Pediatric Cardiology

, Volume 27, Issue 2, pp 259–262

Management Strategy for Very Mild Aortic Valve Stenosis

Authors

    • Division of Pediatric CardiologyMayo Clinic College of Medicine
  • D.J. Driscoll
    • Division of Pediatric CardiologyMayo Clinic College of Medicine
  • J.F. Keane
    • The Children’s Hospital
  • W.M. Gersony
    • Columbia University College of Physicians & Surgeons
  • C.J. Hayes
    • Columbia University College of Physicians & Surgeons
  • J.I. Brenner
    • Johns Hopkins University School of Medicine
  • W.M. O’Fallon
    • Division of Pediatric CardiologyMayo Clinic College of Medicine
  • D.R. Pieroni
    • Children’s Hospital of Buffalo
  • R.R. Wolfe
    • The Children’s Hospital of Denver
  • W.H. Weidman
    • Division of Pediatric CardiologyMayo Clinic College of Medicine
Article

DOI: 10.1007/s00246-005-1134-x

Cite this article as:
Bartz, P., Driscoll, D., Keane, J. et al. Pediatr Cardiol (2006) 27: 259. doi:10.1007/s00246-005-1134-x

Abstract

It is unclear how often patients with very mild aortic stenosis (gradients <25 mmHg) need interval follow-up. The purpose of this study was to define the determinants of disease severity progression and to propose appropriate management strategies. It is known that congenital aortic stenosis is a progressive disease that requires long-term follow-up at consistent intervals. We studied 89 patients with very mild aortic stenosis. Cox proportional hazard modeling was performed to ascertain predictors of morbidity and mortality. Events were defined as valve surgery or death. Of the original 89 patients, 7 died (92% survival); one death was sudden and unexplained and six were noncardiac. Eighteen individuals were lost to follow-up (10 not located and 8 refused participation). Twelve (17%) had valve surgery. The minimum time interval between initial diagnosis of very mild aortic stenosis and surgery was 4.6 years (mean, 14.0). Age at diagnosis, gender, initial gradient, initial gradient/age, and aortic regurgitation were found not to be predictive of outcome. However, the slope of the transaortic gradient [change of gradient/time (years)] was predictive of outcome (hazard ratio of 1.69; confidence interval, 1.4–2.2). At least 17% of these patients progress to require operation. For patients with a gradient slope <1.1, evaluation every 4 or 5 years is recommended. For patients with a gradient slope >1.2, evaluation every 1 or 2 years seems prudent.

Key words

Congenital heart diseaseDisease progressionOutcome study

The incidence of congenital aortic valve stenosis is 1.1–4.3/10,000 live births [2, 4, 7]. In children, the valve is usually bicommissural or bicuspid. Congenital aortic valve stenosis is a progressive disease that can be associated with significant stenosis and/or insufficiency and aortic root dilatation. All patients with congenital aortic valve stenosis need ongoing evaluation to detect the presence and progression of clinically significant disease. However, it is unclear how often patients with very mild aortic valve stenosis (gradients <25 mmHg) need interval examinations and echocardiographic evaluations.

The natural history has been previously well documented. The overall 25-year survival of aortic valve stenosis is 85.1%, with 58.7% of patients ultimately requiring operation. In aortic stenosis with gradients <50 mmHg, the clinical course is less malignant (92.4% 25-year survival with 33.1% need for operation) [3, 8].

The rate of progression of very mild stenosis is not well-defined. Hence, it is unclear how often patients with very mild aortic valve stenosis need interval examination and echocardiographic evaluation. In an ever-increasing cost-conscious medical system, reducing consultation and echocardiography charges without affecting the quality of care is imperative. The purpose of this study was to define the determinants of disease severity progression for patients initially identified as having very mild aortic stenosis and to propose appropriate management strategies.

Materials and Methods

We utilized a patient cohort originally identified in the First Natural History Study (NHS-1). From 1958 to 1969, investigators in the NHS-1 enrolled 462 patients diagnosed with congenital aortic stenosis. Inclusion criteria consisted of clinical diagnosis of aortic stenosis in patients younger than 21 years of age as well as a resting peak systolic gradient ≥10 mmHg between the left ventricle and the ascending aorta or peripheral artery. Cardiac catheterization was conducted at admission and most patients (68.2%) had repeat catheterization 4–8 years later [5, 8].

