PERSPECTIVE

The $ 640 Billion Question

natives to most medical products, firms seek to create the perception that their products are unique in order to justify high prices. Marketing to consumers and physicians will be much less successful if purchasing and prescribing decisions are made by organizations such as managed-care plans or accountable care organizations that are motivated to provide cost-effective care. Such organizations have the incentive and the ability to evaluate competing products and can negotiate with suppliers for the best value. To preserve the present system, manufacturers of health care products spend heavily on federal lobbying. Do these barriers condemn the United States to financial Armageddon or diminished health care for less-affluent Americans? One escape route is tax-supported universal coverage, the finance method that works best in most other high-income countries. Another is disciplined managed competition among health insurers, enhanced by national harmonization of the way in which commercial

insurers, the Centers for Medicare and Medicaid Services, and other payers compare providers on value and of the weight they place on value when tiering network providers and paying them. Both solutions require payment reform. Neither solution is politically feasible without robust physician support: the publics visceral distrust of policies aimed at improving the cost-effectiveness of health care can be neutralized only by their confidence in what their physicians support. The Physician Charter, a modern version of the Hippocratic Oath, has been adopted by physicians organizations that include a majority of U.S. physicians. It ethically commits physicians to working toward the wise and cost-effective management of limited clinical resources. There is not much that physicians can do directly to change the behavior of insurance companies, employers, or other stakeholders, but physicians are the most influential element in health care. The publics trust in them makes physicians the only

plausible catalyst of policies to accelerate diffusion of cost-effective care. Are U.S. physicians sufficiently visionary, public-minded, and well led to respond to this national fiscal and ethical imperative? Its a $640 billion question.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. From Stanford University, Stanford, CA (V.R.F., A.M.); and the Pacific Business Group on Health, San Francisco (A.M.). This article (10.1056/NEJMp1104675) was published on May 18, 2011, at NEJM.org. 1. Chernew ME, Baicker KR, Hsu J. The specter of financial Armageddon health care and federal debt in the United States. N Engl J Med 2010;362:1166-8. 2. Milstein A, Gilbertson E. American medical home runs. Health Aff (Millwood) 2009; 28:1317-26. 3. McGinnis JM. Taking stock: numbers and policies, a look at the numbers. In: Institute of Medicine. The healthcare imperative: lowering costs and improving outcomes: workshop series summary. Washington, DC: National Academies Press, 2010:585-98. 4. Temin P, Maxwell J, Watts C. Corporate health care purchasing among Fortune 500 firms. Health Aff (Millwood) 2001;20(3):181-8. 5. Emanuel EJ, Fuchs VR. Who really pays for health care? The myth of shared responsibility. JAMA 2008;299:1057-9.
Copyright 2011 Massachusetts Medical Society.

How Staphylococcus aureus Adapts to Its Host

Franklin D. Lowy, M.D. s a pathogen, Staphylococcus aureus is well adapted to humans. It can live as a commensal but, provided a suitable opportunity, can initiate severe infection at various body sites. Structural components functioning in concert with secreted products of S. aureus can efficiently target human tissues and evade host defense mechanisms. As a result, S. aureus continues to cause diverse invasive, life-threatening infections despite the availability of effective antimicrobial agents.

Why S. aureus, like some other bacterial pathogens, is able to target some animals and not others remains a puzzle. Recently, Pishchany et al. provided new insight into this question, when they hypothesized that the manner in which S. aureus acquires iron from its human host plays a role in its host specificity.1 Their work further illustrates the breadth of S. aureus pathogenesis. Iron is an essential bacterial nutrient required for survival and virulence; it plays a critical role

in numerous cellular biosynthetic pathways, including DNA replication and bacterial respiration. Animals defend against infection by severely restricting bacterial access to iron. Iron is bound to highaffinity proteins such as transferrin, making free iron effectively inaccessible. Virtually all microbial pathogens have evolved mechanisms for scavenging iron from these iron-binding proteins. S. aureus utilizes hemoglobin acquired from red blood cells as its primary iron source. Hemoglobin,
1987

n engl j med 364;21

nejm.org

may 26, 2011

The New England Journal of Medicine Downloaded from nejm.org on May 29, 2011. For personal use only. No other uses without permission. Copyright 2011 Massachusetts Medical Society. All rights reserved.

