Nitrous Oxide and Serious Long-term Morbidity and Mortality in the Evaluation of Nitrous Oxide in the Gas Mixture for Anaesthesia (ENIGMA)-II Trial

HEALTHCARE research is increasingly focused on longer-term outcomes that are important to patients and the community.¹  For example, the long-term follow-up of the Evaluation of Nitrous Oxide in the Gas Mixture for Anaesthesia (ENIGMA) trial revealed that nitrous oxide increased risk of myocardial infarction in noncardiac surgery patients (odds ratio [OR], 1.59; 95% CI, 1.01 to 2.51; P = 0.04).²  This effect had not been evident at the 30-day follow-up (OR, 0.58; 95% CI, 0.22 to 1.50; P = 0.26).³  However, because the patients in the ENIGMA trial were not selected on the basis of the risk for cardiovascular events and because the event rates were low, both the 30-day and long-term results required confirmation in suitably powered studies.

Therefore, we conducted the ENIGMA-II trial to explore the risks and benefits of nitrous oxide in noncardiac surgery patients with known or suspected cardiovascular disease.⁴  There were no significant differences between patients receiving and not receiving nitrous oxide in terms of their 30-day risk of a composite primary outcome of death or major cardiovascular events.⁵  The aim of the current study was to determine the effect of nitrous oxide administration on the composite primary outcome and secondary outcomes (death, disability, myocardial infarction, and stroke) at 1 yr after surgery in the ENIGMA-II patients.

The ENIGMA-II trial was an international, parallel-group, patient- and observer-blinded, randomized trial. A total of 7,112 noncardiac surgery patients, older than 45 yr and at risk of perioperative cardiovascular complications, were enrolled. ENIGMA-II was registered at ClinicalTrials.gov (number: NCT00430989; principal investigator: P.S.M.; date of registration: January 31, 2007). The protocol⁴  and results⁵  of the 30-day follow-up have been published. The trial steering committee prospectively approved the protocol for the 1-yr follow-up, including the detailed statistical analysis plan. Ethics committee approval for the 1-yr follow-up study was obtained at 39 of the 45 participating sites, and patients at these 39 sites consented to the 30-day and 1-yr follow-ups before randomization. Participating sites, investigators, and names and locations of ethics committees are listed in the appendix.

Participating patients were randomly assigned to 70% N₂O in 30% O₂ or 70% N₂ in 30% O₂ groups. Perioperative care was otherwise at the discretion of the attending anesthesiologists. Patients were monitored with 12-lead electrocardiographs preoperatively and on postoperative days 1 and 3 and with cardiac biomarkers (troponin or, if unavailable, creatine kinase-myocardial band) at 6 to 12 h and 1 to 3 days after surgery. Other investigations were ordered as clinically indicated during the 1-yr follow-up period.

One-year follow-up was conducted via medical record review and telephone interview. The medical record was interrogated for the date and cause of death and the occurrence of myocardial infarction or stroke, any time between 30 days and 1 yr after surgery. The telephone interview was conducted with the patient or their relatives or doctors if the patient had died, was incapacitated, was unavailable, or was unsure about the occurrence of myocardial infarction or stroke.

The primary outcome for the 1-yr follow-up was a composite of death and cardiovascular events (nonfatal myocardial infarction, cardiac arrest, pulmonary embolism, and stroke). The main preprescribed secondary endpoint was disability or death (the inverse of disability-free survival). Disability was defined as a Katz index of independence in activities of daily living score less than 8.⁶ 

Other secondary outcomes were death, fatal or nonfatal myocardial infarction, and fatal or nonfatal stroke. These were defined by patient, relative, or doctor report or by fulfillment of the following criteria, as recorded in the medical record: myocardial infarction was defined by an increased cardiac biomarker level plus at least one of the ischemic symptoms, pathological Q waves, electrocardiographic changes indicative of ischemia, coronary artery intervention, new wall motion abnormality on echocardiography or a fixed defect on radionucleotide scanning, or autopsy finding of new or recent myocardial infarction.⁷  The troponin threshold that was considered abnormal was each site’s laboratory’s 99th percentile (upper reference limit) of a healthy reference population.⁸  Stroke was defined as a new neurologic deficit persisting for 24 h or longer, confirmed by a neurologist or computed tomography or magnetic resonance imaging.

