David Boulware

Correlation of Osteopontin protein expression and pathological stage across a wide variety of tumor histologies

Purpose: Osteopontin (OPN) is an integrin-binding protein overexpressed in various experimental models of malignancy and appears to be involved in tumorigenesis and metastasis. Although various studies have assessed OPN protein levels in several tumor types, a broad survey of OPN expression in human neoplasia under the same experimental conditions has not been carried out. Experimental Design: We used immunohistochemistry to detect OPN in a selection of 350 human tumors and 113 normal tissues, from a variety of body sites, using stage-oriented human cancer tissue arrays. Tumors included malignancies from breast (26), ovary (22), endometrium (14), esophagus (10), stomach (11), pancreas (16), bile duct (1), liver (9), colon (20), kidney (53), bladder (33), prostate (28), head and neck (60), salivary glands (14), lung (17), skin (6), and brain (10). Results: High cytoplasmic OPN staining was observed in 100% of gastric carcinomas, 85% of colorectal carcinomas, 82% of transitional cell carcinomas of the renal pelvis, 81% of pancreatic carcinomas, 72% of renal cell carcinomas, 71% of lung and endometrial carcinomas, 70% of esophageal carcinomas, 58% of squamous cell carcinomas of the head and neck, and 59% of ovarian carcinomas. Although OPN expression was identified in a good number of bladder, prostate, and brain tumors, the majority of 6 skin cancers, 11 of 14 salivary gland cancers, 2 thyroid carcinomas, and 23 of 26 breast cancers revealed low OPN positivity or were negative. When considering all sites, OPN expression significantly correlated with tumor stage (Spearman’s correlation coefficient, P = 0.0002). OPN score and stage were also significantly correlated for specific cancer sites including bladder (P = 0.01), colon (P = 0.004), kidney (P = 0.0001), larynx (P = 0.035), mouth (P = 0.046), and salivary gland (P = 0.011). Conclusions: This study reports the broad distribution of OPN in human tumors from different body sites, suggesting involvement of this protein in tumor formation. The strong correlation between pathological stage and OPN across multiple tumor types suggests a role for OPN in tumor progression.

Factors Affecting Morbidity, Mortality, and Survival in Patients Undergoing Ivor Lewis Esophagogastrectomy

OBJECTIVES To examine the safety of transthoracic esophagogastrectomy (TTE) in a multidisciplinary cancer center and to determine which clinical parameters influenced survival and the rates of death and complications. SUMMARY BACKGROUND DATA Although the incidence of cancer at the gastroesophageal junction has been rising rapidly in the United States, controversy still exists about the safety of surgical procedures designed to remove the distal esophagus and proximal stomach. Alternatives to TTE have been proposed because of the reportedly high rates of death and complications associated with the procedure. METHODS Data from 143 patients treated by TTE by one author (1989-1999) were entered into a computerized database. Preoperative clinical parameters were tested for effect on death, complications, and survival. RESULTS The patient population consisted of 127 men and 16 women. One hundred twenty-one patients had a history of tobacco abuse, and 118 reported the regular ingestion of alcohol. One hundred fifteen patients had adenocarcinoma, 16 had squamous cell cancer, 6 had another form of esophageal tumor, and 6 had high-grade dysplasia associated with Barrett epithelia. Fifty-six patients had adenocarcinomas arising in Barrett epithelium. Twenty-eight patients were treated with neoadjuvant chemoradiation before surgery. Three patients died within 30 days of surgery (mortality rate 2.1%). Five patients (3.5%) had a documented anastomotic leak; three died). Overall, 42 patients had complications (29%). Twenty-six had pulmonary complications (19%). The mean length of stay in the intensive care unit was 3.35 days; the mean hospital length of stay was 13.54 days. The overall 3-year survival rate was 29.6%. CONCLUSIONS A high ASA score and the development of complications predicted an increased length of stay. The presence of diabetes predicted the development of complication and an increased length of stay. None of the other parameters tested predicted perioperative death or complications. Only disease stage, diabetes, and blood transfusion affected overall survival. From these results with a large series of patients with gastroesophageal junction cancers, TTE can be performed with a low death rate (2.1%), a low leak rate (3. 5%), and an acceptable complication rate (29%).

