10 February 2001
10,000 words

JAMA theme issue:
Medical research in the 21st century

NEW YORK--The 7 February 2001 Journal of the American Medical Association is a special issue on the topic, "Opportunities for Medical Research in the 21st Century." (Vol. 285, No. 5)

Since I missed the AMA press conference, I read the issue instead, which took me a week of evenings. I wrote these notes to understand it myself, and I'm distributing it to my email list because it should be useful to some subscribers. (My own comments are sometimes, but not always, in parentheses.)

In 23 review articles, experts in each field profiled their specialty, described current research, reviewed the progress of 25 years and predicted the next 25 years. This research is the latest achievement in (when you stop to think about it) the greatest scientific accomplishment of human civilization: the understanding of our own bodies, down to the level of molecular forces that can be modeled on a computer workstation, and the best knowledge of how to cure disease. This issue makes that accomplishment accessible to any intelligent layman.

The topics include the medical disciplines of cardiology, cancer, neurology and psychiatry, infectious diseases, lung disease, orthopedics, endocrinology, obstetrics, hematology, and autoimmune diseases, as well as introductory essays and reviews of cross-disciplinary topics like gene therapy, pharmacology, biomedical engineering, imaging, and transplantation. My own introduction is immediately below, and following are my notes on each review article.

Most of of these articles deal with disease on the molecular and cell biology level. 25 years ago, medical research would be illustrated with a cross-section of an atherosclerotic coronary artery, blocked by a fatal thrombus. This issue is illustrated with diagrams of molecular cascades leading to inflammation. 25 years ago, biologists were starting out with a few peptide sequences and structures, such as hemoglobin and insulin , laboriously working them out with chromatography gels and x-ray crystallography, and a general idea of cell membranes penetrated by active proteins. Now, 25 Nobel prizes later, those rough blackboard sketches have led to tangled flow charts, and occasionally the mass of details, instead of getting more complicated, falls together into simple principles.

The January 2001 JAMA issue is a good companion to the September 1996 special issue of Scientific American on cancer. The JAMA issue is about all of medicine, the Scientific American issue is specifically on cancer -- but both are really about molecular biology. Both are free on the Internet (although it's hard to imagine anyone reading them through for the first time online -- journalists can get the JAMA issue in the press kit from the JAMA PR office.) Another good companion is the Nobel Prize web site, which is another good way to organize and understand modern biology. To fill in the clinical background, I occasionally referred to the Merck Manual Home Edition or Standard Edition , which is also free on the Internet. Several of the footnotes in the article references cite reports that are available free on the Internet. Most of the articles also link to abstracts of articles in Science, the New England Journal of Medicine and other journals, although the full text is not available on-line to non-subscribers.

The job of reading this JAMA issue, like the job of biomedicine generally, is to find the general, understandable principles in the mass of details. Faced with a huge jigsaw puzzle, you first collect together all the pieces of the same color, then find the edges, then see how it fits together, then fill in the gaps. In molecular biology today, we do seem to have enough of the pieces to get a reasonably good picture of the mechanisms. It's gratifying to find the same mechanisms repeating in different diseases, because that indicates that we've identified most of the mechanisms, in the same way that biologists can be sure they've banded most of the wolves in a population if most of the wolves they recapture are banded.

The celebrated milestone of this issue is the sequencing of the human genome, with its 3 billion base pairs and (according to results of the Celera and Human Genome Project released after this issue was published) 26,000 to 40,000 genes. (This issue has an article by Francis S. Collins, director of the National Human Genome Research Institute, but no mention of J. Craig Venter, president of Celera Genomics.) But (now they tell us) these data give us more questions than answers. Now that we have the sequence, we still have to figure out exactly where the genes are, and we have to figure out how they relate to the (possibly 300,000) proteins.

But there are ways to simplify these data, just as you would sort the pieces of a jigsaw puzzle by color. For example, this JAMA issue discusses a bewildering array of genes, immune cells and molecular signals. These components are much easier to understand when you realize that they are all part of cascades, which eventually ("downstream") control cell functions like proliferation and attachment.

A paradigm of the insights of cell and molecular biology is retinoblastoma (RB), a pediatric retinal cancer normally held in check by the rb gene, which produces the RB protein (pRB). RB cancer is caused by a flaw in a basic cell mechanism, which is conserved from yeast to fruit flies to mammals. pRB is one of the brakes on the cell cycle.

(Another important brake is the p53 protein, which is defective in half of all cancers.)

When a cell reproduces, it cycles through growth, resting, DNA synthesis for a duplicate set of chromosomes, and finally division into 2 new cells. The transition from 1 step to the next is regulated by molecular signals, or switches, which biologists commonly refer to as "accelerators" and "brakes". The main stages of the cell cycle are: G1 (gap), G0 (resting), S (synthesis), G2, and finally M (mitosis) in which the cell divides into 2 daughter cells. (Within mitosis, the cell goes through prophase, prometaphase, metaphase, anaphase, and telophase). [Illustration] [More detailed illustration]

pRB is the brake that holds the cell in the G1 phase. Without pRB, the cell keeps going into S phase and dividing. As JAMA explains, with a beautiful illustration, the pRB protein is part of a cascade, and if there is a flaw anywhere on the cascade, the cell will slip uncontrollably into S phase, and into proliferation and migration, classically in the retina and down the optic nerve. The fatal molecular flaw is the addition of excess phosphate groups onto the pRB protein. Phosphate groups flip molecular switches on and off, [Nobel] an important concept. Defects in this cascade are necessary (but not sufficient) for most human cancers. This genetic defect is responsible for proliferation.

But proliferation, and other cancer mechanisms such as migration, and phosphorylation, are also mechanisms of coronary artery disease. Just as metalloproteinases dissolve the intracellular matrix in atherosclerotic plaque formation, metalloproteinases also dissolve the intracellular matrix in cancers. In bone, osteoporosis can be treated as a defect in apoptosis, another basic cell mechanism also defective in cancer. In autoimmune diseases, defects in control of the immune system appear repeatedly as the cause of the inflammatory component of atherosclerosis, asthma, diabetes, inflammatory bowel disease, and rheumatoid arthritis.

So modern biomedicine is dauntingly complicated, but it does come together into simpler basic patterns.

