James MICHAELSON - Academia.edu (original) (raw)

Papers by James MICHAELSON

Breast Cancer Research, 2004

Therapeutic advances in cardiovascular disease, 2008

Thiazolidinediones are currently indicated for the treatment of Type 2 Diabetes. This class of dr... more Thiazolidinediones are currently indicated for the treatment of Type 2 Diabetes. This class of drugs has been associated with several adverse reactions associated with volume overload. This report describes a case of a 65 year old African-American female with a history of hypertension and obesity, and taking rosiglitazone (Avandia) for her Type 2 Diabetes whose evaluation for chest pain resulted in the incidental finding of pulmonary hypertension noted on echocardiogram. The subsequent evaluation, follow-up and treatment are discussed along with potential pitfalls and implications for clinical care.

Protein Engineering, Design and Selection, 2010

Single-chain Fvs (scFvs) are commonly used building blocks for creating engineered diagnostic and... more Single-chain Fvs (scFvs) are commonly used building blocks for creating engineered diagnostic and therapeutic antibody molecules. Bispecific antibodies (BsAbs) hold particular interest due to their ability to simultaneously bind and engage two distinct targets. We describe a technology for producing stable, scalable IgG-like bispecific and multivalent antibodies based on methods for rapidly engineering thermally stable scFvs. Focused libraries of mutant scFvs were designed using a combination of sequence-based statistical analyses and structure-, and knowledge-based methods. Libraries encoding these designs were expressed in E. coli and culture supernatants-containing soluble scFvs screened in a highthroughput assay incorporating a thermal challenge prior to an antigen-binding assay. Thermally stable scFvs were identified that retain full antigen-binding affinity. Single mutations were found that increased the measured T m of either the V H or V L domain by as much as 148 8 8 8 8C relative to the wild-type scFv. Combinations of mutations further increased the T m by as much as an additional 128 8 8 8 8C. Introduction of a stability-engineered scFv as part of an IgG-like BsAb enabled scalable production and purification of BsAb with favorable biophysical properties.

Breast Cancer Research and Treatment, Feb 15, 2011

British Journal of Cancer, Nov 1, 2005

There has been much uncertainty as to whether metastasis requires mutation at the time of spread.... more There has been much uncertainty as to whether metastasis requires mutation at the time of spread. Here, we use clinical data to calculate the probability of the spread of melanoma and breast cancer cells. These calculations reveal that the probability of the spread of cancer cells is relatively high for small tumours (B1 event of spread for every 500 cells for melanomas of 0.1 mm) and declines as tumours increase in size (B1 event of spread for every 10 8 cells for melanomas of 12 mm). The probability of spread of breast cancer cells from the lymph nodes to the periphery is B1 event of spread for every 10 8 cells in the nodal masses, which have a mean diameter of 5 mm, while the probability of spread of cancer cells from the breast to the periphery when the primary masses are 5 mm is also B1 event of spread for every 10 8 cells. Thus, the occurrence of an event of spread from the breast to the lymph nodes appears not to increase the propensity of the progeny of those cells to spread from the lymph nodes to the periphery. These values indicate that the spread of human breast cancer and melanoma cells is unlikely to occur by a mechanism requiring mutation at the time of spread.

Selected key pathways up-regulated or down-regulated by ACEi treatment alone in Gene Set Enrichme... more Selected key pathways up-regulated or down-regulated by ACEi treatment alone in Gene Set Enrichment Analysis.

Purpose: Angiotensin system inhibitors (ASI) can improve prognosis in multiple cancer types, incl... more Purpose: Angiotensin system inhibitors (ASI) can improve prognosis in multiple cancer types, including pancreatic ductal adenocarcinoma (PDAC). However, no study has examined the effect of ASIs alone or combined with adjuvant chemotherapy in resected PDAC patients. Experimental Design: We performed an analysis of the records of ASI users and nonuser patients with PDAC seen at Massachusetts General Hospital (Boston, MA) between January 2006 and December 2010. To identify mechanisms of ASIs in PDAC, we performed RNA sequencing (RNA-Seq) of resected primary lesions. Results: A total of 794 consecutive patients were included. In 299 resected patients, ASI users experienced longer overall survival (OS) in both univariate (median OS, 36.3 vs. 19.3 months, P ¼ 0.011) and adjusted multivariate [HR, 0.505; 95% confidence interval (CI), 0.339-0.750; P ¼ 0.001] analyses. Propensity score-adjusted analysis also showed a longer median OS for chronic ASI users. In unresected patients, the beneficial effect of ASIs was significant in patients with locally advanced disease, but not in metastatic patients. RNA-Seq analysis revealed in tumors of ASI users (lisinopril) a normalized extracellular matrix, a reduced expression of genes involved in PDAC progression (e.g., WNT and Notch signaling), and an increased expression of genes linked with the activity of T cells and antigen-presenting cells. Finally, chronic use of ASI was associated with a gene expression signature that is predictive of survival in independent validation cohorts. Conclusions: In patients with nonmetastatic PDAC, chronic ASI use is associated with longer OS independently of chemotherapy. Our RNA-Seq analysis suggests that ASIs reduce the malignant potential of cancer cells and stimulate the immune microenvironment in primary PDAC. Clin Cancer Res; 23(19); 5959-69. Ó2017 AACR.

