Steven N. Goodman

Inactivation of the DNA-Repair Gene MGMT and the Clinical Response of Gliomas to Alkylating Agents

BACKGROUND The DNA-repair enzyme O6-methylguanine-DNA methyltransferase (MGMT) inhibits the killing of tumor cells by alkylating agents. MGMT activity is controlled by a promoter; methylation of the promoter silences the gene in cancer, and the cells no longer produce MGMT. We examined gliomas to determine whether methylation of the MGMT promoter is related to the responsiveness of the tumor to alkylating agents. METHODS We analyzed the MGMT promoter in tumor DNA by a methylation-specific polymerase-chain-reaction assay. The gliomas were obtained from patients who had been treated with carmustine (1,3-bis(2-chloroethyl)-1-nitrosourea, or BCNU). The molecular data were correlated with the clinical outcome. RESULTS The MGMT promoter was methylated in gliomas from 19 of 47 patients (40 percent). This finding was associated with regression of the tumor and prolonged overall and disease-free survival. It was an independent and stronger prognostic factor than age, stage, tumor grade, or performance status. CONCLUSIONS Methylation of the MGMT promoter in gliomas is a useful predictor of the responsiveness of the tumors to alkylating agents.

A Comparison of Low-Molecular-Weight Heparin with Unfractionated Heparin for Unstable Coronary Artery Disease

BACKGROUND Antithrombotic therapy with heparin plus aspirin reduces the rate of ischemic events in patients with unstable coronary artery disease. Low-molecular-weight heparin has a more predictable anticoagulant effect than standard unfractionated heparin, is easier to administer, and does not require monitoring. METHODS In a double-blind, placebo-controlled study, we randomly assigned 3171 patients with angina at rest or non-Q-wave myocardial infarction to receive either 1 mg of enoxaparin (low-molecular-weight heparin) per kilogram of body weight, administered subcutaneously twice daily, or continuous intravenous unfractionated heparin. Therapy was continued for a minimum of 48 hours to a maximum of 8 days, and we collected data on important coronary end points over a period of 30 days. RESULTS At 14 days the risk of death, myocardial infarction, or recurrent angina was significantly lower in the patients assigned to enoxaparin than in those assigned to unfractionated heparin (16.6 percent vs. 19.8 percent, P=0.019). At 30 days, the risk of this composite end point remained significantly lower in the enoxaparin group (19.8 percent vs. 23.3 percent, P=0.016). The need for revascularization procedures at 30 days was also significantly less frequent in the patients assigned to enoxaparin (27.1 percent vs. 32.2 percent, P=0.001). The 30-day incidence of major bleeding complications was 6.5 percent in the enoxaparin group and 7.0 percent in the unfractionated-heparin group, but the incidence of bleeding overall was significantly higher in the enoxaparin group (18.4 percent vs. 14.2 percent, P=0.001), primarily because of ecchymoses at injection sites. CONCLUSIONS Antithrombotic therapy with enoxaparin plus aspirin was more effective than unfractionated heparin plus aspirin in reducing the incidence of ischemic events in patients with unstable angina or non-Q-wave myocardial infarction in the early phase. This benefit of enoxaparin was achieved with an increase in minor but not in major bleeding.

Circulating mutant DNA to assess tumor dynamics

The measurement of circulating nucleic acids has transformed the management of chronic viral infections such as HIV. The development of analogous markers for individuals with cancer could similarly enhance the management of their disease. DNA containing somatic mutations is highly tumor specific and thus, in theory, can provide optimum markers. However, the number of circulating mutant gene fragments is small compared to the number of normal circulating DNA fragments, making it difficult to detect and quantify them with the sensitivity required for meaningful clinical use. In this study, we applied a highly sensitive approach to quantify circulating tumor DNA (ctDNA) in 162 plasma samples from 18 subjects undergoing multimodality therapy for colorectal cancer. We found that ctDNA measurements could be used to reliably monitor tumor dynamics in subjects with cancer who were undergoing surgery or chemotherapy. We suggest that this personalized genetic approach could be generally applied to individuals with other types of cancer (pages 914–915).

Pancreatic intraepithelial neoplasia: a new nomenclature and classification system for pancreatic duct lesions.

