Author + information
- Received March 3, 2010
- Revision received June 10, 2010
- Accepted July 6, 2010
- Published online December 14, 2010.
- Luc W.M. Eurlings, MD⁎,
- Petra E.J. van Pol, MD†,
- Wouter E. Kok, MD, PhD‡,
- Sandra van Wijk, MD⁎,
- Cara Lodewijks-van der Bolt, MD§,
- Aggie H.M.M. Balk, MD, PhD∥,
- Dirk J.A. Lok, MD, PhD¶,
- Harry J.G.M. Crijns, MD, PhD⁎,
- Dave J.W. van Kraaij, MD, PhD#,
- Nicolaas de Jonge, MD, PhD⁎⁎,
- Joan G. Meeder, MD, PhD††,
- Martin Prins, MD, PhD⁎ and
- Yigal M. Pinto, MD, PhD‡,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Yigal M. Pinto, Heart Failure Research Center, University of Amsterdam, Meibergdreef 9 1105 AZ Amsterdam, the Netherlands
Objectives The purpose of this study was to assess whether management of heart failure (HF) guided by an individualized N-terminal pro–B-type natriuretic peptide (NT-proBNP) target would lead to improved outcome compared with HF management guided by clinical assessment alone.
Background Natriuretic peptides may be attractive biomarkers to guide management of heart failure (HF) and help select patients in need of more aggressive therapy. The PRIMA (Can PRo-brain-natriuretic peptide guided therapy of chronic heart failure IMprove heart fAilure morbidity and mortality?) study is, to our knowledge, the first large, prospective randomized study to address whether management of HF guided by an individualized target NT-proBNP level improves outcome.
Methods A total of 345 patients hospitalized for decompensated, symptomatic HF with elevated NT-proBNP levels at admission were included. After discharge, patients were randomized to either clinically-guided outpatient management (n = 171), or management guided by an individually set NT-proBNP (n = 174) defined by the lowest level at discharge or 2 weeks thereafter. The primary end point was defined as number of days alive outside the hospital after index admission.
Results HF management guided by this individualized NT-proBNP target increased the use of HF medication (p = 0.006), and 64% of HF-related events were preceded by an increase in NT-proBNP. Nevertheless, HF management guided by this individualized NT-proBNP target did not significantly improve the primary end point (685 vs. 664 days, p = 0.49), nor did it significantly improve any of the secondary end points. In the NT-proBNP–guided group mortality was lower, as 46 patients died (26.5%) versus 57 (33.3%) in the clinically-guided group, but this was not statistically significant (p = 0.206).
Conclusions Serial NT-proBNP measurement and targeting to an individual NT-proBNP value did result in advanced detection of HF-related events and importantly influenced HF-therapy, but failed to provide significant clinical improvement in terms of mortality and morbidity. (Effect of NT-proBNP Guided Treatment of Chronic Heart Failure [PRIMA]; NCT00149422)
Current management of patients with heart failure (HF) is mainly based on clinical signs and symptoms. This approach allows clinicians to respond to worsening HF once it is recognized, but does not allow selection of individuals who are most likely to progress to increased morbidity and mortality and are thus in need of more intensive treatment. Plasma levels of B-type natriuretic peptide (BNP) or N-terminal pro-B-type natriuretic peptide (NT-proBNP) are established indicators of decompensated HF (1,2) and predictors of HF morbidity and mortality (3,4). Natriuretic peptides may therefore be attractive biomarkers to guide management of HF and help select patients in need of more aggressive therapy.
Troughton et al. (5) were the first to suggest in a small pilot study that guiding HF management by aiming for a target NT-proBNP level may improve outcome. In this study, the investigators aimed to achieve NT-proBNP levels of 200 pmol/l (1,700 pg/ml) or lower, a goal that is difficult to achieve in many patients with established HF. The clinical value of such stringent (NT-pro)BNP levels has recently been addressed in several other clinical outcome studies (6–8). These studies failed to show an overall reduction in mortality, but did suggest improved outcome in HF patients under the age of 75 years. Also, the (NT-pro)BNP target value was achieved only in a minority of patients. In the STARS-BNP (Systolic Heart Failure Treatment Supported by BNP) Multicenter study, only 33% reached the BNP target (8), whereas in TIME-CHF (Trial of Intensified vs Standard Medical Therapy in Elderly Patients With Congestive Heart Failure) and the BATTLESCARRED (NT-proBNP-Assisted Treatment to Lessen Serial Cardiac Readmissions and Death) trial, at least 50% of the subjects had not achieved the desired target at the end of the study. At the same time, the TIME-CHF investigators speculated that intensification of therapy might be harmful in the elderly (7).
