Prediction of outcome for patients with cutaneous melanoma

Diagnostics of Thyroid Cancer Using Machine Learning and Metabolomics

The objective of this research is, with the analysis of existing data of thyroid cancer (TC) metabolites, to develop a machine-learning model that can diagnose TC using metabolite biomarkers. Through data mining, pathway analysis, and machine learning (ML), the model was developed. We identified seven metabolic pathways related to TC: Pyrimidine metabolism, Tyrosine metabolism, Glycine, serine, and threonine metabolism, Pantothenate and CoA biosynthesis, Arginine biosynthesis, Phenylalanine metabolism, and Phenylalanine, tyrosine, and tryptophan biosynthesis. The ML classifications' accuracies were confirmed through 10-fold cross validation, and the most accurate classification was 87.30%. The metabolic pathways identified in relation to TC and the changes within such pathways can contribute to more pattern recognition for diagnostics of TC patients and assistance with TC screening. With independent testing, the model's accuracy for other unique TC metabolites was 92.31%. The results also point to a possibility for the development of using ML methods for TC diagnostics and further applications of ML in general cancer-related metabolite analysis.

Machine Learning for Lung Cancer Diagnosis, Treatment, and Prognosis

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.

FAM171B as a Novel Biomarker Mediates Tissue Immune Microenvironment in Pulmonary Arterial Hypertension

The purpose of this study was to uncover potential diagnostic indicators of pulmonary arterial hypertension (PAH), evaluate the function of immune cells in the pathogenesis of the disease, and find innovative treatment targets and medicines with the potential to enhance prognosis. Gene Expression Omnibus was utilized to acquire the PAH datasets. We recognized differentially expressed genes (DEGs) and investigated their functions utilizing R software. Weighted gene coexpression network analysis, least absolute shrinkage and selection operators, and support vector machines were used to identify biomarkers. The extent of immune cell infiltration in the normal and PAH tissues was determined using CIBERSORT. Additionally, the association between diagnostic markers and immune cells was analyzed. In this study, 258DEGs were used to analyze the disease ontology. Most DEGs were linked with atherosclerosis, arteriosclerotic cardiovascular disease, and lung disease, including obstructive lung disease. Gene set enrichment analysis revealed that compared to normal samples, results from PAH patients were mostly associated with ECM-receptor interaction, arrhythmogenic right ventricular cardiomyopathy, the Wnt signaling pathway, and focal adhesion. FAM171B was identified as a biomarker for PAH (area under the curve = 0.873). The mechanism underlying PAH may be mediated by nave CD4 T cells, resting memory CD4 T cells, resting NK cells, monocytes, activated dendritic cells, resting mast cells, and neutrophils, according to an investigation of immune cell infiltration. FAM171B expression was also associated with resting mast cells, monocytes, and CD8 T cells. The results suggest that PAH may be closely related to FAM171B with high diagnostic performance and associated with immune cell infiltration, suggesting that FAM171B may promote the progression of PAH by stimulating immune infiltration and immune response. This study provides valuable insights into the pathogenesis and treatment of PAH.

Novel gene signatures for stage classification of the squamous cell carcinoma of the lung

The squamous cell carcinoma of the lung (SCLC) is one of the most common types of lung cancer. As GLOBOCAN reported in 2018, lung cancer was the first cause of death and new cases by cancer worldwide. Typically, diagnosis is made in the later stages of the disease with few treatment options available. The goal of this work was to find some key components underlying each stage of the disease, to help in the classification of tumor samples, and to increase the available options for experimental assays and molecular targets that could be used in treatment development. We employed two approaches. The first was based in the classic method of differential gene expression analysis, network analysis, and a novel concept known as network gatekeepers. The second approach was using machine learning algorithms. From our combined approach, we identified two sets of genes that could function as a signature to identify each stage of the cancer pathology. We also arrived at a network of 55 nodes, which according to their biological functions, they can be regarded as drivers in this cancer. Although biological experiments are necessary for their validation, we proposed that all these genes could be used for cancer development treatments.

Performance-weighted-voting model: An ensemble machine learning method for cancer type classification using whole-exome sequencing mutation

Background

With improvements in next-generation DNA sequencing technology, lower cost is needed to collect genetic data. More machine learning techniques can be used to help with cancer analysis and diagnosis.

