Transesophageal echocardiography (TEE) is an excellent method with which to detect thrombi in the left atrial appendage [13-19]. Its use has therefore been proposed as a way to allow cardioversion to be done earlier and more safely than would be possible with conventional therapy, which consists of a total of 7 weeks of treatment with warfarin [14-23]. Recent studies [15-19] indicate that TEE-guided cardioversion with short-term anticoagulation therapy may have several advantages over the conventional approach. These advantages include a decreased risk for embolism, which results from the avoidance of cardioversion in patients who have thrombi in the left atrial appendage [15]; a decreased risk for bleeding, which occurs because anticoagulation therapy can be briefer [19]; greater initial conversion to and long-term maintenance of sinus rhythm, which result from doing cardioversion earlier [18, 19]; and greater cost-effectiveness, which results from the decreased incidence of embolic stroke [17].

The ACUTE (Assessment of Cardioversion Using Transesophageal Echocardiography) Pilot Study was a multicenter, randomized clinical trial designed to compare TEE-guided cardioversion with conventional management of cardioversion in patients with atrial fibrillation who have cardioversion [19]. The study had two objectives: to assess the general feasibility of a TEE-guided approach to cardioversion and to determine the general safety of the TEE-guided approach by comparing its clinical outcome with those of conventional management.

Methods

Top Methods Results Discussion Author & Article Info References

Patient Selection

Patients who were candidates for electrical cardioversion were eligible for inclusion if they had atrial fibrillation, or atrial flutter with a history of atrial fibrillation, lasting longer than 2 days. Patients were excluded if they had received anticoagulant therapy for more than 7 days, had required urgent cardioversion as a result of hemodynamic instability, had had a cardioembolic event within the previous month, had contraindications to TEE or warfarin, were women with childbearing potential in whom pregnancy could not be excluded, were unable to give informed consent, or were unable to return for a follow-up visit. Our study protocol was approved by the institutional review boards at all clinical sites, and all patients provided written informed consent in advance.

Study Protocol

Patients who met the inclusion criteria were randomly assigned to receive either a conventional or a TEE-guided approach to cardioversion. Randomization was done using presealed, opaque envelopes that were computer generated and distributed to each clinical site (Figure 1). Random assignments were stratified by site and were generated in blocks of six.

View larger version (29K): [in this window] [in a new window]   Figure 1. The ACUTE (Assessment of Cardioversion Using Transesophageal Echocardiography) study protocol. Patients with atrial fibrillation were randomly assigned to receive either cardioversion with conventional therapy or cardioversion guided by transesophageal echocardiography (TEE) with brief anticoagulation therapy. The conventional therapy group received warfarin therapy for 3 weeks before and 4 weeks after cardioversion. In the TEE group, patients were stratified according to whether a thrombus was detected in either the left atrium or the left atrial appendage. Patients in the TEE group were fully anticoagulated at the time of cardioversion and for 4 weeks after cardioversion. LA = left atrial; LAA = left atrial appendage.

The conventional approach to cardioversion was that recommended by the American College of Chest Physicians: 3 weeks of therapeutic warfarin therapy, then cardioversion, then 4 weeks of warfarin therapy, and then a follow-up examination at the end of the 4 weeks [23]. Prothrombin times were monitored regularly, and the target international normalized ratio (INR) was 2 to 3.

If assigned to the TEE group, patients began receiving anticoagulation therapy at their initial visit. The goal was to have patients therapeutically anticoagulated (therapeutic anticoagulation was defined as a partial thromboplastin time 1.5 to 2.5 times control values or an INR of 2.0 to 3.0) at the time of and after the planned cardioversion, for a total of 4 weeks of therapy. The initial choice of antithrombotic agent was determined by whether the patient was an inpatient or an outpatient at the time of randomization: Heparin was used for inpatients; warfarin was administered to outpatients. Transesophageal echocardiography, with subsequent cardioversion within 24 hours, was then scheduled as soon as stable therapeutic anticoagulation was assured. For example, if a patient was hospitalized and intravenous heparin therapy was administered, TEE was done as soon as a stable therapeutic partial thromboplastin time could be documented (for 24 to 36 hours); subsequent cardioversion was done if the presence of a thrombus was excluded. A 4- to 5-day overlap of warfarin therapy and intravenous heparin therapy was often necessary to maintain adequate anticoagulation after cardioversion. If the patient was to be managed as an outpatient, warfarin therapy was initiated on the day of study enrollment, and TEE and subsequent possible cardioversion were scheduled for at least 5 to 7 days later. Again, cardioversion was done when the patient was therapeutically anticoagulated, and all patients received maintenance therapy with warfarin for 4 weeks after cardioversion [15]. In the TEE group, cardioversion was done immediately after or within 24 hours of TEE because of the potential for thrombus formation in the period between TEE and cardioversion.

