Abbott
Ventricular Assist Devices (VADs)
HeartWare II Ventricular Assist System Addendum to Instructions for Use Rev A Sept 2014
Addendum to Instructions for Use
19 Pages
Preview
Page 1
Thoratec Corporation HEARTMATE II® LEFT VENTRICULAR ASSIST SYSTEM
HeartMate II Post-Approval Updates: Destination Therapy Study and Overview of Risk of Thrombosis Addendum to the HeartMate II LVAS Instructions for Use
Thoratec Corporation (International headquarters) 6035 Stoneridge Drive Pleasanton, CA 94588 USA Telephone: (925) 847-8600 Fax: (925) 847-8574 Emergency HeartLine™ USA: (800) 456-1477 Emergencies outside USA: (925) 847-8600 Website: www.thoratec.com
Thoratec Corporation continually strives to provide the highest quality of products for mechanical circulatory support. Specifications may change without notice. HeartMate II, Thoratec, and the Thoratec logo are registered trademarks, and HeartLine is a trademark of Thoratec Corporation.
©2014 Thoratec Corporation. Document: 10000987.A
Publication Date: 09/2014
ii
Preface HeartMate II Post-Approval Updates: Destination Therapy Study and Overview of Risk of Thrombosis Addendum to the HeartMate II LVAS Instructions for Use This addendum contains important clinical post-approval updates to the Instructions for Use. Specifically, the following information is included: 1. Additional new information regarding the HeartMate II Destination Therapy (DT) Post-Approval Study. 2. Additional new information regarding an Overview of Risk of Thrombosis with HeartMate II in the Post-Approval era.
1
HeartMate II Addendum Destination Therapy (DT) Post-Approval Study Overview A Post-approval Study (PAS) of the HeartMate II (HM II) for the destination therapy indication was conducted as a condition of FDA approval. The purpose of the PAS was to assess whether the commercial use of the device produces results that are comparable to results observed in the HM II Destination Therapy clinical trial. The primary objective of the PAS was to assess two year patient survival free of stroke or reoperation to replace the device following HM II implantation for destination therapy in the commercial setting. A number of secondary objectives were also evaluated, including the incidence of adverse events, quality of life as measured by the EuroQoL instrument and clinical reliability.
Study Design The study was a prospective, multicenter study of patients receiving the HM II for the destination therapy indication. All data was collected via the Interagency Registry of Mechanically Assisted Circulatory Support (INTERMACS) and all oversight committees were per INTERMACS protocol. Patients were followed in the study until study outcome (death, cardiac transplantation or device explantation) or two years, whichever occurred first. Patients who were electively transplanted or explanted for myocardial recovery prior to two years post implant were considered to be a success in the composite endpoint analysis if they survived to two years post implant free of stroke.
Study Population To be enrolled into the INTERMACS registry patients received a commercially marketed LVAS and the patient or their legal representative signed an informed consent for INTERMACS registry participation. Patients who were enrolled in premarket clinical studies of VADs or incarcerated persons (prisoners) could not participate. The first consecutive 247 HM II patients who were identified pre-implant in the INTERMACS database as destination therapy were enrolled in the PAS. The INTERMACS patients comprised the study group. The HM II demonstrated superiority over the HM XVE (the only FDA approved device for destination therapy at the time of the study) in the HM II Destination Therapy PMA clinical trial. It was anticipated that the use of the HM XVE for destination therapy would sharply decline with the approval of the HM II for this indication and that only a very limited number of HM XVE implants would be available as a concurrent control in the PAS study. Therefore, it was decided that the 133 HM II Primary Cohort patients from the HM II Destination Therapy clinical trial would comprise the comparison group. Patient demographic data for both the HM II PAS and comparison group were collected at baseline. Table 1 summarizes the baseline characteristics of the two groups.