As the target NT-proBNP level in the aforementioned studies was not achieved in the majority of HF patients randomized to (NT-pro)BNP–guided therapy, most subjects randomized to the (NT-pro)BNP–guided arms received intensified treatment. These studies therefore show that a more generalized intensification of HF therapy may be beneficial in specific subgroups. However, it was not addressed whether serial assessment of NT-proBNP enables to select patients at risk of increased morbidity and mortality. Therefore, the question remains whether it is beneficial to intensify HF therapy only in those patients most likely to progress towards events.
It is well known that in many HF patients, NT-proBNP levels never normalize, whereas these patients still remain clinically stable over years. This suggests that patients with stable NT-proBNP levels (even when clearly elevated) may have an acceptable prognosis. We hypothesized that elevation of outpatient NT-proBNP levels as compared with the patient's individualized target level, allows selection of those HF patients most likely to progress towards events. The individualized target level was defined as the lowest level at discharge or at 2 weeks follow-up after admission because of HF. We further hypothesized that restricting treatment intensification to these selected patients would be beneficial without additional risk of adverse effects. We therefore performed a prospective randomized study to address whether treatment of HF, guided by an individualized target NT-proBNP level, improves outcome in HF patients.
Study design and study population
To be included, patients had to be hospitalized for decompensated, symptomatic HF, fulfilling the European Society of Cardiology (ESC) diagnostic guideline criteria for acute HF (9). In addition, NT-proBNP levels at admission were required to be at least 1,700 pg/ml, as additional objective evidence of HF (1).
Exclusion criteria were: life-threatening cardiac arrhythmias during the index hospitalization, urgent invasive or surgical intervention performed or planned during the index hospital admission, severe chronic obstructive pulmonary disease with a forced expiratory volume in 1 s (FEV1) of <1 l/s, pulmonary embolism less than 3 months prior to admission, pulmonary hypertension not caused by left ventricular systolic dysfunction, a non–HF-related expected survival of less than 1 year, and patients undergoing hemodialysis or continuous ambulant peritoneal dialysis. A lesser degree of renal dysfunction was not an exclusion criterion. Patients were screened and included during the index admission because of acute HF (Fig. 1). Informed consent was obtained and NT-proBNP levels were measured at hospital discharge. Patients demonstrating a significant decrease in NT-proBNP levels during hospitalization, defined as a decrease of more than 10%, with a drop in NT-proBNP levels of at least 850 pg/ml, were randomized to treatment that was either NT-proBNP guided or clinically guided. Patients in whom NT-proBNP levels decreased <10% during admission were considered not to modulate their NT-proBNP levels enough to allow NT-proBNP–guided treatment. Therefore, these patients were not included. Regular follow-up visits were scheduled at 2 weeks and 1 month, and then every 3 months until the follow-up period of 2 years was completed. Follow-up visits were performed by dedicated HF cardiologists and nurses. The institutional review board or ethics committee at each site approved the protocol, and all patients provided written informed consent before enrollment.
Treatment in the NT-proBNP–guided group was guided by the combination of clinical assessment and NT-proBNP levels. The individual NT-proBNP target value was set at the lowest level at discharge or at 2 weeks follow-up. If at subsequent outpatient visits, NT-proBNP levels were more than 10% with a minimum of 850 pg/ml above this individual target level, NT-proBNP level was considered “off-target,” and therapy was intensified according to the ESC HF treatment guidelines (10). In this treatment group, an electronic case record form indicated at each visit whether NT-proBNP levels were off-target and indicated whether intensification was necessary. Therapy in the clinically-guided treatment group was determined by clinical assessment alone. A therapy advisor, incorporated in the electronic case record form, was designed to give individual treatment advice, depending on several individual variables including the cause of HF (ischemic vs. nonischemic), left ventricular ejection fraction (LVEF), clinical signs of HF, and creatinine clearance. Also, titration schemes for diuretics, angiotensin-converting enzyme (ACE) inhibitors, beta-blockers, angiotensin receptor blockers, and aldosterone receptor blockers were provided. In the clinically-guided treatment group, all cardiologists were blinded to the NT-proBNP levels of the patients during follow-up. At every outpatient visit, vital status was assessed. Quality of life (QOL) was assessed at 3-month intervals by the Minnesota Living with Heart Failure Questionnaire (11). NT-proBNP levels were measured on a Roche Diagnostics Elecsys platform (Roche Diagnostics Ltd., Rotkreuz, Switzerland) at every participating site, except for 1 center where the NT-proBNP levels of patients randomized to the NT-proBNP guided group were measured within 24 h in a participating university hospital nearby.
