Long-term pharmacotherapy for elevated low density lipoprotein levels in children: A retrospective analysis

Abstract

Background

There is limited research detailing low-density lipoprotein cholesterol (LDL-C) trends over the long term in children on various lipid-lowering medications.

Objectives

This study sought to assess factors associated with stability of LDL-C levels in children on long-term pharmacotherapy and their ability to reach the LDL-C goal of ≤ 130 mg/dL while on pharmacotherapy.

Methods

Medical records of children seen in a university pediatric cholesterol clinic between 1998 and 2012 treated with a statin, ezetimibe, or both were reviewed. Aggregate data were obtained to determine the number of children able to reach an LDL-C level of ≤130 mg/dL while on pharmacotherapy. Kaplan-Meier curve and proportional hazard regression analysis were used to examine the propensity for LDL-C levels to stabilize over time while on pharmacotherapy as well as factors affecting this propensity.

Results

Overall, 76 patients who contributed 864 total visits were included. Of the 76 patients, 56 developed a stable LDL-C with median time to stability of 28 months on pharmacotherapy. Younger age at first visit and higher medication potencies/doses were associated with an increased propensity to stabilize. Only 36 patients were able to reach an LDL-C of ≤130 mg/dL, with only 11 of 38 patients with probable familial hypercholesterolemia reaching this goal.

Conclusions

Most children reached LDL-C stability on pharmacotherapy after a median 28-month interval. However, most children had difficulty in reaching the LDL-C goal of ≤130 mg/dL even with aggressive medication titration. This was specifically true for those with probable familial hypercholesterolemia.

Introduction

Dyslipidemia, especially elevated low-density lipoprotein cholesterol (LDL-C), is well known to be a major cardiovascular risk factor in the adult population.¹ Although the risk of premature cardiovascular events in children who have high LDL-C is still not well understood, studies such as the Bogalusa heart study and the Pathological Determinants of Atherosclerosis in Youth study have clearly established a link between development of atherosclerosis and levels of circulating LDL-C in children.^2,3 As a result, the National Cholesterol Education Partnership developed guidelines for screening and treating dyslipidemias in children in 1992.⁴ One of the cornerstones of these guidelines was the use of pharmacologic treatment for children with severe elevations in LDL-C levels. At that time, medication choices were fairly limited, and the guidelines advocated the use of bile acid binding resins as the first line therapy. Given the side effect profile of these medications, compliance suffered and they eventually fell out of favor.⁵

The emergence of 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors, better known as statins, for treatment of elevated LDL-C in the adult population created a brand new potential for treating children given their favorable side-effect profile and proven efficacy for lowering LDL-C in adults. Nearly all the statins have subsequently undergone randomized controlled trials and have shown reasonably good efficacy with favorable side effect profiles in children.^6–15 In 2008, with the modification of the screening guidelines for dyslipidemias in children by the National Cholesterol Education Partnership, treatment guidelines were also modified to denote statins as the first line choice for pharmacologic treatment in children.¹⁶ Additionally, ezetimibe, which acts to decrease absorption of dietary cholesterol at the small intestine brush border, has also been shown to be efficacious in children, both as an adjunct to statin therapy and/or as monotherapy.¹⁷ Although the guidelines for screening children have substantially changed since this time, the treatment guidelines for hyperlipidemia in children have not. Current guidelines suggest starting with a low-dose statin for children who are eligible for treatment and titrating the dose after 3 months if the LDL-C is still greater than 130 mg/dL.¹⁸ This is based on dose-response relationships observed in a few of the randomized clinical trials.^6,10,13 A second titration of the same statin can be done after another 3 months if the goal LDL-C is not met or a second agent could be added. Once the goal level of 130 mg/dL is reached, guidelines recommend checking a fasting lipid profile in 8 weeks and then 3 months to confirm maintenance of the goal level. If maintenance persists, then fasting lipid panels should be checked every 3–4 months in the first year of maintenance followed by every 6 months during the second year and beyond.