PERSPE C T I V E

How staphylococcus aureus Adapts to Its Host

Host Specificity of Staphylococcus aureus. S. aureus uses the iron-regulated surface determinant (Isd) group of proteins to acquire iron from hemoglobin. It secretes a hemolytic toxin that releases hemoglobin from red cells. The released hemoglobin then binds to the staphylococcal receptor, iron surface determinant B (IsdB) on the bacterial cell surface. Heme is extracted from hemoglobin and is transported across the cell wall and cytoplasmic membrane by other Isd proteins. After its release from heme, iron becomes available as a nutrient within the bacterial cell. The increased affinity of the S. aureus IsdB for human hemoglobin (Panel A) versus mouse hemoglobin (Panel B) accounts for the enhanced availability of iron and, in part, for the host specificity of S. aureus.

released from red blood cells by staphylococcal hemolysins, binds to the staphylococcal receptor, iron surface determinant B (IsdB). The bound hemoglobin is then broken down, releasing heme, which is transported into the bacterial cell by transporter proteins and finally degraded into iron (see diagram). There is considerable variability in the amino acid sequences of hemoglobin among different animals, and much of this variation occurs in surface-exposed regions that would most likely interact with IsdB. Examining whether this critical interaction between S. aureus and hemoglobin contributes to the observed host
1988

specificity of this pathogen, Pishchany et al. demonstrated that several different S. aureus strains, including the epidemic methicillin-resistant clone USA300, have a higher affinity for human hemoglobin than for hemoglobin obtained from other mammals.1 S. aureus growth in vitro was also enhanced in the presence of human hemoglobin as compared with mouse hemoglobin. This effect was not found in IsdB mutants. These differences were therefore probably due to human hemoglobins affinity for IsdB. This finding wasnt universal: some bacterial species, such as S. lugdunensis and Corynebacterium

diphtheriae, grew more in the presence of human hemoglobin than in the presence of mouse hemoglobin, whereas others, including Acinetobacter baumanii and Pseudomonas aeruginosa, grew equally well using either animal hemoglobin as a source for iron. Bacteria lacking hemoglobin surface receptors, such as S. epidermidis, failed to grow in the presence of either form of hemoglobin. This enhanced hemoglobinbinding capacity apparently translates into an increased susceptibility to infection.1 Using a mouse model of bacteremic infection, Pishchany et al. found that transgenic mice that were hemizygous for the human hemoglobin gene had more bacteria present in the heart and liver than did control mice. Sometimes strains escape their host-specific constraints and cause disease in other animals. Selected strains of S. aureus not only have a predilection for causing infections in humans but may also cause disease selectively in cattle, pigs, or poultry. This evolution highlights the need to be aware of emerging animal as well as human strains of S. aureus. A methicillin-resistant strain of S. aureus, ST398, classically associated with colonization and infection of pigs and other livestock, has recently gained attention for causing serious infections among pig farmers. S. aureus clone ST5, a strain originally isolated from humans, has undergone extensive genetic modification and is now a major pathogen in poultry. The dissemination of this clone to various geographic regions appears to be due to the worldwide distribution of broiler chickens carrying this strain. Infections due to this poultry-adapted clone have had a major economic impact on the

n engl j med 364;21

nejm.org

may 26, 2011

The New England Journal of Medicine Downloaded from nejm.org on May 29, 2011. For personal use only. No other uses without permission. Copyright 2011 Massachusetts Medical Society. All rights reserved.