Definition of study end points
The primary end point of the PRIMA (Can PRo-brain-natriuretic peptide guided therapy of chronic heart failure IMprove heart fAilure morbidity and mortality?) study was defined as the difference in total number of days alive and outside the hospital between the NT-proBNP–guided and the clinically-guided group. This primary end point replaced the initial end point of reduction in number of events, as the end point of number of days alive outside the hospital included hospital admission as mortality. The primary end point was changed before any patient had been included, in the start-up phase of the study. Major secondary end points encompassed total and cardiovascular mortality, total and cardiovascular hospitalization, and the combined end points of total and cardiovascular morbidity and mortality. Furthermore, renal function, left ventricular systolic function, and age subgroups were analyzed. Additionally, analysis of the use of evidence-based HF medication was performed. Evidence-based HF medication target dose was defined as the recommended maintenance dose approved for the treatment of HF in Europe (10). Finally, it was predefined to analyze the prognostic impact of NT-proBNP levels above the individually set target level at outpatient visits. All events were adjudicated by a blinded event committee, consisting of medical specialists in cardiology, nephrology, vascular medicine, pulmonology, and neurology. Serious adverse events included admissions to the emergency room, hospital admissions, and death.
Treatment in the NT-proBNP group was considered to be protocol adherent when 1 of the following actions was undertaken upon an elevated outpatient NT-proBNP level: starting or intensifying HF medication according to the ESC guidelines, all therapeutic and diagnostic actions searching for underlying causes of HF such as hypertension, ischemic heart disease, valvular heart disease, anemia, and cardiac arrhythmias; hospital admission (for decompensated HF); or registering for heart transplantation.
Based on previous studies and observations, it was estimated that, with an event rate of 20%, 480 patients would be needed to reach a relative risk reduction in number of events of 50% in the NT-proBNP–guided group compared with the clinically-guided group at an α level of 0.05 and a power level of 0.80. Although the primary end point changed during the start-up phase of the study, power analysis remained the same. One year after the first patient being included, a pre-specified interim analysis was performed, with a difference in events (p < 0.01) as criterion to preliminary stop according to lan-DeMets alpha spending rule. The interim analysis demonstrated a pooled event rate of 65%. Thereupon the power analysis was re-evaluated. It was calculated that 364 patients were needed to demonstrate a minimum reduction in pooled events of 30%.
Results are presented as frequencies, mean (SD), or median (interquartile range [IQR]), where appropriate. Between-group comparisons were performed using the t test, Mann-Whitney U test, or chi-square test where appropriate. Event rates for all-cause mortality were estimated by the Kaplan-Meier method. Hazard ratios were calculated using Cox regression analysis. Time-dependent Cox regression analysis was performed to analyze the prognostic impact of elevated NT-proBNP levels above target value at the outpatient clinic. All calculations were performed with the use of the SPSS statistical package version 15.0 (SPSS Inc., Chicago, Illinois).
PRIMA is a prospective, randomized, single-blind study executed in 12 Dutch university and large general hospitals. Patients were recruited between June 2004 and September 2007. In total, 345 patients were randomized: 174 patients to the NT-proBNP–guided treatment group and 171 patients to the clinically-guided group.
Baseline characteristics are shown in Table 1, which did not reveal significant differences between the 2 study groups except for the number of transient ischemic attacks (4.6% in the NT-proBNP–guided vs. 14.6% in the clinically-guided group, p = 0.002). Patients were elderly with a mean age of 72 years, more than 40% were female, and a large proportion had cardiovascular risk factors such as hypertension and diabetes mellitus. History of HF was present in 37% of subjects, and almost a quarter had a LVEF above 45%. After the index admission, most patients were in New York Heart Association functional class II at discharge. The median NT-proBNP target value in the NT-proBNP–guided treatment group was 2,491 pg/ml. The median follow-up was 702 days (IQR: 488 to 730 days). In 34 patients (17 in both randomization groups, see Fig. 1), outpatient follow-up visits after 1-year follow-up were not completed.