Blood, Nov 18, 2011

Abstract 4741 BACKGROUND. Patients with chronic lymphocytic leukemia (CLL) have an increased risk... more Abstract 4741 BACKGROUND. Patients with chronic lymphocytic leukemia (CLL) have an increased risk of developing second primary malignancies, including squamous cell cancers (SCC) of the Head and Neck (HNC). While cutaneous SCC have been shown to behave more aggressively in the setting of CLL, little information exists on the clinical course and outcome of SCC of the oral cavity and pharynx. METHODS. The Massachusetts General Hospital Tumor Registry was queried for all patients with ICD-O-3 codes corresponding to both CLL/SLL and HNC from January 1989 to July 2010, returning seventeen patients. The data collected included patient demographics, blood work and cytogenetics (when available) at the time of CLL diagnosis, primary HNC tumor histology and stage, and all pertinent treatment data, including types of treatment and clinical response. Survival was calculated using the Kaplan-Meier method. RESULTS. Seventeen patients had a diagnosis of both CLL and HNC. Of those, three were excluded from the study as treatment data was unavailable. The fourteen remaining included 12 males and 2 females. Eight patients had their CLL diagnosed first, and had a median interval to HNC diagnosis of 80 months (2-210 months; mean 89 months), four had their HNC diagnosed first and had a median interval to CLL of 1 month (1-100 months; mean 26 months), and two patients had concurrent diagnoses. The median age at diagnosis was 58.5 years (CLL) and 64 years (HNC). Nine patients (64%) were former or current smokers, all with at least a 20 pack year smoking history. Seven patients presented with Rai stage 0 CLL, three patients with Rai stage 1, one patient each with Rai stage 2 and 3, and two with unknown presentations. HNC tumor presentations included 3 T1 tumors, 3 T2 tumors, 3 T3 tumors, 2 T4 tumors, and 3 TX tumors. Three patients had HPV assessed in their tumor and all were positive. Out of the fourteen patients, nine (64%) received first-line chemotherapy for their CLL: 4 with chlorambucil, 1 with fludarabine/rituximab (FR), 1 with FR/lenalidomide, 1 with rituximab, and 1 each with cytoxan/vincristine and steroid monotherapy. Eight patients received second line therapy, three received third line therapy and one received 7 lines of therapy. Six patients (43%) received chemotherapy and radiation as their primary treatment for their Head and Neck tumor, three (21%) received surgery and radiation, three (21%) received only radiation, one received surgery, and one received radiation, chemotherapy, and surgery. One patient required hospitalization due to neutropenic fever during their radiation treatment, and one patient had a week of chemotherapy withheld due to leukopenia, but both were able to complete their full course of therapy. Three patients (21%) received chemotherapy for their CLL before any HNC treatment. These three patients were able to tolerate full HNC therapy without infectious complications or unanticipated cytopenias. First-line therapy for HNC had a median progression-free survival of 26.9 months (range 10.7 – 145.2 months; 95% CI, 12 to 42 months) and a median overall survival of 45.1 months (range 1 to 256.1 months; 95% CI, 9 to 81 months). The median overall survival from HNC for treated CLL patients was 22.6 months and for untreated 59.2 months (p= 0.596). Five patients died from their HNC (36%), two patients died from their CLL (14%), two patients died from unrelated causes, and five patients remain in remission. There were no significant differences in HNC presentation or survival between treated and untreated CLL. CONCLUSIONS. Prior or concurrent CLL therapy did not affect HNC treatment or induce any unexpected toxicities. Interestingly, seven out of fourteen patients (including 2 out of 3 HPV positive) were diagnosed with HNC within 6 months of CLL diagnosis, suggesting that the risk of secondary malignancies may be independent of the degree of immunosuppression attributable to progressive CLL or its therapy. This hypothesis should be explored in a larger population. Disclosures: No relevant conflicts of interest to declare.

Radiology, Aug 1, 2006

To retrospectively determine the long-term risk of falsepositive mammographic assessments and to ... more To retrospectively determine the long-term risk of falsepositive mammographic assessments and to evaluate the effect of screening regularity on the risk of false-positive events.

Blood, Nov 16, 2007

FDG-PET scanning has demonstrated utility in the staging and prognostication of Hodgkin lymphoma,... more FDG-PET scanning has demonstrated utility in the staging and prognostication of Hodgkin lymphoma, but PET characteristics of discrete histologic subsets of Hodgkin lymphoma have never been reported. Further, while studies in this disease have focused on classical Hodgkin lymphoma, the PET features of nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) have not been addressed to date. NLPHL represents less than 5% of all cases of Hodgkin lymphoma and has distinct clinicopathologic features, natural history, and treatment from classical subtypes. Diagnosis may be difficult as it exists in a “gray zone” between Hodgkin and non-Hodgkin lymphomas (NHL), and shares immunophenotypic and histologic features with the T-cell rich variant of diffuse large B-cell lymphoma. Clinically, NLPHL follows a natural history more akin to indolent B-cell NHLs than to classical Hodgkin lymphomas. We retrospectively identified 7 cases of NLPHL diagnosed at our institution since 2003 with available PET staging data. Consistent with prior series of this disease, the majority of patients were male (6:1) and the median age was 32 years. All 7 patients presented with limited stage disease (3 stage I, 4 stage II) and without “B” symptoms. All 7 cases demonstrated increased FDG avidity on PET scanning. The mean Standard Uptake Value (SUV) max was 6.1 (range 1.1–8.8), and mean SUV mean was 5.6 (range 1.0–8.3). We compared these results to recently diagnosed patients with nodular sclerosis and mixed cellularity subtypes of classical Hodgkin lymphoma. Among 13 patients with nodular sclerosis Hodgkin lymphoma (NSHL), the mean SUV max was 10.1 (range 4.8–15.8), while among the 7 patients with mixed cellularity histology (MCHL), the mean SUV max was 5.6 (range 4.0–9.8). Patients with NLPHL had significantly decreased SUV max compared to NSHL patients (p=0.022), but were similar to patients with MCHL (Figure 1). Mixed cellularity histology similarly demonstrated significantly decreased SUV max compared to patients with nodular sclerosing disease (p=0.005). PET intensity therefore highlights differences between classical subtypes of Hodgkin lymphoma, as well as between classical and non-classical variants. When compared to published PET SUV data in aggressive B-cell lymphomas, the maximal SUVs we observe in NLPHL fall well below those reported in aggressive B-cell lymphomas, which may aid in the diagnosis of these occasionally similar appearing neoplasms. We conclude that PET scans identify differences in FDG avidity across classical and non-classical Hodgkin lymphoma subtypes, likely reflecting underlying biological differences. We also demonstrate for the first time that despite its indolent behavior, NLPHL is a uniformly FDG-avid neoplasm, and so PET scans may have a role in the staging and surveillance of this uncommon disease variant. Figure Figure