Although pharmacologic treatment of elevated LDL-C in children has clearly evolved and become more commonplace, longer term data on maintenance therapy have not been as well described in the literature. Furthermore, there are few studies in children looking at escalation of maintenance pharmacotherapy and effect on LDL-C levels. The authors undertook this retrospective study to look at the long-term use of pharmacologic therapy in children for treatment of elevated LDL-C. We specifically sought to determine how quickly LDL-C levels in children typically stabilize when placed on pharmacotherapy, the effect of pharmacotherapy titration on LDL-C stabilization and ability of children to reach the LDL-C goal of 130 mg/dL while on long-term pharmacotherapy.

Methods

This study was approved by the West Virginia University Institutional Review Board. A retrospective chart review was conducted using data gathered from clinic visits to the pediatric cholesterol clinic based in rural Appalachia at West Virginia University. This clinic sees patients from Pennsylvania, West Virginia, Maryland, and Ohio. This clinic began seeing patients in 1998 and data were available from this time period; the study period spanned from 1998 to the end of 2012. Data from patients were eligible for inclusion if, at any point in time, they were placed on pharmacotherapy to specifically lower LDL-C levels. Patients had to have at least two follow-up visits after addition of pharmacotherapy in addition to their baseline visit for inclusion. Patients with baseline missing data were excluded from the analysis. Other than the referral baseline LDL-C- levels, all subsequent LDL-C levels were obtained at the West Virginia University laboratory in the fasting state.

A total of 76 patients were included in this study who contributed a total of 864 visits. Follow-up intervals were variable and at the discretion of the individual provider. On average, patients were typically seen every 3–4 months. Achievement and maintenance via titration of medications were based around a goal of LDL-C ≤130 mg/dL. To account for varying doses and potencies of different statin medications, all statins used in the study (pravastatin, simvastatin, atorvastatin, rosuvastatin) were converted to an equivalent simvastatin dose based on the percentage of LDL-C reduction that has been typically observed for each statin/dose combination.¹⁹

Baseline demographic and clinical variables are summarized using the mean, standard deviation, min, max, and median. Body mass index percentiles were calculated using the sex and age Centers for Disease Control and Prevention growth charts for the United States, body mass index-for-age growth charts.²⁰ Comparisons between groups were made using the nonparametric Mann–Whitney–Wilcoxon test. Stabilization is defined as 2 sequential visits where there is <20% change from the previous LDL-C value after medication was begun. Time to stabilization is defined as the time between the initiation of medication specifically used to lower LDL-C and when the stabilization event occurs.

We used the Kaplan–Meier (KM) curve, with Greenwood's standard error, to describe the experience of LDL-C stabilization over the observation period. Confidence intervals were constructed using the log-log transform. For comparing KM curves, we used the log-rank test (Mantel–Haenszel) to assess differences in the time to LDL-C stabilization. The median event time, the time to which 50% of the participants experience the event of interest, is reported.

The proportional hazard regression model was used to assess adjusted relationships. The considered variables included sex, age, body mass index percentile, ratio of systolic to diastolic blood pressure, the common statin dose, the baseline LDL-C level, the ratio of triglycerides to HDL, use of ezetimibe, use of supplements (omega-3 fatty acids), and if the patient has probable familial hypercholesterolemia (FH), defined as a baseline LDL-C > 190 mg/dL. The use of a statin was not included in the proportional hazard model as the common statin dose variable contains the same information as well as the dose information. The model was stratified on use of supplements.