The following preoperative and intraoperative characteristics were selected prospectively as covariates in the models: age, sex, body mass index (BMI), American Society of Anesthesiologists’ (ASA) physical status, good exercise capacity (≥ 4 metabolic equivalents), history of coronary artery disease, emergency surgery, vascular surgery, randomized treatment (nitrous oxide or no nitrous oxide), propofol maintenance, regional local anesthetic block, volatile anesthetic administration (minimum alveolar concentration [MAC] equivalents), bispectral index (BIS) monitoring, and duration of anesthesia.

The patients followed up at 1 yr were not necessarily a truly representative sample of the original ENIGMA-II cohort, so a logistic regression model was fitted to estimate the probability that each patient was followed up at 1 yr. For each outcome model, observations were weighted by the inverse of these probabilities. Patients who died or experienced myocardial infarction or stroke within the 30-day follow-up of ENIGMA-II were given a weight of 1. Unweighted models were fitted as sensitivity analyses.

Mortality rates were computed for each category of each covariate and were expressed as deaths per 1,000 person-years. Univariate Cox proportional hazard models were used to define hazard ratios (HRs) and 95% CIs for death. Multivariable Cox proportional hazard models for death were constructed, and assessments of proportionality of hazard functions were performed. Preoperative variables were first adjusted for each other. Then, nitrous oxide, propofol maintenance, regional local anesthetic block, and BIS monitoring were adjusted for each other and preoperative variables. Finally, volatile anesthetic administration less than the median MAC value in patients receiving volatile anesthetic maintenance and duration of anesthesia were adjusted for each other, preoperative variables, nitrous oxide, and BIS monitoring (propofol maintenance was not included because patients who received propofol for maintenance had missing data for volatile anesthetic administration).

Because the date of outcomes other than death was sometimes imprecise or missing, logistic regression was used to compute ORs and 95% CIs for the primary composite outcome, the main secondary outcome of disability or death, and further secondary outcomes of death, myocardial infarction, and stroke. Multivariable models were constructed using the technique described in the previous paragraph. The preplanned assessment of the interaction of each variable with nitrous oxide was performed by using interaction terms in the weighted and unweighted regression models. An additional analysis for stroke was conducted that included a history of previous stroke or transient ischemic attack.

Analyses were conducted using Stata 12 (Stata Corporation, USA). All P values are two sided, and P value less than 0.05 was considered statistically significant.

One-year follow-up occurred between June 2009 and October 2014, with a median follow-up time of 386 days (interquartile range, 366 to 458 days). The centers participating in the 1-yr follow-up study recruited 6,651 (95%) of the 6,992 patients who were assessed for the primary outcome at 30 days.⁵  Follow-up data were available for 5,844 (88%) of these patients (fig. 1).

A total of 435 patients (7.4%) died; 99 died before 30 days and 336 subsequently (fig. 2) (between treatment P = 0.18). The causes of death were cancer (37.5%), myocardial infarction (4.4%), stroke (4.4%), other cardiovascular death (14.3%), respiratory failure (7.8%), sepsis (13.0%), other causes (8.7%), and unknown (9.9%). In total 641 patients (10.5%) were recorded as having disability or had died (206 [3.5%] had disability), 514 patients (8.8%) were recorded as having a fatal or nonfatal myocardial infarction, and 111 patients (1.9%) were recorded as having a fatal or nonfatal stroke during the 1-yr follow-up period.