Insight into basic science of disease process is wonderful, but the goal of medicine is to cure patients. The relationship between molecular biology and cure of disease is this: molecular biology establishes cascades of chemical messages. Chemical messengers are bounced along from one cell to another, and from one cell structure to another, like a basketball being fought over and passed along on the way to the hoop. These cascades are usually in balance, with accelerators and brakes, with one cascade promoting inflammation, another cascade inhibiting inflammation; one promoting cell growth, another inhibiting cell growth. Typically, chemical messages will go through the blood stream or intercellular matrix, into a receptor on the cell membrane, through a sequence of messengers under the cell membrane, then into the cell nucleus, where it can bind to DNA, unravel the helix, copy a section of one DNA strand onto some RNA, and send the RNA off to synthesize new proteins. Now that these cascades have been figured out, we can look for drugs to accelerate or brake one side of the balance. That's the process described below.

After an introduction, the JAMA issue is divided into 2 sections, "Opportunities for Medical Research" and "Research Opportunities for Specific Diseases and Disorders." The articles have convenient tables summarizing key research "opportunities" and forecasts.

In the editorial, "Opportunities for Medical Research in the 21st Century," David G. Nathan, Phil B. Fontanarosa and Jean D. Wilson explain their goal, to forecast major advances in medical research for the first quarter of the 21st century. The Human Genome Project will provide immediate insight into single-gene mutations, but most major diseases are polygenic. The protein defect of sickle cell disease has been known for 40 years, but we still don't have satisfactory treatments. (Gene therapy has turned out to be difficult and slow.) The Human Genome Project gives us nucleotide sequences, but that's not enough; what we really need is protein structures. So (now they tell us) we need more grants for structural biology. Over $50 billion will be invested this year in U.S. private and public medical research, and the returns on that investment have been "exceptional."

In one of the major themes of this issue, the editorialists emphasize that most major diseases result from an interaction between genes and the environment. Many of the environmental causes are due "to on the one hand affluence and overabundance, and on the other to poverty, racism, overcrowding and other social ills," and the inequitable distribution of health care, that require "political solutions" rather than more research.

(Among the recurring themes, the most surprising and disappointing is that in the last 25 years, several major diseases are getting worse. Diabetes is getting worse. The death rate for cancer in the U.S. has remained level, with tobacco addiction, particularly among women, wiping out gains in colon cancer. Infectious diseases are actually accelerated by third-world economic development and improved transportation, in the absence of wealth distribution and health care. Antibiotics, fed to farm animals, are becoming less effective, with no new drugs on the immediate horizon. AIDS is established in the U.S. and has devastated sub-Saharan Africa. Tuberculosis causes 2 million deaths worldwide, and drug-resistant TB is poised to attack the U.S. Malaria in some places is getting worse. Worldwide, infant and maternal mortality have remained level -- in many developing countries 1% of mothers die in childbirth. In industrial countries, the survival of low-weight infants (under 1 kg) has "strikingly" improved -- but 50% have permanent neurological disabilities, 17% with cerebral palsy. Sickle cell disease shortens life expectancy in the U.S. by 30 years.)

(Another comparison to this issue would be the September 1973 single-topic issue of Scientific American, "Life and Death and Medicine," which set forth goals which we can now evaluate. The scientific accomplishments have been spectacular: a 1973 illustration shows the amino acid sequence of human growth hormone; 10 years later it was a recombinant-DNA drug. The goals of health care delivery have arguably failed. A 1973 article approvingly describes the Kaiser-Permanente health maintenance organization; this idea was finally implemented by lobbying from Fortune 500 corporations (formed into the Washington Business Group on Health) and insurance companies that scaled it up into a system that lost the efficiency and benefits of the original, added an additional 50% on the dollar for administrative costs and profits, leaves 43 million uninsured and, as Blendon has repeatedly reported in JAMA, those 43 million are frequently unable to get health care even when they are suffering from life-threatening disease. Kaiser-Permanente itself now has trouble competing in the "managed care environment". Themes ignored in this JAMA issue are: research into the organization and distribution of health care, evidence-based medicine, the assessment of outcomes, and the rational adaption of medical research.)

Another concern of the editorialists is the countercurrent of "anti-intellectualism," manifest in "the popularity of alternative medicine," which has grown despite the "paucity of evidence" of efficacy. In contrast, childhood immunizations are rejected by some parents. "There seems to be a widespread belief that miraculous cures can be effected by relatively simple means, such as an adjustment in diet or self-administration of certain remedies, thereby creating unrealistic expectations for many with serious illness."

(Wilson may be familiar to you as one of the editors of Harrison's Internal Medicine and Williams Textbook of Endocrinology, 2 of the textbooks where some of these articles will presumably be rewritten into chapters. Harrison's is a good source for reviews of medical topics not covered here.)

This issue was supported by the Albert and Mary Lasker Foundation, who were also patrons of Henri Matisse, whose cover illustration, Ivy in Flower, was a design for the masoleum of Albert Lasker.

In a commendable practice, several authors credit medical writers or editors for "manuscript preparation." Several authors do not credit medical writers, and it shows. (To be fair, some medical researchers are very good writers, but not all, unfortunately.)

Below are my article-by-article summaries, first the more general section on "Opportunities for Medical Research," and then the section on "Research Opportunities for Specific Diseases and Disorders." Each summary links to the original article.

Opportunities for Medical Research

"Burden of Disease -- Implications for Future Research," Catherine M. Michaud, Christopher J. L. Murray, and Barry R. Bloom. These authors from the Harvard School of Public Health continue on the theme of misallocated resources. In the U.S., life expectancy was 49 years in 1900, 66 years by 1950 [and 76 today]. The highest life expectancy is for women in Japan (84 years), and the lowest for men in Sierra Leone (33 years), which reflects the decreasing life expectancy due to HIV/AIDS in sub-Saharan Africa. In the U.S., major risk factors for death and disability are ischemic heart disease, unipolar major depression, road traffic collisions, cerebrovascular disease, respiratory cancers, and HIV/AIDS. World-wide, major risk factors are pneumonia, diarrhea and AIDS/HIV. One of the surprising findings is that major depression worldwide causes a major reduction in disability-adjusted life years. They conclude: (1) Socioeconomic determinants affect mortality. (2) "Perception varies greatly from reality" despite the "enormous barrage of health information." For example, the public overestimates the "minuscule" effect of drug and child abuse on children's health, compared to the truly important factors of poverty and access to care. (3) The U.S. health care system was ranked only 37th in overall performance.

