Screening for valve disease in patients with hyperprolactinaemia disorders prescribed cabergoline: a service evaluation and literature review

Abstract

Objectives:

The indication for screening for valvular heart disease in patients taking cabergoline is based on evidence from patients with Parkinson’s disease on high-dose medication. However, current patients take much lower doses for indications such as hyperprolactinaemia disorders. Contemporary guidelines for echocardiogram monitoring in patients taking cabergoline are conflicting. This study aimed to review current clinical practice in our area regarding echocardiographic screening and to review the literature examining the evidence of valvular heart disease in patients taking lower dose cabergoline.

Methods:

This was a retrospective study of all patients with hyperprolactinaemia disorders prescribed cabergoline in a single UK NHS health board between January 2014 and July 2015. The proportion of patients receiving baseline and follow-up echocardiograms was recorded. A review of the published literature was carried out using the databases EMBASE and Medline to examine the current evidence for the effect cabergoline has on cardiac valves in patients treated for hyperprolactinaemia disorders.

Results:

The mean age was 51.7 ± 16.5 years with a 64.4% female predominance. The mean duration of therapy was 5.9 years ± 4.1 years. Of the total cohort (n = 45), two (4.4%) patients had an initial baseline echocardiogram and five (13.2%) had follow-up echocardiograms every 24 months. Of the 25 articles identified, 12 showed no clinically significant evidence of valvular dysfunction in the cabergoline group groups. Of the remaining 13 articles, evidence for valvular changes was confined to high cumulative dose cabergoline patients and there was only one confirmed case of ‘cabergoline associated valvulopathy’ described.

Conclusions:

Clinically significant valvular dysfunction is uncommon and generally only reported in high cumulative dose treatment groups. We propose that clearer national guidelines are required and that echocardiogram screening be reserved for patients who are high risk, are taking a high weekly dose (≥2 mg cabergoline weekly) or high cumulative dose.

Introduction

Cabergoline is an ergot derived orally active synthetic dopamine receptor agonist (D2) used in the treatment of prolactinomas, other hyperprolactinaemic states and previously in higher doses in patients with Parkinson’s disease.¹

Early studies of cabergoline reported an association with cardiac valve incompetence and raised sufficient concerns to lead to a recommendation for valve screening.^2,3 Agonist activity at the (5-HT)2B serotonin receptor leading to fibroblast proliferation and fibrosis has been proposed as a pathological mechanism of this adverse effect.⁴ In general, these studies showed a dose-related effect leading to moderate to severe regurgitation, principally of the left heart.^2,5 The dose of cabergoline in these studies was typically greater than 3 mg per day with a cumulative dose of 3 g.⁶

However, subsequent studies have failed to show convincing evidence of valve dysfunction in patients treated with lower doses of cabergoline, such as used in the treatment of prolactinomas and other hyperprolactinaemia disorders.⁷ The dose now used in these conditions is significantly lower due to its high potency for suppressing prolactin and prolonged elimination half life, allowing for once weekly or even less frequent administration.^8,9 Doses typically range from 0.5 to 4 mg weekly compared with doses of up to 3 mg daily historically used in Parkinson’s disease.¹⁰

Guidelines on the need for echocardiographic surveillance for acquired cardiac valve dysfunction in patients treated with cabergoline vary.

The manufacturer’s summary of product characteristics (SPC) for cabergoline states that valvulopathy has been associated with cumulative doses and prolonged use. It recommends that ‘All patients must undergo a cardiovascular evaluation, including echocardiogram to assess the potential presence of asymptomatic valvular disease’ and that ‘following treatment initiation, the first echocardiogram must occur within 3–6 months, thereafter, the frequency of echocardiographic monitoring … must occur at least every 6–12 months’.³ The US Food and Drugs Administration (FDA) also follow this recommendation.¹¹

However, in the UK there are no National Institute for Clinical Excellence guidelines on echocardiogram monitoring for patients taking cabergoline. The Medicines and Healthcare Products Regulatory Agency (MHRA) published a public assessment report on cabergoline, following the same recommendations as the FDA and the SPC.¹²