If thrombi were detected in the left or right atrial appendages or atrial cavities, cardioversion was postponed and the patient received warfarin therapy for 4 weeks. After 4 weeks, TEE was repeated and, if no thrombus was detected, cardioversion was done. If a thrombus was still present, another 4-week course of warfarin therapy was administered and cardioversion was not done [15].

Clinical Outcomes

Our feasibility outcomes were frequency of cardioversion, frequency of cardioversion occurring as scheduled, time to cardioversion, and time to sinus rhythm. Our clinical safety outcomes were clinically apparent ischemic stroke, transient ischemic attack, systemic embolization, deaths related to cardioversion or episodes of bleeding, and detected episodes of clinical hemodynamic instability (worsening congestive heart failure or hypotension) that rendered the patient unable to complete the protocol. Other outcome variables were the prevalence of thrombi, the number of patients without thrombi who had early cardioversion, and the immediate and follow-up rhythms after cardioversion.

These outcomes were assessed for as long as 4 weeks after cardioversion but for no longer than 8 weeks after randomization. In the patients who did not have cardioversion and who spontaneously reverted to sinus rhythm, the variables were assessed at 4 weeks after spontaneous reversion.

Study Organization and Procedures

The administrative organization of the pilot study is described in the Appendix.

Echocardiographic Examination

Conventional transthoracic echocardiography was done in both study groups using commercially available equipment. In the TEE group, TEE was done according to standard techniques using phased-array biplane or multiplane transducers [24-26]. Complete transesophageal echocardiographic examination was done, and special attention was paid to imaging the left and right atria and left and right atrial appendages to assess the presence or absence of thrombi and spontaneous echo contrast.

Echocardiographic Data Analysis

Two-dimensional directed M-mode transthoracic echocardiography was used to derive the left ventricular septal and posterior wall thicknesses and the end-diastolic, end-systolic, and left atrial dimensions. Ejection fraction was calculated using standard techniques [27, 28]. The maximal left atrial and right atrial areas were planimetered on-line, and the severity of mitral regurgitation was qualitatively graded from 0 to 4+ by using color-flow mapping [29].

A thrombus was considered to be present if a mass detected in the appendage or body of the atrium appeared to be distinct from the underlying endocardium, was not caused by pectinate muscles, and was detected in more than one imaging plane. The presence or absence of spontaneous echo contrast was analyzed and defined as dynamic intracavitary echoes with a characteristic swirling pattern distinct from artifact. The degree of spontaneous echo contrast was categorized independently as absent, mild, or severe [30, 31].

Quality Control Measures

Standard definitions of echocardiographic measurements were available to all of the clinical centers as part of a pilot operations manual. Echocardiograms at each clinical center were interpreted locally by a single physician who was highly experienced in echocardiography. Videotapes that showed the results of the first five echocardiographic examinations and all videotapes that showed thrombi were forwarded from the clinical centers to a central laboratory and overread by three experienced reviewers for consensus [19].

Electrical Cardioversion

Cardioversion was done by using the standard method of Lown and associates [4] with an initial energy of at least 40 J for atrial flutter and 200 J for atrial fibrillation.

Statistical Analysis

Summaries of clinical, echocardiographic, and outcome data are expressed as means or frequencies with 95% CIs. Data that were not normally distributed were log-transformed and presented as geometric means. Outcomes were compared for the TEE and conventional therapy groups, for patients with and without thrombus (in the TEE group only), and for patients in the TEE and conventional therapy groups who had cardioversion. These analyses were done using the t-test for independent groups for continuous variables and the Fisher exact test for categorical variables. StatXact (Cytel Software, Cambridge, Massachusetts) was used to compute binary CIs; SAS software (SAS Institute, Cary, North Carolina) was used for all other statistical calculations. Analyses were done using the intention-to-treat principle-that is, patients were analyzed in the group to which they were randomly assigned, regardless of the treatment they actually received. A P value less than 0.05 was considered statistically significant. All tests were two tailed.