"Implications of the Human Genome Project for Medical Science," Francis S. Collins and Victor A. McKusick. A public-private partnership, between the Human Genome Project and He Who Must Not Be Named, led to the sequencing of all 3 billion base pairs of the human genome. More than 100 single-gene disorders have been identified, but polygenic diseases like diabetes involves 5, 10 or more genes, a "daunting challenge." But the poster boy of DNA sequencing is STI-571, now named imatinib mesylate (Gleevec, Novartis; "Glivec" outside the U.S.). In chronic myelogenous leukemia, there is a translocation between chromosomes 9 and 22, producing a chimeric protein partly from chromosome 9 and partly from chromosome 22. [Nobel] This produces an enzyme, bcr-abl kinase. Kinases add a phosphate group (that mechanism again!), which is a switch for enzymes. That's the accelerator. Once they had the bcr-abl kinase protein sequence and structure, they could design a new drug -- named STI-571 -- to be a kinase inhibitor, to prevent that bcr-abl kinase from adding a phosphate. That's the brake. STI-571 (imatinib) turned out to be a dramatic clinical success.

"Gene and Stem Cell Therapies," Eugene H. Kaji and Jeffrey M. Leiden. As many of the other articles point out, gene therapy would be the perfect therapy for many diseases, but it turned out to be a lot more difficult than it looked back in 1994-5, when the first trials for cystic fibrosis , and Duchenne muscular dystrophy were tried and failed. Successes include the improvement of familial hypercholesterolemia by inserting a low density lipoprotein receptor into the liver , curing 2 children of severe combined immunodeficiency after ex vivo infection of their hematopoietic stem cells with a retroviral vector expressing the common gamma chain of the interleukin-2 receptor , and using naked DNA for VEGF to increase profusion in myocardial ischemia patients ) Progress continues slowly. [The Scientist] (requires free registration); [Press release] I couldn't figure out from this JAMA article which studies were tried in humans and didn't work, which did work, and which are possible for the future. (These authors could have used a medical writer.) After the death of a patient, the field suffers from timidity. There is a long, cautious discussion of ethics which avoids specifics, avoids offending anyone, avoids taking a stand on anything that anyone could possibly disagree with, and consists entirely of platitudes.

"Genetic Information, Genomic Technologies, and the Future of Drug Discovery," Thomas F. Bumol and August M. Watanabe. Drug treatments today involve only 500 targets, they say. The human genome contains 35,000 genes, which should provide at least another 5,000 targets. (This repeats a point made by Craig Venter at the Gene Media Forum .) Suppose you have a disease involving the human epithelium, they say. You could read out the messenger RNA display of the normal tissue, and the diseased tissue, and compare them to see which genes are expressed in the normal and healthy tissue, and find the disease-causing genes (once you separate them from the normal polymorphic variations). They give the example of developing protein C as an anti-inflammatory drug. (The New England Journal of Medicine said that their report on protein C (Drotrecogin alfa, Lilly) for treatment of severe sepsis was so important that they posted Bernard's paper on their web site before its 8 March 2001 publication. Protein C reduced mortality from 31% to 25%. Sepsis is an inflammatory and thrombosis reaction to bacterial infection of the blood, and causes 225,000 deaths a year in the U.S. Protein C puts the brakes on inflammation and thrombosis.) Bumol and Watanabe cite the Pharmaceutical Research and Manufacturers of America an on-line survey of new medicines in development, but the link doesn't work. Try this instead.

"Advances in Biomedical Engineering," Linda G. Griffith and Alan J. Grodzinsky. "There's been lots of advances in medicine, thanks to war," said John Lennon, in "How I Won the War." Less well known are the advances in medicine thanks to the Beatles. Computed tomographic scanning, developed at EMI Research Labs, Middlesex, funded in part by EMI's Beatles records, provided the first high-quality cross-sectional images of the human body in the 1970s, the authors say. The great contribution of bioengineering has been in obtaining molecular information, and "the most stunning success" has been in DNA microarrays. However, the authors are not visionaries, but engineers, who get stuck with the problem ("challenge") of making things actually work, so this article actually deals in a hard-nosed way with the practical problems they will have to overcome before the drugs work. So what's the catch (er, challenge)? "Many drugs developed through molecular-level assays prove to be ineffective," they say. It's a reasonable approximation to assume that mRNA levels correspond to protein expression -- but they don't. It's reasonable to assume that a compound with high affinity to a ligand will be an effective modulator of that ligand -- but it isn't, in the case of transforming growth factor alpha, and their discussion of that example is worth reading. This article explains the problems that come up in actually implementing the exciting preliminary results, and how engineers solve them. Or, to quote Dilbert: "I'm an engineer. I can make things better!"

(This article was the subject of a JAMA press release, "New Advancements in Biomedical Engineering Will Revolutionize Disease Diagnosis and Treatment" .)

"Advances in Biomedical Imaging," Clare M. C. Tempany and Barbara J. McNeil. Everybody agrees that modern medical imaging creates awesome pictures. The question is whether those pictures actually improve outcomes, given their awesome cost. The authors point out that HCFA authorized Medicare coverage for PET studies of staging certain cancers. They also admit that HCFA added, "there is still a need for high-quality clinical studies." I've written about this very subject, and I've interviewed Anthony D'Amico, so I understand what they're saying. Once again, they must steer a path between the Scylla of exciting, promising technology and the Charybdis of technical difficulties and rigorous outcomes assessment. Remember, the purpose of medical technology is to improve medical outcomes. The way to read an article like this is to be alert for the word "may," as in "may provide a means for early diagnosis of Alzheimer's." You should also ask yourself (and the doctors you interview), "How will this image help the outcome?" They have made wonderful advances in measuring surrogate markers, but that may or may not correlate with disease. They can layer up to 20 images on top of each other, but that may or may not result in better outcomes. The authors use the example of implanting radioactive seeds in prostate cancer, one of the great successes of imaging technology. Brachytherapy was a terrible treatment until ultrasound guidance finally allowed seeds to be placed accurately. MRI guidance would seem to be even better -- but no one has demonstrated that MRI guidance improves placement, much less outcome. They propose combining molecular imaging and image-guided therapy: a 30-year-old man goes to the doctor, and it "may be possible" to do a prostate scan, detect a small, still-confined but aggressive cancer with several advanced imaging techniques, and treat it immediately with a direct injection of radioactive seeds or some other ablative technology. Unfortunately, current imaging techniques will allow you to visualize the prostate, so that you know where to put the needles, but you can't visualize the tumor. I could make a good living writing about the ongoing assessments of each of these technologies. (And they always have great illustrations.)