None of the above guidance considers dose, however the 2011 endocrine society guidelines state that echocardiogram surveillance is probably not needed in those taking less than 2 mg per week.¹³ The UpToDate topic on the management of hyperprolactinaemia takes a similar standpoint, suggesting ‘using the lowest dose of cabergoline necessary to lower prolactin to normal’ and that ‘cardiac ultrasonography should be ordered approximately every 2 years in patients who take larger than typical doses of cabergoline (e.g. >2 mg per week)’.¹⁴

There is inconsistent advice resulting in uncertainty within clinical practice surrounding which patients should receive echocardiogram screening, especially given that high-dose cabergoline is rarely used in clinical practice due to availability of alternative therapies for Parkinson’s disease.

Aims

This study aimed to review current clinical practice in our area to determine if patients with hyperprolactinaemia disorders were receiving echocardiographic screening in line with national and international recommendations and to review the literature examining the current evidence for the effect cabergoline has on cardiac valves in patients treated for hyperprolactinaemia disorders. Furthermore, we sought to discover if other centres have systems in place to undertake echocardiographic screening in patients taking cabergoline.

Methods

Design and setting of service evaluation

This was a retrospective study of all patients with hyperprolactinaemia disorders prescribed cabergoline in NHS Highland (North) between January 2014 and July 2015 inclusive. This health board provides all public health care for a population in excess of 250,000 people in a remote and rural setting. All 14 health boards in NHS Scotland were contacted to establish if they had local guidelines or protocols on echo surveillance for patients taking cabergoline.

Case selection

Patients were identified who had been prescribed cabergoline from the prescribing information system (PIS). PIS holds the full dataset for community pharmacy prescribing and dispensing in Scotland. It is populated from payment processing data produced by the Practitioner Services Division, comprising scanned prescription forms and electronic message data. A clinical case notes review was undertaken to determine the underlying clinical condition and dose of cabergoline. Clinical notes and review of EchoPac (GE healthcare Vs 113) were used to determine echocardiogram indication, scan frequency and results.

Data handling and statistics

Data were entered into an Excel spreadsheet (Microsoft Excel 97-2003). Descriptive statistics were used to show what proportion of baseline and follow-up echocardiograms were performed. Data are expressed as mean ± standard deviation unless otherwise stated. For continuous measures a T-test difference in means between groups was performed. The cohort was subdivided into those on ‘high-dose’ (≥2 mg weekly) and ‘low-dose’ (<2 mg a week) cabergoline, those patients that received echocardiogram surveillance and those that did not. This distinction was based on the available evidence on risk for valvular heart disease in this population and the clinical practice guidelines published by the Endocrine Society.¹³

The duration of follow up for patients currently taking cabergoline was from the date of their first prescription until the last date of medical contact. For those patients not currently taking cabergoline the duration of follow up was from the date of their first prescription until their last prescription.

Patients were considered to have had an appropriate baseline echocardiogram if an echocardiogram was performed up to 6 months prior to the commencement of cabergoline therapy. The interval between echocardiograms was taken from the dates the procedures were performed or from the date of the first cabergoline prescription until the date of the first echocardiogram. In the absence of clear and consistent guidelines we took a pragmatic (and lenient) approach to set the standard for frequency of follow-up echocardiogram at every 2 years. If patients had stopped taking cabergoline within 2 years of starting they were not included in the follow-up echocardiogram analysis.

Ethics

This was a service evaluation and therefore ethical approval was not required.

Literature review

A review of the published literature was carried out using the databases EMBASE (1980–2016) and Medline (1946–2016). The search was undertaken using the terms ‘cabergoline and heart valve’, ‘cabergoline and cardiac’, ‘ergolines and cardiac’ and ‘ergolines and heart valve’. All of the relevant data returned were published between 2008 and 2016. A grey literature search was performed and reference lists of identified articles were scrutinized.

Studies from peer reviewed journals published in English (or with English translation) examining the association between cardiac valve abnormalities and cabergoline in patients with hyperprolactinaemia disorders were included. The relevant data were collected from the published full text. Case reports were excluded. Databases were last searched 1 July 2016. This is shown in Figure 2.