Fractional polynomial (FP) methodology was used for variable transformation and model selection, permitting variable scaling before the estimation of the polynomial power. Degree 2 fractional polynomials and deviance-based closed testing model selection was used with α-levels for both FP transformation selection and model selection set at 0.10.²¹ The hazard ratio (HR) is reported for proportional hazard models, and the HR 95% confidence interval is reported. Graphical interpretation of the HR is used when there are nonlinear variable transformations.²²

Any KM comparison where one treatment group had <5 individuals in a group was not included in any formal testing. Statistical significance will be taken at the α = 0.05 level. All statistical analyses were performed using the R software environment for statistical computing and graphics.²³

Results

The mean baseline age of the sample was 10.8 ± 3.32 years, and the mean number of visits was 11.3 visits ± 6.4. Of note, the sample was balanced both by gender (at baseline: Male: 48.7%; Female: 51.3%; P = .908) and by the presence or absence of probable familial hypercholesterolemia (at baseline: FH present: 50%; FH absent 50%, P = 1.00). All the patients were Caucasian except for 2 (1 Hispanic and 1 African-American) as is typical for rural Appalachia. Of the patients, 54% used either Medicaid as their form of insurance or had no health insurance. Of 76 total patients, 67 (88%) were prescribed statin therapy at some point in time, and 49 (64%) were prescribed ezetimibe therapy at some point in time. Two patients who were on a statin required a change of medication to ezetimibe due to side effects. Selected baseline demographic variables are depicted in Table 1.

Table 1. Selected baseline demographic variables (N = 76).

Variable	Mean ± standard deviation
Age	10.8 ± 3.3 y
BMI percentile	86.34 ± 19.0
LDL-C	216.6 ± 71.4 mg/dL
Total cholesterol	284.2 ± 75.5 mg/dL
Triglycerides	133.7 ± 118.7 mg/dL
HDL cholesterol	42.7 ± 13.1 mg/dL
Non-HDL cholesterol	241.2 ± 71.5 mg/dL

Overall, 56 patients (73.7%) of the total sample of 76 patients ultimately reached a stable LDL-C at some point in time after being placed on pharmacotherapy. This group was compared to the 20 patients that did not reach stability and the only significant differences between them were the mean age at first visit (13.3 ± 3.1 years for those not reaching stability and 9.8 ± 2.8 years for those reaching stability, P = .004), as well as number of visits (4.9 ± 3.4 for those not reaching stability and 6.02 ± 2.8 for those reaching stability, P = .015). Of note, the baseline LDL-C difference between the two groups was not significant (209.55 ± 82.2 mg/dL for those not reaching stability and 219 ± 67.7 mg/dL for those reaching stability, P = .306).

Of the 56 individuals who stabilized, 29 (51.8%) were prescribed statin monotherapy, 18 (32.1%) were prescribed ezetimibe monotherapy, and 9 (16.1%) were prescribed both a statin and ezetimibe. In general, the Kaplan–Meier curve has a moderate descent and generally consistent, indicating that stabilization events are happening at a consistent and moderate rate (Fig. 1). The 25th percentile, 50th percentile, and 75th percentile of first stabilization times are 18 months (95% confidence interval [CI] = 16– 20), 28 months (95% CI = 13–43), and 41 months (95% CI = 3–80), respectively. These are the times at which 25%, 50%, and 75% of the individuals are expected to experience a <20% change in LDL-C levels for 2 sequential visits.

Figure 1.

Kaplan–Meier estimate of the time to LDL-C stabilization. Median time to LDL-C stabilization (red dashed lines) is 28 months. LDL-C, low-density lipoprotein cholesterol.

When stratifying by the presence of FH, children achieving LDL-C stabilization while taking a statin (either monotherapy or with ezetimibe) were doing so at a similar rate than those who were not taking a statin (P = .378, Fig. 2). Children who were not taking a statin and who did not have FH were stabilizing with a median time of 27 months, whereas those who were not taking a statin with FH were stabilizing with a median time of 29 months. Children who were taking a statin and did not have FH were stabilizing with a median time of 36 months while those who were taking a statin and had FH were stabilizing with a median time of 28 months.

Figure 2.