"Minimally Invasive and Robotic Surgery," Michael J. Mack. Within a few years, gallbladder surgery has changed to a laparoscopic procedure. But this led to "unrealistic expectations" in other procedures. Surgical procedures are excisional, ablative or reconstructive, and the excisional and ablative procedures are easier to perform endoscopically than reconstructive. High volume procedures are easier to perform than low volume, because of the faster learning curve and "market opportunity." Endoscopic procedures are difficult in cardiac surgery, but because of the mortality and morbidity of stopping the heart, research continues on minimally invasive procedures, involving a beating heart. One approach might be to simplify the anastomosis with quick connectors and glues, a snap-in coronary bypass. Motion scaling allows previously-unattainable precision, in for example retinal vein cannulation.

"Prospects for Organ and Tissue Replacement," Laura E. Niklason and Robert Langer. Donor organs and tissues are in inadequate supply, and mechanical devices are inadequate. Artificial tissues and organs grown in culture are under development. Engineered skin and cartilage are currently in clinical use. Urological tissue is being tested in advanced clinical trials. Engineered blood vessels and bladder are functional in preclinical studies. (This is one of those fields where the investigators just got lucky, because to an amazing degree the tissues assemble themselves in culture, generating blood vessels and even nerves.) Strategies for improving immune tolerance, such as clonal T-cell deletion, are being developed.

"Research Opportunities in Transfusion Medicine," Leslie E. Silberstein and Pearl Toy. Since the implementation of nucleic acid testing, the "minuscule" risk from blood transfusion of HIV is 1-2 per 1 million units; of hepatitis C, 1-3 per million units; and of hepatitis B, 1 per 63,000 units. Solvent-detergent treatment has eliminated HIV and hepatitis risk from clotting factor and other plasma derivatives. Inactivation methods are being investigated for other viral and bacterial products. Newly emerging pathogens, such as prions, must be monitored. Blood group antigens may now be typed with genetic methods, and biological functions for some antigens have been found, e.g., as receptors for pathogens or ion transporters in the cell membrane. Hematopoietic stem cells and their derivatives are being studied, such as T cells for anti-tumor activity, and dendritic cells for tumor vaccines. Harmful immune responses may be reduced or eliminated by costimulatory molecules, such as anti-CD40. Alloimmunization is a major clinical problem for patients who require repeated transfusions, occurring in for example 25-30% of sickle cell patients.

Research Opportunities for Specific Diseases and Disorders

Below are articles on cardiology, cancer, neurology and psychiatry, infectious diseases, lung disease, orthopedics, endocrinology, obstetrics, hematology, and autoimmune diseases.

"Prospects for Cardiovascular Research," Robert J. Lefkowitz and James T. Willerson. Cardiovascular disease is the great success of medicine. Death rates from cardiovascular disease have declined dramatically from 1980 to 1992, though they've tapered off (as illustrated on p.629), for reasons that are enumerated in a nice table . Cardiovascular disease is the leading cause of death in the U.S., almost 1 million deaths per year, costing $327 billion. Risk factors include hypercholesterolemia, hypertension, diabetes mellitus, tobacco, obesity and inactivity. Vascular physiology (including lipoprotein levels, coagulation proteins, blood pressure, the immune system, and the biology of vessel walls) is regulated by multiple genes, and they are being identified. Preventable factors include homocysteine and lipoprotein (a). Newly recognized risk factors for atherosclerosis include polymorphisms in the genes for fibrinogen and platelet glycoprotein IIb/IIIa receptors. 4 regions of the human genome influence systolic blood pressure, including 2 polymorphisms of the beta-2 adrenergic receptor gene.

One of the great accomplishments is the current understanding of the mechanism of atherosclerosis, which involves immunology, thrombic homeostasis and cell proliferation. This is beautifully illustrated and summarized in a figure (These figures are not always easy to read in the HTML file, but you can read them in the PDF file which can be enlarged as much as necessary. The PDF file can also print the format of the printed JAMA page. (But for many browsers, it would be easier to just click onto it from the table of contents .)

"Chronic inflammation appears to be central to initiation and progression of the atherosclerotic process," to instability of plaque, and to development of acute coronary syndromes, the authors say.

Chronic endothelial injury can be caused by oxidized low-density lipoprotein, tobacco exposure, free radicals resulting from oxidation of homocysteine and other substances, bacterial infection such as chlamydia, and injury caused by interventional therapies such as angioplasty and stenting.

Injury attracts macrophages to the intimal layer of the artery, and promotes adherence of inflammatory macrophages, T cells and mast cells. This results in platelet aggregation and proliferation of smooth muscle cells in the neointima. Nitric oxide (NO), on the other hand, which is formed by nitric oxide synthase, inhibits platelet aggregation, leukocyte adherence, and muscle proliferation. [Nobel]

(Once again, as with Protein C, nitric oxide synthase is a mediator on a cascade that pulls towards one side of a homeostatic balance which is dysregulated in disease, and that mediator gives pharmaceutical chemists a therapeutic target to correct the balance.)

Turbulent flow also stimulates atherosclerosis, in locations like the carotid artery. Steady laminar flow, which produces steady shear stress, upregulates genes like nitric oxide synthase which protects against atherosclerosis.

Attractive therapeutic targets for up- or down-regulation are NO donors, antioxidants, inflammatory modulators, and adhesion molecules. Statins, which are inhibitors of hydroxymethyl glutaryl coenzyme A synthetase, had a "major impact" on disease.