PERSPECTIVE

How staphylococcus aureus Adapts to Its Host

Factors That May Affect Host-Specific Staphylococcus aureus Virulence Iron-scavenging proteins (e.g., those that scavenge iron from hemoglobin in humans or from transferrin in cows) Host-specific cytolytic toxins (e.g., PantonValentine leukocidin, a cytotoxin directed at polymorphonuclear leukocytes found in the human strain USA300, contributes to morbidity in necrotizing pneumonia) Loss of function (gene decay) for genes that no longer contribute to staphylococcal virulence (e.g., protein A in poultry strain ET5) Alterations in global regulatory genes affecting expression of cassettes of virulence-associated genes in different animal hosts (e.g., strain ET3 in bovine mastitis) Enhanced ability of selected strains to acquire genes through horizontal transfer of mobile genetic elements Antibiotic-resistance determinants (e.g., exposure of pigs to tetracycline may have enhanced spread of strain ST398)

poultry industry. The evolution of this virulent strain, its successful adaptation to a new animal host, and its resultant global spread illustrate the importance of this adaptive capacity of successful bacterial pathogens.2 The genetic modifications that allow cross-species transfer have been clarified through comparisons of the complete genomes of S. aureus isolates obtained from different animals. Several common themes have emerged from these studies. Much of the genetic machinery needed to facilitate host specificity and virulence is acquired through horizontal gene transfer. Mobile genetic elements, such as transposons, plasmids, or bacteriophages, obtained from strains already adapted to specific hosts, often provide the genes necessary for selective virulence. Once acquired, these genes are conserved among the more successful clones. It is possible that the more successful clones are also ones that can more readily acquire these mobile genetic elements. These newly acquired genes often target critical pathways such as those required for the acquisition of iron, those that enable binding to unique tissue sites, or those that facilitate expression of toxins that are selec-

tive for host-specific leukocytes (see box). They provide strains with a survival advantage and permit adaptation to the new animal environment. Genes that no longer contribute to virulence in the new host are often not expressed. HerronOlson et al.3 found that the ironscavenging genes found in strains associated with bovine mastitis were divergent from those found in human strains. Protein A, an S. aureus surface protein that plays a role in human staphylococcal infections, was not expressed in the poultry strain ST5.2 Protein A nonspecifically binds the Fc portion of IgG in humans and interferes with phagocytosis. In poultry, protein A is unable to bind the Fc portion of IgY (the avian equivalent of IgG) and is therefore no longer antiphagocytic.2 Animal models of infection clearly have potential limitations. Because of host specificity, different strains of S. aureus are more or less virulent depending on the animal host. Animal models are therefore sometimes unreliable predictors of the potential success of therapeutic or preventive interventions or of the roles played by specific determinants of bacterial virulence in infection. Studies of potential S. aureus vaccines

have illustrated this problem: despite the success of several vaccine candidates in in vivo studies, no preparation has yet proved successful in human clinical trials. Similarly, recent studies addressing the role of PantonValentine leukocidin (PVL) toxin in the pathogenesis of necrotizing S. aureus pneumonia have been compromised by the animal model chosen. PVL is a cytotoxin that targets polymorphonuclear leukocytes, the primary host cellular defense against S. aureus infection. There are compelling epidemiologic data linking the epidemic PVL-positive strains of communityassociated methicillin-resistant S. aureus with necrotizing pneumonia. Despite these data, mouse models of infection provided conflicting results regarding the contribution of PVL to this infection. Subsequent studies demonstrated that leukocytes from different animals were variably susceptible to the lytic effects of PVL.4 Whereas mouse leukocytes were relatively insensitive to PVL, rabbit leukocytes had susceptibility similar to that of human leukocytes. When a rabbit pneumonia model was used, the results provided evidence of a critical role for PVL in this infection.4,5 There are alternative in vivo approaches that might be more predictive of outcomes in humans, as the recent study by Pishchany et al. suggests. The use of different animals5 or animal models involving transgenic mice expressing human phenotypes may help to provide more reliable insight into the pathogenesis of specific infections, as well as identifying potential targets for therapeutic or preventive approaches to human infections. S. aureus strains isolated from humans and other animals rep1989

n engl j med 364;21

nejm.org

may 26, 2011

The New England Journal of Medicine Downloaded from nejm.org on May 29, 2011. For personal use only. No other uses without permission. Copyright 2011 Massachusetts Medical Society. All rights reserved.