Kaplan–Meier estimate of the time to LDL-C stabilization for individuals stratified on statin use and presence of FH. LDL-C, low-density lipoprotein cholesterol; FH, familial hypercholesterolemia.

A similar analysis was conducted looking solely at ezetimibe. When stratifying by the presence of FH, children who were taking ezetimibe (either monotherapy or with a statin) were stabilizing at a similar rate than those who were not taking ezetimibe (P = .696, Fig. 3). Children who were not taking ezetimibe and did not have FH were stabilizing with a median time of 36 months, while those who were not taking ezetimibe who had FH were stabilizing with a median time of 28 months. Children who were taking ezetimibe with no FH were stabilizing with a median time of 27 months, and those who were taking ezetimibe with FH were stabilizing with a median time of 29 months.

Figure 3.

Kaplan–Meier estimate of the time to LDL-C stabilization for individuals stratified on ezetimibe use and presence or absence of FH. LDL-C, low-density lipoprotein cholesterol; FH, familial hypercholesterolemia.

When applying all possible combinations of FH, ezetimibe usage, and statin usage, there was no statistically significant difference between any of the curves with regards to time to reach a stable LDL-C (Fig. 4P = .579).

Figure 4.

Kaplan–Meier estimate of the time to LDL-C stabilization for individuals stratified by all combinations of FH, ezetimibe use, and statin use. LDL-C, low-density lipoprotein cholesterol; FH, familial hypercholesterolemia.

A proportional hazard model was used to further characterize the time to reaching a stable LDL-C. An optimal model was identified using FP methodology; a model containing the common statin dose, age, ezetimibe use indicator, sex, and FH indicator was selected as optimal (Table 2). The FP methodology transformed the common dose using 2 common dose terms: a shifted and scaled term (common dose + 5)/10 and a product term (common dose + 5)/10 × log ((common dose + 5)/10).

Table 2. Hazard ratio (HR) estimate, 95% confidence interval, and P value for each variable retained in the proportional hazard model.

	HR	95% CI lower	95% CI upper	P
((Common dose + 5)/10) log((common dose + 5)/10)	2.154	1.340	3.461	.002
(Common dose + 5)/10	0.235	0.085	0.653	.005
Age	0.850	0.763	0.946	.003
Ezetimibe use: yes	0.915	0.445	1.883	.810
Sex: female	0.785	0.434	1.421	.425
FH: yes	0.942	0.500	1.775	.853

As the common statin dose enters the model in a complex, nonlinear manner, the HR associated with each term should not be interpreted independently of one another. The total effect of the common dose on the HR is expressed graphically (Fig. 5). The vertical gray line is the median dose at the time of stabilization (20-mg common dose). For the typical individual in the study, Figure 5 shows that for a, a small change in the common statin dose (5 and 10 mg) will have a nominal increase of the HR. In contrast, a 15 mg increase in the common dose will increase the rate of LDL-C stabilization to 1.4 times the previous rate. For a 20 mg increase, stabilization of LDL-C levels will occur at 1.7 times the previous rate (Fig. 5, gray line). The HR is a function of both the current common dose (x-axis) and the change in the common dose. The higher the current dose of the individual, the greater the effect of a dose change on the HR. For the remaining variables, only age had a statistically significant effect on stabilization events, whereby for every year age increased, the rate of stabilization events by decreased by 15%.

Figure 5.

Effect of titration of statin dosing on propensity for LDL-C to stabilize based on current common simvastatin dose. LDL-C, low-density lipoprotein cholesterol.