After an atherosclerotic plaque forms in a coronary artery, it's stable for a while, protected by a fibrous cap, but inflammatory cells can secrete metalloproteinases (remember them?) which dissolve the collagen and cause fissuring or ulceration of the cap. That obstructs the artery and causes unstable angina and acute myocardial infarction. Transient thrombosis and vasoconstriction also cause unstable angina and myocardial infarction. Plaques likely to ulcerate have thin fibrous caps, many inflammatory cells, and an adjacent lipid core. The molecular responses to thrombosis have been characterized. Several markers are associated with worse prognosis. Current research is trying to identify vulnerable plaques.

Arteries often close again after angioplasty and bypass surgery, and clinical trials are beginning for the use of radiation and chemical inhibitors to prevent this. Vascular growth factors, such as VEGF, may stimulate collateral circulation, and have shown promise in clinical trials.

Congestive heart failure has a 10-year mortality of 70%. As you probably know, in congestive heart failure, the heart muscle's pumping ability is reduced, most commonly from coronary artery disease. The compensatory mechanism is to secrete epinepherine and noradrenaline, which causes the heart to pump harder, in the short term, although in the long term this causes the heart to deteriorate. The JAMA authors explain the molecular mechanism behind this: After repeated stimulation, the beta-adrenergic receptors become less responsive. This molecular understanding suggests molecular targets for therapy. For example, in transgenic mice, overexpression of beta-1 adrenergic receptors damages heart muscle and causes heart failure, whereas overexpression of beta-2 adrenergic receptors increases the strength of heart muscles without damage. But in humans, a particular beta-2 adrenergic receptor polymorphism confers a "particularly unfavorable prognosis."

Other promising cellular targets include downregulating cardiac myocyte apoptosis, regulating myocyte calcium metabolism, and cell regulators generally. Tumor necrosis factor may be responsible for several aspects of heart disease, and a protein to inactivate tumor necrosis factor has been tested in patients.

Once you understand the cascade of chemical signals that starts with the receptor and moves downstream into the cell, you can find steps that seem vulnerable to therapeutic intervention. The first messenger is the epinepherine or noradrenaline that stimulates the beta-adrenergic receptor. The second messenger is cyclic adenosine monophosphate (cAMP), which passes the signal downstream. [Nobel] [Nobel] Cyclic adenosine monophosphate has several isoforms, and some of them, in transgenic mice, improve heart contractions without harm. Further downstream, the signal reaches the calcium channels in the muscle, which finally regulates the contractile process. In yet another transgenetic mouse, knocking out an enzyme involved in calcium regulation also improves heart contraction without harm. Even more amazing, this calcium regulation has been improved in rats with adenovirus gene therapy.

Links: American Heart Association, American College of Cardiology, Merck Manual home edition professional edition, Medscape Cardiology

"Toward Mechanism-Based Cancer Care," David M. Livingston and Ramesh Shivdasani. (Despite the 1971 "War on Cancer," the death rate from cancer has remained about level for the last 30 years, as the illustration on p.629 indicates. The death rate from colon cancer seems to have declined, but the death rate from lung cancer has made up for it. For a skeptical look at the War on Cancer, see Chicago Tribune and ACSH.)

Figure 1 and 2 in this article illustrate these new insights in 2 well-chosen areas of cancer research: retinoblastoma and angiogenesis. Usually 4 to 5 genetic defects are necessary for a cell to acquire all the characteristics of cancer, such as proliferation (Figure 1), loss of cell-cell adhesion, invasion of the basement membrane, metastasis, and vascularization (Figure 2). They also illustrate homeostasis: competing processes are normally in balance, and an imbalance, usually from a mutation, throws them out of balance.

Retinoblastoma (RB) is a cancer that occurs when the cellular brake fails. Normally, cells go through a cycle of growth, division and rest. Normally, a protein called E2F is an accelerator -- E2F moves the cells from the resting (G1 or Gap) phase into the growth (S or Synthesis) phase. Normally, the RB protein, pRB, is a brake -- pRB binds to E2F and prevents it from moving the cell into the growth phase. In retinoblastoma, the RB protein doesn't work right. The brakes are off. As you recall, there are proteins called kinases which add phosphate groups, which act as switches. As Figure 1 shows, the protein cyclin D-dependent kinase (cdk4) moves several phosphate groups onto pRB. This inactivates pRB, activates E2F, and moves the cell into S phase growth, thereby fulfilling one necessary step for cancer -- unchecked proliferation. Most human tumors have a defect in the RB pathway. The important point is that a disruption anywhere along the RB pathway (p16, Cyclin D, cdk4, pRB) releases the brakes and leads to proliferation.

Angiogenesis is a normal mechanism that the body uses to respond to lack of oxygen. When cells become hypoxic, they switch to anaerobic glycosis, and send out cell factors that stimulate the growth of new capillaries. Cancer cells also take advantage of angiogenesis. Figure 2 illustrates the latest understanding of the molecular cascades within the cell that stimulate angiogenesis. Hypoxia-inducible factor (HIF) 1-alpha accumulates in the cytoplasm under conditions of hypoxia and is destroyed by enzymes that are active under normal oxygen pressure. When HIF 1-alpha accumulates, it moves into the nucleus, binds to another protein, and activates genes that synthesize proteins for anaerobic glycosis and vascular endothelial growth factor (VEGF). VEGF stimulates endothelial cell proliferation, migration of epithelial cells, and the formation of capillaries that supply the tumor bed. Inhibiting VEGF, among other approaches, should stop or slow tumor growth.

Apoptosis is a process that normally destroys cells when they are no longer needed. One protein that promotes apoptosis is Bax. In follicular cell lymphoma, a gene called Bcl-2 (B-cell lymphoma 2), which inhibits Bax, is overexpressed. This gene is overexpressed when a fragment from chromosome 14 is transferred to chromosome 18, and the brakes are removed.

Another cancer that results from translocation is chronic myelogenous leukemia (CML), which results when a fragment from chromosome 9 is translocated to chromosome 22, and an abnormal protein is produced. In this case, an accelerator is added.

The original function of many cancer genes was to maintain the integrity of the genome of the cell, including the chromosome number. So once a gene is damaged, and becomes a cancer gene, the cell is likely to get new mutations. An examination of these cells can be helpful in prognosis.

Tumors are actually small organs with many of the nutritional and other features of normal organs, the authors say. In fact, tumors develop under the control of genes, such as sonic hedgehog and patched, which are normally active during embryogenesis, where they create organs, and then shut off.