Results

Top Methods Results Discussion Author & Article Info References

Baseline Characteristics

From March 1993 to January 1994, 126 patients at 10 clinical sites were enrolled and randomly assigned to either the TEE group (n = 62) or the conventional therapy group (n = 64) (Appendix). The patients in the two groups were well matched with regard to baseline clinical and echocardiographic variables (Table 1). Patients in the TEE group were more likely to have been receiving heparin (mean duration of therapy, 1.4 days) or heparin and warfarin at the time of cardioversion, whereas patients in the conventional therapy group were more likely to have been receiving warfarin (mean duration of therapy, 36 days) (Table 2). The primary disorders associated with the atrial arrhythmias in the two groups were hypertension and coronary artery disease (Table 3).

Transesophageal Echocardiography

Of the 62 patients assigned to the TEE group, 56 (90%) had TEE and 6 (10%) did not (Figure 2). Of these 56 patients, 45 (80%) had electrical cardioversion: Forty-four (96%) had it at the scheduled time, and 1 had it after a delay of more than 3 weeks that resulted in a protocol violation. Thirty-eight (84%) of the 45 patients having cardioversion had successful early cardioversion without embolization. Two of the 6 patients who did not have TEE had successful cardioversion (Figure 2).

View larger version (22K): [in this window] [in a new window]   Figure 2. Outcome of the 62 patients in the group assigned to receive transesophageal echocardiography-guided cardioversion. Numbers in parentheses are numbers of patients.

In contrast, of the 11 patients who had TEE but did not have cardioversion, 7 had thrombi, 3 spontaneously reverted to sinus rhythm (2 after receiving antiarrhythmic therapy and 1 without receiving this therapy), and 1 converted to sinus rhythm after receiving overdrive pacing (Figure 2).

Thus, 42 (75%) of the 56 patients who did not have thrombi on TEE had successful cardioversion to sinus rhythm using electrical cardioversion or other methods without prolonged anticoagulation therapy. None had an embolic event.

Mild (n = 30) or severe (n = 14) spontaneous echo contrast was present in the left atrium or the left atrial appendage, or both, in 44 (79%) of the 56 patients who had TEE.

Of the seven detected thrombi (13% of the 56 patients who had a thrombus), six were in the left atrial appendage and one was in the right atrial appendage (Table 4, Figure 3). No thrombi were detected on transthoracic echocardiography. The thrombi varied from 7 mm to 22 mm in length. Six patients had protruding or mobile thrombi, and 1 had a sessile thrombus. Mild or severe left atrial spontaneous echo contrast in the left atrial appendage and cavity was detected in 5 (83%) of the 6 patients with left atrial thrombi. Severe spontaneous contrast was detected in the right atrium of the patient who had a thrombus in the right atrial appendage. The 6 patients with left atrial thrombi had decreased left ventricular ejection fractions and a trend toward increased left atrial size compared with the 50 patients who did not have left atrial thrombi (Table 5).

View larger version (86K): [in this window] [in a new window]   Figure 3. Transesophageal echocardiograms. Top. An 8.5-mm protruding thrombus (arrow) in the right atrial appendage of a 53-year-old man. Bottom. A 9-mm left atrial appendage thrombus (arrow) in a 70-year-old man. LAA = left atrial appendage; RAA = right atrial appendage.

In three of the six patients who had thrombi in the left atrial appendage, follow-up TEE showed no residual thrombus (after a mean of 1.5 months) and cardioversion was subsequently successful. In the other three patients, residual thrombi persisted for a mean of 3 months and cardioversion was not attempted. The patient who had the thrombus in the right appendage had no residual thrombus on TEE at 3 months. Electrical cardioversion in this patient was not successful, but subsequent pharmacologic conversion with procainamide restored sinus rhythm.

Feasibility, safety, and other clinical outcomes are summarized in Table 5. Forty-seven patients (76%) had cardioversion. The time to cardioversion was much shorter in the TEE group, as was the time to sinus rhythm (Table 5). No patient in the TEE group had a clinically apparent ischemic stroke, transient ischemic attack, systemic embolization, or episode of bleeding (0% [95% CI, 0% to 5%]). Two deaths were unrelated to cardioversion. One of the patients who died had a thrombus in the left atrial appendage that resulted in postponement of cardioversion. He subsequently died of progressive heart failure. The other patient did not have a thrombus but died of respiratory failure that occurred after knee surgery. He had had successful cardioversion 2 weeks earlier. Repeated TEE was not done in either patient.