PERSPE C T I V E

How staphylococcus aureus Adapts to Its Host

Disclosure forms provided by the author are available with the full text of this article at NEJM.org. From the Columbia University College of Physicians and Surgeons, New York. 1. Pishchany G, McCoy AL, Torres VJ, et al. Specificity for human hemoglobin enhances Staphylococcus aureus infection. Cell Host Microbe 2010;8:544-50. 2. Lowder BV, Guinane CM, Ben Zakour NL, et al. Recent human-to-poultry host jump, adaptation, and pandemic spread of Staphylococcus aureus. Proc Natl Acad Sci U S A 2009;106:19545-50. 3. Herron-Olson L, Fitzgerald JR, Musser JM, Kapur V. Molecular correlates of host specialization in Staphylococcus aureus. PLoS ONE 2007;2(10):e1120. 4. Lffler B, Hussain M, Grundmeier M, et al. Staphylococcus aureus Panton-Valentine leukocidin is a very potent cytotoxic factor for human neutrophils. PLoS Pathog 2010; 6(1):e1000715. 5. Diep BA, Chan L, Tattevin P, et al. Polymorphonuclear leukocytes mediate Staphylococcus aureus Panton-Valentine leukocidin-induced lung inflammation and injury. Proc Natl Acad Sci U S A 2010;107:558792.
Copyright 2011 Massachusetts Medical Society.

resent a diverse genetic pool that allows for adaptation and evolutionary change. Further insight into how these adaptations occur may enhance our ability to predict the emergence of new, more capable bacterial pathogens, the development of more predictive animal models, and the potential identification of critical pathways that might serve as therapeutic targets.

Are We Making Progress in Maternal Mortality?

Anne Paxton, Dr.P.H., and Tessa Wardlaw, Ph.D. he number of women who die during pregnancy or childbirth has decreased by more than a third globally since 1990, according to new estimates from the United Nations1 from nearly 550,000 deaths in that year to roughly 350,000 in 2008. Yet progress has been uneven: while some countries have seen significant improvements, others have seen marked increases in maternal mortality. Furthermore, the overall downward trend is insufficient to achieve the Millennium Development Goal (MDG) of a 75% reduction in maternal mortality between 1990 and 2015 (see table). In the United States, where womens chances of surviving pregnancy and childbirth are far greater than in other parts of the world (see map, and interactive map, available with the full text of this article at NEJM.org) the lifetime risk of pregnancy-related death for a U.S. woman is 1 in 2100, as compared with 1 in 31 for a woman in sub-Saharan Africa maternal mortality actually increased during this period, according to United Nations estimates. Despite spending more money on health
1990

care than any other country, the United States has higher maternal mortality than many other developed countries. How do we interpret these data and conflicting impressions of progress and decline? First, we must recognize that the main complications that lead to death during pregnancy or childbirth are fairly common among all women, regardless of where they live. The women who die from these complications are generally those who lack access to treatment. Globally, the leading cause of death (responsible for 35% of all maternal deaths, according to the World Health Organization [WHO]) is hemorrhage, usually occurring immediately after delivery. But hemorrhage doesnt occur only in countries with high maternal mortality: in 2000, it was the second most frequently seen pregnancy-related complication among U.S. women,2 yet the vast majority of these women were promptly treated and not in danger of dying. In developing countries, by contrast, women may give birth at home, unattended or attended by someone unskilled, or in a

poorly equipped health center or hospital where they cant be treated quickly and effectively. The second-leading cause of maternal death globally (according to the WHO) is hypertensive disorders, another common medical problem during pregnancy that, again, leads to death primarily if there is no access to treatment. Countries with high maternal mortality also have a large burden of pregnancy-related complications, the most devastating of which is obstetrical fistula. In this sense, maternal death is just the tip of the iceberg in terms of the impact of poor availability and quality of obstetrical services. It is estimated that for every woman who dies from a pregnancy-related cause, about 20 more roughly 7 million women yearly experience injury, infection, disease, or disability.3 These facts suggest that a key common factor in the trends in maternal mortality, both globally and in the United States, is access to good obstetrical care. As the table shows, the greatest decrease in maternal mortality has occurred in the East Asia and Pacific region, and much of that decrease

n engl j med 364;21

nejm.org

may 26, 2011

The New England Journal of Medicine Downloaded from nejm.org on May 29, 2011. For personal use only. No other uses without permission. Copyright 2011 Massachusetts Medical Society. All rights reserved.