3
HeartMate II Addendum Table 1 DT Post-Approval Study: Baseline Demographics HM II Post-Approval Study (n=247)
HM II Patients in Primary IDE Study Cohort (ITT) (n=133)
Males
204 (83%)
107 (80%)
Females
43 (17%)
26 (20%)
0–18 years
0 (0%)
0 (0%)
19–39 years
6 (2%)
7 (5%)
40–59 years
69 (28%)
43 (32%)
60–79 years
170 (69%)
83 (62%)
80+ years
2 (1%)
0 (0%)
Caucasian
185 (75%)
98 (74%)
Black
45 (18%)
26 (20%)
Other
17 (7%)
9 (7%)
Height (cm)
(243/247) 173.2 ± 12.3
175.1 ± 9.7
Weight (kg)
(247/247) 85.1 ± 22.2
86.2 ± 19.8
BSA (m2)
(243/247) 2.01 ± 0.29
2.03 ± 0.26
BMI (kg/m2)
(243/247) 28.7 ± 10.7
28.1 ± 5.5
BSA < 1.5 m2 (Small Cohort)
10 (4%)
0 (0%)
CABG
(247/247) 87 (35%)
52 (39%)
Valve Surgery
(247/247) 26 (11%)
14 (11%)
Cancer
(237/247) 37 (16%)
24 (18%)
Diabetes
(247/247) 105 (43%)
63 (47%)
Variable Gender
Age
Race
History
4
HeartMate II Addendum Data Source All data were collected via INTERMACS, the Interagency Registry of Mechanically Assisted Circulatory Support. The patients included the first consecutive 247 HeartMate II patients who gave their consent for inclusion in the INTERMACS registry and were considered destination therapy patients at their initial implant.
Key Study Endpoints As specified in the HM II PAS protocol, the HM II destination therapy PAS would be considered a success if the success rate of the HM II PAS patients was equivalent or better than the success rate of the 133 patients in the clinical trial comparison group. The margin of non-inferiority was set at 10%.
Total Number of Enrolled Study Sites and Subjects, Follow-up Rate A total of 247 patients were enrolled in the PAS at 61 institutions across the United States from Jan 20, 2010 to September 30, 2010. Two hundred and forty-six patients (246/247, 99.6%) were followed for at least 24 months or to an outcome (transplant, death, or explant), whichever occurred first. One patient was followed for only 16 months, after which the medical center withdrew from the study.
Study Visits and Length of Follow-up Study visits were assessed at 1 week and 1, 3, 6, 12, 18 and 24 months post implant. After 24 months, if the patients remained ongoing they were followed in the INTERMACS registry until outcome (transplant, explant or death). After 24 months, the patient’s status was assessed every six months.
5
HeartMate II Addendum Final Effectiveness Findings Overall Patient Outcomes As specified in the HM II Destination Therapy PAS protocol, patients in the study were judged to have reached a successful outcome when they: •
Survived 2 years post implant free of any stroke or reoperation to replace the device
•
Were electively transplanted prior to 2 years post implant and survive to two years post implant free of stroke
•
Were explanted due to myocardial recovery prior to 2 years post implant and survive to two years post implant free of stroke
Patients were considered a failure if they: •
Expired on HM II support prior to 2 years post implant
•
Experienced a stroke, regardless of severity, prior to 2 years post implant
•
Required a reoperation to replace the HM II prior to 2 years post implant
•
Were urgently transplanted due to device failure prior to 2 years post implant
•
Were explanted due to any reason other than myocardial recovery prior to 2 years post implant
•
Failed to survive to two years post elective transplant free of stroke
•
Failed to survive to two years post explant for myocardial recovery free of stroke
The HM II destination therapy PAS would be considered a success if the success rate of the HM II PAS patients is equivalent or better than the success rate of the 133 patients in the clinical trial comparison group. The margin of non-inferiority was set at 10%. As seen in Table 2, the patients in the post-approval study achieved a significantly higher success rate than those followed in the HM II clinical trial.
6
HeartMate II Addendum Table 2 Primary Composite Endpoint: HM II Post-Approval vs. HM II IDE Primary Cohort HM II Post-approval Study
HM II patients in Primary IDE Study Cohort (ITT)
247
134
134 (54%)
57 (43%)
Any ischemic or hemorrhagic stroke
24 (10%)
20 (15%)
Replace or repair device
14 (6%)
13 (10%)
Survival < 2 years
75 (30%)
44 (33%)
Total Failures
113 (46%)
77 (57%)
Patients Total Success Reason for Failure
Fifty-four percent (54%) of HM II PAS patients achieved success while 43% of the control group achieved success on the primary composite endpoint. The difference between proportions is 11.7% with 95% confidence interval = 1.2 to 21.8%. Since the lower confidence interval is greater than the non-inferiority interval of -10% the PAS is non-inferior to the IDE cohort. The difference is also superior to the IDE cohort. Its p-value (using Fisher Exact test) is 0.0321. The 2 year survival for the HM II PAS was 61%, compared to 58% for the control (Kaplan-Meier analysis, patients censored at the time of transplant, explant or last known follow up).