Between 1985 and 1990, as part of the Second Natural History Study (NHS-2), patients from NHS-1 were located and reevaluated. All located patients were asked to complete a detailed health status and medical history questionnaire. In addition, patients were invited to have a history and physical examination, electrocardiagram (ECG), and echocardiography with Doppler [6].

The current study includes all patients with a transaortic gradient between 10 and 25 mmHg (hereafter referred to as very mild aortic valve stenosis) at the time of the catheterization upon admission to NHS-1. Original data sets were reanalyzed. The outcome events of interest were aortic valve surgery or cardiac death. Univariant analysis using Cox proportional hazard modeling was performed. The independent variables of interest were age at admission to NHS-1, initial transaortic gradient upon entry to NHS-1, gender, gradient upon entry to NHS-1/age upon entry to NHS-1, the presence of aortic regurgitation, and the gradient slope. This slope was determined by the difference in the transaortic gradient at the initial and final cardiac catheterizations during NHS-1 divided by the time interval in years.

Results

Participation

Of the original cohort of 89 patients identified as having very mild aortic stenosis during NHS-1, 2 deaths were known to have occurred during NHS-1, and 5 additional deaths were discovered during NHS-2. Six of these deaths were noncardiac (e.g., car accident and gunshot). The seventh individual died of cardiac-related causes at 35 years of age. Catheterization 14 years prior to death had revealed a gradient of 15 mmHg, and autopsy exposed calcified valvular aortic stenosis with a prior myocardial infarction and current acute extension. Overall survival was 92.1% at the time of NHS-2. Of the 82 survivors, 49 (59.8%) participated fully in NHS-2, 15 (18.3%) only completed a questionnaire, 8 (9.6%) refused to participate, and 10 (11.2%) could not be located. Therefore, new data were collected on 64 members (78.0%) of the original cohort during NHS-2 (Fig. 1). In comparing with and without follow-up, odds proportional modeling was performed. No significant differences could be demonstrated between the two groups in regard to age at enrollment, gradient/age, the presence of aortic regurgitation, gender, or gradient slope. In patients without follow-up, initial gradient was slightly higher (19.7 mmHg) as compared to that of patients with NHS-2 follow-up (17.4 mmHg; p = 0.03) (Table 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00246-005-1134-x/MediaObjects/246_2005_1134_f1.gif
Fig. 1

Breakdown of very mild aortic stenosis (AS) cohort.

Table 1

Comparison of patients with and without NHS-2 follow-up using odds proportional modeling

Variable

Follow-up

No follow-up

p

Age at admission (years)

9.2

9.6

0.73

Initial gradient (mmHg)

17.4

19.7

0.03

Initial gradient/age (mmHg/years)

2.4

2.4

0.87

Aortic regurgitation

23 (53%)

3 (30%)

0.18

No. of men

60 (85%)

16 (89%)

0.63

Gradient slope

1.6

1.7

0.91

Morbidity

No patients were diagnosed with infective endocarditis during the follow-up period. Twelve patients (17%) required aortic valve surgery subsequent to the end of NHS-l. Aortic valvotomy was performed in 6 patients, aortic valve replacement was performed in 5 patients, and in 1 patient the type of aortic valve surgery was unknown. The minimum interval between initial diagnosis of very mild aortic stenosis and valve surgery was 4.6 years (range, 4.6–23.2; mean, 14.0).

Mortality/Morbidity Predictors

Age at diagnosis, gender, initial transaortic gradient, initial gradient/age, and the presence of aortic regurgitation were not predictive of outcome. However, the gradient slope was predicative of outcome (hazard ratio, l.69; confidence interval, 1.4–2.2) (Table 2, Fig. 2). Patients with an event had a mean gradient slope of 4.8 (range, 1.2–14.9), whereas patients without an event had a mean of 0.9 (range, −2.3 to 4.7). This implies that as the gradient slope increases, patients had an increased risk of sudden death or surgery.
Table 2

Mortality/morbidity predictors ascertained via Cox proportional hazard modeling

Variable

Event

No event

Hazard ratio (CI)

p

Age at admission (years)