Methods

We developed an ensemble machine learning system named performance-weighted-voting model for cancer type classification in 6,249 samples across 14 cancer types. Our ensemble system consists of five weak classifiers (logistic regression, SVM, random forest, XGBoost and neural networks). We first used cross-validation to get the predicted results for the five classifiers. The weights of the five weak classifiers can be obtained based on their predictive performance by solving linear regression functions. The final predicted probability of the performance-weighted-voting model for a cancer type can be determined by the summation of each classifier's weight multiplied by its predicted probability.

Results

Using the somatic mutation count of each gene as the input feature, the overall accuracy of the performance-weighted-voting model reached 71.46%, which was significantly higher than the five weak classifiers and two other ensemble models: the hard-voting model and the soft-voting model. In addition, by analyzing the predictive pattern of the performance-weighted-voting model, we found that in most cancer types, higher tumor mutational burden can improve overall accuracy.

Conclusion

This study has important clinical significance for identifying the origin of cancer, especially for those where the primary cannot be determined. In addition, our model presents a good strategy for using ensemble systems for cancer type classification.

Machine learning and its potential applications to the genomic study of head and neck cancer-A systematic review

BACKGROUND

Machine learning (ML) is powerful tool that can identify and classify patterns from large quantities of cancer genomic data that may lead to the discovery of new biomarkers, new drug targets, and a better understanding of important cancer genes. The aim of this systematic review was to evaluate the existing literature and assess the application of machine learning of genomic data in head and neck cancer (HNC).

MATERIALS AND METHODS

The addressed focused question was "Does machine learning of genomic data play a role in prognostic prediction of HNC?" PubMed, EMBASE, Scopus, Web of Science, and gray literature from January 1990 up to and including May 2018 were searched. Two independent reviewers performed the study selection according to eligibility criteria.

RESULTS

A total of seven studies that met the eligibility criteria were included. The majority of studies were cohort studies, one a case-control study and one a randomized controlled trial. Two studies each evaluated oral cancer and laryngeal cancer, while other one study each evaluated nasopharyngeal cancer and oropharyngeal cancer. The majority of studies employed support vector machine (SVM) as a ML technique. Among the included studies, the accuracy rates for ML techniques ranged from 56.7% to 99.4%.

CONCLUSION

Our findings showed that ML techniques for the analysis of genomic data can play a role in the prognostic prediction of HNC.

Immunohistochemical Study of Pepsinogen C Expression in Cutaneous Malignant Melanoma: Association with Clinicopathological Parameters

Background

The aim of this study was to evaluate the pepsinogen C expression in malignant cutaneous melanomas and analyze its possible relationship to clinical and pathological parameters. Pepsinogen C is an aspartyl proteinase primarily involved in the digestion of proteins in the stomach and represents one of the main androgen-inducible proteins in breast cancer cells.

Method

Tumoral pepsinogen C expression was retrospectively analyzed in 35 paraffin-embedded tissues from patients with primary malignant cutaneous melanoma and in 10 samples from 10 benign lesions (4 dermal melanocytic nevi, 4 compound melanocytic nevi and 2 dysplastic melanocytic nevi), using immunohistochemical methods.

Results

The benign lesions were consistently negative for pepsinogen C, whereas 20 of the 35 malignant melanomas (57%) showed positive immunostaining for pepsinogen C. The percentage of pepsinogen C-positive tumors was significantly higher in men than in women (p=0.01) and in epithelioid melanomas than in fusocellular or mixed type melanomas (p=0.003). In addition, the percentage of pepsinogen-C positive tumors was positively and significantly correlated with lesion thickness (p=0.003), Clark's level of invasion (p=0.028) and tumor stage (p<0.001).

Conclusion

Pepsinogen C could be a new prognosticator of unfavorable outcome in cutaneous malignant melanoma.