Up to now, diagnosis usually requires at least hundreds of millions of cells, which often does not happen until the cancer is metastasized and difficult to cure. Today, in animal models, it is possible to recognize tens of thousands of cells, which corresponds to a much earlier stage.

In treatment, CML has been successfully treated by a new drug, imatinib (STI-571). The protein which forms the accelerator is a tyrosine kinase, which means that it moves a phosphate group onto another protein (more precisely, onto the tyrosine peptide of another protein). Imatinib is a tyrosine kinase inhibitor. Most leukemias have translocated fragments like CML, so this strategy might work for them too. Trans-retinoic acid has induced remission in acute promyelotic leukemia, apparently by inducing apoptosis.

Other treatments try to restore the function of p53, a brake, which is mutated in most solid tumors. p53 examines DNA as it is being duplicated, and stops the cell, or destroys the cell by apoptosis, if the duplication is not accurate.

Links: American Cancer Society, American Society of Clinical Oncology, Medscape Hematology/Oncology

"Prospects for Neurology and Psychiatry," by W. Maxwell Cowan, Eric R. Kandel. Disorders of the nervous system, which include stroke, depression and Alzheimer, have higher morbidity than "nearly all other disorders combined," the authors say, in an epidemiological review. Information has grown exponentially, and the growth of the Society for Neuroscience has been "astonishing." As an example of molecular approaches, the gene for Huntington Disease was located near the short arm of chromosome 4, which led to a diagnostic test. Huntington disease patients have a large number of repeats of a CAG sequence, which encodes the amino acid glutamine, and glutamine accumulation within the neurons may destroy them. Normal genes have fewer than 40 CAG repeats, but Huntington disease patients have 40 to 80 or more, and the more repeats, the worse the disease. The number of repeats increases in succeeding generations. CAG repeats were discovered in other neurological diseases, and similar trinucleotide repeats were discovered in yet other neurological diseases. The Huntington disease gene has been inserted into transgenic mice and fruit flies. Huntington disease is however a monogenic disorder, while psychiatric diseases like schizophrenia are polygenic. With the human genome sequenced, similar mechanisms should be discovered for polygenic diseases within 20 years. Deletions on chromosome 22q11 seem to confer a much greater susceptibility to schizophrenia. In the same 1.5 million base pair region are genes for catechol-O-methyltransferase and monoamide oxidase, which confer a susceptibility for obsessive-compulsive disorder, particularly in men who also suffered major depression. Elsewhere, imaging studies have localized brain areas altered in schizophrenia. Fetal dopaminergic neurons have successfully treated Parkinson patients, and the use of brain stem cells is promising, especially with neuronal growth factors.

Links: Nobel, Nobel Lecture on video.

(This article was the subject of a JAMA press release, "Closer Partnership Between Neurology and Psychiatry Likely in the 21st Century" .

"Development of Antimicrobial Agents in the Era of New and Reemerging Infectious Diseases and Increasing Antibiotic Resistance," Gail H. Cassell and John Mekalanos. The authors are pessimistic. In 25 years, things have gotten worse. There is a popular idea that infectious disease will continue to decrease in the U.S. to be replaced by diseases of lifestyle and the environment, but actually many diseases of unknown etiology are actually caused by infectious agents. International travel, the breakdown of public health measures, and other factors have caused tuberculosis, malaria, and cholera to spread, "often in more virulent forms." Multi-drug resistant microbes are spreading, even to vancomycin, "and there appear to be few, if any, new classes of drugs currently in clinical development." One theoretical way to create more effective antimicrobial compounds would be to find non-natural compounds that have not been found in the biosphere before. If no microbe has ever been exposed to a compound, it is less likely that microbes have enzymes to destroy it. All efflux pumps, the major resistance mechanism, are derived from a common ancestor, and therefore vulnerable. Natural antibiotics target only 20 gene products, limited perhaps because the producing organisms had to develop their own resistance mechanism. But drug developers can use "forbidden fruit" antibiotics, which could not evolve in microbes because microbes could not develop a resistance mechanism. More than 30 bacterial genomes have been sequenced, blueprints for attack. Several powerful antibiotics attack multiple targets, for example, quinolones inhibit both DNA gyrases and topoisomerases (both of which normally untwist bacterial DNA during replication). So one strategy would be to find common motifs in essential bacterial proteins, in those bacterial genomes.

"Prospects for Vaccines to Protect Against AIDS, Tuberculosis, and Malaria," Norman L. Letvin, Barry R. Bloom, and Stephen L. Hoffman.

AIDS has killed 19 million, and 34 million are currently infected worldwide. CD8 cytotoxic T lymphocytes and CD4 T lymphocytes seem to be responsible for containing HIV and SIV infection in monkeys. Several vaccines elicit high CD4 and CD8 T-cell responses. But they must elicit antibody responses to different isolates.

Tuberculosis causes 2 million deaths annually worldwide. Drug-resistant TB is a major threat to the U.S. Most infected individuals develop an effective immune response. 100 vaccines have been tested in animals. DNA vaccines are cheap, easy to produce and induce long-lasting responses. A live, genetically engineered BCG virus was effective in a UK trial.

Malaria causes 1-3 million deaths annually worldwide. The mortality rate of children has not been reduced in 25 years, and in some places it has gotten more common. Malaria organisms express different proteins at each stage of their life cycle, and surface proteins can express 50 to 100 variants. 2 vaccines could be developed: 1 for nonimmune travelers could prevent infection, while 1 for children would limit replication of the erythrocytic stage. Human volunteers are available for testing, but current vaccines have been short-acting or ineffective. The Plasmidium falciparum genome will be sequenced by the end of 2002. Effective vaccines should be available in 10 to 25 years.

