|
|
Background Alternative therapies for Staphylococcus aureus bacteremia and endocarditis are needed.
Methods We randomly assigned 124 patients with S. aureus bacteremia with or without endocarditis to receive 6 mg of daptomycin intravenously per kilogram of body weight daily and 122 to receive initial low-dose gentamicin plus either an antistaphylococcal penicillin or vancomycin. The primary efficacy end point was treatment success 42 days after the end of therapy.
Results Forty-two days after the end of therapy in the modified intention-to-treat analysis, a successful outcome was documented for 53 of 120 patients who received daptomycin as compared with 48 of 115 patients who received standard therapy (44.2 percent vs. 41.7 percent; absolute difference, 2.4 percent; 95 percent confidence interval, –10.2 to 15.1 percent). Our results met prespecified criteria for the noninferiority of daptomycin. The success rates were similar in subgroups of patients with complicated bacteremia, right-sided endocarditis, and methicillin-resistant S. aureus. Daptomycin therapy was associated with a higher rate of microbiologic failure than was standard therapy (19 vs. 11 patients, P=0.17). In 6 of the 19 patients with microbiologic failure in the daptomycin group, isolates with reduced susceptibility to daptomycin emerged; similarly, a reduced susceptibility to vancomycin was noted in isolates from patients treated with vancomycin. As compared with daptomycin therapy, standard therapy was associated with a nonsignificantly higher rate of adverse events that led to treatment failure due to the discontinuation of therapy (17 vs. 8, P=0.06). Clinically significant renal dysfunction occurred in 11.0 percent of patients who received daptomycin and in 26.3 percent of patients who received standard therapy (P=0.004).
Conclusions Daptomycin (6 mg per kilogram daily) is not inferior to standard therapy for S. aureus bacteremia and right-sided endocarditis. (ClinicalTrials.gov number, NCT00093067
[ClinicalTrials.gov]
.)
Daptomycin is a cyclic lipopeptide antibiotic that is rapidly bactericidal in vitro against most clinically relevant gram-positive bacteria, including S. aureus.11,12,13,14 Daptomycin is approved for the treatment of complicated skin and skin-structure infections attributable to gram-positive organisms at a dose of 4 mg per kilogram of body weight intravenously per day,15 but its efficacy in the treatment of S. aureus bacteremia and endocarditis and its safety at higher doses have not been established. We assessed the efficacy and safety of daptomycin as compared with standard therapy for S. aureus bacteremia and endocarditis.
Methods
Study Design
This open-label, randomized trial was conducted between August 28, 2002, and February 16, 2005. The institutional review board at each site approved the protocol, and all patients or their authorized representatives provided written informed consent. Eligible patients were 18 years of age or older and had one or more blood cultures that were positive for S. aureus within two calendar days before initiating study medication. Patients were ineligible if they had a creatinine clearance of less than 30 ml per minute, known osteomyelitis, polymicrobial bacteremia, or pneumonia. Full exclusion criteria are listed in the Supplementary Appendix, available with the full text of this article at www.nejm.org.
The adjudication committee of five infectious-disease experts (Drs. Abrutyn, Corey, Cosgrove, Fowler, and Karchmer) reviewed the data from each patient to establish the diagnosis and outcome. The committee was unaware of patients' treatment assignments.
Definitions
The diagnosis at baseline was determined with the use of the modified Duke criteria.16 The adjudication committee determined final diagnoses (Table 1). Uncomplicated bacteremia was defined by the isolation of S. aureus from enrollment blood cultures in patients without endocarditis and without evidence of hematogenous spread. Patients with such infections received medication for a minimum of 10 to 14 days. In patients without definite endocarditis, complicated bacteremia was defined by the isolation of S. aureus from blood cultures on at least two days through study day 5, the presence of spread of infection, or infection involving prostheses not removed within four days. Such patients received medication for a minimum of 28 to 42 days. Uncomplicated right-sided endocarditis was defined as definite or possible16 methicillin-susceptible S. aureus (MSSA) endocarditis in the absence of predisposing abnormalities or active infection of the mitral or aortic valve in a patient with active injection-drug use, a serum creatinine level of less than 2.5 mg per deciliter (221 µmol per liter), and no evidence of extrapulmonary sites of infection. Patients with uncomplicated MSSA right-sided endocarditis received medication for a minimum of 14 to 28 days. Complicated right-sided endocarditis was defined as definite or possible16 endocarditis in the absence of predisposing abnormalities or active infection of the mitral or aortic valve, with extrapulmonary sites of infection, a serum creatinine level of at least 2.5 mg per deciliter, MRSA bacteremia, or the absence of injection-drug use. Patients with such infections were treated for a minimum of 28 to 42 days. Left-sided endocarditis was defined according to the criteria of Li et al.16 Patients with such infections received medication for a minimum of 28 to 42 days.