Proliferative epithelial lesions in the smaller caliber pancreatic ducts and ductules have been the subject of numerous morphologic, clinical, and genetic studies; however, a standard nomenclature and diagnostic criteria for classifying these lesion have not been established. To evaluate the uniformity of existing systems for grading duct lesions in the pancreas, 35 microscopic slides with 35 representative duct lesions were sent to eight expert pathologists from the United States, Canada, and Europe. Kappa values for interobserver agreement could not be calculated initially because more than 70 different diagnostic terms were used by the eight pathologists. In several cases, the diagnoses rendered for a single duct lesion ranged from “hyperplasia,” to “metaplasia,” to “dysplasia,” to “carcinoma in situ.” This review therefore demonstrated the need for a standard nomenclature and classification system. Subsequently, during a working group meeting, the pathologists agreed to adopt a single standard system. The terminology pancreatic intraepithelial neoplasia (or PanIN) was selected, and diagnostic criteria for each grade of PanIN were established (http://pathology.jhu.edu/pancreas_panin). This new system was then evaluated by having the eight pathologists rereview the original 35 cases. Only seven different diagnoses were rendered, and kappa values of 0.43, 0.14, and 0.42 were obtained for PanINs 1, 2, and 3 respectively. Cases assigned other diagnoses (e.g., squamous metaplasia) collectively had a kappa value of 0.41. These results show both the potential of the classification system, and also the difficulty of classifying these lesions even with a consistent nomenclature. However, even when there is lack of consensus, having a restricted set of descriptions and terms allows a better understanding of the reasons for disagreement. It is suggested that we adopt and apply this system uniformly, with continued study of its reliability and use, and possibly further refinement. The acceptance of a standard classification system will facilitate the study of pancreatic duct lesions, and will lead ultimately to a better understanding of their biologic importance.

Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients

ObjectiveThis single-institution study examined the outcome after pancreaticoduodenectomy in patients with adenocarcinoma of the head of the pancreas. Summary of Background DataIn recent years, pancreaticoduodenectomy for adenocarcinoma of the head of the pancreas has been associated with decreased morbidity and mortality and, in some centers, 5-year survival rates in excess of 20%. MethodsTwo hundred one patients with pathologically verified adenocarcinoma of the head of the pancreas undergoing pancreaticoduodenectomy at The Johns Hopkins Hospital between 1970 and 1994 were analyzed (the last 100 resections were performed between March 1991 and April 1994). This is the largest single-institution experience reported to date. ResultsThe overall postoperative in-hospital mortality rate was 5%, but has been 0.7% for the last 149 patients. The actuarial 5-year survival for all 201 patients was 21%, with a median survival of 15.5 months. There were 11 5-year survivors. Patients resected with negative margins (curative resections: n = 143) had an actuarial 5-year survival rate of 26%, with a median survival of 18 months, whereas those with positive margins (palliative resections: n = 58) fared significantly worse, with an actuarial 5-year survival rate of 8% and a median survival of 10 months (p < 0.0001). Survival has improved significantly from decade to decade (p < 0.002), with the 3-year actuarial survival of 14% in the 1970s, 21% in the 1980s, and 36% in the 1990s. Factors significantly favoring long-term survival by univariate analyses included tumor diameter < 3 cm, negative nodal status, diploid tumor DNA content, tumor S phase fraction < 18%, pylorus-preserving resection, <800 mL intraoperative blood loss, <2 units of blood transfused, negative resection margins, and use of postoperative adjuvant chemotherapy and radiation therapy. Multivariate analyses indicated the strongest predictors of long-term survival were diploid tumor DNA content, tumor diameter < 3 cm, negative nodal status, negative resection margins, and decade of resection.