Figure 2.

Flow diagram for literature review.

When available, the variables included in the summary of relevant articles included:

• date of publication,

• type of study,

• study participant numbers,

• age,

• duration of treatment,

• cumulative dose of cabergoline,

• weekly dose of cabergoline.

When available, the outcomes included in the summary of relevant articles included:

• difference in prevalence of valvular dysfunction,

• difference in means,

• odds ratio (OR),

• relative risk (RR).

Studies were not individually assessed for bias. The study summaries were presented in tabulated form. This was not registered as a systematic review.

Results

A total of 56 patients were identified as having been prescribed cabergoline between January 2014 and July 2015. Of these, 11 patients were excluded leaving 45 patients eligible for analysis (Figure 1). Clinicians in our institution report variation in their approach to echocardiogram surveillance. Some are influenced by MHRA guidance and the Society For Endocrinology 2011 position, aiming to perform echocardiograms at baseline and 1–2 yearly. Others are influenced by the available literature and perform baseline echocardiograms and only perform surveillance follow-up echocardiograms in at-risk patients or those on higher-dose cabergoline.

Figure 1.

Flow diagram for patient selection.

Patient demographics, indication and dose

The mean age was 51.7 ± 16.5 years, with a 64.4% female predominance. The reasons for cabergoline use are shown in Table 1 along with the recommended British National Formulary (BNF) doses. The mean duration of therapy was 5.9 years ± 4.1 years.

Table 1.

Indication and dosing of cabergoline-treated patients.

Indication	Proportion of patient’s, n (%)	Dosing
Prolactinoma	33 (73.3) n = 2 missing data on dose	250 μg twice weekly – 2 mg twice weekly
Nonfunctioning pituitary adenoma	5 (11.1)	0.5 mg twice weekly – 1 mg twice weekly
Hyperprolactinaemia	3 (6.7)	250 μg twice weekly – 1 mg weekly
Ongoing galactorrhoea despite normal prolactin	1 (2.1)	0.25 mg weekly
Treated acromegaly	1 (2.1)	500 μg twice weekly
Missing data (indication)	2 (4.3)	1 mg weekly

Of the total cohort (n = 45) only two (4.4%) patients with hyperprolactinaemia disorders had an appropriate initial baseline echocardiogram performed before commencing cabergoline. Furthermore, only five (13.2%) patients had appropriate follow-up echocardiograms every 24 months (average interval 0.8 ± 0.5 years). A total of seven patients were excluded from all follow-up analysis as the duration of treatment was less than 24 months and they would not have required follow up. A total of 80 echocardiograms were performed for all patients during the follow-up time period. The vast majority of echocardiograms performed showed no valvular abnormalities. Mild to moderate mitral, tricuspid and aortic regurgitation were reported in a small number of studies, as was slight mitral leaflet calcification, mildly thickened aortic valve, mild aortic stenosis and aortic regurgitation, which were of no clinical significance.

Patients who received echocardiogram assessment compared with those who did not

Of the patients who had appropriate baseline echocardiograms (n = 2) these were all undertaken specifically for cabergoline monitoring; that is, no patients had concomitant valve disease. The mean age was 45.5 years ± 14.8 years. The mean age of those who did not have any baseline echos (n = 43) was 52.0 years ± 16.7 years. There was no statistically significant difference between the age of the two groups (p = 0.59).

Of the patients who had at least one follow-up echocardiogram (n = 25) the majority were for cabergoline monitoring with other indications, including general valve assessment, assessment of heart murmurs, presurgery assessment, left ventricular assessment and post non ST elevation myocardial infarction valve replacement. The mean age was 55.8 ± 16.5 years. The patients who had no follow-up echocardiograms (n = 13) had a mean age of 49.5 ± 17.2 years. There was no statistically significant difference between the age of the two groups (p = 0.26).