Nitrous oxide did not increase the risk of the primary outcome (OR, 1.08; 95% CI, 0.94 to 1.25; P = 0.27) (table 1). Age, BMI, ASA physical status, exercise capacity, coronary artery disease, emergency surgery, propofol maintenance, regional local anesthetic block, and duration of anesthesia were significant predictors of the primary outcome. The adjusted ORs were largely unaffected by weighting to adjust for missing data (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 1). The interaction between nitrous oxide and age was statistically significant (P = 0.046) in the weighted model for the primary outcome, but this did not otherwise change the main result (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 2).

Nitrous oxide did not increase the risk of disability or death (OR, 1.07; 95% CI, 0.90 to 1.27; P = 0.44) (table 2). Age, BMI, exercise capacity, emergency surgery, and duration of anesthesia were significant predictors of death. The adjusted ORs were largely unaffected by weighting to adjust for missing data (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 3). There was no significant interaction between nitrous oxide administration and any covariate in either the weighted or unweighted models for disability or death (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 4).

Nitrous oxide did not increase the risk of death (HR, 1.17; 95% CI, 0.97 to 1.43; P = 0.11) (table 3). Age, BMI, ASA physical status, exercise capacity, emergency surgery, vascular surgery, and duration of anesthesia were significant predictors of death. The adjusted HRs were largely unaffected by weighting to adjust for missing data (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 5). The interactions between nitrous oxide and exercise capacity (P = 0.035), and nitrous oxide and MAC equivalents (P = 0.026), were statistically significant in the unweighted model for death (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 6).

Nitrous oxide did not increase the risk of myocardial infarction (OR, 0.97; 95% CI, 0.81 to 1.17; P = 0.78) (table 4). Age, ASA physical status, exercise capacity, coronary artery disease, emergency surgery, vascular surgery, regional local anesthetic block, BIS monitoring, and duration of anesthesia were significant predictors of myocardial infarction. The adjusted ORs were largely unaffected by weighting to adjust for missing data (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 7). The interaction between nitrous oxide and BIS monitoring was statistically significant in the weighted model for myocardial infarction (P = 0.045) (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 8).

Nitrous oxide did not increase the risk of stroke (OR, 1.08; 95% CI, 0.74 to 1.58; P = 0.70) (table 5). Age and vascular surgery were significant predictors of stroke. The adjusted ORs were largely unaffected by weighting to adjust for missing data (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 9). There was no significant interaction between nitrous oxide administration and any covariate in either the weighted or unweighted models for stroke (Supplemental Digital Content 1, https://links.lww.com/ALN/B204, table 10). A history of previous stroke or transient ischemic attack was a significant predictor of stroke within 1 yr of surgery (adjusted OR, 2.43; 95% CI, 1.64 to 3.61; P < 0.005). Adjusted ORs for other covariates were largely unaffected by inclusion of a history of previous stroke or transient ischemic attack (results not shown).

We found that nitrous oxide did not increase the incidence of a composite of death or major cardiovascular complications at 1 yr after surgery in patients enrolled in the ENIGMA-II trial. These results are consistent with our findings at 30 days postoperatively and further support the safety of nitrous oxide administration in patients with known or suspected cardiovascular disease.

The ENIGMA-II trial was established on the premise that the pathophysiological effects of nitrous oxide might lead to major cardiovascular events after noncardiac surgery.^4,9  These effects include the inhibition of methionine synthase with resulting hyperhomocysteinemia^10,11  and endothelial dysfunction,^12,13  especially in patients with genetic¹⁴  or dietary¹⁵  predispositions. However, the recent Vitamins in Nitrous Oxide (VINO) randomized trial¹⁶  did not confirm earlier findings¹⁷  of myocardial ischemia among patients exposed to nitrous oxide or an ameliorating effect of vitamin B₁₂ and folate administration on postoperative troponin increases. The VINO study and our ENIGMA-II analyses provide strong evidence to refute the hypothesis that the hyperhomocysteinemia associated with nitrous oxide administration leads to adverse cardiovascular outcomes.