Toward Evidence-Based Medical Statistics. 1: The P Value Fallacy

The past decade has seen the rise of evidence-based medicine, a movement that has focused attention on the importance of using clinical studies for empirical demonstration of the efficacy of medical interventions. Increasingly, physicians are being called on to assess such studies to help them make clinical decisions and understand the rationale behind recommended practices. This type of assessment requires an understanding of research methods that until recently was not expected of physicians. These research methods include statistical techniques used to assist in drawing conclusions. However, the methods of statistical inference in current use are not evidence-based and thus have contributed to a widespread misperception. The misperception is that absent any consideration of biological plausibility and prior evidence, statistical methods can provide a number that by itself reflects a probability of reaching erroneous conclusions. This belief has damaged the quality of scientific reasoning and discourse, primarily by making it difficult to understand how the strength of the evidence in a particular study can be related to and combined with the strength of other evidence (from other laboratory or clinical studies, scientific reasoning, or clinical experience). This results in many knowledge claims that do not stand the test of time (1, 2). A pair of articles in this issue examines this problem in some depth and proposes a partial solution. In this article, I explore the historical and logical foundations of the dominant school of medical statistics, sometimes referred to as frequentist statistics, which might be described as error-based. I explicate the logical fallacy at the heart of this system and the reason that it maintains such a tenacious hold on the minds of investigators, policymakers, and journal editors. In the second article (3), I present an evidence-based approach derived from Bayesian statistical methods, an alternative perspective that has been one of the most active areas of biostatistical development during the past 20 years. Bayesian methods have started to make inroads into medical journals; Annals, for example, has included a section on Bayesian data interpretation in its Information for Authors section since 1 July 1997. The perspective on Bayesian methods offered here will differ somewhat from that in previous presentations in other medical journals. It will focus not on the controversial use of these methods in measuring belief but rather on how they measure the weight of quantitative evidence. We will see how reporting an index called the Bayes factor (which in its simplest form is also called a likelihood ratio) instead of the P value can facilitate the integration of statistical summaries and biological knowledge and lead to a better understanding of the role of scientific judgment in the interpretation of medical research. An Example of the Problem A recent randomized, controlled trial of hydrocortisone treatment for the chronic fatigue syndrome showed a treatment effect that neared the threshold for statistical significance, P=0.06 (4). The discussion section began, hydrocortisone treatment was associated with an improvement in symptoms This is the first such study to demonstrate improvement with a drug treatment of [the chronic fatigue syndrome] (4). What is remarkable about this paper is how unremarkable it is. It is typical of many medical research reports in that a conclusion based on the findings is stated at the beginning of the discussion. Later in the discussion, such issues as biological mechanism, effect magnitude, and supporting studies are presented. But a conclusion is stated before the actual discussion, as though it is derived directly from the results, a mere linguistic transformation of P=0.06. This is a natural consequence of a statistical method that has almost eliminated our ability to distinguish between statistical results and scientific conclusions. We will see how this is a natural outgrowth of the P value fallacy. Philosophical Preliminaries To begin our exploration of the P value fallacy, we must consider the basic elements of reasoning. The process that we use to link underlying knowledge to the observed world is called inferential reasoning, of which there are two logical types: deductive inference and inductive inference. In deductive inference, we start with a given hypothesis (a statement about how nature works) and predict what we should see if that hypothesis were true. Deduction is objective in the sense that the predictions about what we will see are always true if the hypotheses are true. Its problem is that we cannot use it to expand our knowledge beyond what is in the hypotheses. Inductive inference goes in the reverse direction: On the basis of what we see, we evaluate what hypothesis is most tenable. The concept of evidence is inductive; it is a measure that reflects back from observations to an underlying truth. The advantage of inductive reasoning is that our conclusions about unobserved states of nature are broader than the observations on which they are based; that is, we use this reasoning to generate new hypotheses and to learn new things. Its drawback is that we cannot be sure that what we conclude about nature is actually true, a conundrum known as the problem of induction (5-7). From their clinical experience, physicians are acutely aware of the subtle but critical difference between these two perspectives. Enumerating the frequency of symptoms (observations) given the known presence of a disease (hypothesis) is a deductive process and can be done by a medical student with a good medical textbook (Figure 1, top). Much harder is the inductive art of differential diagnosis: specifying the likelihood of different diseases on the basis of a patients signs, symptoms, and laboratory results. The deductions are more certain and objective but less useful than the inductions. Figure 1. The parallels between the processes of induction and deduction in medical inference ( top ) and statistical inference ( bottom ). The identical issue arises in statistics. Under the assumption that two treatments are the same (that is, the hypothesis of no difference in efficacy is true), it is easy to calculate deductively the frequency of all possible outcomes that we could observe in a study (Figure 1, bottom). But once we observe a particular outcome, as in the result of a clinical trial, it is not easy to answer the more important inductive question, How likely is it that the treatments are equivalent? In this century, philosophers have grappled with the problem of induction and have tried to solve or evade it in several ways. Karl Popper (8) proposed a philosophy of scientific practice that eliminated formal induction completely and used only the deductive elements of science: the prediction and falsification components. Rudolf Carnap tried an opposite strategyto make the inductive component as logically secure as the deductive part (9, 10). Both were unsuccessful in producing workable models for how science could be conducted, and their failures showed that there is no methodologic solution to the problem of fallible scientific knowledge. Determining which underlying truth is most likely on the basis of the data is a problem in inverse probability, or inductive inference, that was solved quantitatively more than 200 years ago by the Reverend Thomas Bayes. He withheld his discovery, now known as Bayes theorem; it was not divulged until 1762, 20 years after his death (11). Figure 2 shows Bayes theorem in words. Figure 2. Bayes theorem, in words. As a mathematical equation, Bayes theorem is not controversial; it serves as the foundation for analyzing games of chance and medical screening tests. However, as a model for how we should think scientifically, it is criticized because it requires assigning a prior probability to the truth of an idea, a number whose objective scientific meaning is unclear (7, 10, 12). It is speculated that this may be why Reverend Bayes chose the more dire of the publish or perish options. It is also the reason why this approach has been tarred with the subjective label and has not generally been used by medical researchers. Conventional (Frequentist) Statistical Inference Because of the subjectivity of the prior probabilities used in Bayes theorem, scientists in the 1920s and 1930s tried to develop alternative approaches to statistical inference that used only deductive probabilities, calculated with mathematical formulas that described (under certain assumptions) the frequency of all possible experimental outcomes if an experiment were repeated many times (10). Methods based on this frequentist view of probability included an index to measure the strength of evidence called the P value, proposed by R.A. Fisher in the 1920s (13), and a method for choosing between hypotheses, called a hypothesis test, developed in the early 1930s by the mathematical statisticians Jerzy Neyman and Egon Pearson (14). These two methods were incompatible but have become so intertwined that they are mistakenly regarded as part of a single, coherent approach to statistical inference (6, 15, 16). The P Value The P value is defined as the probability, under the assumption of no effect or no difference (the null hypothesis), of obtaining a result equal to or more extreme than what was actually observed (Figure 3). Fisher proposed it as an informal index to be used as a measure of discrepancy between the data and the null hypothesis. It was not part of a formal inferential method. Fisher suggested that it be used as part of the fluid, non-quantifiable process of drawing conclusions from observations, a process that included combining the P value in some unspecified way with background information (17). Figure 3. The bell-shaped curve represents the probability of every possible outcome under the null hypothesis. P P It is worth noting one widely prevalent