The follow-up rhythm was sinus rhythm in 34 (56% [CI, 42% to 68%]) of the 62 patients in the TEE group (Table 5).

Conventional Therapy

Of the 64 patients assigned to receive conventional therapy, 37 (58%) had electrical cardioversion (Figure 4). Twenty-six of these 37 patients (70%) had cardioversion at the scheduled time, 3 had it earlier than the scheduled time because of acute hemodynamic instability (hypotension and congestive heart failure), and 8 had it after a delay caused by bleeding complications or subtherapeutic INRs. One of these patients required a transfusion of 3 units of packed red blood cells, and cardioversion was delayed for 2 weeks. Of the 37 patients who had cardioversion, 28 (76%) had an immediately successful cardioversion that resulted in sinus rhythm.

View larger version (19K): [in this window] [in a new window]   Figure 4. Outcome of the 64 patients assigned to receive conventional therapy. Numbers in parentheses are numbers of patients.

In one patient who had electrical cardioversion as scheduled, a peripheral embolism to an upper extremity occurred 3 days after cardioversion (2% [CI, 0% to 8%]), while the patient's INR was 1.8. This patient was hospitalized for an additional 5 days and had peripheral angiography that documented the embolus, which was removed surgically. The disorder associated with the atrial fibrillation was hypertension, and the patient had normal left ventricular function and mild left atrial enlargement.

Of the 27 patients (42%) in the conventional therapy group who did not have cardioversion, 20 (74%) reverted to sinus rhythm an average of 16 days (CI, 9 to 23 days) after randomization. Of these 20 patients, 9 were receiving antiarrhythmic therapy (18 days [CI, 7 to 28 days]) and 11 converted spontaneously (14 days [CI, 4 to 25 days]). The other 7 patients had other reasons, including bleeding (traumatic hematoma), for not having cardioversion (Figure 4).

The follow-up rhythm was sinus rhythm in 37 (59% [CI, 45% to 70%]) of the 64 patients in the conventional therapy group. Table 5 summarizes the feasibility, safety, and other outcomes.

Discussion

Top Methods Results Discussion Author & Article Info References

Our randomized, controlled pilot study of patients with atrial fibrillation who have electrical cardioversion shows that TEE-guided cardioversion with short-term anticoagulation therapy is feasible and safe compared with conventional therapy. It allows for early and more convenient cardioversion and may decrease the risk for embolism, may reduce clinical instability, and may allow tracking of the resolution of thrombi. In addition, our study evaluated the conventional approach to cardioversion recommended by the American College of Chest Physicians [23].

Rationale for Transesophageal Echocardiography-Guided Cardioversion

Interest in TEE as a way to screen for thrombi before cardioversion and thus decrease the risk for embolism has recently been increasing [15]. Some studies initially advocated the use of TEE to obviate the need for anticoagulation therapy because TEE has such an excellent rate of thrombi detection, but this approach, however attractive, is flawed because it cannot exclude thrombi that may form after cardioversion [32-34]. Recent retrospective studies [35, 36] have confirmed the need for anticoagulation therapy in the pericardioversion period.

The ACUTE study, which was based on the results of these studies, was designed to allow TEE to exclude thrombi and have patients fully anticoagulated at the time of cardioversion and for 4 weeks thereafter [15-1934, 35].

Rationale for Conventional Therapy

Embolic stroke remains a serious complication of cardioversion; its reported prevalence is 0.6% to 5.6% [4, 12, 15]. On the basis of previous studies that were neither randomized nor controlled [10, 22, 37-39], the American College of Chest Physicians has recommended that a total of 7 weeks of anticoagulation therapy (3 weeks before and 4 weeks after cardioversion) be given to patients with atrial fibrillation lasting longer than 2 days. However, the major limitations of the conventional approach are the need to delay cardioversion for 3 weeks, the increased risk for bleeding, the inconvenience of re-admission for cardioversion, the lack of controlled studies showing efficacy, and the low rate of stroke among patients not receiving anticoagulant therapy [9, 10, 21, 40, 41].