Breast Cancer Research and Treatment, Jan 13, 2007

Background The anatomy of the nipple has become clinically relevant. Diagnostic techniques access... more Background The anatomy of the nipple has become clinically relevant. Diagnostic techniques access the breast through nipple ducts and surgeons offer nipplesparing mastectomy. There is variation in the number of ducts reported and little is known about the spatial location of ducts, their size, and their relationship to orifices on the surface. Methods Nipple specimens were taken from 129 consecutive mastectomies. Each was sectioned coronally into 3 mm blocks and one section was prepared from each block. The number of ducts and cross-sectional areas of nipple and duct 'bundle' were recorded. Three nipples were sectioned at 50 lm intervals and digitally reconstructed in three dimensions. Results The median number of ducts was 23 (interquartile range 19-28). Reconstructions and summary data from 25 nipples show a central duct bundle narrowing to form a 'waist' as the ducts enter breast parenchyma. A three-dimensional reconstruction focusing on one nipple tip demonstrated 29 ducts arising from 15 orifices. Beneath the skin, most ducts are very narrow, gradually becoming larger deeper within the nipple. Conclusions This work demonstrates that many ducts share a few common openings onto the surface of the nipple, explaining the observed discrepancy between number of ducts and of orifices. Neither duct diameter nor position predicts whether a duct system will terminate close to the nipple or pass deeper into the breast. These new insights into nipple anatomy will be of use in considering the reliability of a ductal approach to diagnosis and in planning nipple-sparing mastectomy.

Journal of The American College of Surgeons, Jun 1, 2007

Background. Liver resection is performed with increasing frequency. Nearly all of the published i... more Background. Liver resection is performed with increasing frequency. Nearly all of the published information on operative mortality and morbidity rates associated with liver resection is derived from studies that rely on retrospective data collection from single centers. The goal of this study is to use audited data from private sector of the National Surgical Quality Improvement Program (NSQIP) Patient Safety Study (PSS) to characterize complications following liver resection and to identify variables that are associated with 30-day morbidity and mortality. Study Design. Prospectively collected data on liver resection patients from 14 hospitals were collected using NSQIP PSS methodology. Rates of occurrence of 21 defined post-operative complications were measured. Bivariate analyses and stepwise logistic regression were used to identify factors associated with 30-day morbidity and mortality. Results. At least one complication occurred in 22.6% of patients within 30 days. Stepwise logistic regression identified several preoperative factors associated with morbidity including serum albumin, SGOT > 40, previous cardiac surgery, operative work RVU, severe COPD, and open wound or wound infection at time of operation. Mortality within 30 days was observed in 2.6% of patients. Factors associated with mortality were found to be male gender, ASA 3 or higher, presence of ascites, dyspnea, and severe COPD. Only 0.7% of patients without any complications died, compared to 9.0% of patients with at least one complication (p < 0.0001). Conclusions. Use of prospectively collected, audited, multi-institutional data in this study increases the likelihood that these results are representative of U.S. medical centers. Several preoperative and intraoperative factors associated with morbidity and mortality after liver resection were identified, and these factors should be considered during patient selection and peri-operative management.

Journal of clinical & experimental pathology, Aug 1, 2015

Journal of clinical & experimental pathology, Aug 3, 2015

Blood, Nov 16, 2007

Secondary CNS lymphoma (SCNSL) is a rare complication of diffuse large B cell non-Hodgkins lympho... more Secondary CNS lymphoma (SCNSL) is a rare complication of diffuse large B cell non-Hodgkins lymphoma (DLBCL) which is almost universally fatal. Several groups have analyzed risk factors for the development of SCNSL and published risk estimates for both individual anatomic sites of DLBCL as well as combinations of clinical risk factors. Although the rates of SCNSL for all DLBCL patients range from 2.5% to 5% subgroups can be identified with risks as high as 33%. High risk scenarios where CNS prophylaxis is most commonly offered include either a combination of an elevated LDH and more than one extranodal site of disease, or disease involving the bone marrow, testicle, sinus, orbit, or a para-spinal location. The type of prophylaxis used varies widely from intrathecal to high-dose intravenous methotrexate as well as intrathecal use of cytarabine. Despite the advances made in the systemic therapy of DLBCL there is very limited data on the efficacy of these chemopreventive regimens in SCNSL. We undertook a retrospective review at our institution of all patients with DLBCL treated with inpatient intermediate dose methotrexate (3.5 grams per meter squared) followed by leucovorin rescue between 2002 and 2006 for CNS prophylaxis. We found 37 patients. Of 15 patients initially evaluated with diagnostic lumbar punctures (LP), no patient had a positive CSF cytology. The median age of the patients was 59 (26–79). The average IPI was 3. All patients were tested for HIV and 3 were found to be positive. The indications for CNS prophylaxis were primarily an elevated LDH and more than 1 extranodal site of disease (20) but also included patients with bone marrow (9), testicular (3), orbital (5), para-spinal (2) and sinus (6) involvement, with 6 patients having more than one site of involvement. All patients received an anthracycline containing chemotherapy regimen with curative intent. Thirty-six patients received rituximab. Patients received a median of 6 cycles of systemic chemotherapy and 3 courses of MTX given on day 15 of alternating cycles (usually 2, 4 and 6). In addition 2 patients received a single dose of 12 mg of intrathecal methotrexate at the time of their initial LP and 5 patients received intrathecal liposomal cytarabine. We calculated the expected number of cases of SCNSL using the estimates from van Besien et al. Blood ‘98, Hollender et al. Ann Onc ‘02, and Boehme et al. Ann Onc ‘06 as appropriate depending on the sites of disease for each patient. For this population the expected number of cases of SCNSL was 7.98 and the observed number of cases was 1 (p=.002 two-sided binomial probability). In this population of 37 patients with an elevated risk of SCNSL, the use of intravenous intermediate dose methotrexate substantially reduced their risk of this fatal complication.

Background: While randomized trials have shown that mammographic screening saves lives, critical ... more Background: While randomized trials have shown that mammographic screening saves lives, critical questions concerns the optimal age of screening initiation and screening interval. Methods: A computer simulation model of breast cancer growth, spread, and detection, which could calculate the reductions in death and cost resulting from various screening intervals and ages of screening initiation. Results: The simulation revealed that breast cancer survival of 90+% appears to be achievable if women are screened annually. The American Cancer Society recommendation of yearly screening from age 40 appears to be a highly effective strategy, yielding a population-wide breast cancer survival of 88%, which corresponds to a 66% population-wide reduction in death compared to no screening. Less intensive screening strategies, such as screening every second or third year, not screening until age 50, or ending screening at 70 should yield markedly worse outcomes, while the pattern of screening used in the United Kingdom of every 36 months from age 50 to 70 is very ineffective, achieving only a 12% reduction in death. There appears to be no upper age limit for screening. A small additional reduction in the breast cancer death rate appears to be achievable by decreasing the screening interval to every six months, and by decreasing the age of screening initiation to age 30, but there appears to be little additional benefit from going beyond this. Conclusions: Prompt annual mammographic screening from age 40 appears to be capable of leading to considerable reductions in breast cancer lethality, with a small additional benefit to be derived from screening as frequently as twice a year from age 30.

