Bibliometric analysis and research trends of artificial intelligence in lung cancer

Background

Due to the rapid advancement of technology, artificial intelligence (AI) has become extensively used for the diagnosis and prognosis of various diseases, such as lung cancer. Research in the field of literature has demonstrated that artificial intelligence (AI) can be valuable in the timely detection of lung cancer and the formulation of an effective treatment plan. This study aims to conduct a bibliometric analysis to examine and illustrate the specific areas of focus, research frontiers, evolutionary processes, and trends in existing research on artificial intelligence in the context of lung cancer.

Methods

Publications on AI in lung cancer were selected from the SCIE and ESCI indexes on September 19, 2023. The examination of nations, academic publications, organizations, writers, citations, and terms in this domain was visually analyzed with InCites and VOSviewer.

Results

In this study, a total of 4275 publications were selected and analyzed. Artificial intelligence-related lung cancer publications have increased significantly in the last 5 years. China and the USA have contributed the most to the literature in this field (1418 publications with 13.92 citation impacts and 1117 publications with 37.34 citation impacts, respectively). The institution with the highest contribution was "Chinese Academy of Sciences," with 118 publications and 29.09 citation impacts. Among the research categories, "Radiology, Nuclear Medicine & Imaging", "Oncology", and "Engineering, Biomedical" were in first place.

Conclusion

The USA and China have always been leaders in this field and will continue to be for some time. Research in countries such as the Netherlands is increasing. However, research collaboration has to be strengthened in developing countries.

Artificial Intelligence in Lung Cancer Imaging: Unfolding the Future

Lung cancer is one of the malignancies with higher morbidity and mortality. Imaging plays an essential role in each phase of lung cancer management, from detection to assessment of response to treatment. The development of imaging-based artificial intelligence (AI) models has the potential to play a key role in early detection and customized treatment planning. Computer-aided detection of lung nodules in screening programs has revolutionized the early detection of the disease. Moreover, the possibility to use AI approaches to identify patients at risk of developing lung cancer during their life can help a more targeted screening program. The combination of imaging features and clinical and laboratory data through AI models is giving promising results in the prediction of patients’ outcomes, response to specific therapies, and risk for toxic reaction development. In this review, we provide an overview of the main imaging AI-based tools in lung cancer imaging, including automated lesion detection, characterization, segmentation, prediction of outcome, and treatment response to provide radiologists and clinicians with the foundation for these applications in a clinical scenario.

Artificial intelligence in clinical applications for lung cancer: diagnosis, treatment and prognosis

Artificial Intelligence (AI) is a branch of computer science that includes research in robotics, language recognition, image recognition, natural language processing, and expert systems. AI is poised to change medical practice, and oncology is not an exception to this trend. As the matter of fact, lung cancer has the highest morbidity and mortality worldwide. The leading cause is the complexity of associating early pulmonary nodules with neoplastic changes and numerous factors leading to strenuous treatment choice and poor prognosis. AI can effectively enhance the diagnostic efficiency of lung cancer while providing optimal treatment and evaluating prognosis, thereby reducing mortality. This review seeks to provide an overview of AI relevant to all the fields of lung cancer. We define the core concepts of AI and cover the basics of the functioning of natural language processing, image recognition, human-computer interaction and machine learning. We also discuss the most recent breakthroughs in AI technologies and their clinical application regarding diagnosis, treatment, and prognosis in lung cancer. Finally, we highlight the future challenges of AI in lung cancer and its impact on medical practice.

COVID-19 and its impact on cancer, HIV, and mentally ill patients

Severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2) and its disease, COVID-19 is a global pandemic creating an unprecedented medical as well economic havoc across the world. Despite the wide spread global infection rates, the death rate is low for COVID-19. However, COVID-19 patients with other comorbid conditions face severe health complications irrespective of their gender or age. As the management of COVID-19 patients is taking up health resources, it is getting difficult to treat patients suffering from other dreadful diseases like cancer, HIV, and mental health issues. In this chapter, we discuss the effects of COVID-19 and management of cancer patients of main cancer subtypes (e.g., breast, lung, blood cancers), and patients affected with HIV and mental health issues. Finally, we also add a perspective on Ayurvedic treatment and its efficacy on COVID-19 patients.

