Support vector machine-based nomogram predicts postoperative distant metastasis for patients with oesophageal squamous cell carcinoma

Background:

We aim to develop effective models for predicting postoperative distant metastasis for oesophageal squamous cell carcinoma (OSCC) for the purpose of guiding tailored therapy.

Methods:

We used data from two centres to establish training (n=319) and validation (n=164) cohorts. All patients underwent curative surgical treatment. The clinicopathological features and 23 immunomarkers detected by immunohistochemistry were involved for variable selection. We constructed eight support vector machine (SVM)-based nomograms (SVM1–SVM4 and SVM1'–SVM4'). The nomogram constructed with the training cohort was tested further with the validation cohort.

Results:

The outcome of the SVM1 model in predicting postoperative distant metastasis was as follows: sensitivity, 44.7% specificity, 90.9% positive predictive value, 81.0% negative predictive value, 65.6% and overall accuracy, 69.5%. The corresponding outcome of the SVM2 model was as follows: 44.7%, 92.1%, 82.9%, 65.9%, and 70.1%, respectively. The corresponding outcome of the SVM3 model was as follows: 55.3%, 93.2%, 87.5%, 70.7%, and 75.6%, respectively. The SVM4 model was the most effective nomogram in prediction, and the corresponding outcome was as follows: 56.6%, 97.7%, 95.6%, 72.3%, and 78.7%, respectively.Similar results were observed in SVM1', SVM2', SVM3', and SVM4', respectively.

Conclusion:

The SVM-based models integrating clinicopathological features and molecular markers as variables are helpful in selecting the patients of OSCC with high risk of postoperative distant metastasis.

Quantitative evaluation of viable tissue perfusion changes with contrast-enhanced greyscale ultrasound in a mouse hepatoma model following treatment with different doses of thalidomide

OBJECTIVE

This study aimed to quantify intratumoural viable tissue perfusion with contrast-enhanced greyscale ultrasound to evaluate tumour response to anti-angiogenic treatment.

METHODS

H22 hepatoma-bearing mice were treated with low-dose thalidomide (Group B), high-dose thalidomide (Group C) or 0.5% carboxylmethylcellulose (Group A). Contrast-enhanced greyscale ultrasound was performed after 7 days of treatments to evaluate the percentage of non-enhanced area for each tumour; regions of interest within the enhanced area were analysed offline to determine the area under the curve (AUC), maximum intensity (IMAX), perfusion index (PI), mean transit time (MTT), time to peak (TTP) and quality of fit (QOF). Immunohistochemical analysis was performed for evaluation of microvascular density (MVD).

RESULTS

The percentage of non-enhanced area was significantly larger in Group C than in Groups A and B (p<0.05); however, there was no significant difference between Groups A and B. Treatment with thalidomide resulted in a significant decrease in AUC, PI and IMAX compared with Group A (p<0.05). Immunohistochemistry showed significant decreases in MVD in Groups B and C compared with Group A (p<0.05); however, there was no significant difference in MVD between Groups B and C. MVD was positively correlated with IMAX (r = 0.419, p = 0.023) and PI (r = 0.455, p = 0.013).

CONCLUSION

Quantitatively analysing intratumoural viable tissue perfusion enables early evaluation of tumour response to anti-angiogenic therapy before apparent changes in tumour necrosis.

Evaluation of epidermal growth factor receptor (EGFR) in nasopharyngeal carcinoma and lymph node metastases: Implications for EGFR targeted therapy