Background: While screening and chemoprevention offer the potential for reducing the rate of brea... more Background: While screening and chemoprevention offer the potential for reducing the rate of breast cancer death before the appearance of disease, critical questions concern the actual saving in life and morbidity that can be expected from these interventions, their cost, interaction, and best mix. Methods: A computer simulation model of breast cancer growth, spread, and detection, which calculates the savings in life and cost/benefit ratios expected from various usages of mammography screening and chemoprevention. Results: Women age 50 to 75 can expect to gain ~4-5 days of life for every year of screening, slightly less than 1 day of life for every year of tamoxifen chemoprevention, and slightly more than 1 day of life for every year of Letrozole chemoprevention. A group of women beginning screening and chemoprevention at age 40 who live until age 85 can expect to gain of ~7 months of life for annual screening. In contrast, women may be expected to gain about ~7 days of life for 5 years of tamoxifen chemoprevention and ~10 days of life for 5 years of Letrazole chemoprevention. Screening's markedly lower cost and greater potential for reducing death makes it a far more economical way to save life than chemoprevention; for women in their fifties, screening costs ~$7000/year of life saved, while tamoxifen chemoprevention costs ~$200,00/year of life saved, and Letrozole chemoprevention costs ~$1,00,000/year of life saved. Identifying women at high risk has little impact on these values. Conclusions: Chemoprevention can be expected to yield a modest saving in life, although at considerable cost, while mammographic screening should lead to a considerable savings in life, at modest cost. The small benefit of breast cancer chemoprevention may justify its high cost once widespread use of prompt annual screening has been achieved. C:\Triple modified april25\code to send out modifed for Prevention 4 07 #4 YYL Maybe 5th try\Raw Data From My Dump 2nd Try May 15 YLS 543\ [[All three YLS #3.xls]BenfitOfScreeningNochemFIG

Predicting the risk of breast carcinoma death is both a challenge, and an essential requirement f... more Predicting the risk of breast carcinoma death is both a challenge, and an essential requirement for arriving at the best treatment for each patient. METHODS: CancerMath.net calculators were created with HTML, JavaScript, PHP, using the XML/SWF Charts v5.07 package and Adobe Flash to animate and display the graphs. The risk of cancer death was calculated by the SNAP (Size+Nodes+PrognosticFactor) method of the binary biological model of cancer metastasis, executed in Javascript, from information on tumor size, nodal status, and other prognostic factors. Accuracy was tested against two large breast carcinoma datasets: 7,907 patients seen at two academic hospitals and 362,491 patients from the SEER national dataset. RESULTS: We describe a series of web-based calculators available at http://www.CancerMath.net: 1) An outcome calculator, which calculates expected survival information at the time of diagnosis given the current standard of care treatment; 2) a conditional survival calculator, which calculates expected survival information at various times after the time of diagnosis given the current standard of care treatment; 3) a treatment calculator, which calculates expected survival information at the time of diagnosis with and without a range of adjuvant therapy options; 4) A nodal status calculator, which calculates chance that cancer will be present in the nodes. The CancerMath calculators also provide the patient's classification (T, N, and M) and stage. The CancerMath calculators were found to be highly accurate and specific, as seen by their capacity for stratifying patients into groups differing by as little as a 2% risk of death CONCLUSIONS: The CancerMath.net breast carcinoma calculators provide accurate and useful estimates of the risk of death, which can aid in a analysis of the various adjuvant therapy options available to each patient. METHODS Construction of the calculators The calculators were written in JavaScript, PHP, and HTML, using XML/SWF Charts v5.07 package along with Adobe Flash to animate and display the graphs. Patient data Two datasets were used to test both the underlying mathematics behind our calculators and the Adjuvant! Online breast cancer calculator: SEER Dataset: Breast cancer data was extracted from the US Surveillance, Epidemiology, and End Results (SEER) dataset, which is provided by the National Cancer Institute. It consists of 17 regional registries across the US, with data from 1973 to 2004. An extensive effort was made to clean the data, for example by considering only first malignant primary tumors. The total number of patients is 508,861. Analysis was restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes. The number of these patients was 362,491. Testing of the breast carcinoma outcome calculator whose parameters were designed to capture survival of women with breast carcinoma after 1987, when most eligible patients were receiving adjuvant chemotherapy of hormone therapy, is restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes, diagnosed after 1987, since this was the target population for this calculator. The number of these patients was 293,576. Partners Healthcare Breast Cancer Database: This database consists of 24,771 breast cancer patients diagnosed at the Massachusetts General Hospital or the Brigham and Women's hospital between 1968 and 2007. Analysis was restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes. The number of these patients was 7,907. Testing of the breast carcinoma outcome calculator whose parameters were designed to capture survival of women with breast carcinoma after 1987, when most eligible patients were receiving adjuvant chemotherapy of hormone therapy, is restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes, diagnosed after 1987, since this was the target population for this calculator. The number of these patients was 6,415. Mathematical methods The essential mathematical framework used by the calculators, the binary biological model of cancer metastasis, includes a series of linked equations, the SNAP (Size+Nodes+PrognosticMarkers) method, which can be used to integrate information on tumor size, nodal status, and other prognostic factors into an estimate of the risk of cancer death for each patient (TABLE I), as well as an expression, NodalSizeOnly, Equation, for relating tumor size to the chance that cancer in present in the lymph nodes. Details can be found in reference 15 and the various technical reports available at http://www.lifemath.net/cancer/about/techreports/index.php). The impact of adjuvant therapy on outcome is identical to Adjuvant! Online, from the reductions in death reported by the metaanalyses 10,11 , as summarized by Radvin et al 5,6 , and the values were determined by decompiling the java file that drives the Adjuvant! Online calculator.