Glypican-1 is a novel immunohistochemical marker to differentiate poorly differentiated squamous cell carcinoma from solid predominant adenocarcinoma of the lung

Background

The histological classification of non-small cell lung cancer (NSCLC) is essential in determining new cancer-specific targeted therapies. However, the accurate typing of poorly differentiated is difficult, particularly for poorly differentiated squamous cell carcinoma and adenocarcinoma of the lung with limited immunohistochemical markers. Thus, novel immunohistochemical markers are required. We assumed the possibility of the immunohistochemical expression of glypican-1 in lung squamous cell carcinoma.

Methods

The microarray dataset GSE43580 from Gene Expression Omnibus database were analyzed for confirming the gene expression of glypican-1 in lung squamous cell carcinoma. We immunohistochemically investigated the use of glypican-1 as a novel positive diagnostic marker for lung squamous cell carcinoma. Glypican-1 expression in 63 cases of poorly differentiated lung squamous cell carcinoma and 60 cases of solid predominant lung adenocarcinoma was investigated by immunohistochemistry. Additionally, we compared glypican-1 expression with the expressions of p40, cytokeratin 5/6, thyroid transcription factor-1 (TTF-1), and napsin A.

Results

All 63 cases of lung squamous cell carcinoma showed glypican-1 expression. In contrast, only 2 cases of lung adenocarcinoma showed glypican-1 expression. The sensitivity, specificity, and diagnostic accuracy of glypican-1 expression for differentiating lung squamous cell carcinoma from lung adenocarcinoma were 100%, 96.7%, and 98.4%, respectively. These were similar to those of p40 and significantly better than those of CK 5/6.

Conclusions

We recommend the use of glypican-1 as an additional positive marker of lung squamous cell carcinoma.

Histopathologic and molecular approach to staging of multiple lung nodules

Distinguishing multiple primary lung cancers from intrapulmonary metastases in patients with synchronous multifocal lung adenocarcinomas can be challenging. The most recent 8th edition American Joint Committee on Cancer staging manual (AJCC staging manual) distinguishes four disease patterns in patients with multiple lung nodules: (I) two or more distinct and histologically different masses (considered unrelated and staged as individual cancers); (II) multiple ground-glass or part-solid nodules, histologically of with lepidic growth pattern (considered separate tumors, T staged based on highest T stage lesion); (III) patchy areas of ground-glass and consolidations, histologically often invasive mucinous adenocarcinomas (considered single tumor with diffuse "pneumonic-type" involvement); and (IV) separate nodules with the same histologic features based on comprehensive histologic subtyping (considered intrapulmonary metastases). Histologic and molecular features, in conjunction with clinical and radiological information, can all be tools to assist with staging of multiple nodules. Histologic features of adenocarcinomas are best characterized using comprehensive histologic subtyping (percentage of lepidic, acinar, solid, papillary and micropapillary pattern). Genomic alterations are commonly assessed using fluorescence in-situ hybridization and next generation sequencing (NGS). The AJCC considers exactly matching breakpoints by comparative genomic hybridization (CGH) as the only evidence for intrapulmonary metastases, and clearly different histologic types or subtypes as the only evidence for separate primary tumors. Similar histologic subtypes or the same biomarker pattern are considered merely relative arguments in favor of a single tumor source. When assessing multifocal lung cancer, pathologists should consider, and carefully weigh the importance of, molecular testing results in addition to the tumor's histologic features. For many cases encountered in routine clinical practice, absolute certainty cannot be reached as to whether they represent multiple primary cancers or intrapulmonary metastases. Classification of difficult cases often benefits from multidisciplinary discussion.

The impact of immunohistochemistry on the classification of lung tumors

INTRODUCTION

To highlight the role of immunohistochemistry to lung cancer classification on the basis of existing guidelines and future perspectives.