8.6

9.4

0.95 (0.81–1.09)

0.47

Initial gradient (mmHg)

17.6

17.9

0.99 (0.87–1.16)

0.94

Initial gradient/age (mmHg/years)

2.8

2.3

1.39 (0.91–2.00)

0.12

Aortic regurgitation

2 (67%)

24 (48%)

1.32 (0.12–29.06)

0.82

No. of men

11 (85%)

57 (87%)

2.18 (0.50–2.00)

0.40

Gradient slope

4.8

0.9

1.69 (1.36–2.23)

0.00001

https://static-content.springer.com/image/art%3A10.1007%2Fs00246-005-1134-x/MediaObjects/246_2005_1134_f2.gif
Fig. 2

Point plot comparing individual gradient slopes of patients with and without an event.

Discussion

Aortic valve stenosis with a transaortic gradient <25 mmHg traditionally has been considered “trivial.” This may be a misnomer since 17% of the patients in this study ultimately required valve operation. Clearly, these patients require lifelong cardiac evaluation and monitoring, but it is unclear how frequently follow-up is necessary.

Financially, the interval between evaluations has significant implications. According to U.S. Census data, there were approximately 4 million 2-year-olds in the United States in 2000. Using a moderate estimate of the incidence of aortic stenosis, one would predict that approximately 800 2-year-olds were diagnosed with aortic stenosis in 2000. Of those, 154 (19.3%) will have very mild aortic stenosis. Currently, many practitioners would evaluate these patients with an annual examination and echocardiogram at an average cost of approximately $1,000. If examined annually, over a 25 follow-up period, these 154 patients would accrue 3080 annual examinations. Using this protocol, by 2020, the 154 children with very mild aortic stenosis diagnosed in 2000 would accrue total charges of $3.1 million.

Since the shortest interval between diagnosis and operation was 4.6 years, one could argue that extending the follow-up interval to 4 years seems reasonable. All patients who had an event had gradient slopes >1.2. Therefore, for patients with a gradient slope <1.1, follow-up every 4 or 5 years seems prudent. In the current cohort, 49% had gradient slopes <1.1. Applying this formula to the 154 patients diagnosed in 2000 would result in savings of at least $1.2 million.

Additionally, all events occurred in patients with a gradient slope >1.2. In that group, the shortest interval from the measurement of a gradient <25 mmHg to the time of an event was 4.6 years. Hence, one might conclude that patients with a gradient slope >1.2 should have evaluation at intervals of <4.6 years. Since the number of patients in this study is relatively small, it may be safest to reevaluate these patients every 2 or 3 years. Approximately one-third of the individuals would have a slope gradient >1.2 and never progress to need an intervention. Unfortunately, based on the current data there is no accurate method to predict which patients with gradient slope >1.2 will progress to intervention. Therefore, practitioners will need to exercise appropriate clinical judgment.

It should be noted that gradient slope in this data set was derived from data collected at cardiac catheterization. Currently, the slope would have to be calculated from Doppler-derived transaortic gradients, Bengur et al. [1] showed that Doppler mean gradient underestimated catheterization peak-to-peak gradient, whereas Doppler peak instantaneous overestimated this value; however, both showed a linear relationship. This implies that the slope would be valid for both Doppler mean gradient and Doppler peak instantaneous gradient.

A limitation of the current study is that NHS-1 included no patients younger than 2 years old with very mild aortic stenosis. Therefore, a financial model and recommendations should be applied only to patients 2 years of age or older. In addition, Doppler gradients for aortic valve stenosis in children can be variable depending on the operator, where the gradient was obtained, and the level of agitation of the child.

In conclusion, although very mild aortic stenosis requires lifelong evaluation, many patients can be evaluated safely at intervals of more greater than 1 year. Specifically, for patients with a gradient slope <1.1, evaluation every 4 to 5 years seems reasonable, whereas for patients with a gradient slope >1.2, evaluation every 1 or 2 years seems judicious.

Acknowledgment

We dedicate this article to Dr. William H. Weidman, who died September 2004. Dr. Weidman was the principal investigator of NHS-2.

Copyright information

© Springer Science+Business Media, Inc. 2006