Table 3 depicts the rates of stabilization, LDL-C value changes over time, and whether a child reached the LDL-C goal stratified by both presence or absence of probable FH and medication types. As far as reaching a goal LDL-C of ≤130 mg/dL, a total of 36 patients (47%) were at the LDL-C goal at the end of their observation period, defined as the LDL-C level at stabilization if the patient stabilized or the LDL-C at the most recent clinic visit if they did not stabilize. Most of these patients (n = 25) were identified as not having probable FH and most of these patients (n = 25) had stabilization of their LDL-C levels at the goal level. Among those who did not have probable FH, both ezetimibe monotherapy (11 of 14 patients) and statin monotherapy (14 of 22) allowed most patients to reach the LDL-C goal. However, ezetimibe monotherapy appeared better at maintaining stability at the goal level (9 of 14 patients) compared to the statin monotherapy group (8 of 22 patients). Both of the statin and ezetimibe monotherapy groups had similar drops in their LDL-C levels from baseline (43 mg/dL in the ezetimibe group and 39.5 mg/dL in the statin group). Interestingly, the 2 patients who did not have FH and prescribed combination therapy were both unable to reach the goal LDL-C level and stabilized at a level above the goal.

Table 3.

Summary statistics for time to stabilization of low-density lipoprotein cholesterol (LDL-C) levels stratified by baseline familial hypercholesterolemia (FH), statin use, and achievement of LDL-C goal at the time of stabilization: sample size (n), number of individuals who achieved stable LDL-C levels, the median months to achieve stabilization and 95% confidence interval, the median LDL-C at baseline, the median LDL-C at the end of observation, the P value for the difference between the baseline and end of observation LDL-C levels using the Mann–Whitney–Wilcoxon test for difference and the counts of individuals in the four combinations of stabilization and achieving goal LDL-C levels

	Probable FH
	No				Yes
	Statin				Statin
	No		Yes		No		Yes
	Ezetimibe		Ezetimibe		Ezetimibe		Ezetimibe
	No	Yes	No	Yes	No	Yes	No	Yes
N	—	14	22	2	—	7	23	8
n stable		12	13	2		7	15	7
Median months to stable LDL-C	—	27 (16–36)	36 (18–58)	59 (2–NA)	—	29 (11–39)	28(17–40)	25 (15–41)
Median, mo (95% CI)
Median LDL-C baseline	—	160	167.5	144.5	—	221	276	274
Median LDL-C at end of obs.		117	128	142		156	163	133
Diff in LDL-C P value		.006	<.001	1.000		.016	<.001	.008
Not stable, LDL-C >130	—	0	3	0	—	0	6	0
Not stable, LDL-C < 130		2	6	0		0	2	1
Stable, LDL-C >130		3	5	2		5	12	4
Stable, LDL-C < 130		9	8	0		2	3	3

Among patients that were diagnosed as having probable FH, only 11 of 38 (28.9%) were able to reach the LDL-C goal, whereas 29 of 38 patients (76.3%) developed a stable LDL-C level. Eight of the 11 patients who were able to reach the LDL-C goal ultimately stabilized at the goal. Notably, 21 of the 38 patients with probable FH reached a stable LDL-C level above the goal. When comparing each treatment within the probable FH category, combination therapy seemed to have the best chance at both getting to the goal LDL-C level (4 of 8 patients) and staying stable at the goal (3 of 8 patients). All patients prescribed ezetimibe monotherapy who had probable FH developed a stable LDL-C. Of note, combination therapy led to the greatest drop in LDL-C from baseline (141 mg/dL or 51.4%) when compared with ezetimibe monotherapy (65 mg/dL or 29.4%) and statin monotherapy (113 mg/dL or 40.9%) although all 3 categories did lead to statistically significant declines in LDL-C levels.

Discussion

In this particular study, the use of pharmacologic agents used to lower LDL-C levels in children were studied retrospectively to determine their ability to keep LDL-C levels stable over time, to determine the effects of statin titration on LDL-C levels, and the ability to get children to the goal level of 130 mg/dL.

