Preoperative prediction of hepatocellular carcinoma tumour grade and micro-vascular invasion by means of artificial neural network: a pilot study.

BACKGROUND & AIMS: Hepatocellular carcinoma (HCC) prognosis strongly depends upon nuclear grade and the presence of microscopic vascular invasion (MVI). The aim of this study was to develop an artificial neural network (ANN) that is able to predict tumour grade and MVI on the basis of non-invasive variables. METHODS: Clinical, radiological, and histological data from 250 cirrhotic patients resected (n=200) or transplanted (n=50) for HCC were analyzed. ANN and logistic regression models were built on a training group of 175 randomly chosen patients and tested on the remaining testing group of 75. Receiver operating characteristics curve (ROC) and k-statistics were used to analyze model accuracy in the prediction of the final histological assessment of tumour grade (G1-G2 vs. G3-G4) and MVI (absent vs. present). RESULTS: Pathologic examination showed G3-G4 in 69.6% of cases and MVI in 74.4%. Preoperative serum alpha-fetoprotein (AFP), tumour number, size, and volume were related to tumour grade and MVI (p<0.05) and were used for ANN building, whereas, tumour number did not enter into the logistic models. In the training group, ANN area under ROC curves (AUC) for tumour grade and MVI prediction were 0.94 and 0.92, both higher (p<0.001) than those of logistic models (0.85 for both). In the testing group, ANN correctly identified 93.3% of tumour grades (k=0.81) and 91% of MVI (k=0.73). Logistic models correctly identified 81% of tumour grades (k=0.55) and 85% of MVI (k=0.57). CONCLUSION: ANN identifies HCC tumour grades and MVI on the basis of preoperative variables more accurately than the conventional linear model and should be used for tailoring clinical management.

Diagnostic role of circulating free plasma DNA detection in patients with localized prostate cancer.

To analyze the potential diagnostic relevance of free plasma DNA (FPDNA), we enrolled 64 patients with localized prostate cancer (CaP). FPDNA was quantified by real-time polymerase chain reaction assessment of the HTERT gene in blood samples from 64 patients with CaP and 45 healthy males. Methylation of the GSTP1 gene was used to confirm the neoplastic origin of FPDNA in selected cases. The mean +/- SD levels of FPDNA were higher in patients with CaP (15.4 +/- 10.9 ng/mL) than in control subjects (5.5 +/- 3.5 ng/mL; P <.001). By using the best cutoff value, the sensitivity of the test was 80%, the specificity was 82%, the area under the receiver operating characteristic curve, 0.881. High FPDNA values were significantly associated with pathologic T3 stage (P = . 035). Methylation of the GSTP1 gene was found in 4 (25%) of 16 FPDNA samples and 15 (94%) of 16 tissue samples. Quantification of FPDNA discriminates between patients with CaP and healthy subjects and correlates with pathologic tumor stage. FPDNA is a candidate biomarker for early diagnosis and monitoring of CaP.

Quantification of free plasma DNA before and after chemotherapy in patients with advanced epithelial ovarian cancer.

OBJECTIVES: A nonrandomized trial was planned to investigate the role of free plasma DNA (FPDNA) in patients with epithelial ovarian cancer before and after chemotherapy. Twenty-two patients with advanced stage ovarian cancer not suitable for debulking were treated with a neoadjuvant platinum/taxanes chemotherapy. Patients with clinical complete or partial response underwent radical hystero-oophorectomy, omentectomy, and lymphadenectomy and were followed up every 3 to 6 months. METHODS: Blood samples were obtained from each patient before chemotherapy, before each cycle, before and after surgery. FPDNA was quantified by real-time quantitative polymerase chain reaction using the Quantifiler Human Quantification Kit and expressed in ng/mL. Fifty female healthy blood donor volunteers were used as controls. RESULTS: Median FPDNA quantities discriminated between patients before chemotherapy (29.6+/-22.7 ng/mL) and controls (6.4+/-4.0 ng/mL) using a 14.5 ng/mL cutoff with 77% sensitivity and 96% specificity (P<0.001). Mean FPDNA concentrations significantly decreased after chemotherapy (17.9+/-14.5 ng/mL, P=0.001). A peak of FPDNA levels (66.2+/-45.2 ng/mL) was observed in association with surgery (P<0.001). Median follow-up and median progression-free survival time were 13.4+/-5.1 and 11.7+/-5.6 months, respectively. Eight patients with FPDNA values above the cutoff after chemotherapy showed disease progression or died, whereas 7 patients with FPDNA below the cutoff were free from disease. Patients with FPDNA levels above and below the cutoff showed significantly separated progression-free survival curves (P=0.007, log-rank test). CONCLUSIONS: FPDNA quantification significantly discriminates between cancer patients and controls and correlates with response to chemotherapy. Although performed in a limited series, we demonstrated a correlation between FPDNA values and clinical behavior of ovarian cancer patients.