AREAS COVERED

Four orienting key-issues were structured according to an extensive review on the English literature: a) cancer subtyping; b) best biomarkers and rules to follow; c) negative and positive profiling; d) suggestions towards an evidence-based proposal for lung cancer subtyping. A sparing material approach based on a limited number of specific markers is highly desirable. It includes p40 for squamous cell carcinoma ('no p40, no squamous'), TTF1 for adenocarcinoma, synaptophysin for neuroendocrine tumors and vimentin for sarcomatoid carcinoma. A close relationship between genotype and phenotype also supports a diagnostic role for negative profiles. Expert commentary: Highly specific and sensitive IHC markers according to positive and negative diagnostic algorithms seem appropriate for individual patients' lung cancer subtyping.

Expression of p63 and CK5/6 in early-stage lung squamous cell carcinoma is not only an early diagnostic indicator but also correlates with a good prognosis

Background

Non-small cell lung cancer (NSCLC) accounts for 80% of lung cancers, and lung squamous cell carcinoma (SQCC) is one of the main types. Advances in the treatment of lung SQCC are lacking when compared to lung adenocarcinoma. The main treatment for early-stage SQCC is surgery. However, factors affecting the efficacy of surgical treatments for early-stage lung SQCC remain unclear. In this study, we examined the significance of commonly used lung SQCC diagnostic markers p63, p40, and cytokeratin (CK)5/6 in prognosis.

Methods

Seventy-six cases of early-stage lung SQCC (N₀) were obtained from our lung cancer database (January 2000 to December 2009). Tissue microarray and immunohistochemical (IHC) staining were used to detect the expression of p63, p40, and CK5/6. The effect of the expression level of each marker on patients' survival was examined.

Results

Sensitivity and specificity of each marker for detecting lung SQCC was 87.0% and 81.0% for p63, 75.9% and 97.9% for p40, and 78.9% and 97.7% for CK5/6. Survival rates of patients with high expression levels of p63 or CK5/6 or both were higher than in patients with low expression levels (P < 0.05). Expression levels of p40 had no effect on survival (P > 0.05). Multivariate analysis showed that high levels of p63 expression p63+CK5/6 co-expression were independent prognostic factors for good survival.

Conclusion

IHC staining detection of p63 and CK5/6 in specimens should be routinely performed in postoperative early-stage lung SQCC patients. Its significance lies not only in differential diagnosis, but also in determining prognosis.

Large cell carcinoma of the lung: a tumor in search of an author. A clinically oriented critical reappraisal

Large cell carcinoma (LCC) is a merely descriptive term indicating a subtype of lung cancer with no specific features of small-cell lung cancer (SCLC), adenocarcinoma (ADC) or squamous cell carcinoma (SQC). This diagnosis is allowed on surgical specimens only, whereas its counterpart in biopsy/cytology samples is non-small-cell lung carcinoma (NSCLC), not otherwise specified (NOS). Although these two terms do not fulfill the same concept, they can be interchangeable synonyms at the clinical level, reflecting, in different ways, the inability to define a specific subtype. Immunohistochemistry (IHC), next generation sequencing (NGS) analysis and, historically, electron microscopy have been unveiling diverse cell differentiation lineages in LCC, resulting in LCC-favor ADC, LCC-favor SQC and LCC-favor large-cell neuroendocrine carcinoma (LCNEC), the latter hopefully to be included into the neuroendocrine tumor (NET) group in the future. Paradoxically, however, the interpretation issues of LCC/NSCLC-NOS are not diminishing, but even increasing albeight an accurate diagnosis is oncologically required and crucial. Also, rare LCC/NSCLC-NOS cases exhibiting null/unclear phenotype, are difficult to classify, and this terminology could be maintained for the sake of classification (basically these tumors are serendipitous ADC, as also confirmed by the lack of p40). In this review article, seven relevant issues to LCC have been addressed by using a question-answer methodology, with final key points discussing major interpretation issues. In conclusion, most LCC/NSCLC-NOS may be eventually re-classified and addressed by exploiting IHC and/or molecular testing to satisfy the criteria of precision medicine (the right drug, to the right patient, at the right time).

Advances in lung adenocarcinoma classification: a summary of the new international multidisciplinary classification system (IASLC/ATS/ERS)

Due to advances in the understanding of lung adenocarcinoma since the advent of its 2004 World Health System classification, an international multidisciplinary panel [sponsored by the International Association for the Study of Lung Cancer (IASLC), American Thoracic Society (ATS), and European Respiratory Society (ERS)] has recently updated the classification system for lung adenocarcinoma, the most common histologic type of lung cancer. Here, we summarize and highlight the new criteria and terminology, certain aspects of its clinical relevance and its potential treatment impact, and future avenues of research related to the new system.