|
Patients were considered to have clinical failure if they had no response to the study drug on the basis of ongoing signs and symptoms of infection. Patients were considered to have microbiologic failure if they had persistent or relapsing S. aureus infection, defined as either ongoing positive cultures leading to discontinuation of the study drug or subsequent isolation of S. aureus of the same strain type after apparent clinical improvement.
Randomization, Treatment, and Monitoring
Patients were randomly assigned with the use of a block randomization schedule to receive either daptomycin or standard therapy; this centralized computer-generated schedule was designed to achieve a 1:1 ratio of patients, stratified according to investigative site.
Patients received daptomycin (Cubicin, Cubist Pharmaceuticals) (6 mg per kilogram of body weight intravenously once daily) or standard therapy with either vancomycin (1 g every 12 hours with appropriate dose adjustment) or an antistaphylococcal penicillin (nafcillin, oxacillin, or flucloxacillin) (2 g every 4 hours), depending on the susceptibility of the causative strain to methicillin. The duration of therapy was determined by the investigator on the basis of the working diagnosis. All patients who were randomly assigned to standard treatment and patients with left-sided endocarditis who were assigned to daptomycin were to receive gentamicin (1 mg per kilogram intravenously every eight hours or adjusted on the basis of renal function) for the first four days.
The investigators evaluated patients at baseline, at the end of therapy, and 42 days after the end of therapy (at the test-of-cure visit). Blood cultures were obtained daily until they were negative, as well as at the end of therapy and 42 days after the end of therapy (Figure 1 of the Supplementary Appendix). Patients were required to undergo transesophageal echocardiography within five days after starting the study medication. All echocardiograms were reviewed by an independent expert unaware of the patient's treatment assignment; these results were used in all analyses.
An independent data and safety monitoring committee was established to regularly review the data and make recommendations regarding continuation of the study.
Clinical Outcomes
The primary outcome was the clinical success rate in each of the two treatment groups in the modified intention-to-treat population at the visit 42 days after the end of therapy. Failure at this visit was defined as clinical failure, microbiologic failure, death, failure to obtain blood culture, receipt of potentially effective nonstudy antibiotics, or premature discontinuation of the study medication because of clinical failure, microbiologic failure, or an adverse event.
Microbiologic Methods
The species of all isolates was identified, and each isolate was tested for antimicrobial susceptibility by a central laboratory with the use of accepted interpretative criteria.17 A minimal inhibitory concentration (MIC) of at least 2 µg of daptomycin per milliliter was considered to indicate a nonsusceptible isolate of S. aureus. S. aureus isolates from patients with persistent or recurrent infection were evaluated with repeated susceptibility testing and pulsed-field gel electrophoresis.
Statistical Analysis
Efficacy Assessments
In this noninferiority trial, the 95 percent confidence interval for the difference in success rates (daptomycin minus standard therapy) was calculated on the basis of the normal approximation to the binomial distribution. The noninferiority test was based on the lower bound of the confidence intervals being within the prespecified noninferiority margin of 20 percent and the upper bounds containing 0 percent. Assuming 65 percent efficacy in both treatment groups, a statistical power of 80 percent, and a one-sided significance level of 0.025, we estimated that 90 patients would need to be enrolled in each treatment group in the modified intention-to-treat population in order to test the null hypothesis (that the treatments differed by at least 20 percent). All reported P values were two-sided and not adjusted for multiple analyses.