The ENIGMA-II 1-yr follow-up results contrast with our finding in the long-term follow-up of the ENIGMA study²  of an increased long-term risk of myocardial infarction in patients who were randomly assigned to nitrous oxide group. There were important differences between ENIGMA and ENIGMA-II: (1) the inclusion criteria (unselected in ENIGMA and selected for cardiovascular risk factors in ENIGMA-II); (2) the percentage of inspired oxygen administered (30% in the nitrous oxide group and 80% in the no nitrous group in ENIGMA and 30% in both groups in ENIGMA-II); and (3) the duration of follow-up (3.5 yr [range, 0 to 5.7 yr] in ENIGMA and 1.06 yr [range, 0 to 3.89 yr] in ENIGMA-II). Most likely, though, differences in long-term outcomes simply reflect the smaller sample size of the ENIGMA follow-up (n = 1,660), leading to a spurious finding in that study.¹⁸ 

The incidences of death (7.4%), disability (3.5%), myocardial infarction (8.8%), and stroke (1.9%) in this study are consistent with the inclusion criteria for ENIGMA-II and previously published studies of noncardiac surgery patients with cardiovascular disease.^19–23  In relation to myocardial infarction in particular, these studies point to substantial scope for improved outcomes through primary prevention, early detection, treatment, and prevention of complications.²⁴  Unfortunately, no primary prevention measures for perioperative myocardial infarction in noncardiac surgical patients are conclusively proven to be both effective and safe,^25–28  including the omission of nitrous oxide.

The World Health Organization defines disability as “difficulties in any area of functioning as they relate to environmental and personal factors.”²⁹  We chose to use the Katz score, which measures physical functioning, to determine the disability in the ENIGMA-II trial.⁶  Among the widely used scales, we now recommend the World Health Organization Disability Assessment Schedule for this purpose, as it includes cognition, interpersonal relationships, participation in society, self-care, work and household roles, and mobility.³⁰ 

Many covariates were associated with increased risk of the primary and/or secondary outcomes of this study, including increasing age,^2,19,31–33  low BMI,^19,31,32,34  higher ASA physical status,^2,19,31–33  lower exercise capacity,³⁵  coronary artery disease,^2,32  emergency surgery,^2,19,33  propofol maintenance,²  regional local anesthetic block,³⁶  lack of BIS monitoring, and longer duration of anesthesia.^2,19  Although analyses were sequentially adjusted for pre- and intraoperative factors, it is possible that some of these associations resulted from selection bias or residual confounding, in particular the interventions at the discretion of the attending anesthesiologist where the factors that impacted on the decision to use the intervention may not have been captured completely. For example, lack of BIS monitoring was associated with an increased long-term risk of myocardial infarction in this follow-up study but was not associated with major cardiovascular events in longer-term follow-up studies of patients who were randomly assigned to receive or not receive BIS-guided anesthesia.^19,32  Therefore, this result should be interpreted with caution.

This study is one of the largest randomized controlled trials to follow noncardiac surgery patients with risk factors for major cardiovascular events over an extended postoperative period.^2,19  Our results provide a robust estimate of the incidence of death and effect of nitrous oxide on survival after surgery. The different directions of effect on death at the 30-day⁵  and 1-yr follow-ups (toward a protective effect at 30 days and toward a harmful effect at 1 yr) likely arise from random error. The study is limited with respect to myocardial infarction and stroke because patients were not specifically screened for these events during the period between 30 days and 1 yr, and we relied on medical record, patient or surrogate reports. The number and scope of covariates available for adjustment in the model further limit the study. Finally, a number of centers were unable to participate in the 1-yr follow-up study due to limited resources, resulting in a 5% rate of missing 1-yr outcomes. Although we adjusted for missing 1-yr outcomes in our analyses, and results were similar in weighted and unweighted analyses, a subtle effect cannot be ruled out.

In conclusion, nitrous oxide administration did not increase the incidence of a composite of death or major cardiovascular events during a 1-yr follow-up period in patients randomly assigned to the ENIGMA-II trial. Nitrous oxide can be safely administered to patients with known or suspected cardiovascular disease undergoing noncardiac surgery.