Molecular assessment of histopathological staging in squamous-cell carcinoma of the head and neck.

BACKGROUND Surgical oncologists rely heavily on the histopathological assessment of surgical margins to ensure total excision of the tumor in patients with head and neck cancer. However, current techniques may not detect small numbers of cancer cells at the margins of resection or in cervical lymph nodes. METHODS We used molecular techniques to determine whether clonal populations of infiltrating tumor cells harboring mutations of the p53 gene could be detected in histopathologically negative surgical margins and cervical lymph nodes of patients with squamous-cell carcinoma of the head and neck. RESULTS We identified 25 patients with primary squamous-cell carcinoma of the head and neck containing a p53 mutation who appeared to have had complete tumor resection on the basis of a negative histopathological assessment. In 13 of these 25 patients, molecular analysis was positive for a p53 mutation in at least one tumor margin. In 5 of 13 patients with positive margins by this method (38 percent), the carcinoma has recurred locally, as compared with none of 12 patients with negative margins (P = 0.02 by the log-rank test). Furthermore, molecular analysis identified neoplastic cells in 6 of 28 lymph nodes (21 percent) that were initially negative by histopathological assessment. CONCLUSIONS Among specimens initially believed to be negative by light microscopy, a substantial percentage of the surgical margins and lymph nodes from patients with squamous-cell carcinoma of the head and neck contained p53 mutations specific for the primary tumor. Patients with these positive margins appear to have a substantially increased risk of local recurrence. Molecular analysis of surgical margins and lymph nodes can augment standard histopathological assessment and may improve the prediction of local tumor recurrence.