Safety Assessments
The incidence of adverse events was compared between treatment groups with the use of Fisher's exact test. Changes in laboratory values, in particular, changes in serum creatine kinase and creatinine clearance, were evaluated. Creatinine clearance was calculated at baseline and at regular intervals thereafter according to the Cockcroft–Gault formula.18
The study was designed by the sponsor (Cubist Pharmaceuticals) with the assistance of Drs. Chambers, Fowler, and Karchmer. Data were held and analyzed by the sponsor. However, the authors had complete and unfettered access to the data and vouch for the veracity and completeness of the data and analyses used for this article.
Results
Patients
Of the 246 randomized patients, 10 (6 assigned to daptomycin and 4 assigned to standard therapy) did not receive the study drug (Figure 1 of the Supplementary Appendix); thus, 236 — 120 assigned to daptomycin and 116 assigned to standard therapy — were treated at 44 sites in four countries and represent the population used in the safety analyses. These 10 patients plus 1 patient who was treated with standard therapy and who enrolled before the amendment to allow patients with a high likelihood of left-sided endocarditis were excluded from the modified intention-to-treat population (120 assigned to daptomycin and 115 assigned to standard therapy). Risk factors for S. aureus infection were similar in the two groups (Table 1). Clinical evidence of the systemic inflammatory response syndrome was present in 89 of 120 patients treated with daptomycin (74.2 percent) and in 87 of 115 patients treated with standard therapy (75.7 percent). The median duration of therapy was 14 days for daptomycin and 15 days for standard therapy (including early discontinuations). Gentamicin was administered to 107 of 115 patients in the standard-therapy group (93.0 percent) and to 1 patient in the daptomycin group (0.8 percent) for a median of four days. At entry, 75.0 percent of the daptomycin group and 79.1 percent of the standard-therapy group met modified Duke criteria for definite or possible endocarditis. The final diagnoses were distributed similarly in the treatment groups (Table 1). MRSA was isolated from 45 of 120 patients who were treated with daptomycin (37.5 percent) and 44 of 115 patients who were treated with standard therapy (38.3 percent).
Outcome
A successful outcome was documented 42 days after the end of therapy for 53 of 120 patients in the daptomycin group, as compared with 48 of 115 patients in the standard-therapy group (44.2 percent vs. 41.7 percent; absolute difference, 2.4 percent; 95 percent confidence interval for the difference, –10.2 to 15.1 percent). The lower limit of the 95 percent confidence interval was within the prespecified margin of –20 percent; thus, daptomycin was not inferior to standard therapy. Differences in success rates were similar across all of the prespecified analyses (Table 2 and Figure 1). Success rates favored daptomycin over vancomycin among patients who were infected with MRSA (44.4 percent for daptomycin vs. 31.8 percent for standard therapy, P=0.28) but were higher among patients receiving standard therapy for MSSA infection (44.6 percent for daptomycin vs. 48.6 percent for standard therapy, P=0.74). At the end of therapy, the success rates were 61.7 percent in the daptomycin group, as compared with 60.9 percent in the standard-therapy group (74 of 120 patients vs. 70 of 115 patients; absolute difference, 0.8 percent; 95 percent confidence interval, –11.7 to 13.3 percent). Eighty-five percent of patients treated with daptomycin and 84 percent of patients treated with standard therapy survived to the end of the study (P=0.98).
|
|
The median length of time to clearance of MRSA bacteremia (eight days and nine days, respectively; P=0.25) and MSSA bacteremia (four days and three days, respectively; P=0.28) did not differ significantly between the daptomycin and standard-therapy groups.
Reasons for Treatment Failure
The overall rates of failure of daptomycin and standard therapy were similar (55.8 percent and 58.3 percent, respectively) (Table 3). The reasons for failure, however, were different. Failure of daptomycin treatment was more frequently attributable to persistent or relapsing S. aureus infection (19 of 120 patients [15.8 percent] vs. 11 of 115 patients [9.6 percent] in the standard-therapy group, P=0.17). In contrast, failure of standard therapy was somewhat, but not significantly more frequently, associated with treatment-limiting adverse events (occurring in 17 patients [14.8 percent] vs. 8 patients [6.7 percent] in the daptomycin group, P=0.06).
|
A sensitivity analysis was performed to define the relative contribution of individual reasons for failure. This analysis demonstrated similar success rates in the two groups when each reason for failure was considered, including failure of efficacy (e.g., microbiologic failure, death, or clinical failure) (Table 4). In addition, the overall success rates for daptomycin (49.2 percent) and standard therapy (48.7 percent) were similar when treatments that failed only because of treatment-limiting adverse events were classified as successes (absolute difference, 0.5 percent; 95 percent confidence interval, –12.3 to 13.3 percent) or were excluded from the analysis (data not shown).