7
HeartMate II Addendum
100% 90%
Percent Survival
80%
HMII PAS
70% 60% 50%
HMII IDE
40% 30% 20% 10% 0%
0
3
6
9 12 15 18 21 24 27 30 33 36 39 42 45 48 Months
Figure 1 Kaplan-Meier Analysis of Survival: HM II Post-Approval vs. HM II IDE Primary Cohort
Safety: All Cause Adverse Events Adverse events, as defined by INTERMACS protocol, were collected and are reported in Table 3. Since the adverse event definitions were different from those used in the HM II Destination Therapy clinical trial, a comparison of adverse events is not possible.
8
HeartMate II Addendum Table 3 DT Post-Approval Study: Adverse Events HM II Post-approval Study (n=247)
Adverse Events
Patients Percent
UCL
LCL
# Events
Events/Pt-Yr
Arterial Non-CNS Thromboembolism
2
1%
2%
0%
2
0.01
Bleeding
133
54%
60%
48%
325
0.84
31
13%
17%
8%
35
0.09
Cardiac Arrhythmia
92
37%
43%
31%
154
0.40
Device Malfunction
45
18%
23%
13%
55
0.14
Hemolysis
16
6%
10%
3%
22
0.06
Hepatic Dysfunction
18
7%
11%
4%
20
0.05
Hypertension
18
7%
11%
4%
32
0.08
Infection*
141
57%
63%
51%
340
0.88
Driveline
39
16%
20%
11%
75
0.19
Pump Pocket
8
3%
5%
1%
11
0.03
Pump Interior
3
1%
3%
0%
3
0.01
Blood
48
19%
24%
14%
71
0.18
Line Sepsis
10
4%
7%
2%
10
0.03
Other Infection
111
45%
51%
39%
215
0.56
Myocardial Infarction
1
0%
1%
0%
1
0.00
Neurological Dysfunction
31
13%
17%
8%
37
0.10
Pericardial Drainage
21
9%
12%
5%
25
0.06
Psychiatric Episode
34
14%
18%
9%
40
0.10
Renal Dysfunction
45
18%
23%
13%
57
0.15
Respiratory Failure
66
27%
32%
21%
88
0.23
Right Heart Failure
45
18%
23%
13%
62
0.16
Stroke
34
14%
18%
9%
40
0.10
Hemorrhagic Stroke
19
8%
11%
4%
20
0.05
Ischemic Stroke
6
2%
4%
1%
8
0.02
Venous Thromboembolism
14
6%
9%
3%
15
0.04
Wound Dehiscence
3
1%
3%
0%
4
0.01
Bleeding with Surgery
†
* Each event may have multiple sites of infection. † Duration of all patients is 385.79 patient-years; data provided as of September 30, 2012.
9
HeartMate II Addendum Secondary Objectives Quality of life, as measured by EuroQol, was collected as a secondary objective. Table 4 displays the quality of life data. The thermometer is the patient’s self-assessment of their quality of life. A higher score indicates an improved quality of life. The Total Score is a sum of the 5 quality of life domains tested: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each domain is scored with the higher score representing more extreme problems. The lower the total score the better the quality of life. Functional status of patients in the PAS was assessed using the Six Minute Walk test and the New York Heart Association (NYHA) classification. Table 5 and Table 6 are the results of these tests. The results of the secondary assessments demonstrate improved and sustained quality of life and functional status in HM II destination therapy PAS patients.