A comprehensive geriatric intervention detects multiple problems in older breast cancer patients

UNLABELLED Studies of comprehensive geriatric assessment (CGA) have shown the importance of follow-up for effectiveness, but this has not been tested in an oncology clinic. In this pilot study, we enrolled 15 early breast cancer patients, aged 70 and older. They received a multidisciplinary CGA every 3 months and structured follow-up from the SAOP nurse practitioner, dietitian, social worker, and pharmacist according to risk. Total follow-up was 6 months. Median age of evaluable patients was 79 years (range 72-87). Median number of comorbidities by Cumulative Index Rating Scale-Geriatric (CIRS-G) was 5 (3-9) at baseline. Ten patients were at pharmacological risk, five at psychosocial risk, and eight at nutritional risk. Patients presented on average six problems initially, and three new problems during follow-up. The intervention directly influenced oncological treatment in four cases. It ensured continuity/coordination of care in seven cases. Success rate in addressing problems was 87%. Mean Functional Assessment of Cancer Treatment-Breast (FACT-B) scores improved from 110.5 (S.D. 16.7) to 116.3 (S.D. 16.5) (t=0.025). Function and independence were maintained. CONCLUSIONS Older patients with early breast cancer have a high prevalence of comorbidity. A CGA with follow-up has potential for improving the treatment and prognosis of these patients and is feasible in an academic oncology setting.

Multi-Platform, Multi-Site, Microarray-Based Human Tumor Classification

The introduction of gene expression profiling has resulted in the production of rich human data sets with potential for deciphering tumor diagnosis, prognosis, and therapy. Here we demonstrate how artificial neural networks (ANNs) can be applied to two completely different microarray platforms (cDNA and oligonucleotide), or a combination of both, to build tumor classifiers capable of deciphering the identity of most human cancers. First, 78 tumors representing eight different types of histologically similar adenocarcinoma, were evaluated with a 32k cDNA microarray and correctly classified by a cDNA-based ANN, using independent training and test sets, with a mean accuracy of 83%. To expand our approach, oligonucleotide data derived from six independent performance sites, representing 463 tumors and 21 tumor types, were assembled, normalized, and scaled. An oligonucleotide-based ANN, trained on a random fraction of the tumors (n = 343), was 88% accurate in predicting known pathological origin of the remaining fraction of tumors (n = 120) not exposed to the training algorithm. Finally, a mixed-platform classifier using a combination of both cDNA and oligonucleotide microarray data from seven performance sites, normalized and scaled from a large and diverse tumor set (n = 539), produced similar results (85% accuracy) on independent test sets. Further validation of our classifiers was achieved by accurately (84%) predicting the known primary site of origin for an independent set of metastatic lesions (n = 50), resected from brain, lung, and liver, potentially addressing the vexing classification problems imposed by unknown primary cancers. These cDNA- and oligonucleotide-based classifiers provide a first proof of principle that data derived from multiple platforms and performance sites can be exploited to build multi-tissue tumor classifiers.

Randomized Phase III Trial of Gemcitabine-Based Chemotherapy With In Situ RRM1 and ERCC1 Protein Levels for Response Prediction in Non–Small-Cell Lung Cancer

PURPOSE We evaluated the efficacy of gemcitabine versus gemcitabine and carboplatin in patients with advanced non-small-cell lung cancer (NSCLC) and a performance status (PS) of 2 and assessed if tumoral RRM1 and ERCC1 protein levels are predictive of response to therapy. PATIENTS AND METHODS A randomized phase III trial was conducted in community-based oncology practices. Tumor specimens were collected a priori and shipped to a single laboratory for blinded determination of in situ RRM1 and ERCC1 protein expression levels by an automated quantitative immunofluorescent-based technology. RESULTS One hundred seventy patients were randomly assigned. Overall median survival was 5.1 months for gemcitabine and 6.7 months for gemcitabine and carboplatin (P = .24). RRM1 (range, 5.3 to 105.6; median, 34.1) and ERCC1 (range, 5.2 to 131.3; median, 34.7) values were significantly and inversely correlated with disease response (r = -0.41; P = .001 for RRM1; r = -0.39; P = .003 for ERCC1; ie, response was better for patients with low levels of expression). A model for response prediction that included RRM1, ERCC1, and treatment arm, was highly predictive of the treatment response observed (P = .0005). We did not find statistically significant associations between survival and RRM1 or ERCC1 levels. CONCLUSION Single-agent chemotherapy remains the standard of care for patients with advanced NSCLC and poor PS. Quantitative analysis of RRM1 and ERCC1 protein expression in routinely collected tumor specimens in community oncology practices is predictive of response to gemcitabine and gemcitabine and carboplatin therapy. Oncologists should consider including in situ expression analysis for these proteins into their therapeutic decisions.