Management guided by an individualized NT-proBNP target did not significantly improve the primary end point, the number of days alive outside the hospital: median number of days alive outside the hospital was 685 versus 664 days, p = 0.49 (Fig. 2). In the NT-proBNP–guided group, mortality was lower, as 46 patients died (26.5%), versus 57 (33.3%) in the clinically-guided group (Fig. 3), but this was not statistically significant. The number of scheduled visits did not differ between the NT-proBNP–guided and the clinically-guided group (mean 7.1 ± 3.1 vs. 6.9 ± 3.0, p = 0.424). However, there was a trend towards an increase in unscheduled visits in the NT-proBNP–guided group (mean 1.4 ± 1.9 vs. 1.1 ± 1.7, p = 0.063). The total number of cardiovascular and HF-related admissions between the NT-proBNP–guided and the clinically-guided groups were not different (mean 1.11 ± 2.20 vs. 1.05 ± 1.47, p = 0.552, and 0.70 ± 1.89 vs. 0.60 ± 1.25, p = 0.989). In addition, none of the other pre-specified end points was statistically significantly different between the groups (Table 2).
In the subset of patients with left ventricular systolic dysfunction (ejection fraction below 45%), total mortality tended to be lower in the NT-proBNP–guided group, which however did not attain statistical significance (mortality: 25% in NT-proBNP–guided vs. 33% in usual care, p = 0.164). In contrast, in patients with preserved left ventricular systolic function, mortality was identical in both groups (31%). Furthermore, in patients under 75 years of age, a trend was seen towards improved outcome in the NT-proBNP–guided treatment arm (number of days alive outside hospital as percentage of total follow-up: 87.4% vs. 82.8%, p = 0.114). Therapy guided by NT-proBNP levels also tended to be favorable in patients with lower discharge creatinine levels, but again these differences did not attain statistical significance (number of days alive outside hospital as percentage of total follow-up: 92.7% vs. 87.9%, p = 0.076).
NT-proBNP levels and use of medication
After 1-year follow-up, 80% of patients were at or below their individual target level. In 23% of all outpatient visits, NT-proBNP levels were above the individualized target level. In 79% of all outpatient visits with an off-target NT-proBNP level, protocol adherent action was undertaken (Table 3). Evidence-based medication for HF was extensively used in both groups (Table 4). Renin-angiotensin inhibition was used significantly more frequently after 1 year in the NT-proBNP–guided group. Management guided by an individualized NT-proBNP target also led to an overall increased use of HF medication (Table 5). An increased NT-proBNP value most often prompted physicians to intensify diuretic therapy. The lack of significant beneficial outcomes suggests that intensifying diuretic therapy may not be adequate to prevent events. We therefore compared the effects of intensifying evidence-based HF medication (i.e., increase renin-angiotensin system blockade, beta-blockade, and spironolactone) to the effect of intensifying diuretics in response to an increased NT-proBNP level. However, no differences were found between the 2 types of treatment in ability to increase the number of patients on or below target level at the next outpatient follow-up (40% vs. 33%, p = 0.369). Moreover, in general, intensifying HF medication (evidence-based HF medication or diuretics) compared with no pharmacological HF intervention was not associated with a significantly higher number of patients who reached their NT-proBNP target (47% vs. 36%, p = 0.117).
During follow-up, NT-proBNP levels and levels of urea and creatinine did not significantly differ between both treatment groups, although there was a trend for increased creatinine in the NT-proBNP–guided group (Table 6). Also, no difference was seen in QOL between the NT-proBNP–guided and the clinically-guided groups: median QOL at discharge: 47 (IQR: 34 to 62) versus 48 (IQR: 36 to 60), p = 0.95, at 6-month follow-up: 23 (IQR: 10 to 39) versus 25 (IQR: 11 to 42), p = 0.64, and at 12-month follow-up: 20 (IQR: 6 to 36) versus 23 (IQR: 10 to 38), p = 0.6.
The individualized NT-proBNP target value appeared to be of prognostic importance. Most HF-related events (64%) were preceded by off-target NT-proBNP levels at previous outpatient follow-up. Outpatient elevation of NT-proBNP levels above this individualized target value indicated an increased risk for major end points such as total mortality (hazard ratio [HR]: 1.84, p = 0.007), cardiovascular mortality (HR: 2.53, p < 0.001), and HF-related mortality (HR: 3.69, p < 0.001) (Table 7).