"Research Opportunities and Advances in Lung Disease," Ronald G. Crystal. The lung, which has a surface area the size of a tennis court, is an organ that "mediates gas exchange and defends against a hostile environment." 2 concepts are the "cornerstones" of thinking about pulmonary disease: (1) susceptibility for lung disease is the result of an interaction among environment, genetics and host responses; and (2) inappropriate inflammatory processes are responsible for most lung disease other than cancer. In cystic fibrosis, for example, the mutant CFTR gene results in lower levels, or at least impaired function, of the cystic fibrosis transmembrane receptor (CFTR) in airway epithelial cells. The CFTR protein transports sodium and chloride ions in the epithelial cells on the airway surface of the lung. But the genetic defect itself doesn't cause much direct harm. Rather, the damage is done by the infection -- the environmental cause -- and the inappropriate inflammatory process. The CFTR defect can be transiently corrected in humans by gene therapy, which will probably cure cystic fibrosis in the next 25 years. Many individuals are exposed to environmental causes of lung disease, such as asbestos, but only a few develop disease, which indicates a variation in immune response. Asthma is an inflammatory disorder of the airways, with the extent dependent on environmental factors. Current research is directed at characterizing the different cell types and their genetic expression, and regrowing lung tissue. Monoclonal antibodies may treat asthma. The 10 major and over 20 minor cell types in the lung must be grown in cell culture and characterized with gene chips.

(This article was the subject of a JAMA press release, Future Lung Disease Treatment May Include Gene Therapy, "Designer" Antibiotics. Crystal is co-editor of The Lung: Scientific Foundations, Lippencott-Raven, 1997.)

"Musculoskeletal Disorders and Orthopedic Conditions," Adele L. Boskey. Many of the problems with joint implants have been solved (although infection remains a problem), and total joint arthroplasty for young patients should last a lifetime. Elderly patients have fractures because of variations in genes that regulate cell proliferation, matrix protein production, and the homeostasis between osteoblasts and osteoclasts. Master genes controlling the development of bone and cartilage have been identified, and the understanding of specific biochemical pathways has already led to new drugs. For example, osteorotegerin ligand, which controls differentiation and activation of bone resorbing osteoclasts, was identified, cloned, and synthesized, and soluble antagonists are being evaluated in clinical trials to prevent bone loss. 2 trials of gene therapy for arthritis are in progress. Tissue engineering for bone and cartilage is progressing. New computer technology, such as imaging and robotic tools, are improving orthopedic surgery.

"Prospects for Research for Disorders of the Endocrine System," Jean D. Wilson. Endocrinology is a discipline of the 20th century, "some of the most dramatic applications of organic chemistry to medicine." [Nobel] Oral contraceptives were a "crowning achievement." One of the first accomplishments was to identify and characterize major hormones. Then, hormone deficiency diseases were identified, and treated by hormone replacement. During the last 25 years, researchers have realized that some hormone diseases are due not to hormone deficiency, but to hormone resistance, by cellular receptors or intracellular messenger systems, for example. Endocrinologists have found hormone-resistant diseases for every known human hormone, possibly including obesity and type 2 diabetes mellitus. Immunoassays have identified additional classes of low-level hormones such as DHEA, and genetic sequencing improved diagnostics even more. Transport mechanisms, cellular receptors and messenger systems have been characterized. The role of autoimmune mechanisms are better understood. But treatment of hormone excess has been "imperfect". Agonists, superagonists, and antagonists have been developed, and endocrinologists have realized that constant administration of agonists may cause antagonistic effects. [Nobel] (That's why urologists give lutenizing hormone releasing hormone agonists like leuprolide and goserelin to turn off testosterone in prostate cancer, and that's why athletes who take steroids become sterile.) Better characterizations of receptors should lead to more specific inhibitors and agonists, and possibly gene therapy.

"Prospects for Research in Diabetes Mellitus," Jerrold M. Olefsky. Unlike other major diseases, the death rate from diabetes is growing sharply. (Figure 1) Major studies have shown that elevated blood glucose is a direct cause of diabetic complications. Type 1 diabetes, which usually occurs in children, is caused by auto-immune destruction of the pancreatic beta cells. The genes, immune mechanism and environmental factors must be identified, particularly the specific beta-cell auto-antigens. Within the next 25 years, cadaveric islet cell transplants will be largely perfected, without the problems of immunosuppressive drugs. Advanced glucose monitoring systems and insulin delivery systems will be available. Type 2 diabetes is a disease in which the body's attempt to compensate for declining function leads to a further decline (as in heart failure): Interactions between genetic factors, on the one hand, and obesity and other environmental factors, on the other hand, lead to insulin resistance. To compensate, beta cells overproduce insulin, until they decline. Thus, insulin resistance combined with beta-cell failure leads to diabetes.

"Prospects for Research in Reproductive Health and Birth Outcomes," Robert L. Goldenberg and Alan H. Jobe. Things have actually not been getting much better, and social factors have frustrated some of the gains of research.

Maternal mortality rates worldwide have not decreased in the last 10 to 20 years, and deaths in the immediate neonatal period have decreased slightly or not at all. There are striking differences in maternal and child health in industrialized versus developing countries. In industrialized countries, infant mortality is 5-8 per 1,000; in developing countries infant mortality is 50-150 per 1,000. In industrialized countries women have 1.5-2 births; in developing countries women have 3-7 births (which increases the incidence of adverse events per woman). In some countries, nearly half the women have experienced the death of a child. In industrialized countries fewer than 10 per 100,000 women die in childbirth; in many developing countries 1% die. In the U.S., 1 in 5 pregnant women smoke.

With better understanding of physiology, new reproductive technologies have been more successful. Cryopreservation of ovarian tissue could delay menopause for many years. Litigation has prevented access to contraceptive options like IUDs and long-acting hormonal contraceptives. Medical pregnancy termination by methotrexate and mifepristone, and the morning-after pill, are not widely available but should be over the next 25 years.

Prenatal diagnosis will improve, including microchip screening, permitting selective termination or pre-implantation diagnosis.

"Although survival rates have been strikingly improved, about 50% of surviving infants with birth weights less than 1 kg have permanent neurological disabilities." 17% had cerebral palsy. "The major goal should be an improvement in neurodevelopmental outcomes, with a societal consensus about the limits of viability." Identifying factors that injure and protect the developing brain is crucial. Chronic chorioamnionitis is associated with proinflammatory cytokines and cerebral palsy. Evidence-based medicine, muti-center randomized controlled trials, and insights from basic research and developmental biology are cause for optimism.