Fall in C-Peptide During First 2 Years From Diagnosis: Evidence of at Least Two Distinct Phases From Composite Type 1 Diabetes TrialNet Data

Interpretation of clinical trials to alter the decline in β-cell function after diagnosis of type 1 diabetes depends on a robust understanding of the natural history of disease. Combining data from the Type 1 Diabetes TrialNet studies, we describe the natural history of β-cell function from shortly after diagnosis through 2 years post study randomization, assess the degree of variability between patients, and investigate factors that may be related to C-peptide preservation or loss. We found that 93% of individuals have detectable C-peptide 2 years from diagnosis. In 11% of subjects, there was no significant fall from baseline by 2 years. There was a biphasic decline in C-peptide; the C-peptide slope was −0.0245 pmol/mL/month (95% CI −0.0271 to −0.0215) through the first 12 months and −0.0079 (−0.0113 to −0.0050) from 12 to 24 months (P < 0.001). This pattern of fall in C-peptide over time has implications for understanding trial results in which effects of therapy are most pronounced early and raises the possibility that there are time-dependent differences in pathophysiology. The robust data on the C-peptide obtained under clinical trial conditions should be used in planning and interpretation of clinical trials.

Fall in C-peptide During First 2 Years From Diagnosis Evidence of at Least Two Distinct Phases From Composite TrialNet Data

Interpretation of clinical trials to alter the decline in β-cell function after diagnosis of type 1 diabetes depends on a robust understanding of the natural history of disease. Combining data from the Type 1 Diabetes TrialNet studies, we describe the natural history of β-cell function from shortly after diagnosis through 2 years post study randomization, assess the degree of variability between patients, and investigate factors that may be related to C-peptide preservation or loss. We found that 93% of individuals have detectable C-peptide 2 years from diagnosis. In 11% of subjects, there was no significant fall from baseline by 2 years. There was a biphasic decline in C-peptide; the C-peptide slope was −0.0245 pmol/mL/month (95% CI −0.0271 to −0.0215) through the first 12 months and −0.0079 (−0.0113 to −0.0050) from 12 to 24 months (P < 0.001). This pattern of fall in C-peptide over time has implications for understanding trial results in which effects of therapy are most pronounced early and raises the possibility that there are time-dependent differences in pathophysiology. The robust data on the C-peptide obtained under clinical trial conditions should be used in planning and interpretation of clinical trials.

Significance of Sentinel Lymph Node Micrometastases in Human Breast Cancer

BACKGROUND The significance of micrometastatic disease in the sentinel lymph nodes (SLN) of patients with invasive breast cancer has been questioned. The objective of our study was to review the impact of micrometastatic carcinoma detected by SLN biopsy. STUDY DESIGN Between January 1997 and May 2004, 2,408 patients with invasive breast cancer and an SLN with micrometastatic (N0[i+], N1mi) or no metastatic (N0[i-]) disease were identified through our breast database. Slide review was performed and reclassified by the 6(th) edition of the American Joint Committee on Cancer Staging Manual. Of these, 27 were excluded from analysis because of evidence of macrometastatic disease on slide review or enrollment in the American College of Surgeons Oncology Group Z10 study. RESULTS Of 2,381 patients, 2,108 were N0(i-), 151 were N0(i+), and 122 were N1mi. Overall and disease-free survivals of patients with an N1mi SLN were substantially worse than those in patients with an N0(i-) SLN (p < 0.001 and p=0.006, respectively). Additional positive non-SLNs were identified in 15.5% (15 of 97) of N1mi patients and 9.3% (10 of 107) of N0(i+) patients undergoing completion axillary lymph node dissection. Overall survival of the N0(i+) SLN patients not undergoing axillary dissection was substantially less than those undergoing axillary dissection (p=0.02). CONCLUSIONS Detection of micrometastatic carcinoma (N1mi) in the SLNs of invasive breast cancer patients is a major indicator of poorer survival compared with N0(i-) patients. Although survival of patients with an N0(i+) SLN does not statistically differ from that of N0(i-) patients, 9.3% of these patients had additional axillary nodal disease on axillary dissection, and N0(i+) patients had a decreased survival when axillary dissection was omitted.