Patients on high- and low-dose cabergoline

High-dose (≥2 mg/week) cabergoline was prescribed in five (11.1%) patients. The mean duration of treatment was 7.1 ± 3.4 years. The mean age was 63.0 ± 11.6 years. Of these five patients, 0 (0.0%) had an initial echocardiogram before starting treatment. Regarding appropriate follow up, one (20%) patient had no follow-up echocardiogram and one (20%) patient had follow up at least every 24 months with three echocardiograms performed with an average interval of 0.7 ± 0.4 years. Of the remaining three patients (60%), 12 echocardiograms were performed with an average interval of 2.0 ± 1.0 years. A total of 15 echocardiograms were performed during the follow-up period. If guidelines had been followed, this should have been 21.

Low-dose (<2 mg/week) cabergoline was prescribed in 40 (88.9%) patients. The average duration of treatment was 5.5 ± 4.1 years. The mean age was 50.3 ± 16.6 years. Of these patients, two (5.0%) had an initial echocardiogram before starting treatment. Regarding appropriate follow up, four (12.1%) had follow-up echocardiograms consistent with 24 months between scans, with seven echocardiograms performed at an average interval of 0.8 ± 0.5 years; 12 (36.4%) had no follow-up echocardiograms. Of the remaining 17 (51.5%), 56 echocardiograms were performed with an average interval of 2.3 ± 2.0 years. A total of 65 echocardiograms were performed during the follow-up period. If guidelines had been followed, this should have been 135. There was no statistically significant difference between the age of the two groups (p = 0.11). Table 2 shows a summary of the total number of recommended and observed echocardiograms.

Table 2.

Number of recommended and observed echocardiograms performed.

	Recommended*	Observed
Baseline	45	2
Follow up based on frequency of 2 years	111	80
Difference	Recommended – observed	31

^*Based on cohort of 45 patients followed up for an average of 5.9 years.

Literature review

From the database search, 439 articles were identified with 334 remaining after duplicates were removed. Of these, 286 were excluded on title and abstract. This left 48 articles included in the full article review. Of these, 23 were excluded leaving 25 relevant articles for analysis. Table 3 shows a summary of relevant published studies to date. There were no randomized control studies identified. Of the 25 articles reviewed, 12 showed no valve changes or dysfunction as a result of cabergoline use. Of the remaining 13 articles describing valve changes in patients taking cabergoline, only one confirmed case in one paper was defined as ‘cabergoline-associated valvulopathy’.

Table 3.

Effects of treatment with cabergoline for hyperprolactinaemia on valvular heart disease: overview of the literature.