CRC Press eBooks, Oct 28, 2020

Breast Cancer Research, 2004

Therapeutic advances in cardiovascular disease, 2008

Thiazolidinediones are currently indicated for the treatment of Type 2 Diabetes. This class of dr... more Thiazolidinediones are currently indicated for the treatment of Type 2 Diabetes. This class of drugs has been associated with several adverse reactions associated with volume overload. This report describes a case of a 65 year old African-American female with a history of hypertension and obesity, and taking rosiglitazone (Avandia) for her Type 2 Diabetes whose evaluation for chest pain resulted in the incidental finding of pulmonary hypertension noted on echocardiogram. The subsequent evaluation, follow-up and treatment are discussed along with potential pitfalls and implications for clinical care.

Protein Engineering, Design and Selection, 2010

Single-chain Fvs (scFvs) are commonly used building blocks for creating engineered diagnostic and... more Single-chain Fvs (scFvs) are commonly used building blocks for creating engineered diagnostic and therapeutic antibody molecules. Bispecific antibodies (BsAbs) hold particular interest due to their ability to simultaneously bind and engage two distinct targets. We describe a technology for producing stable, scalable IgG-like bispecific and multivalent antibodies based on methods for rapidly engineering thermally stable scFvs. Focused libraries of mutant scFvs were designed using a combination of sequence-based statistical analyses and structure-, and knowledge-based methods. Libraries encoding these designs were expressed in E. coli and culture supernatants-containing soluble scFvs screened in a highthroughput assay incorporating a thermal challenge prior to an antigen-binding assay. Thermally stable scFvs were identified that retain full antigen-binding affinity. Single mutations were found that increased the measured T m of either the V H or V L domain by as much as 148 8 8 8 8C relative to the wild-type scFv. Combinations of mutations further increased the T m by as much as an additional 128 8 8 8 8C. Introduction of a stability-engineered scFv as part of an IgG-like BsAb enabled scalable production and purification of BsAb with favorable biophysical properties.

Breast Cancer Research and Treatment, Feb 15, 2011

British Journal of Cancer, Nov 1, 2005

There has been much uncertainty as to whether metastasis requires mutation at the time of spread.... more There has been much uncertainty as to whether metastasis requires mutation at the time of spread. Here, we use clinical data to calculate the probability of the spread of melanoma and breast cancer cells. These calculations reveal that the probability of the spread of cancer cells is relatively high for small tumours (B1 event of spread for every 500 cells for melanomas of 0.1 mm) and declines as tumours increase in size (B1 event of spread for every 10 8 cells for melanomas of 12 mm). The probability of spread of breast cancer cells from the lymph nodes to the periphery is B1 event of spread for every 10 8 cells in the nodal masses, which have a mean diameter of 5 mm, while the probability of spread of cancer cells from the breast to the periphery when the primary masses are 5 mm is also B1 event of spread for every 10 8 cells. Thus, the occurrence of an event of spread from the breast to the lymph nodes appears not to increase the propensity of the progeny of those cells to spread from the lymph nodes to the periphery. These values indicate that the spread of human breast cancer and melanoma cells is unlikely to occur by a mechanism requiring mutation at the time of spread.

Selected key pathways up-regulated or down-regulated by ACEi treatment alone in Gene Set Enrichme... more Selected key pathways up-regulated or down-regulated by ACEi treatment alone in Gene Set Enrichment Analysis.

Purpose: Angiotensin system inhibitors (ASI) can improve prognosis in multiple cancer types, incl... more Purpose: Angiotensin system inhibitors (ASI) can improve prognosis in multiple cancer types, including pancreatic ductal adenocarcinoma (PDAC). However, no study has examined the effect of ASIs alone or combined with adjuvant chemotherapy in resected PDAC patients. Experimental Design: We performed an analysis of the records of ASI users and nonuser patients with PDAC seen at Massachusetts General Hospital (Boston, MA) between January 2006 and December 2010. To identify mechanisms of ASIs in PDAC, we performed RNA sequencing (RNA-Seq) of resected primary lesions. Results: A total of 794 consecutive patients were included. In 299 resected patients, ASI users experienced longer overall survival (OS) in both univariate (median OS, 36.3 vs. 19.3 months, P ¼ 0.011) and adjusted multivariate [HR, 0.505; 95% confidence interval (CI), 0.339-0.750; P ¼ 0.001] analyses. Propensity score-adjusted analysis also showed a longer median OS for chronic ASI users. In unresected patients, the beneficial effect of ASIs was significant in patients with locally advanced disease, but not in metastatic patients. RNA-Seq analysis revealed in tumors of ASI users (lisinopril) a normalized extracellular matrix, a reduced expression of genes involved in PDAC progression (e.g., WNT and Notch signaling), and an increased expression of genes linked with the activity of T cells and antigen-presenting cells. Finally, chronic use of ASI was associated with a gene expression signature that is predictive of survival in independent validation cohorts. Conclusions: In patients with nonmetastatic PDAC, chronic ASI use is associated with longer OS independently of chemotherapy. Our RNA-Seq analysis suggests that ASIs reduce the malignant potential of cancer cells and stimulate the immune microenvironment in primary PDAC. Clin Cancer Res; 23(19); 5959-69. Ó2017 AACR.