A Gene Expression Model of Intrinsic Tumor Radiosensitivity: Prediction of Response and Prognosis After Chemoradiation

PURPOSE Development of a radiosensitivity predictive assay is a central goal of radiation oncology. We reasoned a gene expression model could be developed to predict intrinsic radiosensitivity and treatment response in patients. METHODS AND MATERIALS Radiosensitivity (determined by survival fraction at 2 Gy) was modeled as a function of gene expression, tissue of origin, ras status (mut/wt), and p53 status (mut/wt) in 48 human cancer cell lines. Ten genes were identified and used to build a rank-based linear regression algorithm to predict an intrinsic radiosensitivity index (RSI, high index = radioresistance). This model was applied to three independent cohorts treated with concurrent chemoradiation: head-and-neck cancer (HNC, n = 92); rectal cancer (n = 14); and esophageal cancer (n = 12). RESULTS Predicted RSI was significantly different in responders (R) vs. nonresponders (NR) in the rectal (RSI R vs. NR 0.32 vs. 0.46, p = 0.03), esophageal (RSI R vs. NR 0.37 vs. 0.50, p = 0.05) and combined rectal/esophageal (RSI R vs. NR 0.34 vs. 0.48, p = 0.001511) cohorts. Using a threshold RSI of 0.46, the model has a sensitivity of 80%, specificity of 82%, and positive predictive value of 86%. Finally, we evaluated the model as a prognostic marker in HNC. There was an improved 2-year locoregional control (LRC) in the predicted radiosensitive group (2-year LRC 86% vs. 61%, p = 0.05). CONCLUSIONS We validate a robust multigene expression model of intrinsic tumor radiosensitivity in three independent cohorts totaling 118 patients. To our knowledge, this is the first time that a systems biology-based radiosensitivity model is validated in multiple independent clinical datasets.

Systems Biology Modeling of the Radiation Sensitivity Network: A Biomarker Discovery Platform

PURPOSE The discovery of effective biomarkers is a fundamental goal of molecular medicine. Developing a systems-biology understanding of radiosensitivity can enhance our ability of identifying radiation-specific biomarkers. METHODS AND MATERIALS Radiosensitivity, as represented by the survival fraction at 2 Gy was modeled in 48 human cancer cell lines. We applied a linear regression algorithm that integrates gene expression with biological variables, including ras status (mut/wt), tissue of origin and p53 status (mut/wt). RESULTS The biomarker discovery platform is a network representation of the top 500 genes identified by linear regression analysis. This network was reduced to a 10-hub network that includes c-Jun, HDAC1, RELA (p65 subunit of NFKB), PKC-beta, SUMO-1, c-Abl, STAT1, AR, CDK1, and IRF1. Nine targets associated with radiosensitization drugs are linked to the network, demonstrating clinical relevance. Furthermore, the model identified four significant radiosensitivity clusters of terms and genes. Ras was a dominant variable in the analysis, as was the tissue of origin, and their interaction with gene expression but not p53. Overrepresented biological pathways differed between clusters but included DNA repair, cell cycle, apoptosis, and metabolism. The c-Jun network hub was validated using a knockdown approach in 8 human cell lines representing lung, colon, and breast cancers. CONCLUSION We have developed a novel radiation-biomarker discovery platform using a systems biology modeling approach. We believe this platform will play a central role in the integration of biology into clinical radiation oncology practice.