Effect of Early Cardioversion

The need to wait 3 weeks before cardioversion may be a major limitation of the conventional strategy [19]. According to our study design, cardioversion could be done earlier with the TEE-guided approach than with the conventional approach (0.6 weeks compared with 4.8 weeks). In addition, more patients in the TEE group than in the conventional therapy group had electrical cardioversion (76% compared with 58%) and had electrical cardioversion as originally scheduled (71% compared with 41%).

On the other hand, one unexpected benefit of the 3 weeks of delay associated with the conventional approach was that 20 (31%) of the 64 patients in the conventional therapy group who did not have cardioversion converted spontaneously (n = 11) or pharmacologically (n = 9) to sinus rhythm. However, the risk for embolism during spontaneous or pharmacologic conversion may be similar to that during electrical cardioversion, and atrial "stunning" may still occur after conversion from atrial fibrillation to sinus rhythm [4, 12, 42-44].

Potential Advantages of the Transesophageal Echocardiography-Guided Approach

We found that the TEE-guided approach to cardioversion may have advantages over the conventional approach. Thus, although the numbers of patients and events are small, clinically apparent embolism (stroke, transient ischemic attacks, and systemic embolization) did not occur in the TEE group but was detected in one patient in the conventional therapy group (Table 5). The patient who had an embolic event 3 days after cardioversion required both an embolectomy and 5 additional days in the hospital. This suggests that the TEE-guided approach may be more cost-effective than the conventional approach [17].

Our pilot study suggests that bleeding caused by prolonged anticoagulation therapy with warfarin or heparin may represent potential morbidity for the conventional approach to cardioversion, which mandates 7 weeks of anticoagulation therapy (the TEE-guided approach requires only 4 weeks of this therapy). Bleeding complications occurred in the conventional therapy group but not in the TEE group. Bleeding is an important complication of anticoagulation therapy, as has been shown in recent randomized trials of long-term anticoagulation therapy in patients with nonvalvular atrial fibrillation [45-47].

Effect of Thrombi Detection

Seven thrombi were detected in the TEE group: Six were in the left atrial appendage, and one was in the right atrial appendage. Thrombi were found in 13% of patients having TEE; this rate is similar to the rates of 10% to 15% recently reported in other series [16, 18, 48]. Whether these thrombi would be present after 3 weeks of anticoagulation therapy (as in the conventional group) is unknown, but at least 50% of the thrombi were still present after prolonged anticoagulation therapy (6 weeks) in the TEE group.

Follow-up Rhythm

The TEE-guided approach allows cardioversion to sinus rhythm to be done earlier, and this may be important because the duration of atrial fibrillation may influence the success of cardioversion and the maintenance of sinus rhythm [49, 50]. However, patients in the TEE group, despite having earlier cardioversion, showed immediate rates of conversion to sinus rhythm that were similar to those of patients in the conventional therapy group (85% compared with 76%). Similarly, during the follow-up period, the prevalence of sinus rhythm in the TEE group was similar to that in the conventional therapy group (55% and 56%, respectively), further emphasizing that many variables influence the course of atrial fibrillation [3].

Previous Studies of Transesophageal Echocardiography-Guided Cardioversion

Clinical data to support the use of the TEE-guided approach to cardioversion are emerging [16, 18, 51, 52]. Manning and colleagues did an earlier study on the safety of this approach in 94 patients [16], followed by a report of their larger experience [18]. They used TEE to study 230 patients who had atrial fibrillation that lasted longer than 2 days or was of unknown duration, and they identified 40 patients (15%) with atrial thrombi [18]. Of the 196 patients without thrombi, 186 (95%) had successful electrical or pharmacologic cardioversion without prolonged anticoagulation therapy and none had an embolic event. Manning and colleagues concluded that the TEE-guided approach to cardioversion was useful and may help to screen for thrombi, thus decreasing the risk for embolism [16, 18]. However, their studies did not include a control group and evaluated a mixture of approaches to cardioversion, including antiarrhythmic therapy and various anticoagulation regimens [53, 54]. Stoddard and associates [51] also recently described the role of TEE-guided cardioversion with brief anticoagulation therapy in 206 patients with atrial fibrillation. Although encouraging, the results of these studies must be tempered by the realization that the studies' sample sizes were too small to show clinical efficacy or cost-effectiveness.