Most children in our study did achieve stability of their LDL-C, as we defined it, at some point during their treatment course. Children who were seen at earlier ages in the clinic (and hence, likely to be a proxy for earlier treatment) were much more likely to develop stability than those who were seen later. It took the average child slightly .2 years to achieve this stability, and there was no significance noted between the use of statins and ezetimibe on how quickly a stable LDL-C value was reached. There potentially is a pubertal influence contributing to older children being less likely to stabilize as LDL-C levels are well known to be fairly volatile during puberty. Furthermore, older adolescent children who are prescribed treatment are more likely to behave differently than their younger counterparts. They may be less compliant with treatment as their parents may task them with more independent responsibility as they are older. Their dietary choices may also be different as well, which may be likely to contribute. Given the retrospective nature of this study, data on compliance and specific dietary habits were extremely difficult to capture.

On the other hand, most children in our study did not reach the goal when accounting for stabilization. Children who did reach the goal of 130 mg/dL typically had a lower baseline LDL-C, as evidenced by the fact that most children with probable FH did not reach the goal. Nearly half of the children in the study reached a stable LDL-C above the goal level. It is possible that children who were defined as reaching stability could have either destabilized later and ultimately developed an LDL-C ≤130 mg/dL, or they could have had several visits by, which their LDL-C levels dropped by no more than 20% and reached the goal that way. The latter condition may not have been captured if a patient had 2 sequential visits of <20% change after starting medication, which would have ended their observation period due to reaching stability. In fact, 63 of the 76 patients did in fact reach an LDL-C of ≤130 mg/dL at 1 visit. However, an analysis was done to look at the number patients who were at the goal LDL-C at the last clinic visit, independent of reaching LDL-C stability. Interestingly, 25 patients without probable FH and 11 patients with probable FH were at the goal, which are the same numbers when accounting for stabilization. Thus, it appears that several patients may have sporadically reached the goal but were clearly unable to maintain it. Furthermore, it appears that initially stabilizing above the goal may be a reasonable predictor for the inability not only to reach the goal LDL-C level later on but also to maintain levels at the goal.

One of the more interesting things to note was the effect of titrating statins during maintenance treatment of elevated LDL-C levels. As our data show, increasing the simvastatin common dose in children (in effect, equivalent to changing to a more potent statin and/or increasing the actual dose) made stabilization of the LDL-C levels appear to occur much more quickly. One likely reason for this is that since our definition of stabilization required 3 sequential visits, the higher doses or higher potency statins would have likely been prescribed at the third visit, once it was clear that the LDL-C changed very little from the previous 2 visits. For example, a patient with a baseline LDL-C of 197 may have initially been started on simvastatin 5 mg and now has an LDL-C of 147 at their first follow-up visit. A titration to 10 mg is completed since the patient remains above the LDL-C goal of 130 mg/dL and the LDL-C drops to 140. As the patient is still above the goal, another titration to 20 mg occurs with a repeat LDL-C of 139. This patient would meet our definition of a stable LDL-C level, and it would be associated with the 20-mg dose of simvastatin. This is very clinically relevant as it would appear from our data that when a child is started on a specific statin at a specific dose, they will get most significant LDL-C reduction solely related to that specific medication and dose. Continuously titrating the statin after the full effect of LDL-C reduction has been realized may not add any additional significant LDL-C reduction. In our patients with probable FH, this appeared very apparent as over half of these children reached a stable LDL-C above the goal of 130 mg/dL. In fact, this group appeared to do better in terms of reaching the goal LDL-C when ezetimibe was added instead of titrating the statin.