The PRIMA study is a prospective randomized study to address whether HF therapy, guided by an individualized NT-proBNP level, improves outcome in HF patients.
The PRIMA study randomized 345 patients to HF therapy guided by an individually set NT-proBNP target level in addition to clinical signs, or by clinical signs only. It addressed the benefit of selective intensification of therapy only when NT-proBNP increases beyond the individually defined “optimal” NT-proBNP level. Assessing the optimum natriuretic peptide target level is most challenging (12). As such, the PRIMA study complements the recent studies on the benefits of a more general intensification of therapy by aiming for absolute NT-proBNP targets (6,7).
PRIMA showed that selective intensification by an individualized NT-proBNP target did not significantly improve any of the pre-specified primary or secondary outcome measures. Although treatment guided by an individualized NT-proBNP target slightly improved the number of days alive outside the hospital, and improved overall mortality, these changes were not statistically significant.
Individualized NT-proBNP–guided therapy resulted in significantly intensified pharmacological HF therapy reflected by an increased use of diuretics and ACE inhibitors or angiotensin II receptor antagonists in the NT-proBNP–guided group.
We hypothesized that individualized NT-proBNP–guided treatment would improve outcome. Our first assumption was that stability of NT-proBNP would portend an improved prognosis, even when the stable NT-proBNP level is well above normal. This first assumption was confirmed in this study. Patients who maintained their individual NT-proBNP target level indeed had a highly significantly better outcome. Most events occur in patients with an unstable NT-proBNP. Indeed, increases of NT-proBNP level above each individual optimum was a strong predictor of HF-related events with hazard ratios up to 4.17, p < 0.001. Our second assumption was that treatment of HF guided by an individualized target NT-proBNP level could avert HF events. Despite the ability to detect 64% of the imminent events, and despite the subsequent intensification of HF medication, events were not significantly averted. This suggests that although NT-proBNP measurement can help detect worsening HF, current standard-of-care HF therapy is unable to avert subsequent events. NT-proBNP levels react upon ventricular wall stretch (13). Therefore, deterioration of HF is needed before levels rise. Elevated marker levels before an increase in cardiac pressures takes place might help us identify patients at risk for events; in such an early phase, medical intervention can still avert worse outcome. NT-proBNP appears to be a “passenger-seat” marker: you can see which direction your car is heading, but you are not in control of the steering wheel. In order to reduce morbidity and mortality, a “driver-seat” marker is urgently needed.
A number of reasons may account for this lack of significant improvement. First, not enough events during follow-up may have been detected, as 36% of events were undetected by measurement of NT-proBNP at 3-month intervals. Second, intensification of treatment against a background of high use of evidence-based HF therapy may not suffice to avert an imminent event. Third, current “gold-standard” HF therapy may altogether be inadequate in preventing HF-related events in patients with deterioration of HF.
The effects of NT-proBNP guidance seemed mitigated in patients with preserved systolic function, although differences did not reach statistical significance. The lack of well-established medical or other intervention measures for patients with preserved ejection fraction HF may limit the success of interventions prompted by off-target NT-proBNP in this group.
The effects of NT-proBNP guidance seemed more favorable in younger patients (age under 75 years) and those with better renal function, although these differences also did not reach statistical significance. Both younger patients and patients with preserved renal function have less comorbidity, and are expected to better tolerate intensification of therapy. It is these type of patients that have been included in most landmark trials that yielded the evidence for HF therapy. Elderly patients with severe renal dysfunction and patients with preserved left ventricular systolic function HF form the majority of HF patients, but they are underrepresented in landmark trials (14). Therefore, speculatively, it seems that the intensified use of evidence-based HF therapy is mainly effective in the subgroups where this evidence was obtained.
Previous studies have demonstrated that NT-proBNP levels fall in response to optimizing HF therapy (15–17). In PRIMA, intensifying HF therapy in patients with rising NT-proBNP levels failed to lower these levels.
The lack of benefit seen in our study is in line with the overall lack of benefit seen with studies of NT-proBNP–guided therapy by an absolute NT-proBNP target (6,7). PRIMA shows that treatment of HF guided by an individualized NT-proBNP target against a background of optimal HF therapy does not have additional beneficial effects.