"Prospects for Research in Hematologic Disorders: Sickle Cell Disease and Thalassemia," William C. Mentzer and Yuet Wai Kan. Sickle cell anemia and thalassemia [Nobel] are the most common genetic diseases in the world. Sickle cell disease, occurring in 70,000 to 80,000 Americans, shortens life expectancy by 30 years. The median life expectancy for patients with severe beta-thalassemia treated aggressively is 30 years. In Italy and Greece, prenatal DNA testing and elective termination has reduced the incidence significantly; it should be possible to isolate fetal cells from maternal circulation for PCR diagnosis. New drugs could increase fetal hemoglobin levels, increase red blood cell hydration, and inhibit the sickling process. Stem cell transplantation, which cures both diseases, could be improved. The most desirable treatment would be gene therapy -- that tantalizing goal again -- which would correct defective genes in hematopoietic stem cells.

(Olefsky is co-editor of Diabetes Mellitus: A Fundamental and Clinical Text, Lippincott Williams & Wilkins, and the co-author of a chapter in DeGroot Textbook of Endocrinology.)

"Prospects for Research in Inflammatory Bowel Disease," Richard S. Blumberg and Warren Strober. This is an example of how the study of immunology and immunomodulator cascades has given researchers a good understanding of the broad outlines of the mechanisms of a major disease, and suggested several lines of therapeutic attack. There are drugs that could interrupt those cascades, and should be more effective than the non-specific treatments used today. (Once again, they're reporting research, not human therapy.)

Inflammatory bowel disease is a collective term for ulcerative colitis, which affects the surface of the colon, and Crohn disease, which affects the full thickness of any part of the gastrointestinal tract, usually the terminal ilium and colon. Both are associated with a predisposition to cancer and a 2-fold increase in mortality rate. Both are apparently polygenic, with loci associated on chromosomes 12 and 16, and possibly 1, 3, 6 and 7. Both are apparently due to an abnormal mucosal immune function, combined with an urban environment, and possibly infection. They seem to be an inappropriate reaction to normal intestinal microflora. The defect seems to be in the T helper cells. T cells are supposed to tell the difference between pathogens and harmless microorganisms; these T cells can't.

3 figures illustrate this inflammatory process in beautiful detail. (Figure 1) (Figure 2) (Figure 3) Macrophages or dendritic cells take an antigen (presumably a bacterial fragment), pass it on to a T helper cell, which becomes activated, enters a venule, rolls along the inner wall like a wheel until it finally sticks, and recruits leukocytes, which squeeze out of the venule, gather at the bowel epithelial lumen, and produce inflammatory mediators that destroy epithelial tissue. (This process bears an uncanny resemblance to Pac-Man.) Compounds that interrupt any part of this cascade could theoretically prevent inflammation. Monoclonal antibodies can theoretically inactivate any protein. Cytokines in these cascades usually have cytokines with the opposite effect, and they could be used as therapeutic drugs. Cytokines have to fit into a receptor, like a key into a lock, and there are compounds that can jam the receptor. Sometimes chemists can look at the structure of a cytokine and find a simple drug to block it. And eventually gene therapy could create a permanent source of the right drug.

In mouse models, a disease that resembles Crohn disease is associated with a response from T helper-1 cells, and a disease that resembles ulcerative colitis is associated with a response from T helper-2 cells. Both responses end up on a "final common pathway." This is important because (1) It suggests that these diseases can be treated at the final common pathway. (2) TH1 and TH2 pathways are understood well enough to suggest pathways that can be interrupted by therapeutic drug.

(Blumberg is the co-author of the chapter on inflammatory bowel disease in Harrison's.)

Also see "Update on Inflammatory Bowel Disease Therapy," from Digestive Disease Week 2000, or more generally [Medscape Gastroenterology]

"Prospects for Autoimmune Disease." This article is about 3 different diseases, by separate authors:

"JAMA Patient Page: Participating in Medical Research Studies." This is an explanation of clinical trials written for patients, on a 6th grade level. I think it avoids the tough issue: The purpose of clinical trials is to help future patients, not the patients in the trial; to help society, not individuals. Reportedly, Jessie Gelsinger understood and accepted this. This page encourages patients to seek out clinical trials on the Internet, as if clincal trials were the best treatment available. Actually, patients should understand that therapies tested in clinical trials are in equipoise; that's why they're being tested. (See Clinical hype: Don't buy it, Marcia Angell, USA Today, 30 July 2001.) This kind of patient information encourages patients to enter a trial with unrealistic hopes of being cured even of a terminal disease. As A-M The, et al. showed in the British Medical Journal, patients persist in false beliefs that further treatments will cure them, even after doctors try to explain that their condition is terminal. Some doctors argue that it is compassionate to allow such patients to continue in their delusion, and some patients say that they wouldn't want to know. But deluded patients can't give informed consent. This harms not only the individual patient, by deceiving them, but also society, because it promotes a misunderstanding of scientific method, which makes it more difficult to design and implement clinical trials. Gelsinger died because researchers were unwilling to confront the need for studies with infants. This intellectual cowardice creates a world in which congressmen pressure the National Institutes of Health to provide experimental treatments for influential constituents, in which congressmen pressure the NIH to conduct studies of crackpot therapies that failed scientific review, and in which desperate patients call doctors begging for inappropriate drugs like endostatin.

Furthermore, this patient page does not explain to patients how they can evaluate news reports on new research. We who write medical news know that it is an increasingly competitive, low-paid market, in which reporters often aren't held to traditional journalism standards, such as fact-checking, getting second sources on controversies, and disclosure of financial conflicts of interest.

The public trusts the AMA on clinical medicine, and the AMA is uniquely positioned to give intelligent patients the information they need to make important decisions, and yet the AMA web site adds nothing to the Merck Manual. To take the BMJ's example, if a patient with small cell lung cancer (or again, pancreatic cancer) tried to find out his prognosis from the AMA's patient "Health Information", he couldn't find out.

While this generally excellent issue begins with an editorial calling for rationality and condemning "unrealistic expectations" and "miraculous cures," it ironically ends in a patient page that, as always, forfeits the opportunity to educate patients on exactly that. Doctor, heal thyself.


These are my personal notes and have not been prepared for publication, fact-checked or reviewed. I apologize to those whose work I criticized in a spirit of academic discussion.

Copyright 2001 Norman Bauman. You may distribute this free for personal use to your friends and colleagues, including e-mail mailing lists, but not for commercial or promotional use, or print publication, without permission.