One could argue clinically, based on the data regarding stabilization, statin titration, and propensity to reach the LDL-C goal that the current treatment guidelines can be slightly modified for treatment of elevated LDL-C levels based on whether a patient has FH: If the patient does not have probable FH, starting with either ezetimibe mono-therapy or statin therapy is reasonable as both appear to be equally efficacious in bringing down LDL-C levels significantly, although ezetimibe is superior in leading to stable LDL-C levels as well. If a statin is used, the lowest dose should be started as per the current recommendation and 1 dose titration could be considered if the patient is not at the LDL-C goal. McCrindle et al were able to increase the amount of patients meeting the LDL-C goal at 26 weeks of treatment by nearly one-third with a 10 mg to 20 mg increase in atorvastatin after an initial 4 weeks of treatment.⁹ However, caution should be used with further dose titrations beyond this initial (either via increasing statin potency or increasing the dose) given that this would not appear to lead to more significant LDL-C decline and potentially lead to more side effects. Stein et al appeared to demonstrate this phenomenon with their study using lovastatin 10 mg and titrating to 20 mg and 40 mg. Initially, a 17% drop from baseline LDL-C was observed at 10 mg followed by an additional 7% at 20 mg, but only an additional 3% was found at 40 mg.¹³ To our knowledge, no prospective studies exist evaluating a switch from a lower potency statin to a higher potency statin in children. We cannot say for certain that potentially using dual therapy with a statin and ezetimibe would be of any benefit given that only 2 patients who did not have probable FH were in this category. For patients who have probable FH with severe elevations in LDL-C (>250 mg/dL) initial statin monotherapy with either 1-dose titration or immediate addition of ezetimibe if the LDL-C goal is not met may be the more beneficial rout in this group. The IMPROVE-IT trial in adults has clearly shown the benefit of adding ezetimibe in further reducing LDL-C levels after maximal LDL-C decrease has occurred on a statin.²⁴

Our data specifically on patients who have probable FH bring about another important issue that have recently been explored in the treatment of elevated LDL-C levels in adults and need further clarification in children. Goal-directed therapy had been the mainstay of adult treatment guidelines for several years, until recently. This paradigm was abandoned in favor of treatment based purely on risk stratification. If an adult belongs to one of 4 high risk groups (including LDL-C . 190), then a moderate potency statin/dose combination is recommended, and no dose titrations are necessary as the guidelines do not recommend obtaining additional LDL-C measurements.²⁵ Given the difficulty in reaching the LDL-C goal of #130 mg/dL for our probable FH patients, it makes sense to compare the effects on specific cardiovascular outcomes of titrating medications based around the current pediatric LDL-C goal to simply starting a moderate potency statin (or ezetimibe), independent of its effects on LDL-C levels.

The strengths of this study include the large number of data points and variables, high mean length of time spent on medications, variety of medications, and dosages studied, and the ability to specifically analyze the effects of not only titrating statin dosages but also titrating statin potency. This allowed for a detailed analysis permitting an in-depth time-to-event analysis for LDL-C stabilization and its relationship to medication titration and ability to reach the LDL-C goal of 130 mg/dL. Limitations include the retrospective nature of the analysis, the potential limitation for generalizability due to the geographic area from which the clinic draws patients, the relatively small sample size limiting power for some of the groups, and the inability to accurately measure potential noncompliance with medication and the potential effect of dietary choices on LDL-C levels.

Conclusion

This retrospective study looked at the use of lipid-lowering pharmacotherapy over the long term in children in terms of looking at stability of LDL-C levels over time and its relationship to medication titration and ability to reach the LDL-C goal of 130 mg/dL. Most children reached stable LDL-C levels on average after 2 years of pharmacotherapy, and they typically did so if they were seen in clinic and treated at earlier ages, and this was independent of the type of pharmacotherapy they received and whether they had probable FH. However, most children had difficulty in meeting the LDL-C goal level, especially if they were found to have probable FH. Furthermore, titration of statin medication beyond the initial dose really did not contribute any further significant LDL-C decline and simply appeared to promote stability of LDL-C levels. The addition of ezetimibe to a statin did appear to help children with severe familial hypercholesterolemia achieve the LDL-C goal. Continuing studies are needed to address whether using pharmacotherapy to lower LDL-C levels in children will impact their cardiovascular morbidity and mortality when they reach adulthood. This particular study allows us to begin to address the issue of using pharmacotherapy in children over a relatively long term.