Blood, Nov 18, 2011

Abstract 4741 BACKGROUND. Patients with chronic lymphocytic leukemia (CLL) have an increased risk... more Abstract 4741 BACKGROUND. Patients with chronic lymphocytic leukemia (CLL) have an increased risk of developing second primary malignancies, including squamous cell cancers (SCC) of the Head and Neck (HNC). While cutaneous SCC have been shown to behave more aggressively in the setting of CLL, little information exists on the clinical course and outcome of SCC of the oral cavity and pharynx. METHODS. The Massachusetts General Hospital Tumor Registry was queried for all patients with ICD-O-3 codes corresponding to both CLL/SLL and HNC from January 1989 to July 2010, returning seventeen patients. The data collected included patient demographics, blood work and cytogenetics (when available) at the time of CLL diagnosis, primary HNC tumor histology and stage, and all pertinent treatment data, including types of treatment and clinical response. Survival was calculated using the Kaplan-Meier method. RESULTS. Seventeen patients had a diagnosis of both CLL and HNC. Of those, three were excluded from the study as treatment data was unavailable. The fourteen remaining included 12 males and 2 females. Eight patients had their CLL diagnosed first, and had a median interval to HNC diagnosis of 80 months (2-210 months; mean 89 months), four had their HNC diagnosed first and had a median interval to CLL of 1 month (1-100 months; mean 26 months), and two patients had concurrent diagnoses. The median age at diagnosis was 58.5 years (CLL) and 64 years (HNC). Nine patients (64%) were former or current smokers, all with at least a 20 pack year smoking history. Seven patients presented with Rai stage 0 CLL, three patients with Rai stage 1, one patient each with Rai stage 2 and 3, and two with unknown presentations. HNC tumor presentations included 3 T1 tumors, 3 T2 tumors, 3 T3 tumors, 2 T4 tumors, and 3 TX tumors. Three patients had HPV assessed in their tumor and all were positive. Out of the fourteen patients, nine (64%) received first-line chemotherapy for their CLL: 4 with chlorambucil, 1 with fludarabine/rituximab (FR), 1 with FR/lenalidomide, 1 with rituximab, and 1 each with cytoxan/vincristine and steroid monotherapy. Eight patients received second line therapy, three received third line therapy and one received 7 lines of therapy. Six patients (43%) received chemotherapy and radiation as their primary treatment for their Head and Neck tumor, three (21%) received surgery and radiation, three (21%) received only radiation, one received surgery, and one received radiation, chemotherapy, and surgery. One patient required hospitalization due to neutropenic fever during their radiation treatment, and one patient had a week of chemotherapy withheld due to leukopenia, but both were able to complete their full course of therapy. Three patients (21%) received chemotherapy for their CLL before any HNC treatment. These three patients were able to tolerate full HNC therapy without infectious complications or unanticipated cytopenias. First-line therapy for HNC had a median progression-free survival of 26.9 months (range 10.7 – 145.2 months; 95% CI, 12 to 42 months) and a median overall survival of 45.1 months (range 1 to 256.1 months; 95% CI, 9 to 81 months). The median overall survival from HNC for treated CLL patients was 22.6 months and for untreated 59.2 months (p= 0.596). Five patients died from their HNC (36%), two patients died from their CLL (14%), two patients died from unrelated causes, and five patients remain in remission. There were no significant differences in HNC presentation or survival between treated and untreated CLL. CONCLUSIONS. Prior or concurrent CLL therapy did not affect HNC treatment or induce any unexpected toxicities. Interestingly, seven out of fourteen patients (including 2 out of 3 HPV positive) were diagnosed with HNC within 6 months of CLL diagnosis, suggesting that the risk of secondary malignancies may be independent of the degree of immunosuppression attributable to progressive CLL or its therapy. This hypothesis should be explored in a larger population. Disclosures: No relevant conflicts of interest to declare.

Radiology, Aug 1, 2006

To retrospectively determine the long-term risk of falsepositive mammographic assessments and to ... more To retrospectively determine the long-term risk of falsepositive mammographic assessments and to evaluate the effect of screening regularity on the risk of false-positive events.

Blood, Nov 16, 2007

FDG-PET scanning has demonstrated utility in the staging and prognostication of Hodgkin lymphoma,... more FDG-PET scanning has demonstrated utility in the staging and prognostication of Hodgkin lymphoma, but PET characteristics of discrete histologic subsets of Hodgkin lymphoma have never been reported. Further, while studies in this disease have focused on classical Hodgkin lymphoma, the PET features of nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) have not been addressed to date. NLPHL represents less than 5% of all cases of Hodgkin lymphoma and has distinct clinicopathologic features, natural history, and treatment from classical subtypes. Diagnosis may be difficult as it exists in a “gray zone” between Hodgkin and non-Hodgkin lymphomas (NHL), and shares immunophenotypic and histologic features with the T-cell rich variant of diffuse large B-cell lymphoma. Clinically, NLPHL follows a natural history more akin to indolent B-cell NHLs than to classical Hodgkin lymphomas. We retrospectively identified 7 cases of NLPHL diagnosed at our institution since 2003 with available PET staging data. Consistent with prior series of this disease, the majority of patients were male (6:1) and the median age was 32 years. All 7 patients presented with limited stage disease (3 stage I, 4 stage II) and without “B” symptoms. All 7 cases demonstrated increased FDG avidity on PET scanning. The mean Standard Uptake Value (SUV) max was 6.1 (range 1.1–8.8), and mean SUV mean was 5.6 (range 1.0–8.3). We compared these results to recently diagnosed patients with nodular sclerosis and mixed cellularity subtypes of classical Hodgkin lymphoma. Among 13 patients with nodular sclerosis Hodgkin lymphoma (NSHL), the mean SUV max was 10.1 (range 4.8–15.8), while among the 7 patients with mixed cellularity histology (MCHL), the mean SUV max was 5.6 (range 4.0–9.8). Patients with NLPHL had significantly decreased SUV max compared to NSHL patients (p=0.022), but were similar to patients with MCHL (Figure 1). Mixed cellularity histology similarly demonstrated significantly decreased SUV max compared to patients with nodular sclerosing disease (p=0.005). PET intensity therefore highlights differences between classical subtypes of Hodgkin lymphoma, as well as between classical and non-classical variants. When compared to published PET SUV data in aggressive B-cell lymphomas, the maximal SUVs we observe in NLPHL fall well below those reported in aggressive B-cell lymphomas, which may aid in the diagnosis of these occasionally similar appearing neoplasms. We conclude that PET scans identify differences in FDG avidity across classical and non-classical Hodgkin lymphoma subtypes, likely reflecting underlying biological differences. We also demonstrate for the first time that despite its indolent behavior, NLPHL is a uniformly FDG-avid neoplasm, and so PET scans may have a role in the staging and surveillance of this uncommon disease variant. Figure Figure