|
The overall incidence of adverse events in the two treatment groups was similar (Table 5). Most events were considered by the investigators to be unrelated to study treatment and were mild to moderate in severity. Of the 120 patients who received daptomycin, 62 (51.7 percent) had a serious adverse event, as did 52 of 116 patients who received standard therapy (44.8 percent). Adverse events that occurred in at least 5 percent of patients in either treatment group are listed in Table 2 of the Supplementary Appendix. All differences in the rates of adverse events between groups reaching a significance level of P
0.05 are provided in Table 3 of the Supplementary Appendix. No statistical adjustments for multiple comparisons were made.
|
Eleven of 120 patients who received daptomycin had adverse events related to the peripheral nervous system (e.g., paresthesias, dysesthesias, and peripheral neuropathies) (9.2 percent), as compared with 2 of 116 patients who received standard therapy (1.7 percent, P=0.02). All of the events were classified as mild to moderate in severity; most were short-lived and resolved during continued treatment.
As compared with patients who received daptomycin, significantly more patients who received standard therapy had renal impairment as an adverse event, defined by the investigators as interstitial nephritis, toxic nephropathy, acute prerenal failure, acute or chronic renal failure, renal impairment, or renal tubular necrosis (18.1 percent vs. 6.7 percent, P=0.009) or on the basis of worsening creatinine clearance (46.8 percent vs. 19.8 percent, P<0.001). The incidence of renal impairment was similar among patients who received gentamicin and vancomycin (20.4 percent) and patients who received gentamicin and an antistaphylococcal penicillin (18.6 percent). In the safety analysis, renal impairment resulted in the discontinuation of treatment in 5 of 116 patients in the standard-therapy group (4.3 percent) and in 1 of 120 patients in the daptomycin group (0.8 percent). The proportion of patients with clinically significant decreases in renal function during treatment (a decrease in the creatinine clearance to less than 50 ml per minute or a decrease of more than 10 ml per minute from a baseline creatinine clearance of less than 50 ml per minute) was significantly higher in the standard-therapy group than in the daptomycin group by day 7 (14.2 percent vs. 5.2 percent, P=0.03) and through the end of the study (26.3 percent vs. 11.0 percent, P=0.004). Among patients who had clinically significant decreases in renal function, these decreases were reversible (defined by a final creatinine clearance within 10 ml per minute of the baseline value) in 5 of 13 patients who received daptomycin (38.5 percent) and 14 of 30 patients who received standard therapy (46.7 percent; 10 of 19 patients received an antistaphylococcal penicillin [52.6 percent], and 4 of 11 patients received vancomycin [36.4 percent]).
Discussion
Our findings demonstrate that daptomycin is not inferior to standard therapy for the treatment of S. aureus bacteremia and right-sided endocarditis caused by MSSA or MRSA. Our findings were consistent among all efficacy populations and at assessments at both the end of therapy and 42 days after the end of therapy. The study population had high rates of complicated infections, coexisting conditions, and MRSA bacteremia.3,4,5,6
The growing problem of MRSA bacteremia and endocarditis is magnified by reports of clinical failure19,20 and resistance21,22,23,24 to vancomycin. The similar success rates among patients with MRSA bacteremia and right-sided endocarditis treated with daptomycin and such patients who were treated with vancomycin suggest that daptomycin may be considered an alternative to vancomycin in the management of these serious infections.
Although elevations of creatine kinase were documented in one quarter of the patients in the daptomycin group who had normal levels of creatine kinase at baseline, only three patients withdrew from the study for this reason. In our study, clinically significant daptomycin-associated elevations of creatine kinase occurred in 6.7 percent of recipients. Patients who are taking daptomycin should be monitored for elevations in creatine kinase and skeletal-muscle dysfunction.
Patients in the standard-therapy group may have had significantly greater increases in serum creatinine because of the addition of gentamicin to the treatment regimen.25 This observation suggests that even a few days of low-dose gentamicin may result in considerable renal dysfunction in this high-risk population. Our findings provide support for the position of recent guidelines26 that the use of gentamicin in the treatment of native-valve S. aureus endocarditis should be considered optional. However, our study was not designed to examine the contribution of gentamicin to the efficacy or toxicity of standard treatment.