Does Immunohistochemistry Affect Response to Therapy and Survival of Inoperable Non–Small Cell Lung Carcinoma Patients? A Survey of 145 Stage III-IV Consecutive Cases

Whether non–small cell lung carcinoma (NSCLC) unveiled by immunohistochemistry (IHC) has the same clinical outcome as those typed by morphology is still matter of debate. A total of 145 stage III-IV, consecutive inoperable NSCLC patients treated by chemotherapy (133 cases) or EGFR tyrosine kinase inhibitor (12 cases) and including 100 biopsies, 11 surgical specimens, and 34 cytological samples had originally accounted for 120 adenocarcinomas (ADs), 19 squamous cell carcinomas (SQCs), and 6 adenosquamous carcinomas (ADSQCs) by integrating morphology and thyroid transcription factor-1 (TTF1)/p40 IHC. Thirty-two NSCLC–not otherwise specified (NSCLC-NOS) cases were identified by morphology revision of the original diagnoses, which showed solid growth pattern ( P < .001), 22 ADs, 5 SQCs, and 5 ADSQCs by IHC profiling ( P < .001), and 10 gene-altered tumors (3 EGFR, 5 KRAS, and 2 ALK). While no significant relationships were observed between response to therapy and original, morphology or IHC diagnoses, driver mutations and tumor differentiation by TTF1 expression, AD run better progression-free survival (PFS) or overall survival (OS) than other tumor types by morphology ( P = .010 and P = .047) and IHC ( P = .033 and P = .046), respectively. Furthermore, patients with NSCLC-NOS confirmed as AD by IHC tended to have poorer OS ( P = .179) and PFS ( P = .193) similar to that of ADSQC and SQC ( P = .702 and P = .540, respectively). A category of less differentiated AD with poorer prognosis on therapy could be identified by IHC, while there were no differences for SQC or ADSQC. The terminology of “NSCLC-NOS, favor by IHC” is appropriate to alert clinicians toward more aggressive tumors.

Subtyping non-small cell lung cancer: relevant issues and operative recommendations for the best pathology practice

Morphology still remains the cornerstone in lung cancer classification and cytology and small biopsy samples should be interpreted by morphology, whenever feasible, according to shared and widely agreed-upon diagnostic schemes. However, as novel therapy strategies are being offered on the basis of the diverse tumor characteristics, pathologists are now challenged by the need to offer clinicians more detailed typing of non-small cell lung cancer, not otherwise specified (NSCLC-NOS), especially when dealing with limited diagnostic material or poorly differentiated tumors. Close integration of morphology, immunohistochemistry, and clinical data is highly warranted according to a multidisciplinary approach to limit the category of NSCLC-NOS as much as possible or exclude unsuspected metastases, so rendering more definite and clinically useful diagnoses. Among the many proposed immunohistochemical markers, which as a whole are more practical and diagnostically useful than cumbersome and expensive molecular assays, a 2-hit model including thyroid transcription factor-1 (TTF-1) and p40 (the latter more specific for squamous differentiation than p63) seems to be the most effective to basically highlight adenocarcinoma (positivity for TTF-1 regardless of p63) and squamous (always strongly and diffusely positive for p40 or p63 and negative for TTF-1) differentiation. This minimalist 2-hit diagnostic approach paves the way to novel perspectives in clinical trials on lung cancer, and it is also in keeping with the need of strategically preserving diagnostic material for molecular assays that are essential for personalizing therapies.