16510

Background: Epidermal growth factor receptor (EGFR) overexpression is proved to be an independent predictor of poor clinical outcome in Nasopharyngeal Carcinoma (NPC). Expression level of EGFR is also important for EGFR-Targeted therapy but yet unexplored in the primary NPC tumors and lymph node metastases. Aim of this study was to correlate EGFR expression levels in primary tumors and lymph node metastases in NPC to find out whether assessing EGFR status on primary cancer is an effective tool for planning EGFR targeted therapy. Methods: Archived tissue samples of primary tumor and corresponding lymph node metastases of 86 NPC patients (67 males, 19 females; median age = 52 years) diagnosed between 1997 and 2006 were collected from Sun Yat-sen University Cancer Center. Immunohistochemical evaluation of EGFR expression in primary tumors and related lymph node metastasis was performed. According to the UICC 2002 staging system for nasopharyngeal carcinoma, stages Iib, III, IVa, IVb and IVc included 43, 11, 3, 27 and 2 cases respectively. Cervical and supraclavicular lymph nodes were taken in 66 (76.7%) and 20 patients (23.3%) respectively. EGFR status was defined as positive if the percentage of malignant cells stained were =1%. The SPSS 10.0 system was used for statistical analysis and Chi- square test was used to assess the correlation between the EGFR expression in the primary and metastatic tumors. Results: EGFR status was positive in 63 primary tumors (73.2%) and 52 metastatic lymph nodes(60.5%). In 18 EGFR expressing primary tumors (28.6%), the corresponding metastatic lymph node site was negative whereas it was found positive in 7 metastic lymph node (13.5%) of EGFR-negative primary tumors. The difference between these two groups (ie, EGFR-positive to EGFR-negative vs. EGFR-negative to EGFR-positive) was statistically significant (P=0.001). Conclusions: High levels of EGFR protein are frequently observed in NPC and represent potential therapeutic target. However, equivalent expression is not displayed by metastatic nodes, which may limit the effectiveness of anti- EGFR treatment in advanced diseases. These findings have to be considered in future prospective studies.

No significant financial relationships to disclose.

E2F-HDAC complexes negatively regulate the tumor suppressor gene ARHI in breast cancer

ARHI is a maternally imprinted tumor suppressor gene whose expression is markedly downregulated in breast cancer. Reactivation of ARHI expression in breast cancer cells is associated with increased histone H3 acetylation and decreased lysine 9 methylation of histone H3. An ARHI promoter segment that spanned bases -420 to +58 (designated the P2 region) exhibits significantly higher promoter activity in normal cells than in cancer cells. To better understand the molecular mechanisms contributing to this differential transcriptional activity, we sought to identify transcription factors that bind to the P2 region of the ARHI promoter and regulate its activity. Sequence analysis and oligonucleotide competition in electrophoretic mobility shift assays identified an A2 fragment containing an E2F-binding site. Using specific antibodies in supershift assays, we have shown that anti-E2F1 and 4 antibodies can supershift the A2-protein complexes, whereas anti-E2F2 and 6 antibodies cannot, demonstrating that the A2 fragment interacts with specific members of the E2F family proteins. When compared with normal breast epithelial cells, breast cancer cells have significantly elevated expression of E2F1, 4 and increased E2F DNA-binding activity. Moreover, chromatin immunoprecipitation experiments revealed that both E2F1 and 4 bind to the ARHI promoter in breast cancer cells in vivo. This binding was reduced when the cells were treated with the histone deacetylase (HDAC) inhibitor--trichostatin A (TSA). When SKBr3 cells were cotransfected with an ARHI/luciferase reporter and E2F-expression vectors, E2F1 and 4 reduced ARHI promoter activity 2-3-fold, and this reduction could be reversed by TSA treatment. The negative regulation by E2F-HDAC complexes could also be reduced by small interfering RNA of E2F1 and 4. While the retinoblastoma protein, pRB, alone had no effect on ARHI promoter activity, repression by E2F1, but not E2F4, was enhanced by the coexpression of pRB. Taken together, our results suggest that E2F1, 4 and their complexes with HDAC play an important role in downregulating the expression of the tumor suppressor gene ARHI in breast cancer cells.