This study was supported by Australian and New Zealand College of Anaesthetists Project Grants 10/014 and 12/008 (Melbourne, Victoria, Australia). The Evaluation of Nitrous Oxide in the Gas Mixture for Anaesthesia (ENIGMA)-II trial was supported by a National Health and Medical Research Council Project Grant (435015) (Canberra, Australian Capital Territory, Australia); a Research Grant Council General Research Fund grant (461409) (Hong Kong, Special Administrative Area, People's Repubic of China); and a Health and Research Grant Council Medical Research Fund grant (11121051) (Hong Kong, Special Administrative Area, People’s Republic of China).

The authors declare no competing interests.

Full protocol available at: kate.leslie@mh.org.au. Raw data available at: kate.leslie@mh.org.au.

Investigators: P. Myles, S. Wallace, W. Galagher, C. Farrington, and A. Ditoro.

Ethics Committee: Alfred Hospital Ethics Committee, Research and Ethics Unit, Alfred Health, P. O. Box 315, Prahran, Victoria 3181, Australia.

Investigators: P. Peyton, S. Baulch, and S. Sidiropoulos.

Ethics Committee: Austin Health Human Research Ethics Committee, Research Ethics Unit, Austin Hospital, P. O. Box 5555, Heidelberg, Victoria 3084, Australia.

Investigators: R. Bulach and D. Bryant.

Ethics Committee: Southern Health Research Directorate, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia.

Investigators: E. O’Loughlin and V. Mitteregger.

Ethics Committee: South Metropolitan Area Health Service Human Research Ethics Committee, Fremantle Hospital and Health Service, Alma Street, Fremantle, Western Australia 6160, Australia.

Investigators: S. Bolsin and C. Osborne.

Ethics Committee: The Barwon Health Research and Ethics Advisory Committee, The Geelong Hospital, P. O. Box 281, Geelong, Victoria 3220, Australia.

Investigators: R. McRae and M. Backstrom.

Ethics Committee: Human Research Ethics Committee-B, Southern Health 246 Clayton Road, Clayton, Victoria 3168, Australia.

Investigators: K. Leslie and R. Cotter.

Ethics Committee: Melbourne Health Human Research Ethics Committee, Royal Melbourne Hospital, Grattan Street, Parkville, Victoria 3050, Australia.

Investigators: M. Paech and S. March.

Ethics Committee: Royal Perth Hospital Ethics Committee, Royal Perth Hospital Wellington Street, G. P. O. Box X2213, Perth, Western Australia 6847, Australia.

Investigators: B. Silbert and S. Said.

Ethics Committee: Human Research Ethics Committee-D, St. Vincent’s Hospital, Grattan Street, Fitzroy, Victoria 3065, Australia.

Investigators: R. Halliwell and J. Cope.

Ethics Committee: Sydney West Area Health Service Human Research Ethics Committee, Westmead Hospital, Hawkesbury Road, Westmead, New South Wales 2145, Australia.

Investigators: D. Fahlbusch and D. Crump.

Ethics Committee: Calvary Health Care Adelaide Human Research and Ethics Committee, Calvary Health Care Adelaide, 89 Strangways Terrace, North Adelaide, South Australia 5006, Australia.

Investigator: G. Thompson.

Ethics Committee: Peter MacCallum Cancer Centre Ethics Committee, Peter MacCallum Cancer Centre, St. Andrews Place, East Melbourne, Victoria 3002, Australia.

Investigator: A. Jefferies.

Ethics Committee: Melbourne Health Human Research Ethics Committee, Royal Melbourne Hospital, Grattan Street, Parkville, Victoria 3050, Australia.

Investigator: M. Reeves.

Ethics Committee: Tasmanian Human Research Ethics Committee, Health and Medical Office of Research Services, University of Tasmania, Hobart, Tasmania 7001, Australia.

Investigators: N. Buckley and T. Tidy.