Study Limitations

The goal of our pilot study was to assess the feasibility and safety of a TEE-guided approach to cardioversion with anticoagulation therapy, and we believe that we have shown the approach to be both feasible and safe. Because of low rates of events, a multicenter study with 3000 patients and adequate statistical power ( = 0.05 and power = 90%) is needed to definitively compare the TEE-guided approach to cardioversion with the conventional approach. The estimated rates of stroke events for the conventional therapy and TEE groups are 2.9% and 1.2%, respectively. With the addition of major bleeding as a composite end point, the estimated rates of major complications are 3.5% and 1.5%, respectively [55].

One limitation of our pilot study is that the time periods of observation used in comparing the two approaches are unequal, with an end point at 4 weeks after cardioversion but no later than 8 weeks after randomization. Differences in follow-up time from study enrollment resulted from differences between the two approaches in time to cardioversion and rates of early spontaneous or pharmacologic conversion.

A possible limitation of the TEE-guided approach to cardioversion is that TEE may miss tiny thrombi in the left atrial appendage, which may have multiple lobes [56]. It is unlikely that this occurred, however, because all patients in the TEE-guided group had biplane or multiplane imaging [57]. In addition, the patients were therapeutically anticoagulated at the time of cardioversion [52].

Conclusions

Our pilot study suggests that the TEE-guided approach to cardioversion with short-term anticoagulation therapy may be feasible and safe compared with conventional therapy. The TEE-guided approach allows for earlier and more convenient cardioversion and may decrease the risk for embolism by detecting thrombi, reducing hemodynamic instability, and allowing the tracking of resolution of thrombi. Our study also confirms that the conventional approach to cardioversion is relatively safe and effective. However, our findings are preliminary and, by themselves, are not sufficiently strong to serve as the basis for management decisions. More definitive guidance on the management of atrial fibrillation must await the results of a larger trial, which is currently under way.

Appendix

Coordinating Center: The Cleveland Clinic Foundation, Department of Cardiology. Principal Investigator and Medical Program Director: Allan L. Klein, MD. Collaborating Investigators: Richard A. Grimm, DO; Ian W. Black, MD; and Dominic Y. Leung, MB, BS.

Core Echocardiography Laboratory: The Cleveland Clinic Foundation, Department of Cardiology, Cardiovascular Imaging Center. Allan L. Klein, MD; Richard A. Grimm, DO; Ian W. Black, MD; R. Daniel Murray, PhD; Dominic Y. Leung, MB, BS; and Susan E. Vaughn, RN.

Executive Committee: Warren J. Manning, MD; David A. Orsinelli, MD; and Marcus F. Stoddard, MD.

Operating Committee: Ravin Davidoff, MB, BCH; Roberto M. Lang, MD; Timothy P. Obarski, DO; Alan S. Pearlman, MD; Thomas R. Porter, MD; Miguel A. Quinones, MD; Rita I. Redberg, MD; and Miguel Zabalgoitia, MD.

Statistical Coordinating Center: The Cleveland Clinic Foundation. Dave P. Miller, MS (Senior Biostatistician), and Kristopher L. Arheart, EdD (Senior Biostatistician).

Clinical Centers and Collaborating Investigators: The Cleveland Clinic Foundation (Cleveland, Ohio): Allan L. Klein, MD, and Richard A. Grimm, DO; University of Nebraska (Omaha, Nebraska): Thomas R. Porter, MD; Riverside Methodist Hospital (Columbus, Ohio): Timothy P. Obarski, DO; The Ohio State University (Columbus, Ohio): Anthony C. Pearson, MD, and David A. Orsinelli, MD; University of California, San Francisco Medical Center (San Francisco, California): Rita F. Redberg, MD; University of Louisville (Louisville, Kentucky): Marcus F. Stoddard, MD; Texas Heart Institute (Houston, Texas): Susan Wilansky, MD; Prince Henry Hospital (Sidney, Australia): Warren F. Walsh, MD; Bronx Veteran Affairs Medical Center (Bronx, New York): Larry Baruch, MD; University Gesamthochschule-Essen (Essen, Germany): Raimund Erbel, MD.

Dr. Black: Manly Hospital, Darley Road, Manly, New South Wales 2095, Australia.

Mr. Miller and Dr. Arheart: The Cleveland Clinic Foundation, Desk P88, 9500 Euclid Avenue, Cleveland, OH 44195.