Author/s (year)	Study design	Sample size (n) and characteristics	Intervention	Results	Cabergoline-associated valvulopathy demonstrated?
Lancellotti et al.1	Cross-sectional case-controlled study	Cabergoline group (n = 102) Matched control group (n = 51) Mean age cabergoline group 51 ± 14 years, control group 48 ± 12 years	Duration of treatment was 12–228 months The median cumulative dose was 204 mg (range 18– 1718) Weekly dose of cabergoline range 0.25–5.25 mg	No significantly increased risk of clinically relevant valvular regurgitation or leaflet thickening There was a higher mitral valve tenting area in patients (1.34 ± 0.32) compared with controls (1.23 ± 0.22), p = 0.03	No
Wakil et al.15	Cross-sectional comparative study	Cabergoline group (n = 44) Matched control group (n = 566) Mean age cabergoline group 41.8 ± 13.2 years, control group 36.3 ± 8.7 years	Mean duration of treatment 44.8 (median 33) months Mean cumulative dose 311 mg Weekly dose range 0.25–4 mg	An increased risk of mild grade tricuspid (OR 3.1, CI 1.0–9.6, p = 0.04) and pulmonary (OR 7.8, CI 0.8–78.4, p < 0.0001) valve regurgitation was found with cabergoline but this was deemed not clinically significant	No
Bogazzi et al.16	Observational, case-controlled study	Cabergoline group (n = 100) Matched controls group (n = 100) Mean age cabergoline group 41 ± 13 years, control group 38 ± 7 years	Mean treatment duration 67 ± 39 months Mean cumulative dose 279 ± 301 mg (range 15–1327)	There was no statistically significant difference between the two groups for the measured outcomes	No
Colao et al.17	Observational, case-controlled study	Cabergoline group (n = 50) Matched control group (n = 50) De novo patients, diagnosed but untreated (n = 20) Mean age cabergoline group 36.5 ± 10.5 years, control group 36.7 ± 10.4 years, de novo group 28.2 ± 8.7 years	Treatment duration: 32% (12–60 months) and 68% (>60 months) Median cumulative dose cabergoline 280 mg (range 32–1938) Weekly dose of cabergoline range 0.2–7 mg	Statistically significant increase in moderate tricuspid regurgitation in patients taking cabergoline at cumulative doses greater than 280 mg, compared with de novo patients and control subjects (54% versus 0% versus 18%, p = 0.023). No other statistically significant differences found. Only valvular changes were seen without associated thickening thereby not meeting definition of cabergoline-associated valvulopathy	No
Devin et al.18	Retrospective single-arm study	Cabergoline group (n = 45) Mean age 41 ± 10 years	Mean duration of therapy 39 ± 29 months Mean cumulative dose 145.7 mg Mean weekly dose 0.91 ± 0.96 mg	Valve abnormalities in 3 of 45 patients (7%) No cases of moderate or severe valvular lesions Prevalence of valvular heart disease no different to previously reported normal populations	No
Kars et al.19	Cross-sectional study	Dopamine agonist group (n = 78, n = 47, cabergoline patients) Matched control group (n = 78) Mean age cabergoline group 47 ± 1.4 years, control group 48 ± 0.9 years	Mean duration of therapy with cabergoline 5.2 ± 0.4 years Mean cumulative dose 363 ± 55 mg	For pooled dopamine agonist patients, compared with controls, there was a statistically significant increase in aortic valve calcifications (40% versus 18%, p = 0.003) and mild tricuspid regurgitation (41% versus 26%, p = 0.042). However, no clinically relevant valvular heart disease seen	No
Vallette et al.20	Prospective multicentre control matched study	Cabergoline group (n = 70) Control group (n = 70) Mean age in cabergoline group 44 ± 13 years, control group 45 ± 16 years	Mean duration of treatment was 55 ± 22 months Mean cumulative dose 282 ± 271 mg Weekly dose range 0.5–4 mg	No clinically significant association with low-dose cabergoline use and the development of valvulopathies	No
Herring et al.6	Case-control study	Cabergoline group (n = 50) Matched control group (n = 50) Mean age cabergoline group 51.2 ± 2.2 years, control group 50.9 ± 2.2 years	Mean treatment duration 6.6 ± 0.5 years Mean cumulative dose 443 ± 53 mg	No clinically significant association with low-dose cabergoline use and the development of valvulopathies	No