Breast Cancer Research and Treatment, Jan 13, 2007

Background The anatomy of the nipple has become clinically relevant. Diagnostic techniques access... more Background The anatomy of the nipple has become clinically relevant. Diagnostic techniques access the breast through nipple ducts and surgeons offer nipplesparing mastectomy. There is variation in the number of ducts reported and little is known about the spatial location of ducts, their size, and their relationship to orifices on the surface. Methods Nipple specimens were taken from 129 consecutive mastectomies. Each was sectioned coronally into 3 mm blocks and one section was prepared from each block. The number of ducts and cross-sectional areas of nipple and duct 'bundle' were recorded. Three nipples were sectioned at 50 lm intervals and digitally reconstructed in three dimensions. Results The median number of ducts was 23 (interquartile range 19-28). Reconstructions and summary data from 25 nipples show a central duct bundle narrowing to form a 'waist' as the ducts enter breast parenchyma. A three-dimensional reconstruction focusing on one nipple tip demonstrated 29 ducts arising from 15 orifices. Beneath the skin, most ducts are very narrow, gradually becoming larger deeper within the nipple. Conclusions This work demonstrates that many ducts share a few common openings onto the surface of the nipple, explaining the observed discrepancy between number of ducts and of orifices. Neither duct diameter nor position predicts whether a duct system will terminate close to the nipple or pass deeper into the breast. These new insights into nipple anatomy will be of use in considering the reliability of a ductal approach to diagnosis and in planning nipple-sparing mastectomy.

Journal of The American College of Surgeons, Jun 1, 2007

Background. Liver resection is performed with increasing frequency. Nearly all of the published i... more Background. Liver resection is performed with increasing frequency. Nearly all of the published information on operative mortality and morbidity rates associated with liver resection is derived from studies that rely on retrospective data collection from single centers. The goal of this study is to use audited data from private sector of the National Surgical Quality Improvement Program (NSQIP) Patient Safety Study (PSS) to characterize complications following liver resection and to identify variables that are associated with 30-day morbidity and mortality. Study Design. Prospectively collected data on liver resection patients from 14 hospitals were collected using NSQIP PSS methodology. Rates of occurrence of 21 defined post-operative complications were measured. Bivariate analyses and stepwise logistic regression were used to identify factors associated with 30-day morbidity and mortality. Results. At least one complication occurred in 22.6% of patients within 30 days. Stepwise logistic regression identified several preoperative factors associated with morbidity including serum albumin, SGOT > 40, previous cardiac surgery, operative work RVU, severe COPD, and open wound or wound infection at time of operation. Mortality within 30 days was observed in 2.6% of patients. Factors associated with mortality were found to be male gender, ASA 3 or higher, presence of ascites, dyspnea, and severe COPD. Only 0.7% of patients without any complications died, compared to 9.0% of patients with at least one complication (p < 0.0001). Conclusions. Use of prospectively collected, audited, multi-institutional data in this study increases the likelihood that these results are representative of U.S. medical centers. Several preoperative and intraoperative factors associated with morbidity and mortality after liver resection were identified, and these factors should be considered during patient selection and peri-operative management.

Journal of clinical & experimental pathology, Aug 1, 2015

Journal of clinical & experimental pathology, Aug 3, 2015

Blood, Nov 16, 2007

Secondary CNS lymphoma (SCNSL) is a rare complication of diffuse large B cell non-Hodgkins lympho... more Secondary CNS lymphoma (SCNSL) is a rare complication of diffuse large B cell non-Hodgkins lymphoma (DLBCL) which is almost universally fatal. Several groups have analyzed risk factors for the development of SCNSL and published risk estimates for both individual anatomic sites of DLBCL as well as combinations of clinical risk factors. Although the rates of SCNSL for all DLBCL patients range from 2.5% to 5% subgroups can be identified with risks as high as 33%. High risk scenarios where CNS prophylaxis is most commonly offered include either a combination of an elevated LDH and more than one extranodal site of disease, or disease involving the bone marrow, testicle, sinus, orbit, or a para-spinal location. The type of prophylaxis used varies widely from intrathecal to high-dose intravenous methotrexate as well as intrathecal use of cytarabine. Despite the advances made in the systemic therapy of DLBCL there is very limited data on the efficacy of these chemopreventive regimens in SCNSL. We undertook a retrospective review at our institution of all patients with DLBCL treated with inpatient intermediate dose methotrexate (3.5 grams per meter squared) followed by leucovorin rescue between 2002 and 2006 for CNS prophylaxis. We found 37 patients. Of 15 patients initially evaluated with diagnostic lumbar punctures (LP), no patient had a positive CSF cytology. The median age of the patients was 59 (26–79). The average IPI was 3. All patients were tested for HIV and 3 were found to be positive. The indications for CNS prophylaxis were primarily an elevated LDH and more than 1 extranodal site of disease (20) but also included patients with bone marrow (9), testicular (3), orbital (5), para-spinal (2) and sinus (6) involvement, with 6 patients having more than one site of involvement. All patients received an anthracycline containing chemotherapy regimen with curative intent. Thirty-six patients received rituximab. Patients received a median of 6 cycles of systemic chemotherapy and 3 courses of MTX given on day 15 of alternating cycles (usually 2, 4 and 6). In addition 2 patients received a single dose of 12 mg of intrathecal methotrexate at the time of their initial LP and 5 patients received intrathecal liposomal cytarabine. We calculated the expected number of cases of SCNSL using the estimates from van Besien et al. Blood ‘98, Hollender et al. Ann Onc ‘02, and Boehme et al. Ann Onc ‘06 as appropriate depending on the sites of disease for each patient. For this population the expected number of cases of SCNSL was 7.98 and the observed number of cases was 1 (p=.002 two-sided binomial probability). In this population of 37 patients with an elevated risk of SCNSL, the use of intravenous intermediate dose methotrexate substantially reduced their risk of this fatal complication.

Background: While randomized trials have shown that mammographic screening saves lives, critical ... more Background: While randomized trials have shown that mammographic screening saves lives, critical questions concerns the optimal age of screening initiation and screening interval. Methods: A computer simulation model of breast cancer growth, spread, and detection, which could calculate the reductions in death and cost resulting from various screening intervals and ages of screening initiation. Results: The simulation revealed that breast cancer survival of 90+% appears to be achievable if women are screened annually. The American Cancer Society recommendation of yearly screening from age 40 appears to be a highly effective strategy, yielding a population-wide breast cancer survival of 88%, which corresponds to a 66% population-wide reduction in death compared to no screening. Less intensive screening strategies, such as screening every second or third year, not screening until age 50, or ending screening at 70 should yield markedly worse outcomes, while the pattern of screening used in the United Kingdom of every 36 months from age 50 to 70 is very ineffective, achieving only a 12% reduction in death. There appears to be no upper age limit for screening. A small additional reduction in the breast cancer death rate appears to be achievable by decreasing the screening interval to every six months, and by decreasing the age of screening initiation to age 30, but there appears to be little additional benefit from going beyond this. Conclusions: Prompt annual mammographic screening from age 40 appears to be capable of leading to considerable reductions in breast cancer lethality, with a small additional benefit to be derived from screening as frequently as twice a year from age 30.