Table 4 DT Post-Approval Study: Quality of Life – EuroQoL* HeartMate II PAS Cohort Interval: Preimplant Patients at Start of Interval
247
Patients Completing Test/Therm
171 / 169
Total Score (mean ± sd)
8.9 ± 2.1
Total Score (median & range)
9.0 (5.0 - 13.0)
Thermometer (mean ± sd)
42.2 ± 23.8
Thermometer (median & range)
40 (0 - 100)
Interval: 3 Months Patients at Start of Interval
206
Patients Completing Test/Therm
119 / 115
Total Score (mean ± sd)
7.4 ± 1.8
Total Score (median & range)
7.0 (5.0 - 12.0)
Thermometer (mean ± sd)
72.6 ± 18.4
Thermometer (median & range)
75 (2 - 100)
Interval: 6 Months Patients at Start of Interval Patients Completing Test/Therm
120 / 123
Total Score (mean ± sd)
7.2 ± 1.8
Total Score (median & range)
10
191
7.0 (5.0 - 12.0)
Thermometer (mean ± sd)
74.4 ± 19.9
Thermometer (median & range)
80 (3 - 100)
HeartMate II Addendum Table 4 DT Post-Approval Study: Quality of Life – EuroQoL* (Continued) HeartMate II PAS Cohort Interval: 1 Year Patients at Start of Interval
171
Patients Completing Test/Therm
104 / 103
Total Score (mean ± sd)
7.0 ± 1.9
Total Score (median & range)
6.0 (5.0 - 12.0)
Thermometer (mean ± sd)
74.2 ± 19.3
Thermometer (median & range)
80 (7 - 100)
Interval: 18 months Patients at Start of Interval
151
Patients Completing Test/Therm
74 / 73
Total Score (mean ± sd)
6.8 ± 1.8
Total Score (median & range)
6.0 (5.0 - 13.0)
Thermometer (mean ± sd)
74.5 ± 19.9
Thermometer (median & range)
80 (0 - 100)
Interval: 2 years Patients at Start of Interval
130
Patients Completing Test/Therm
66 / 26
Total Score (mean ± sd)
7.0 ± 1.9
Total Score (median & range)
6.0 (5.0 - 11.0)
Thermometer (mean ± sd)
66.3 ± 27.4
Thermometer (median & range)
75 (6 -100)
*Total Score: Lower score indicates better QoL Thermometer: Higher score indicates better QoL
11
HeartMate II Addendum
Table 5 DT Post-Approval Study: Six Minute Walk Test HeartMate II PAS Cohort Interval: Preimplant Patients at Start of Interval
247
Patients Who Walked
47
Feet Walked (mean ± sd)
600.4 ± 319.5
Feet Walked (median & range)
600 (25 - 1394)
Not Done: Too Sick
143
Not Done: Other Reason
45
Not Done: Unknown Reason
12
Interval: 3 Months Patients at Start of Interval
206
Patients Who Walked
82
Feet Walked (mean ± sd)
825.7 ± 416.4
Feet Walked (median & range)
801 (42 - 2160)
Not Done: Too Sick
19
Not Done: Other Reason
77
Not Done: Unknown Reason
28
Interval: 6 Months Patients at Start of Interval
191
Patients Who Walked
72
Feet Walked (mean ± sd)
887.8 ± 388.6
Feet Walked (median & range)
893 (40 - 1634)
Not Done: Too Sick
14
Not Done: Other Reason
76
Not Done: Unknown Reason
29
Interval: 1 Year
12
Patients at Start of Interval
171
Patients Who Walked
63
Feet Walked (mean ± sd)
878.3 ± 359.1
Feet Walked (median & range)
876 (180 - 1571)
Not Done: Too Sick
7
Not Done: Other Reason
69
Not Done: Unknown Reason
32
HeartMate II Addendum Table 5 DT Post-Approval Study: Six Minute Walk Test (Continued) HeartMate II PAS Cohort Interval: 18 months Patients at Start of Interval
151
Patients Who Walked
43
Feet Walked (mean ± sd) Feet Walked (median & range)
964.7 ± 381.8 1000 (240 - 1625)
Not Done: Too Sick
10
Not Done: Other Reason
54
Not Done: Unknown Reason
44
Interval: 2 years Patients at Start of Interval
130
Patients Who Walked
34
Feet Walked (mean ± sd) Feet Walked (median & range)
970.1 ± 386.5 1010 (114 - 1485)
Not Done: Too Sick
7
Not Done: Other Reason
43
Not Done: Unknown Reason
46
13
HeartMate II Addendum Table 6 DT Post-Approval Study: NYHA Class HeartMate II PAS Cohort Interval: Preimplant Patients at Start of Interval
247
NYHA Class I
1 (<1%)
NYHA Class II
0 (0%)
NYHA Class III
40 (16%)
NYHA Class IV