Applications of Machine Learning in Cancer Prediction and Prognosis

Machine learning is a branch of artificial intelligence that employs a variety of statistical, probabilistic and optimization techniques that allows computers to “learn” from past examples and to detect hard-to-discern patterns from large, noisy or complex data sets. This capability is particularly well-suited to medical applications, especially those that depend on complex proteomic and genomic measurements. As a result, machine learning is frequently used in cancer diagnosis and detection. More recently machine learning has been applied to cancer prognosis and prediction. This latter approach is particularly interesting as it is part of a growing trend towards personalized, predictive medicine. In assembling this review we conducted a broad survey of the different types of machine learning methods being used, the types of data being integrated and the performance of these methods in cancer prediction and prognosis. A number of trends are noted, including a growing dependence on protein biomarkers and microarray data, a strong bias towards applications in prostate and breast cancer, and a heavy reliance on “older” technologies such artificial neural networks (ANNs) instead of more recently developed or more easily interpretable machine learning methods. A number of published studies also appear to lack an appropriate level of validation or testing. Among the better designed and validated studies it is clear that machine learning methods can be used to substantially (15–25%) improve the accuracy of predicting cancer susceptibility, recurrence and mortality. At a more fundamental level, it is also evident that machine learning is also helping to improve our basic understanding of cancer development and progression.

Sentinel lymph node mapping and staging in endometrial cancer: A Society of Gynecologic Oncology literature review with consensus recommendations

The emphasis in contemporary medical oncology has been "precision" or "personalized" medicine, terms that imply a strategy to improve efficacy through targeted therapies. Similar attempts at precision are occurring in surgical oncology. Sentinel lymph node (SLN) mapping has recently been introduced into the surgical staging of endometrial cancer with the goal to reduce morbidity associated with comprehensive lymphadenectomy, yet obtain prognostic information from lymph node status. The Society of Gynecologic Oncology's (SGO) Clinical Practice Committee and SLN Working Group reviewed the current literature for preparation of this document. Literature-based recommendations for the inclusion of SLN assessment in the treatment of patients with endometrial cancer are presented. This article examines.

Machine learning applications in cancer prognosis and prediction

Cancer has been characterized as a heterogeneous disease consisting of many different subtypes. The early diagnosis and prognosis of a cancer type have become a necessity in cancer research, as it can facilitate the subsequent clinical management of patients. The importance of classifying cancer patients into high or low risk groups has led many research teams, from the biomedical and the bioinformatics field, to study the application of machine learning (ML) methods. Therefore, these techniques have been utilized as an aim to model the progression and treatment of cancerous conditions. In addition, the ability of ML tools to detect key features from complex datasets reveals their importance. A variety of these techniques, including Artificial Neural Networks (ANNs), Bayesian Networks (BNs), Support Vector Machines (SVMs) and Decision Trees (DTs) have been widely applied in cancer research for the development of predictive models, resulting in effective and accurate decision making. Even though it is evident that the use of ML methods can improve our understanding of cancer progression, an appropriate level of validation is needed in order for these methods to be considered in the everyday clinical practice. In this work, we present a review of recent ML approaches employed in the modeling of cancer progression. The predictive models discussed here are based on various supervised ML techniques as well as on different input features and data samples. Given the growing trend on the application of ML methods in cancer research, we present here the most recent publications that employ these techniques as an aim to model cancer risk or patient outcomes.

Expression of TLR3, 4, 7 and 9 in cutaneous malignant melanoma: relationship with clinicopathological characteristics and prognosis

Toll-like receptors (TLRs) have achieved an extraordinary amount of interest in cancer research due to their role in tumor progression. The aim of this study was to investigate the expression and clinical relevance of TLR3, 4, 7 and 9 in cutaneous malignant melanoma (CMM). The expression levels of TLR3, 4, 7 and 9 were analyzed in tumors from 30 patients with CMM. The analysis was performed by immunohistochemistry, and the results were correlated with various clinicopathological findings and with relapse-free survival. Our results indicate that there was a wide variability in the immunostaining score values for each receptor. Positive staining for TLRs was generally found in tumor cells, especially for TLR4 and TLR9. Nevertheless, a significant percentage of tumors also showed TLR4 expression in mononuclear inflammatory cells (62.1 %) and in fibroblast-like cells (34.5 %). Our results showed no significant association between score values for each TLR and clinicopathological characteristics of patients. However, our results demonstrated that high TLR4 expression was significantly associated with a shortened relapse-free survival (p = 0.001). Therefore, TLR4 expression may be a new prognostic factor of unfavorable evolution in cutaneous malignant melanoma.