Napsin A expression in primary mucin-producing adenocarcinomas of the lung: an immunohistochemical study

Immunohistochemical expression of napsin A in primary pulmonary mucinous tumors is not well established. Napsin A immunoreactivity was evaluated in 43 mucin-producing adenocarcinomas of the lung consisting of 18 tumors formerly classified as mucinous bronchioloalveolar carcinoma, 15 colloid adenocarcinomas, 5 solid predominant adenocarcinomas with mucin production, and 5 adenocarcinomas with signet ring cell features, as well as in 25 extrapulmonary mucinous adenocarcinomas of different anatomic sites. Immunohistochemical expression of thyroid transcription factor 1 (TTF-1) was also compared. Thirty-three percent of mucinous lung tumors exhibited positive immunoreactivity for napsin A, whereas 42% expressed TTF-1. All 25 extrapulmonary mucinous adenocarcinomas lacked expression of napsin A and TTF-1. Mucin-producing neoplasms of the lung infrequently express napsin A, suggesting that immunohistochemical assessment of napsin A may have limited diagnostic usefulness for distinguishing primary and metastatic mucinous adenocarcinomas involving the lung.

CUG-binding protein 1 (CUGBP1) expression and prognosis of non-small cell lung cancer

BACKGROUND AND AIMS

Non-small cell lung cancer (NSCLC) is the leading cause of cancer mortality worldwide. As CUGBP1 may also play a great role in tumor genesis and deterioration, the purpose of this study was to detect the expression of CUGBP1 mRNA and CUGBP1 and assess the prognostic significance of CUGBP1 in NSCLC.

METHODS

Expression of CUGBP1 mRNA and CUGBP was detected by Semi-quantitative PCR and Immunohistochemistry, respectively, from 57 NSCLC patients. The percentage of CUGBP1 mRNA and CUGBP1 expression was correlated with clinical characteristics using χ (2) test. The prognostic significance was assessed by univariate and multivariate analyses in the Cox hazard model.

RESULTS

The expression of CUGBP1 mRNA and CUGBP1 was over-expressed in cancer group and was correlated with TNM stage and Differentiation. By both univariate and multivariate survival analyses, CUGBP1 expression (P = 0.0074, HR = 3.701, 95 % CI 1.420-9.648), TNM-stage (HR = 4.043, 95 % CI 2.098-7.794) and age (HR = 3.207, 95 % CI 1.544-6.664) were noted to be independent indicators of a shorter postsurgical survival.

CONCLUSIONS

The expression of CUGBP1 independently predicted a shorter postsurgical survival in NSCLC.

Multiparametric molecular characterization of pulmonary sarcomatoid carcinoma reveals a nonrandom amplification of anaplastic lymphoma kinase (ALK) gene

BACKGROUND

Genetic alterations for targeting therapy are largely unexplored issues in pulmonary sarcomatoid carcinoma (PSC), a life-threatening tumor subset.

METHODS

EGFR, HER2, KRAS, p53, CTNNB1, BRAF and PIK3CA mutations were assessed by direct sequencing, ALK, EGFR and HER2 gene status by fluorescence in situ hybridization (FISH), and ALK protein expression by immunohistochemistry (IHC) in 20 pleomorphic carcinomas (PLC), two pulmonary blastomas (PB) and one carcinosarcoma (CS). Surgical specimens and, in case of positivity, the corresponding preoperative biopsies were analyzed. Furthermore, 51 consecutive metastatic lung adenocarcinomas (MELAD) were used as controls for FISH and IHC assays of ALK gene.

RESULTS

While no rearrangements of ALK were detected in PSC, relevant amplification was identified 5/23 (22%) surgical specimens and paired biopsies (four PLC and one PB, two females and three males, four current and one never smoker, aged 30-83 years). Considering tumor heterogeneity, the percentage of ALK amplified tumor cells ranged from 11% to 43%, with a mean gene copy gain (GCG ± SD) of 6.9 ± 0.8 and no signal differences between the epithelial (6.5 ± 0.9) and the sarcoma-like components (6.8 ± 0.9) of tumors. In the remaining 18 non-amplified PSC, the relevant value was 2.9 ± 0.5 in 1-80% tumor cells (p<0.001). ALK amplification was closely associated with chromosome 7 (EGFR) or 17 (HER2) polysomy (p<0.001). Out of 51 MELAD, 10 were ALK-rearranged (p=0.026) and only one amplified (p=0.009). No amplified tumors, either PSC or MELAD, expressed the relevant protein by IHC, while the 10 ALK-rearranged MELAD were strongly positive. TP53, KRAS and CTNNB1 mutations accounted for 30%, 22%, and 4% of cases, respectively, with no significant relationship with ALK amplification. No mutations for EGFR, HER2, BRAF or PIK3CA gene were observed.