196 (79%)
Unknown
10 (5%)
Interval: 1 Month Patients at Start of Interval
222
NYHA Class I
1 (<1%)
NYHA Class II
19 (9%)
NYHA Class III
16 (7%)
NYHA Class IV
21 (9%)
Unknown
165 (74%)
Interval: 3 Months Patients at Start of Interval
206
NYHA Class I
42 (20%)
NYHA Class II
71 (34%)
NYHA Class III
30 (15%)
NYHA Class IV
11 (5%)
Unknown
52 (25%)
Interval: 6 Months Patients at Start of Interval
14
191
NYHA Class I
48 (25%)
NYHA Class II
65 (34%)
NYHA Class III
26 (14%)
NYHA Class IV
7 (4%)
Unknown
45 (24%)
HeartMate II Addendum Table 6 DT Post-Approval Study: NYHA Class (Continued) HeartMate II PAS Cohort Interval: 1 Year Patients at Start of Interval
171
NYHA Class I
50 (29%)
NYHA Class II
49 (29%)
NYHA Class III
17 (10%)
NYHA Class IV
11 (6%)
Unknown
44 (26%)
Interval: 18 months Patients at Start of Interval
151
NYHA Class I
34 (23%)
NYHA Class II
45 (30%)
NYHA Class III
13 (9%)
NYHA Class IV
5 (3%)
Unknown
54 (36%)
Interval: 2 years Patients at Start of Interval
130
NYHA Class I
28 (22%)
NYHA Class II
38 (29%)
NYHA Class III
12 (9%)
NYHA Class IV
3 (2%)
Unknown
49 (38%)
Study Strengths and Weaknesses The trial was conducted according to a prespecified protocol approved by the FDA. The study was hypothesis driven and powered to detect a difference between treatment groups. All study patients (100%) were followed for at least 24 months or to an outcome (transplant, death, or explant). Survival and quality of life trends were similar to those seen in the HeartMate II pivotal trial. The study had several limitations. The study data was collected by a national registry called INTERMACS and was not monitored against medical records. Adverse events were not adjudicated by an independent clinical events committee.
15
HeartMate II Addendum Post-Approval Overview of Risk of Thrombosis Pump thrombosis is a serious adverse event that may require intravenous medical therapy or a reoperation to replace the pump. As such, it was specifically tracked in both the Bridge to Transplant and Destination Therapy clinical trials. In addition, as part of post-market surveillance, Thoratec collects and investigates all reports of pump thrombosis for reporting under the U.S. FDA Medical Device Reporting (MDR) regulations and similar reporting requirements in other regulatory jurisdictions. Table 7, below, presents the incidence of pump thrombosis (suspected and/or confirmed) observed in the Bridge to Transplant (BTT) clinical trial, the Destination Therapy (DT) clinical trial, and the world-wide experience (BTT and DT) that was reported to Thoratec as of January 11, 2013.
Table 7 Incidence of Pump Thrombosis Clinical Experience Bridge to Transplant
Dates Mar 2005 – 1
US Clinical Trial
Sep 2007
Destination Therapy
Mar 2005 –
US Clinical Trial
May 20092
World-wide
Sep 2005 –
clinical experience3
Jan 2013
Implants
Reports of Thrombosis
126
2 (confirmed)
133
5 (confirmed)
11, 647
750 (suspected or confirmed) 334 (confirmed)
Percent [95% Conf. Int.] 1.6% [0% - 4%] 3.8% [1% - 7%] 6.4% 2.9%
1
Including 1 year of follow-up after implant Including 2 years of follow-up after implant 3 Analysis of complaint and device tracking records 2
Note that comparisons of rates between the clinical studies and the postmarket experience are confounded by more recently recognized definitions of suspected pump thrombosis. During the clinical trials device thrombosis was defined as, “Any obstructive thrombus in the device or its conduits associated with clinical symptoms of impaired pump performance (e.g. decreased pump flow, need to increase pump speed, increased power, hemolysis) or the need for thrombolytic or surgical intervention.”