The role of mammalian coronins in development and disease

Coronins have maintained a high degree of conservation over the roughly 800 million years of eukaryotic evolution.1,2 From its origins as a single gene in simpler eukaryotes, the mammalian Coronin gene family has expanded to include at least six members (see Chapter 4). Increasing evidence indicates that Coronins play critical roles as regulators of actin dependent processes such as cell motility and vesicle trafficking3,4 (see Chapters 6-9). Considering the importance of these processes, it is not surprising that recent findings have implicated the involvement of Coronins in multiple diseases. This review primarily focuses on Coronin 1C (HGNC symbol: CORO1C, also known as Coronin 3) which is a transcriptionally dynamic gene that is up-regulated in multiple types of clinically aggressive cancer. In addition to reviewing the molecular signals and events that lead to Coronin 1C transcription, we summarize the results of several studies describing the possible functional roles of Coronin 1C in development as well as disease progression. Here, the main focus is on brain development and on the progression of melanoma and glioma. Finally, we will also review the role of other mammalian Coronin genes in clinically relevant processes such as neural regeneration and pathogenic bacterial infections (see Chapter 10).

The molecular basis of melanomagenesis and the metastatic phenotype

The last two decades have seen spectacular advances in our understanding of the biology of melanoma and, in particular, have elucidated the mechanisms operative in disease initiation and progression. With respect to the former, the genetics of melanoma and in particular the impact of genetic defects on dysregulation of the cell cycle are key issues in malignant transformation and are a major focus of this review. With respect to the latter, consideration also is given to the acquisition of growth factor autonomy and the capacity for invasion and metastasis from the standpoint of cell adhesion, motility, and matrix digestion. These events have specific morphologic correlates that will be briefly addressed. Where relevant, we will address certain of the modern pharmacogenetic strategies that flow from these novel observations concerning melanoma biology.

Prediction of metastases in melanoma patients with positive sentinel node: Histological and molecular approach

It is now established that sentinel node (SN) biopsy is a minimally-invasive procedure that accurately indicates the regional nodal status. In our institute, 14 consecutive patients had only one node micrometastases after elective lymph node dissection or positive SN for primary cutaneous melanoma. These 14 patients could be clearly divided into two groups: (i) patients who developed distant metastases or in-transit metastases (metastasized group); and (ii) and patients who remains free from metastases (non-metastasized group). The purpose of this study was to identify the histological and molecular factors that might predict the further dissemination beyond the SN. We assessed the maximum depth from the capsule to the deepest melanoma cells and the maximum diameter of melanoma nests in the lymph nodes as histological parameters and also evaluated the quantitative expression of tyrosinase mRNA as a molecular approach. The mean maximum depth and the maximum diameter were significantly smaller in the metastasized group than those in the non-metastasized group. Tyrosinase mRNA expression was strongly correlated with the histological tumor burden. Tyrosinase mRNA expression was higher in the former group than that in the latter group but there were no significant differences between them. Melanoma patients with small micrometastases (<0.5 mm deep, <1 mm in diameter) and a low level of tyrosinase mRNA had less chances for hematogenous metastases via lymph nodes.

Collagenase-3 (MMP-13) expression in cutaneous malignant melanoma

BACKGROUND

Matrix metalloproteases (MMPs), enzymes with the ability to degrade the extracellular matrix, play an important role in tissue invasion by cutaneous malignant melanoma (CMM). One specific MMP, collagenase-3 (MMP-13), is thought to have a key function in the activation of MMP.

AIMS

To evaluate the expression of MMP-13 in CMM and assess its possible relationship to clinical and pathological parameters.

METHODS

MMP-13 expression was analyzed in 51 paraffin-embedded tumor samples from patients with invasive CMM, ten samples from in situ melanomas, and in eight samples from benign lesions (three dermal melanocytic nevi, three compound melanocytic nevi and two atypical melanocytic nevi) using immunohistochemical techniques. The median follow-up period in patients with invasive CMM was 50 months.

RESULTS

Benign lesions were consistently negative for MMP-13, whereas three of the ten in situ melanomas (30%) and 23 of the 51 invasive CMMs (45%) showed positive immunostaining for MMP-13. The percentage of MMP-13-positive tumors correlated significantly and positively with the mitotic index (p=0.002) in invasive CMM. However, our results did not show any significant association between tumoral MMP-13 expression and relapse-free survival in patients with invasive CMM.