Nineteen of 120 patients who received daptomycin (15.8 percent) had microbiologic failure, defined as persistent or relapsing S. aureus infection. In six of these patients, S. aureus isolated during or after treatment became less susceptible to daptomycin. In contrast, 11 of 115 patients (9.6 percent) who received standard therapy had microbiologic failure, including 9 of 53 patients who received vancomycin and 2 of 62 patients who received an antistaphylococcal penicillin. In four of these nine patients who received vancomycin, the MIC of vancomycin increased to 2 µg per milliliter from baseline values of 0.5 or 1 µg per milliliter. Most patients with persistent or relapsing infection had complicated bacteremia associated with osteomyelitis or indwelling prostheses. Given the complicated nature of these infections, the contribution of decreased antimicrobial susceptibility to clinical failure is difficult to determine. However, these observations, as well as reports of clinical failures associated with reduced susceptibility to vancomycin20 and daptomycin,27 underscore the importance of adjunctive therapy, especially surgical intervention, in optimizing outcomes.
Because of the strict definition for success based on findings 42 days after the completion of therapy, treatment in many cases failed for reasons other than lack of efficacy. For example, not having blood cultures drawn at the visit 42 days after the end of therapy, despite a lack of evidence of persistent or relapsing S. aureus infection, contributed to the high failure rates seen in both groups. In addition, gentamicin therapy may have contributed to the likelihood of both microbiologic success and treatment failure associated with adverse effects in the patients who received standard therapy. However, the overall success rates 42 days after the end of therapy in both groups were similar when treatments that failed only because of treatment-limiting adverse events were classified as successes and when outcomes were assessed according to measures of efficacy (e.g., microbiologic failure, death, or clinical failure).
The open-label nature of the study might have led to bias if investigators were more likely to withdraw a patient in one of the treatment groups than in the other.28 The effect of this potential bias was minimized by including an objective microbiologic end point as part of the assessment and by the use of a blinded adjudication committee to establish the final outcome. Another limitation was the small number of patients with left-sided S. aureus endocarditis and the poor clinical outcome among these patients in both treatment groups. Additional studies are required to optimize medical and surgical therapies in the management of this life-threatening infection.
In conclusion, our results suggest that daptomycin at a dose of 6 mg per kilogram once daily is not inferior to standard therapy for the treatment of bacteremia and right-sided endocarditis caused by MSSA or MRSA.
Supported by Cubist Pharmaceuticals.
Drs. Fowler, Boucher, Corey, Abrutyn, Karchmer, Rupp, Levine, Chambers, Cabell, Link, Price, and Cosgrove report having served as consultants for Cubist Pharmaceuticals. Drs. Tally and Vigliani are employees of Cubist Pharmaceuticals. Dr. Fowler reports having received grant support from Inhibitex, Merck, Nabi, and Theravance; Dr. Corey, from Inhibitex, Merck, and Theravance; Dr. Abrutyn, from Tenet Healthcare Foundation; Dr. Karchmer, from Cubist, Merck, Theravance, and Pfizer; Dr. Rupp, from Becton Dickinson and 3M; Dr. Levine, from Cubist; Dr. Chambers, from Cubist and Ortho-McNeil; Dr. Filler, from Amgen; Dr. Zervos, from Merck and Elan; Dr. Parsonnet, from Procter & Gamble; Dr. Bernstein, from Theravance, Cubist, and Peninsula; Dr. Price, from Denver Health Medical Center, Fujisawa, GlaxoSmithKline, and Cubist; Dr. Rehm, from Cubist; Dr. Tice, from Cubist, Roche, and Merck; and Dr. Cosgrove, from Merck. Dr. Fowler reports having received lecture fees from Cubist, Nabi, and Pfizer; Dr. Boucher, from Pfizer, Schering-Plough, and Cubist; Dr. Karchmer, from Cubist; Dr. Rupp, from Aventis, Ortho-McNeil, Cubist, Schering-Plough, and Bayer; Dr. Levine, from Cubist; Dr. Chambers, from Basilea and Pfizer; Dr. Link, from Cubist; Dr. Zervos, from Pfizer, Aventis, Cubist, and Wyeth; Dr. Bernstein, from Cubist and Ortho-McNeil; Dr. Price, from Ortho-McNeil; Dr. Forrest, from AdvanDx; Dr. Rehm, from Wyeth and Cubist; and Dr. Tice, from Cubist. No other potential conflict of interest relevant to this article was reported.