Background: While screening and chemoprevention offer the potential for reducing the rate of brea... more Background: While screening and chemoprevention offer the potential for reducing the rate of breast cancer death before the appearance of disease, critical questions concern the actual saving in life and morbidity that can be expected from these interventions, their cost, interaction, and best mix. Methods: A computer simulation model of breast cancer growth, spread, and detection, which calculates the savings in life and cost/benefit ratios expected from various usages of mammography screening and chemoprevention. Results: Women age 50 to 75 can expect to gain ~4-5 days of life for every year of screening, slightly less than 1 day of life for every year of tamoxifen chemoprevention, and slightly more than 1 day of life for every year of Letrozole chemoprevention. A group of women beginning screening and chemoprevention at age 40 who live until age 85 can expect to gain of ~7 months of life for annual screening. In contrast, women may be expected to gain about ~7 days of life for 5 years of tamoxifen chemoprevention and ~10 days of life for 5 years of Letrazole chemoprevention. Screening's markedly lower cost and greater potential for reducing death makes it a far more economical way to save life than chemoprevention; for women in their fifties, screening costs ~$7000/year of life saved, while tamoxifen chemoprevention costs ~$200,00/year of life saved, and Letrozole chemoprevention costs ~$1,00,000/year of life saved. Identifying women at high risk has little impact on these values. Conclusions: Chemoprevention can be expected to yield a modest saving in life, although at considerable cost, while mammographic screening should lead to a considerable savings in life, at modest cost. The small benefit of breast cancer chemoprevention may justify its high cost once widespread use of prompt annual screening has been achieved. C:\Triple modified april25\code to send out modifed for Prevention 4 07 #4 YYL Maybe 5th try\Raw Data From My Dump 2nd Try May 15 YLS 543\ [[All three YLS #3.xls]BenfitOfScreeningNochemFIG

Predicting the risk of breast carcinoma death is both a challenge, and an essential requirement f... more Predicting the risk of breast carcinoma death is both a challenge, and an essential requirement for arriving at the best treatment for each patient. METHODS: CancerMath.net calculators were created with HTML, JavaScript, PHP, using the XML/SWF Charts v5.07 package and Adobe Flash to animate and display the graphs. The risk of cancer death was calculated by the SNAP (Size+Nodes+PrognosticFactor) method of the binary biological model of cancer metastasis, executed in Javascript, from information on tumor size, nodal status, and other prognostic factors. Accuracy was tested against two large breast carcinoma datasets: 7,907 patients seen at two academic hospitals and 362,491 patients from the SEER national dataset. RESULTS: We describe a series of web-based calculators available at http://www.CancerMath.net: 1) An outcome calculator, which calculates expected survival information at the time of diagnosis given the current standard of care treatment; 2) a conditional survival calculator, which calculates expected survival information at various times after the time of diagnosis given the current standard of care treatment; 3) a treatment calculator, which calculates expected survival information at the time of diagnosis with and without a range of adjuvant therapy options; 4) A nodal status calculator, which calculates chance that cancer will be present in the nodes. The CancerMath calculators also provide the patient's classification (T, N, and M) and stage. The CancerMath calculators were found to be highly accurate and specific, as seen by their capacity for stratifying patients into groups differing by as little as a 2% risk of death CONCLUSIONS: The CancerMath.net breast carcinoma calculators provide accurate and useful estimates of the risk of death, which can aid in a analysis of the various adjuvant therapy options available to each patient. METHODS Construction of the calculators The calculators were written in JavaScript, PHP, and HTML, using XML/SWF Charts v5.07 package along with Adobe Flash to animate and display the graphs. Patient data Two datasets were used to test both the underlying mathematics behind our calculators and the Adjuvant! Online breast cancer calculator: SEER Dataset: Breast cancer data was extracted from the US Surveillance, Epidemiology, and End Results (SEER) dataset, which is provided by the National Cancer Institute. It consists of 17 regional registries across the US, with data from 1973 to 2004. An extensive effort was made to clean the data, for example by considering only first malignant primary tumors. The total number of patients is 508,861. Analysis was restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes. The number of these patients was 362,491. Testing of the breast carcinoma outcome calculator whose parameters were designed to capture survival of women with breast carcinoma after 1987, when most eligible patients were receiving adjuvant chemotherapy of hormone therapy, is restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes, diagnosed after 1987, since this was the target population for this calculator. The number of these patients was 293,576. Partners Healthcare Breast Cancer Database: This database consists of 24,771 breast cancer patients diagnosed at the Massachusetts General Hospital or the Brigham and Women's hospital between 1968 and 2007. Analysis was restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes. The number of these patients was 7,907. Testing of the breast carcinoma outcome calculator whose parameters were designed to capture survival of women with breast carcinoma after 1987, when most eligible patients were receiving adjuvant chemotherapy of hormone therapy, is restricted to patients with 1-50 mm tumors and 0-7 positive lymph nodes, diagnosed after 1987, since this was the target population for this calculator. The number of these patients was 6,415. Mathematical methods The essential mathematical framework used by the calculators, the binary biological model of cancer metastasis, includes a series of linked equations, the SNAP (Size+Nodes+PrognosticMarkers) method, which can be used to integrate information on tumor size, nodal status, and other prognostic factors into an estimate of the risk of cancer death for each patient (TABLE I), as well as an expression, NodalSizeOnly, Equation, for relating tumor size to the chance that cancer in present in the lymph nodes. Details can be found in reference 15 and the various technical reports available at http://www.lifemath.net/cancer/about/techreports/index.php). The impact of adjuvant therapy on outcome is identical to Adjuvant! Online, from the reductions in death reported by the metaanalyses 10,11 , as summarized by Radvin et al 5,6 , and the values were determined by decompiling the java file that drives the Adjuvant! Online calculator.

CRC Press eBooks, Oct 28, 2020