Ethics Committee: The Hamilton Health Sciences/McMaster Health Sciences Research Ethics Board, 293 Wellington Street, Hamilton, Ontario, L8L 8E7, Canada.

Investigators: T. Schricker, R. Lattermann, and D. Iannuzzi.

Ethics Committee: SDR Committee, McGill University Health Centre, 687 Avenue des Pins, Montreal, Quebec, H3A 1A1, Canada.

Investigators: S. Beattie and J. Carroll.

Ethics Committee: University Health Network Research Ethics Board, 700 University Avenue, Toronto, Ontario, M5G 1Z5, Canada.

Investigators: M. Jacka and C. Bryden.

Ethics Committee: Health Research Ethics Board (Biomedical Panel), Heritage Medical Research Centre, University of Alberta, Edmonton, Alberta, T6G 2S2, Canada.

Investigator: N. Badner.

Ethics Committee: The University of Western Ontario Research Ethics Board for Health Sciences Research Involving Human Subjects, The University of Western Ontario, London, Ontario, N6A 5C1, Canada.

Investigators: M. T. V. Chan (ANZCA Clinical Trials Network member) and M. W. Y. Tsang.

Ethics Committee: The Joint Chinese University of Hong Kong, New Territories East Cluster Clinical Research Ethics Committee, Prince of Wales Hospital, Shatin, Hong Kong.

Investigators: B. C. P. Cheng and A. C. M. Fong.

Ethics Committee: New Territories West Cluster Clinical and Research Ethics Committee, Tuen Mun Hospital, Tuen Mun, New Territories, Hong Kong.

Investigators: L. C. Y. Chu and E. G. Y. Koo.

Ethics Committee: New Territories West Cluster Clinical and Research Ethics Committee, Tuen Mun Hospital, Tuen Mun, New Territories, Hong Kong.

Investigators: N. Mohd and L. E. Ming.

Ethics Committee: Jawatankuasa Etika Perubatan Pusta Perubatan Universiti Malaya, Lembah Pantai, Kuala Lumpur 59100, Malaysia.

Investigators: D. Campbell and D McAllister

Ethics Committee: New Zealand-Health and disability Ethics Committees, Multi-Region Ethics Committee, Ministry of Health, Wellington, New Zealand.

Investigators: S. Walker and S. Olliff.

Ethics Committee: New Zealand-Health and Disability Ethics Committees, Multi-Region Ethics Committee, Ministry of Health, Wellington, New Zealand.

Investigators: R. Kennedy.

Ethics Committee: New Zealand-Health and Disability Ethics Committees, Multi-Region Ethics Committee, Ministry of Health, Wellington, New Zealand.

Investigators: A. Eldawlatly and T. Alzahrani.

Ethics Committee: College of Medicine Research Centre, King Saud University, P. O. Box 2925, Riyadh 1146, Saudi Arabia.

Investigators: N. Chua.

Ethics Committee: National Health Group Domain Specific Review Board, 6 Commonwealth Lane, Singapore 149547, Singapore.

Investigators: R. Sneyd and H. McMillan.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigators: I. Parkinson.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigators: A. Brennan.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigator: P. Balaji.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigators: J. Nightingale.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigators: G. Kunst.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigators: M. Dickinson.

Ethics Committee: NRES Committee, South West-Central Bristol Bristol Research Ethics Committee Centre, Whitefriars Level 3, Block B Lewin’s Mead, Bristol, BS1 2NT, United Kingdom.

Investigators: B. Subramaniam and V. Banner-Godspeed.

Ethics Committee: Committee on Clinical Investigation, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts 02215.

Investigators: D. I. Sessler, J. Liu, A. Kurz, B. Hesler, A. Y. Fu, C. Egan, A. N. Fiffick, M. T. Hutcherson, A. Turan, and A. Naylor.

Ethics Committee: Institutional Review Board, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195.

Investigators: D. Obal and E. Cooke.

Ethics Committee: Institutional Review Board, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195.