16
HeartMate II Addendum Subsequent to regulatory approval and commercial distribution, users of the HeartMate II LVAS have reported suspected pump thrombosis when observing some of the following: •
Clinical hemolysis along with elevated lactate dehydrogenase (LDH) and/or plasma free hemoglobin (PFHgb) levels
•
Low flow as evidenced by echo and other means
•
Elevated LDH and/or PFHgb levels without clinical signs of hemolysis
•
Sustained power elevations
•
Congestive heart failure symptoms
•
Inability of the HeartMate II LVAS to unload the left ventricle with the aortic valve opening at each heartbeat
•
Cardiogenic shock
Additionally, there is overlap between the symptoms of device thrombosis, hemolysis due to other reasons, and other unrelated adverse events. Device thrombosis can only be confirmed through disassembly and examination of an explanted pump. The propensity for pump thrombosis is multi-factorial in nature, including pump-related factors, patient-related factors, and patient management-related factors. There has been variability among centers in the reported incidence of confirmed device thrombosis, ranging from 0% to 15% within 6 months of implantation for centers that have implanted at least 10 HeartMate II LVADs. Starling et al. reported that in their three-center experience, the rate of HeartMate II LVAS pump thrombosis at 3 months increased from 2.2% before March 2011 to 8.4% after March 2011.1 Other reports have not identified this type of increased pump thrombosis rate.2,3,4,5,6, Schmitto et al. reported that 2.2% of patients were confirmed to have pump thrombosis at 3 months in their single-center experience from 2004-2013.5 Hoefer et al. reported no premature losses of HeartMate II devices due to suspected or confirmed pump thrombosis at their center.4
1. Starling RC, Moazami N, Silvestry SC, et al.: Unexpected abrupt increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:33-40. 2. Frazier OH: Increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:1464-5. 3. Goldstein D, John R, Salerno C: Increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:1465. 4. Hoefer D, Velik-Salchner C, Antretter H: Increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:1464. 5. Schmitto JD, Avsar M, Haverich A: Increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:1463-4. 6. Stulak JM, Maltais S: A different perspective on thrombosis and the HeartMate II. N Engl J Med 2014;370:1467-8.
17
HeartMate II Addendum A multi-center analysis of the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) registry reported a decrease in the freedom from pump thrombosis at 6 months from 98% in 2010 to 94% in 2012.7 However, the analysis also showed that there was no decrease in the overall survival with HeartMate II patients during the same time periods, and that the survival rate exceeded the rate observed in the premarket clinical trials. A 2014 review of pump thrombosis indicated that the incidence of pump thrombosis was increasing concurrently with changes in clinical practice, such as modified anticoagulation targets and reduction of pump speeds.8 Although pump thrombosis has been proposed to be multi-factorial in etiology, approximately 25% of the events are subsequently identified to have had a mechanical cause such as inflow cannula/outflow graft obstruction.9 The following key clinical practices have been identified that may help in the prevention of pump thrombosis: •
Optimization of pump and cannula position during implantation to prevent inflow obstruction
•
Optimization of pump speeds to avoid low flows through the pump in order to facilitate pump washing
•
Maintenance of strict anticoagulation with an INR target of over 2.0
Goldstein et al. have developed an algorithm for the diagnosis and management of patients presenting with suspected pump thrombosis.10
7. Kirklin JK, Naftel DC, Kormos RL, et al.: Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) analysis of pump thrombosis in the HeartMate II left ventricular assist device. J Heart Lung Transplant 2014;33:12-22. 8. Mehra MR, Stewart GC, Uber PA: The vexing problem of thrombosis in long-term mechanical circulatory support. J Heart Lung Transplant 2014;33:1-11. 9. Uriel N, Han J, Morrison KA, et al.: Device thrombosis in HeartMate II continuous-flow left ventricular assist devices: a multifactorial phenomenon. J Heart Lung Transplant 2014;33:51-9. 10. Goldstein DJ, John R, Salerno C, et al.: Algorithm for the diagnosis and management of suspected pump thrombus. J Heart Lung Transplant. 2013 Jul;32(7):667-70.
18