Genomic allelotyping for distinction of recurrent and de novo hepatocellular carcinoma after orthotopic liver transplantation.

Distinction between recurrent and de novo hepatocellular carcinoma (HCC) after orthotopic liver transplantation (OLT) bears important clinical and therapeutic implications. Techniques for molecular profiling of clinically suspected de novo and recurrent HCC are required since the histological/clinical discrimination of donor vs. recipient tumor origin is difficult. Multiple PCR amplification of 16 highly polymorphic short tandem repeat (STR) DNA sequences (routinely used for paternity and forensic assays) was applied in two patients who developed a second HCC after OLT. In both patients the technique provided reliable evidence that the two second HCC were recurrences of the primary tumor. Multiple STR genetic allelotyping is an effective tool for clear-cut discrimination of donor/recipient origin of a second HCC after OLT. Its application could be of great therapeutic relevance for such OLT patients.

Application of a fluorescent PCR method for molecular diagnosis of posttransplant lymphoproliferative disorders on routine tissue sections.

Molecular detection of monoclonality can play an important role in the diagnosis of posttransplantation lymphoproliferative disorders (PTLD). To permit accurate molecular diagnosis of PTLD even on very small amounts of DNA extracted from routinely embedded histologic material, we adapted a commercially available PCR protocol (for FR-1, -2 and -3 regions), originally designed for use on fresh/frozen samples. We applied this approach on routine biopsy/surgical material of 10 PTLD (from nine patients). All three FR regions were always amplified, indicating that the extracted DNA was of medium quality. All five PTLD morphologically classified as lymphomas were monoclonal in at least one FR region. Thus, using the WHO histologic, immunohistochemical, and clinical criteria as the reference standard, the approach provided 100% sensitivity for detection of monoclonal malignancies, supporting the validity of the method. Of five specimens classified morphologically as polymorphic PTLD, three displayed a solitary IgH gene rearrangement peak, consistent with the presence of a monoclonal B-cell population (ie, monoclonal polymorphic PTLD). This rapid and straightforward procedure, which allows identification of a wide range of IgH rearrangements, could facilitate molecular analysis of PTLD in routine practice, while limiting consumption of valuable diagnostic material.

Real time RT-PCR approach for the evaluation of ERBB2 overexpression in breast cancer archival samples: a comparative study with FISH, SISH, and immunohistochemistry.

We tested the reliability of real time reverse transcription polymerase chain reactions as an alternative method for the assessment of ERBB2 status in paraffin-embedded tissues of 83 patients with breast cancer and 20 non-neoplastic controls. PCR was also compared with the immunohistochemical (IHC) HercepTest score and with fluorescence (FISH) and silver (SISH) in-situ hybridization, in 42 selected cases. ERBB2 mRNA was overexpressed in 26/83 (31%) breast cancer samples, using a cutoff calculated as the mean value of the controls plus 3 SD or with the receiver operating curve. The PCR test showed a 96% sensitivity and a 100% specificity when compared with FISH, with an area under the receiver operating curve of 98.4%. Overexpression of ERBB2 at PCR was also significantly correlated with amplification in FISH (P<0.001, Mann-Whitney test) and in SISH (P<0.001, Mann-Whitney test), and with the IHC HercepTest scores 2 or 3 (P<0.001, Spearman rank correlation). FISH, SISH, and IHC were also compared with each other. ERBB2 amplification in FISH significantly correlated with that in SISH (P=0.002, chi test with a concordance of the 87%), but not with IHC HercepTest scores (P=0.214, chi test). Real time PCR is a reliable and cost-effective method for the assessment of ERBB2 status in archival breast cancer samples, compared with FISH. Its introduction in routine diagnostic pathology practice is feasible even if it requires amendments to the current clinical oncology protocols.