As opposed to what could be expected from large HF surveys (18), only a minority of patients included in our study had a history of HF. NT-proBNP levels might decrease more in response to HF treatment in patients with de novo HF compared with patients with a history of HF and background therapy before admission. Because we excluded patients if NT-proBNP levels decreased <10%, with a minimum of 850 pg/ml, more patients with a history of HF and prolonged exposure to therapy might be excluded than patients with de novo HF.
Power analysis was based on the initial primary end point of reduction in events. Sample size was calculated to demonstrate a minimum reduction in pooled events of 30%. A post hoc power analysis indicated that a difference of 4% in percentage of time (approximately 28 days) alive outside the hospital could have been detected.
One may surmise that choosing an individual (often elevated) NT-proBNP target is not accurate enough, and that more stringent targets should be aimed for in all subjects. However, the prognostic impact we observed of NT-proBNP levels above the individual target, even when the target is clearly elevated, argues against this notion. Furthermore, the most common reaction to an elevated NT-proBNP level was to increase dosage of diuretics. We have not been able to demonstrate the ability of any subtype of intervention (evidence-based medication, diuretics, or nonpharmacologic) to be more effective in lowering off-target NT-proBNP levels. The use of our electronic therapy advisor might have lead to more intensified treatment in the clinically-guided group than would occur in daily practice where such advices are not generated.
Previous studies have demonstrated that discharge NT-proBNP levels, decrease in NT-proBNP levels during admission because of HF, and outpatient NT-proBNP levels in patients with stable, chronic HF were of prognostic importance (4,19,20). The possible additive value of a combination of static and dynamic NT-proBNP levels for determining individual prognosis still remains to be assessed.
This is the first study to our knowledge that evaluated whether HF therapy guided by an individualized NT-proBNP target level improves outcome. PRIMA shows that unstable NT-proBNP levels indeed indicate imminent events, but that intensification of currently used medication in patients on optimal HF therapy does not prevent further deterioration.
The authors would like to thank the members of the event committee, Drs. Bas Bekkers, Roger Rennenberg, Marjolein Drent, and Karin Faber for their outstanding contribution. Gratitude is also owed to Mr. Vincent Kleinen, analyst at the clinical laboratory of Maastricht University Medical Center.
Continuing Medical Education (CME) is available for this article.
Main funding (>€200,000) for this study was provided by the Netherlands Heart Foundation, Netherlands Organisation for Scientific Research (NWO), and the Royal Netherlands Academy of Arts and Sciences (KNAW)–Interuniversity Cardiology Institute of the Netherlands. Minor funding of an unrestricted research grant (<€70,000 per sponsor) was provided by Pfizer, AstraZeneca, Medtronic, and Roche Diagnostics. Dr. Pinto is a recipient of honoraria and research grants from Roche Diagnostics. All other authors have reported that they have no relationships to disclose.
- Abbreviations and Acronyms
- angiotensin-converting enzyme
- B-type natriuretic peptide
- European Society of Cardiology
- heart failure
- interquartile range
- left ventricular ejection fraction
- N-terminal pro–B-type natriuretic peptide
- quality of life
- Received March 3, 2010.
- Revision received June 10, 2010.
- Accepted July 6, 2010.
- American College of Cardiology Foundation
- Januzzi J.L.,
- van Kimmenade R.,
- Lainchbury J.,
- et al.
- Doust J.A.,
- Pietrzak E.,
- Dobson A.,
- Glasziou P.
- Hartmann F.,
- Packer M.,
- Coats A.J.,
- et al.
- Lainchbury J.G.,
- Troughton R.W.,
- Strangman K.M.,
- et al.
- Jourdain P.,
- Jondeau G.,
- Funck F.,
- et al.
- Nieminen M.S.,
- Bohm M.,
- Cowie M.R.,
- et al.
- Swedberg K.,
- Cleland J.,
- Dargie H.,
- et al.
- Maisel A.
- McCullough P.A.,
- Sandberg K.R.
- Lenzen M.J.,
- Boersma E.,
- Reimer W.J.,
- et al.
- Murdoch D.R.,
- McDonagh T.A.,
- Byrne J.,
- et al.
- Zugck C.,
- Haunstetter A.,
- Kruger C.,
- et al.
- Nieminen M.S.,
- Brutsaert D.,
- Dickstein K.,
- et al.