Nachtigall et al.21	Case-record retrospective study	Cabergoline group (n = 100) Matched control group (n = 100) Mean age cabergoline group 44 ± 13 years, control group 44 ± 13 years	Mean duration of treatment 48 ± 4 months Mean cumulative dose of cabergoline 253 ± 52 mg Mean weekly dose 0.8 ± 0.2 mg	No significant differences in valvular function in cabergoline group compared with controls Female patients had a higher prevalence of mild tricuspid regurgitation than female controls (15.4% versus 1.9%, p = 0.03) Male patients had more trace tricuspid regurgitation than male controls (68.8% versus 45.8%, p = 0.02) Valvular regurgitation was not clinically significant with no correlation to dose or duration	No
Lafeber et al.22	Case-control study	Cabergoline group (n = 115) Matched control group (n = 115) Mean age cabergoline group 50.3 ± 14.7 years	Median treatment duration 115 months (range 7–551) Median cumulative cabergoline dose 277  mg (range 16–3385 ) Median weekly dose of cabergoline of 1.0  mg (range 0.25–12.0)	No significant difference between the prevalence of clinically relevant valvular regurgitation or mitral valve tenting area	No
Tan et al.23	Cross-sectional study	Cabergoline group (n = 72) Matched control group (n = 72) Age cabergoline group 36 years (range 30–45), controls 38 years (range 31–49)	Duration of therapy was 53 months (range 26–96 months) Median cumulative dose of cabergoline was 126 mg (range 58–258) Mean weekly dose of cabergoline 0.8 ± 0.2 mg	No association between cabergoline treatment for hyperprolactinaemia and valvulopathies	No
Gu et al.24	Cases Investigated effect of masking of drug exposure	Cabergoline group (n = 24) Cabergoline and bromocriptine group (n = 13) Bromocriptine group (n = 3) Mean age 49.3 ± 9.6 years	Mean duration of therapy 119.3 ± 52.8 months Mean cumulative dose of cabergoline alone 465 ± 561 mg, both cabergoline and bromocriptine 822 ± 1442 mg, bromocriptine alone 8793 ± 9191 mg	Low-dose dopamine agonist therapy rarely causes cardiac valve abnormalities Evidence of over reporting of valve regurgitation in group B	One case of mitral regurgitation concluded by authors to be due to cabergoline
Boguszewski et al.25	Case control	Cabergoline group (n = 51) Bromocriptine group (n = 19) Control group (n = 59) Mean age cabergoline group 42.3 ± 13.5 years, bromocriptine group 41.8 ± 11.5 years, control group 43.7 ± 13.0 years	Mean duration of therapy of cabergoline group 37.8 ± 21.3 months, bromocriptine group 54.8 ± 30.2 months Cumulative dose of cabergoline group 16–1286.8 mg, bromocriptine group 4687.5–23,478.8 mg	Increased trace mitral regurgitation (49% versus 27.1%; p = 0.02), trace tricuspid regurgitation (45.1% versus 20.3%; p = 0.0003) and mild tricuspid regurgitation (7.8% versus 0%; p = 0.0003) in cabergoline group compared with controls. Not considered clinically significant	No
Delgado et al.26	Two-year single-arm follow up of Kars et al.19 cross-sectional study	Dopamine agonist patients (n = 74, n = 45 cabergoline patients) Mean age cabergoline group 48 ± 1.8 years	At initiation of Kars et al.19 mean duration of therapy with cabergoline 5.2 ± 0.4 years Cumulative dose of cabergoline at 2-year follow up was 401 ± 55 mg	In pooled patients there was a significant increase in the prevalence of valvular calcifications (48–58%, p = 0.004). In particular, aortic valve calcifications (39–53%, p = 0.002) In patients taking cabergoline, there was a significantly higher prevalence of aortic valve calcification compared with the group of patients not treated with cabergoline (63% versus 38%, p = 0.016) These structural changes were not accompanied by an increased prevalence of valvular dysfunction or other changes	No
Maione et al.27	Cross-sectional longitudinal study	Cabergoline group (n = 42, n = 26 followed up longitudinally) Matched control group (n = 42, n = 26 followed up longitudinally Median age cabergoline group 42 years (range 15–73), control group 48 years (range 27–73)	Median duration of treatment 35 months (range 4–132) Median cumulative dose 203 mg (range 15–990) Median weekly dose of 1 mg (range 0.5–3.5)	Treatment group did not have a statistically significant increase in prevalence or incidence of valve abnormalities	No
Yarman et al.28	Retrospective	Cabergoline group (n = 14) Bromocriptine group (n = 8) Mean age 45 years (range 20–67)	Mean duration of therapy 61 (range 12–128) Mean cumulative dose 155 mg (range 50–351)	No valve abnormality in any patients	No