Genomic structure and promoter characterization of an imprinted tumor suppressor gene ARHI

We have recently identified a maternally imprinted tumor suppressor gene, ARHI (aplysia ras homolog I), the expression of which is lost in ovarian and breast cancers. We have now characterized the genomic structure of the gene including its promoter and the methylation status of its upstream CpG islands. The ARHI gene spans approximately 8 kb containing two exons and one intron. Exon 1 contains 81 non-translated nucleotides, connected to exon 2 with a 3.2-kb intron. The entire protein-coding region is located within exon 2 and encodes a 229-residue small GTP-binding protein belonging to the Ras superfamily. Genomic structure analysis has identified three potential CpG islands. Two of them (CpG island I and II) are located within the promoter and adjacent exon 1 of the ARHI gene. Aberrant methylation of these CpG islands has been detected in breast cancer cells but not in normal epithelial cells, supporting the possibility that appropriate methylation status of the CpG islands in the promoter region may play a role in the downregulation of ARHI gene expression. A TATA box is found 27 bp upstream of the transcription start site associated with several putative transcription factor binding sites. Transient transfection with nested deletion constructs of the 2-kb ARHI promoter regions fused to a luciferase reporter indicated a 121-bp sequence upstream of the transcription initiation site is required for basal promoter activity. Interestingly, this is the region where lower promoter activity has been observed in cancer cells than in normal cells.

Mechanism of transmembrane signaling: insulin binding and the insulin receptor

Transmembrane signaling via receptor tyrosine kinases generally requires oligomerization of receptor monomers, with the formation of ligand-induced dimers or higher multimers of the extracellular domains of the receptors. Such formations are expected to juxtapose the intracellular kinase domains at the correct distances and orientations for transphosphorylation. For receptors of the insulin receptor family that are constitutively dimeric, or those that form noncovalent dimers without ligands, the mechanism must be more complex. For these, the conformation must be changed by the ligand from one that prevents activation to one that is permissive for kinase phosphorylation. How the insulin ligand accomplishes this action has remained a puzzle since the discovery of the insulin receptor over 2 decades ago, primarily because membrane proteins in general have been refractory to structure determination by crystallography. However, high-resolution structural evidence on individual separate subdomains of the insulin receptor and of analogous proteins has been obtained. The recently solved quaternary structure of the complete dimeric insulin receptor in the presence of insulin has now served as the structural envelope into which such individual domains were fitted. The combined structure has provided answers on the details of insulin/receptor interactions in the binding site and on the mechanism of transmembrane signaling of this covalent dimer. The structure explains many observations on the behavior of the receptor, from greater or lesser binding of insulin and its variants, point and deletion mutants of the receptor, to antibody-binding patterns, and to the effects on basal and insulin-stimulated autophosphorylation under mild reducing conditions.

The human ARHI tumor suppressor gene inhibits lactation and growth in transgenic mice

ARHI is a novel imprinted tumor suppressor gene. To study its function in vivo, we have developed transgenic mice that overexpress ARHI. Offspring bearing the transgene had significantly lower body weights than did nontransgenic littermates. In addition, strong expression of the ARHI transgene was associated with greatly impaired mammary gland development and lactation, failure of ovarian folliculogenesis resulting in decreased fertility, loss of neurons in the cerebellar cortex, and impaired development of the thymus. Decrease in body size and defects in the mammary glands correlated with the level of transgene expression. Immunohistochemical analysis indicated that expression of prolactin (PRL), but not growth hormone, was lower in the pituitary glands of mice with defective mammary gland development. The defect in pregnancy-associated mammary tissue proliferation was associated with decreased serum PRL and progesterone levels. Moreover, lower levels of estrogen receptor and progesterone receptor were observed in postpartum mammary glands and in the ovaries of mice that overexpressed ARHI. Our data suggest that ARHI can inhibit PRL secretion and act as a negative regulator in murine growth and development.

Quaternary structure of the insulin-insulin receptor complex

The three-dimensional (3D) structure of the intrinsically dimeric insulin receptor bound to its ligand, insulin, was determined by electron cryomicroscopy. Gold-labeled insulin served to locate the insulin-binding domain. The 3D structure was then fitted with available known high-resolution domain substructures to obtain a detailed contiguous model for this heterotetrameric transmembrane receptor. The 3D reconstruction indicates that the two alpha subunits jointly participate in insulin binding and that the kinase domains in the two beta subunits are in a juxtaposition that permits autophosphorylation of tyrosine residues in the first step of insulin receptor activation.