Toward evidence-based medical statistics. 2 : The Bayes factor

The second article on evidence-based statistics explores the inductive Bayesian approach to measuring evidence and combining information and addresses the epistemologic uncertainties that affect al...

Association between cigarette smoking and mutation of the p53 gene in squamous-cell carcinoma of the head and neck.

BACKGROUND Although epidemiologic studies have long associated tobacco and alcohol use with the development of squamous-cell carcinoma of the head and neck, the molecular targets of these carcinogens have yet to be identified. We performed a molecular analysis to determine the pattern of mutations in the p53 gene in neoplasms from patients with squamous-cell carcinoma of the head and neck and a history of tobacco or alcohol use. METHODS Sequence analysis of the conserved regions of the p53 gene was performed in tumor samples from 129 patients with primary squamous-cell carcinoma of the head and neck. We then used statistical analysis to identify any patient characteristics associated with mutation of the p53 gene. RESULTS We found p53 mutations in 42 percent of the patients (54 of 129). Fifty-eight percent of the patients who smoked cigarettes and used alcohol (37 of 64; 95 percent confidence interval, 45 to 70 percent), 33 percent of the patients who smoked but abstained from alcohol (13 of 39; 95 percent confidence interval, 19 to 50 percent), and 17 percent of the patients who neither smoked nor drank alcohol (4 of 24, 95 percent confidence interval, 5 to 37 percent) had p53 mutations (P = 0.001). (Two patients used alcohol but did not smoke, and neither had a p53 mutation.) Furthermore, 100 percent of the mutations in the patients who neither drank nor smoked occurred at sites containing cytidine phosphate guanosine dinucleotides (potentially representing endogenous mutations) within the p53 gene (5 of 5 mutations; 95 percent confidence interval, 48 to 100 percent), whereas only 23 percent of those in cigarette smokers consisted of such changes (12 of 53 mutations; 95 percent confidence interval, 12 to 36 percent; P = 0.001). CONCLUSIONS In our study, a history of tobacco and alcohol use was associated with a high frequency of p53 mutations in patients with squamous-cell carcinoma of the head and neck. Preliminary evidence linked cigarette smoking to p53 mutations at nonendogenous mutation sites. Our findings suggest a role for tobacco in the molecular progression of squamous-cell carcinoma of the head and neck and support the epidemiologic evidence that abstinence from smoking is important to prevent head and neck cancer.

Adalimumab with or without Methotrexate in Juvenile Rheumatoid Arthritis

BACKGROUND Tumor necrosis factor (TNF) has a pathogenic role in juvenile rheumatoid arthritis. We evaluated the efficacy and safety of adalimumab, a fully human monoclonal anti-TNF antibody, in children with polyarticular-course juvenile rheumatoid arthritis. METHODS Patients 4 to 17 years of age with active juvenile rheumatoid arthritis who had previously received treatment with nonsteroidal antiinflammatory drugs underwent stratification according to methotrexate use and received 24 mg of adalimumab per square meter of body-surface area (maximum dose, 40 mg) subcutaneously every other week for 16 weeks. We randomly assigned patients with an American College of Rheumatology Pediatric 30% (ACR Pedi 30) response at week 16 to receive adalimumab or placebo in a double-blind fashion every other week for up to 32 weeks. RESULTS Seventy-four percent of patients not receiving methotrexate (64 of 86) and 94% of those receiving methotrexate (80 of 85) had an ACR Pedi 30 response at week 16 and were eligible for double-blind treatment. Among patients not receiving methotrexate, disease flares (the primary outcome) occurred in 43% of those receiving adalimumab and 71% of those receiving placebo (P=0.03). Among patients receiving methotrexate, flares occurred in 37% of those receiving adalimumab and 65% of those receiving placebo (P=0.02). At 48 weeks, the percentages of patients treated with methotrexate who had ACR Pedi 30, 50, 70, or 90 responses were significantly greater for those receiving adalimumab than for those receiving placebo; the differences between patients not treated with methotrexate who received adalimumab and those who received placebo were not significant. Response rates were sustained after 104 weeks of treatment. Serious adverse events possibly related to adalimumab occurred in 14 patients. CONCLUSIONS Adalimumab therapy seems to be an efficacious option for the treatment of children with juvenile rheumatoid arthritis. (ClinicalTrials.gov number, NCT00048542.)