CONCLUSIONS

MMP-13 appears to be a factor associated with tumor aggressiveness in CMM. It seems to eliminate an important barrier not only against tumoral invasion but also against proliferation.

Validity of medical staff assessment at admission of patient's risk of nosocomial infection: a prospective study in a surgical intensive care unit

OBJECTIVE

To evaluate the ability of a surgical intensive care unit (SICU) medical staff to assess at admission the individual risk of nosocomial infection (NI) during SICU stay in patients admitted for at least 48 h.

DESIGN

Prospective observational study.

SETTING

A tertiary-care university hospital.

PATIENTS AND PARTICIPANTS

201 admissions to the SICU from November 19, 2003, until April 16, 2004.

MEASUREMENTS AND RESULTS

Assessment by medical staff at admission of each patient's estimated risk of NI (pneumonia, venous central catheter-related infection, symptomatic urinary tract infection, and bacteremia) during SICU hospitalization, in order to classify patients into four groups: NI risk very low or absent (group 1), low (group 2), high (group 3), very high or certain (group 4). NI was diagnosed via routine surveillance according to Centers for Disease Control case definitions.

RESULTS

154 patients were assessed; the percentage of patients with NI increased with estimated risk at admission, from 0% in group 1 to 14.3% in group 4. Positive predictive value of medical assessment varied from 8.4% to 14.5%, according to the cutoff value. Negative predictive value varied from 92.1% to 100%.

CONCLUSION

Our study suggests that ICU physicians encounter a major difficulty when informing patients or patients' families about the risk of NI occurrence, as they cannot predict this risk accurately. This limitation should be explained to patients and their families.

New developments in the staging of melanoma

As the incidence of melanoma increases, so does the search for new staging techniques that may provide important prognostic information and aid in the detection of early metastatic disease. The application of molecular techniques may provide powerful new tools in this search. This review summarizes recent findings obtained by means of conventional RT-PCR, cDNA arrays, and proteomics in the investigation of human melanoma. The molecular tools discussed in this review demonstrate how global transcript and protein analysis might contribute not only to the staging of melanoma, but may hold great promise in improving the diagnosis and treatment of this disease.

Pathologic analysis of sentinel lymph nodes in melanoma patients: Current and future trends

Sentinel lymph node dissection (SLND) has become the standard of care for the staging of clinically-node negative melanomas and breast cancers. A large literature documents the efficacy of SLND in the staging of melanoma and breast cancer. The SLND has lower associated patient morbidity in comparison to elective node dissections that remove the closest regional-draining node group. SLND has improved accuracy over traditional regional node dissection for the staging of melanoma. Currently, several multicenter trials are evaluating the prognostic significance of melanoma micrometastases in SLN detected by immunohistochemical and molecular methods. Pending trial outcome analysis, SLND has no proven effect on mortality. However, given the current oncologic emphasis on detection and removal of nodal tumor metastases, the technique has an important role in minimizing the invasiveness of tumor staging for melanoma and breast cancer. As long as lymph node metastases are used for staging solid malignancies, surgical pathologists are likely to encounter SLN excisional biopsies as a part of their routine practice.

Tyrosinase RT-PCR as a Supplement to Histology for Detecting Melanoma and Nevus Cells in Paraffin Sections of Sentinel Lymph Nodes

The detection of tyrosinase mRNA in sentinel lymph nodes (SLNs) by reverse transcription polymerase chain reaction (RT-PCR) is a sensitive indicator for the presence of melanoma or nevus cells, but it does not enable a distinction between both. We have established an efficient method for extraction and reverse transcription of tyrosinase mRNA from paraffin sections that permits the close correlation of the RT-PCR results with (immuno)histologic findings in adjacent sections. One hundred fifty-three SLNs and 6 non-SLN specimens originating from 92 melanoma and 4 nonmelanoma patients were studied to test the reliability of this approach. The predictive value of positive RT-PCR results was 0.98 for the presence of melanoma or nevus cells; the corresponding negative predictive value was 0.83. Furthermore, the detection rate of tyrosinase mRNA significantly correlated with tumor burden. Among the 33 melanoma-positive SLNs without nevus cells, positive RT-PCR results were obtained in all specimens with extended peripheral (S2) or deeply invasive (S3) micrometastases but in only 46% of the cases with few localized melanoma cells in the subcapsular zone (S1). Routine (immuno)histologic evaluation alone had missed microclusters of melanoma cells in one SLN and small nevus cell aggregates in six other SLNs. They were detected only during microscopic reexamination caused by a positive RT-PCR result. We conclude that histology and immunohistochemistry remain the indispensable gold standard for the identification of melanoma and nevus cells in SLNs. Additional molecular analyses using adjacent paraffin sections may further improve the diagnostic accuracy by sensitizing and guiding the microscopist's attention.