Source Information
From Duke University Medical Center, Durham, N.C. (V.G.F., G.R.C., C.H.C.); Tufts New England Medical Center (H.W.B.) and Beth Israel Deaconess Medical Center (A.W.K.) — both in Boston; Drexel University College of Medicine, Philadelphia (E.A.); University of Nebraska Medical Center, Omaha (M.E.R.); Wayne State University School of Medicine, Detroit (D.P.L.); San Francisco General Hospital, San Francisco (H.F.C.); Cubist Pharmaceuticals, Lexington, Mass. (F.P.T., G.A.V.); Forsyth Medical Center, Winston-Salem, N.C. (A.S.L.); Onze Lieve Vrouw Ziekenhuis, Aalst, Belgium (I.D.); Harbor–UCLA Medical Center, Torrance, Calif. (S.G.F.); William Beaumont Hospital, Royal Oak, Mich. (M.Z.); East Carolina University, Greenville, N.C. (P.C.); Dartmouth–Hitchcock Medical Center, Lebanon, N.H. (J.P.); Veterans Affairs Medical Center, Dayton, Ohio (J.M.B.); Denver Health Medical, Denver (C.S.P.); University of Maryland School of Medicine (G.N.F.), Johns Hopkins University School of Medicine (S.E.C.) — both in Baltimore; Klinikum, University of Cologne, Cologne, Germany (G.F.); Lehigh Valley Hospital Trauma and Critical Care Research, Allentown, Pa. (M.G.); Cleveland Clinic Foundation, Cleveland (S.J.R.); Johann Wolfgang Goethe Universität, Frankfurt, Germany (H.R.B.); and University of Hawaii, and Queens Medical Center — both in Honolulu (A.T.).
Address reprint requests to Dr. Fowler at Box 3281, Division of Infectious Diseases, Duke University Medical Center, Durham, NC 27710, or at vance.fowler{at}duke.edu.
References
In addition to the authors, the following persons participated in the S. aureus Endocarditis and Bacteremia Study: Patient Enrollment: Amarillo, Tex. — R. Akins; Atlanta — H. Albrecht; Denver — M. Barron, M. Bessesen; Fargo, N.D. — P. Carson; New York — D. Fierer; Danville, Pa. — M. Foltzer; St. Louis — D. Goodenberger; Winston-Salem, N.C. — K. High; Manhasset, N.Y. — B. Hirsch; Besançon CEDEX, France — B. Hoen; Buffalo, N.Y. — C. Hsiao; San Francisco — H. Lampiris; Philadelphia — M. Levison, B. Suh; Bethlehem, Pa. — T.P. Le; Leuven, Belgium — W. Peetermans; Chicago — D. Pitrak, J.P. Quinn, J. Segreti; Louisville, Ky. — J. Ramirez; Boston — P.A. Rice; Regensburg, Germany — B. Salzberger; Cumberland, Md. — J. Stone; Akron, Ohio — J. Tan; Rochester, Minn. — Z. Temesgen; Data and Safety Monitoring Committee: New York — K. Freeman; Charleston, S.C. — J. John, Jr.; Springfield, Mass. — D. Longworth; Cubist Pharmaceuticals Study Team: J. Alder, D. Anastasiou, R. Arbeit, J. Aziz, K. Blackerby, E. Campanaro, M. Campion, B. Eisenstein, J. Lufkin, S. Luperchio, C. Otradovec, C. Rundlett, U. Stoutenburgh, G. Thorne, P. Tipirneni.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Therapy for Methicillin-Resistant Staphylococcus aureus
Siegman-Igra Y., Torres-Tortosa M., Caballero-Granado F. J., Canueto J., Jetton L., Cosgrove S. E., Fowler V. G. Jr., Boucher H. W., Moran G. J., Talan D. A.
Extract |
Full Text |
PDF
N Engl J Med 2006;
355:2153-2155, Nov 16, 2006.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
