditary lung cancer (HLC). The goal of the current project is to utilize the whole exome sequencing (WES) data in identifying the CNVs in the HLC families (≥3 LC/family) recruited by the Genetic Epidemiology of Lung Cancer Consortium (GELCC). This work uses germline WES data from 203 individuals (60 with a LC diagnosis) from 25 HLC families (≥3 LC affected/family). We limited our investigation to CNVs called by two independent tools using read depth to infer copy number changes and genomic breakpoints: GATK 4 (https://gatk.broadinstitute.org) and XHMM (https://atgu.mgh.harvard.edu/xhmm/). We have identified CNVs (deletions) in the following protein coding genes: KIR2DL1 (19q13.42; OR=1.77), DMBT1 (10q26.13; OR=1.56), SIBPB1 (20p13; OR=1.53), LILRA6 (20p13; OR=1.53), and HLA-DRB5 (6p21.32; OR=1.64) in ≥3 families in at least two individuals/family. All of these genes with the exception of HLA-DRB5 were previously reported in LC studies; HLA-DRB5 was reported in other cancer studies. DMBT1 was reported to be a candidate tumor suppressor gene and frequent loss of heterozygosity (LOH) was observed in several human tumors. It is highly expressed in lung and loss of expression was reported in LC cell lines. All CNVs identified in our study are consistent with Mendelian expectations and appear to be uncommon in the general population. CNVs in these genes may contribute to the lung cancer risk in the HLC families. Further validation of the identified variants is ongoing. We will consider overall CNV burden and disease association with identified deletions and duplications. The enrichment of rare variants in oncogenically associated genes that co-segregate with LC provides specific mutations for future work and improves our understanding of the inheritance and pathogenesis of LC. Citation Format: John Waldron, Kirsten W. Termine, Anthony M. Musolf, Mariza de Andrade, Colette Gaba, Ramaswamy Govindan, Ping Yang, Ming You, Ellen Jaeger, Angelle F. Bencaz, Marshall W. Anderson, Ann G. Schwartz, Susan M. Pinney, Christopher I. Amos, Joan E. Bailey-Wilson, Diptasri M. Mandal. Analyses of whole exome sequencing (WES) data of hereditary lung cancer families identify germline copy number variations (CNVs) in multiple genes [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 1452.

9135 Background: Bronchial carcinoid (BC) is often categorized into multifocal (MBC) or solitary (SBC). MBC, excluding tumorlet and diffuse idiopathic pulmonary neuroendocrine cell hyperplasia, is considered a relatively uncommon subgroup of BC, with much of the MBC literature stemming from case reports/small series. Our study analyzes MBC among a large cohort of 569 patients with BC and argues for change to the current clinical understanding of MBC. We suggest using the term bronchial neuroendocrine tumor (bNET) to more accurately represent its cells of origin and move away from “carcinoid” (historically meaning “carcinoma-like”) and the outdated associated connotation that carcinoids all have a similar, benign clinical and biological behavior. Methods: Using the Mayo Clinic Epidemiology and Genetics of Lung Cancer Database with Institutional Review Board approval, we retrospectively reviewed 569 patients with bNET (204 males, 365 females) presenting to Mayo Clinic Rochester between 1/1997-12/2012. We used univariate and multivariate Cox regression analyses to evaluate factors affecting overall survival. Results: 80 patients (of 569, 14.1%) were diagnosed with multifocal (MbNET) and 489 with solitary (SbNET). Two-sided Fisher’s exact tests found that older age, female gender, never having smoked cigarettes, and tumorlets were associated with MbNET diagnosis. Family lung cancer history, histopathologic grading (pathology: typical vs. atypical), Ki67, and presence of syndromes (carcinoid, Cushing, and MEN1 syndromes) were similar between MbNET and SbNET groups. Most MbNET cases were stage III-IV at the time of diagnosis, while the majority of SbNET cases were stage I. 5-year OS (83%) and 5-year PFS (75%) of MbNET patients were higher than those of their SbNET counterparts (74% and 68%, respectively). Metastasis status was an independent prognostic factor of poor OS in SbNET (P<0.001) but not in MbNET (P=0.71)(Table). Conclusions: Clinical, radiologic, and histopathologic characteristics and prognostic factors differed between SbNET and MbNET. MbNET arose as a new entity often with advanced stage disease, but good prognosis, which does not follow the NSCLC TNM staging system as in the 2009 NCCN guidelines (used during this study) or the updated 2021 NCCN guidelines which continue to stage lung carcinoid similarly to NSCLC. It may be beneficial to consider multifocal lung carcinoid instead as multifocal bNET, a new clinical entity, warranting a novel staging approach that more accurately reflects prognosis.[Table: see text]