Apolipoprotein D expression in cutaneous malignant melanoma

BACKGROUND AND OBJECTIVES

Apolipoprotein D (Apo D) is a protein component of the human plasma lipid transport system, and an androgen-regulated protein in both breast and prostate cancer cell lines. Our goal was to evaluate the expression of Apo D in malignant cutaneous melanomas, as well as to assess its possible relationship to clinical and pathological parameters.

METHODS

Apo D expression was analyzed in 32 paraffin-embedded tissues from patients with invasive cutaneous malignant melanomas, in 8 samples from in situ melanoma, and in 10 samples from 10 benign lesions (4 dermal melanocytic nevi, 4 compound melanocytic nevi, and 2 dysplastic melanocytic nevi), using immunohistochemical techniques.

RESULTS

The benign lesions were consistently negative for Apo D, whereas 3 of the 8 "in situ" melanomas (37.5%) and 12 of the 32 invasive melanomas (37.5%) showed positive immunostaining for Apo D. The percentage of Apo D-positive tumors was significantly higher in nodular than in superficial spreading melanomas (P = 0.011) and in melanomas with vertical growth phase than in melanomas with radial growth phase (P = 0.02). In addition, the percentage of Apo D-positive tumors was positively and significantly correlated with Clark's level of invasion (P = 0.046).

CONCLUSIONS

Apo D may be a new prognostic factor of unfavorable evolution in cutaneous malignant melanoma.

The active fraction of plasmatic plasminogen activator inhibitor type 1 as a possible indicator of increased risk for metastatic melanoma

Plasminogen activator inhibitor type 1 (PAI1) is considered to be the main regulator of fibrinolytic activity in blood and has been identified as a key-enzyme in the metastasis and vascularization of solid tumors. The aim of this study was to determine whether high or low plasma levels and/or activity of PAI1 correlate with the presence of metastatic disease for patients with melanoma. We hypothesized that the presence of metastases could result in a disturbance of the fibrinolytic balance of the blood. To test our hypothesis, we have developed a unique enzyme-linked immunosorbent assay (ELISA) that can measure both the total amount as well as the active fraction of PAI1 in the plasma. We then used this novel assay to analyze the plasmatic PAI1 levels and activity of patients with advanced melanoma (AM, n = 18) and primary melanoma (PM, n = 21) and compare it to a control population (n = 38). We found no statistically significant difference in the total plasmatic PAI1 levels between the controls and patients with PM or AM (P = 0.6199). In contrast, there was a significant difference in the active fraction of PAI1 between the controls and patients with PM or AM (P = 0.0076). The difference between the control and AM groups was highly significant (P = 0.0042). A value of less than 44% active PAI1 was shown to be clinically meaningful by linear discriminant analysis. Surprisingly, the difference between the control and PM groups was also significant--although borderline (P = 0.0488). Of the patients with PM, 19% had PAI1 activity values less than 44%, which strongly supports further investigations to determine whether plasmatic PAI1 activity might be a biological marker of increased metastatic risk.

Overexpression of the cell adhesion molecule L1 is associated with metastasis in cutaneous malignant melanoma

Modulation of cell adhesion molecule expression plays a key role in melanoma metastasis. In particular, the expression of the cell adhesion molecule L1 has been associated with the metastatic phenotype in a murine model of malignant melanoma. However, no such association between L1 expression and metastasis has been investigated in a clinical study. Therefore, L1 expression was determined immunohistochemically in 100 cases of malignant melanoma and correlated with metastasis in a 10-year retrospective study. Furthermore, nine distant metastases and five sentinel lymph node metastases were analysed for their L1 expression. Additionally, the expression of alpha2,3 sialic acid residues, which are recognised by the siglec domain of L1, was determined by Maackia amurensis agglutinin (MAA) lectin histochemistry. The log-rank test between Kaplan-Meier curves revealed a positive association between L1 expression and metastasis (P<0.0001) and multivariate Cox regression analysis adjusted for tumour thickness, ulceration and mitotic rate confirmed the prognostic power of L1 in malignant melanoma. As alpha2,3 sialic acid residues were absent in melanoma cells, homotypic adhesion between melanoma cells via their siglec domain can be excluded, suggesting a different adhesive function of L1 during melanoma metastasis. The functional role of L1 was further stressed by the fact that its expression was preserved in metastatic lesions.