CONCLUSION

ALK gene amplification is a nonrandom and clonally related event in a subset of PSC, but its biologic rationale deserves further investigation. KRAS mutation could represent a novel tool for therapy of such so deadly tumors with MEK inhibitors.

ΔNp63 (p40) and thyroid transcription factor-1 immunoreactivity on small biopsies or cellblocks for typing non-small cell lung cancer: a novel two-hit, sparing-material approach

INTRODUCTION

Diagnosing non-small cell lung cancer on biopsy/cellblock samples by morphology may be demanding. As sparing material for molecular testing is mandatory, a minimalist immunohistochemistry (IHC)-based diagnostic approach is warranted by means of novel, reliable, and easy-to-assess biomarkers.

METHODS

Forty-six consecutive biopsy/cellblock samples and the corresponding resection specimens (as the gold standard for morphology and IHC) from 30 adenocarcinomas (AD), 10 squamous carcinomas (SQC), 5 adenosquamous carcinomas (ADSQC), and 1 sarcomatoid carcinoma (SC) were IHC-evaluated for p40 [corresponding to nontransactivating ΔNp63 isoforms] and thyroid transcription factor-1 (TTF1) by semiquantitative assessment. For p40, also immunodecoration intensity was taken into account and dichotomized as strong or low.

RESULTS

Nonrandom and overlapping distributions of the relevant markers were found in biopsy/cellblock and surgical specimens, which closely correlated with each other and the diverse tumor categories, with no differences in area under curve-receiver-operating-characteristic curves for each marker between any two samples, including p40 and p63. Diagnostic combinations were p40-/TTF1+ or TTF1- for AD (where p40 was negative, apart from 5/30 AD showing at the best 1-2% tumor cells with low intensity); p40+/TTF1- (p40 strong and by far higher than 50%) for SQC; and p40+/TTF1+ or p40+/TTF1- (p40 strong and less than 50%) for ADSQC. The single SC case was p40-/TTF1-, suggesting glandular lineage. Practically, 41/46 (89%) tumors were correctly classified by IHC on small samples, including 30 AD, 10 SQC, 1/5 ADSQC, and no SC. Underdiagnosis of ADSQC was actually because of sampling error of biopsies/cellblocks rather than insufficient biomarker robustness, whereas underdiagnosis of SC was really because of the failure of either marker to highlight epithelial-mesenchymal transition.

CONCLUSIONS

This minimalist IHC-based model of p40 and TTF1 on biopsy/cellblock samples was effective to correctly subtype most cases of lung cancer.

Usefulness of immunohistochemical and histochemical studies in the classification of lung adenocarcinoma and squamous cell carcinoma in cytologic specimens

Histologic subtyping of non-small cell lung carcinoma (NSCLC) is important because the efficacy of new treatments depends on tumor histologic features. We assessed the diagnostic accuracy of classification of lung adenocarcinoma and squamous cell carcinoma (SCC) on cytologic and biopsy specimens based on cytomorphologic studies alone or in combination with ancillary studies compared with resection specimens. Compared with adenocarcinoma, the diagnosis of SCC was based more often on cytomorphologic studies alone (139/185 [75.1%] vs 107/263 [40.7%]). Significantly increased use of immunohistochemical studies in cytology was noted after introduction of targeted lung carcinoma therapies (22/156 [14.1%] for adenocarcinoma and 5/46 [11%] for SCC from 2000-2004 vs 134/156 [85.9%] for adenocarcinoma and 41/46 [89%] for SCC from 2005-2010). Use of immunohistochemical studies resulted in increased diagnostic accuracy for adenocarcinoma (56% [44/78] from 2000-2004 vs 83.2% [154/185] after 2005) but not for SCC (77% [57/74] before 2004 vs 73.9% [82/111] from 2005-2010). Adenocarcinoma showed high expression of cytokeratin (CK)7 (146/146 [100%]), thyroid transcription factor-1 (131/152 [86.2%]), surfactant A (29/36 [81%]), and periodic acid-Schiff with diastase (69/86 [80%]). All SCCs were positive for CK5/6 and p63. Use of immunohistochemical studies on cytologic cell blocks may improve classification of NSCLC.