CEACAM1 expression in cutaneous malignant melanoma predicts the development of metastatic disease

PURPOSE

The cell adhesion molecule CEACAM1 is involved in intercellular adhesion and subsequent signal transduction events in a number of epithelia. CEACAM1 downregulation has been demonstrated in colorectal and prostate carcinomas. This study sought to analyze whether its expression in malignant melanoma is associated with metastasis.

PATIENTS AND METHODS

CEACAM1 expression was immunohistochemically evaluated in 100 primary cutaneous malignant melanomas and correlated with metastasis in a 10-year follow-up. Furthermore, CEACAM1 expression was analyzed in metastatic lesions (11 distant metastases and six sentinel lymph node metastases). Univariate Kaplan-Meier analysis and multivariate Cox proportional hazard regression analysis adjusted for standard prognostic indicators were performed to assess the prognostic relevance of CEACAM1 expression.

RESULTS

A total of 28 of 40 patients with CEACAM1-positive primary melanomas developed metastatic disease, compared with only six of 60 patients with CEACAM1-negative melanomas. Often, the strongest CEACAM1 expression was observed at the invading front. In addition, CEACAM1 expression was preserved in the metastatic lesions. Kaplan-Meier analysis revealed a highly significant association between CEACAM1 expression and metastasis (P <.0001); multivariate Cox regression analysis, including CEACAM1 expression status adjusted for tumor thickness, presence of ulceration, and mitotic rate, confirmed that CEACAM1 is an independent factor for the risk of metastasis and demonstrated that the predictive value of CEACAM1 expression is superior to that of tumor thickness.

CONCLUSION

Expression of the cell adhesion molecule CEACAM1 in the primary tumors in melanoma patients is associated with the subsequent development of metastatic disease. This raises the possibility of a functional role for this cell adhesion molecule in the metastatic spread it indicates.

Can immunohistochemical markers and mitotic rate improve prognostic precision in patients with primary melanoma?

BACKGROUND

In addition to tumor thickness, several other prognostic parameters have been identified in primary human melanomas. Some are available readily (localization, gender, age, and ulceration). Others must be evaluated with a moderate or even substantial amount of work (mitoses and immunohistochemical markers). This study was undertaken to determine whether this extra effort is justified because it actually improves the precision of prognostic statements.

METHODS

Immunohistologic markers were determined on frozen sections from 691 biopsies of human melanomas with the immunoperoxidase method. Univariate and multivariate Cox regression analyses were performed with metastases and with death as endpoints.

RESULTS

Fifteen parameters were related to disease free survival in univariate Cox regression analysis: tumor thickness, ulceration, localization, gender, age, mitoses, and the immunohistochemical markers very late antigen (VLA)-2, human leukocyte antigen (HLA)-ABC, HLA-DR, NKI-beteb, Mel 14, intercellular adhesion molecule (ICAM-1), K-1-2, G-7-E2, and H-2-4-7. Three of the easily available parameters exhibited independent significance in multivariate Cox regression analysis: tumor thickness, ulceration, and localization. If mitotic rate was included in this model, then it had independent prognostic significance but ulceration was no longer significant. However, the model that included tumor thickness, localization, and ulceration had a slightly higher overall chi-square test score, indicating a better performance compared with thickness, localization, and mitoses. The model that included tumor thickness, localization, and mitoses could not be improved by any of the immunohistochemical markers in this study.

CONCLUSIONS

Nine immunohistochemical markers with established prognostic significance for primary human melanoma were not found to improve a prognostic model that included tumor thickness, localization, and mitoses. If mitoses was replaced by ulceration, then the model performed slightly better, although ulceration was not significant in the presence of mitoses.