Elenkova et al.29	Case-control study	Cabergoline group (n = 103) Bromocriptine group (n = 55) Naïve patients group (n = 74) Control group (n = 202) Mean age cabergoline group 38.6 ± 9.93 years, control group 38.0 ± 9.65 years	Mean duration of treatment 46.5 ± 21.1 months Mean cumulative dose 173.9 mg (range 24–1322)	No statistically significant difference in clinically relevant valve regurgitation Subclinical valve fibrosis occurred significantly more often in the cabergoline-treated group compared with controls (40% versus 23.5%; p = 0.004). OR for developing valvular fibrosis with cabergoline was 2.27 (95% CI 1.17–4.41; p = 0.016)	No
Halperin et al.30	Retrospective, multicentre, cross-sectional study	Dopamine agonist group (n = 83) Cabergoline group (n = 62) Matched control group (n = 58) Mean age cabergoline group 37.1 ± 10.6 years, control group 38.6 ± 9.7 years	Mean duration of treatment 204.8 ± 168 weeks Mean cumulative doses of cabergoline 217.4 ± 306.6 mg	No clinically significant increase in valvular dysfunction in cabergoline versus control groups There was a statistically significant increased risk of mild tricuspid regurgitation associated with high cumulative doses of cabergoline >180 mg compared with controls (40% versus 13.8%, p = 0.024)	No
Córdoba-Soriano et al.31	Case-control study	Cabergoline group (n = 32) Controls (n = 32) Mean age cabergoline group 38.8 ± 10.4 years, control group 46.68 ± 12.5 years	Median treatment duration of 46 months (range 30–96) Cumulative dose 158 mg (range 63–363) Mean weekly dose 0.5 mg (range 0.25–1.25)	One case of moderate tricuspid regurgitation in cabergoline group at a cumulative dose of 130 mg. No morphological changes in the valve where noted when examined closer	No
Auriemma et al.32	Single-arm prospective study	Cabergoline group (n = 40) Mean age 38.7 ± 12.5 years	Median cumulative dose 48 mg (range 12– 588) at 24 months and 149 mg (range 48–1260) at 60 months	No increased risk of clinically significant cardiac valve regurgitation from baseline after 5 years of treatment with cabergoline	No
Drake et al.9	Cross-sectional study	Cabergoline group (n = 601) Cabergoline and bromocriptine group (n = 110) Bromocriptine group (n = 32) Median age 42 years (range 34–52)	Median cumulative dose in cabergoline group 152 mg (range 50–348), bromocriptine group 7815 mg (range 1764–20,477). Median daily dose in cabergoline group 0.5 mg (IQR 0.5–1.0)	2.6% (n = 20) had moderate regurgitation or stenosis. 1.7% (n = 13) had moderate thickening, calcification or restricted mobility. One case had myxomatous tricuspid and nine had myxomatous mitral valve. Fourteen cases (1.9%) had moderate mitral valve calcification. 2.6% (n = 20) patients had moderate regurgitation or stenosis. Investigators concluded no presence of the triad of cabergoline associated valvulopathy	No
Caputo et al.7	Single-arm cross-sectional study	Cabergoline group (n = 40) Mean age 44.1 ± 12.8 years	Mean duration of treatment 71.2 ± 66.2 months Mean cumulative dose of cabergoline 391.3 ± 532.7 mg	Five cases (13%) of valvular thickening. Two cases of myxomatous mitral valve disease. Three cases of age-related aortic sclerosis. These patients had normal cardiovascular examination	No
Drake et al.33	Follow up of previously reported cross-sectional study	192 patients Median age 51 years (IQR 42–62)	Median duration of treatment 34 months (IQR 24–42) Median cumulative cabergoline doses at the first and second echocardiograms 97 mg (IQR 20–377) and 232 mg (IQR 91–551)	No associations between cumulative doses of cabergoline and prevalence of any valvular abnormality	No
Vroonen et al.34	Prospective study	Cabergoline group (n = 100) Median age – 41.5 years	Median duration of treatment 124.5 months (IQR 80.75–188.75) Median cumulative dose of cabergoline 277.8 mg (IQR 121.4–437.8)	No clinically relevant changes to cardiac valve function or valvular calcifications following treatment with cabergoline	No
Dogan et al.35	Case-control study	Cabergoline group (n = 32) Control group (n = 28)	Cabergoline cumulative dose 7.8 ± 5.5 mg at 3 months and 31 ± 22 mg at 12 months	No change in the prevalence of any valve regurgitations	No

CI, confidence interval; IQR, interquartile range; OR, odds ratio.

Discussion

This service evaluation highlighted significant variation in practice with regard to echocardiographic surveillance of patients with hyperprolactinaemia disorder treated with cabergoline in our area. Very few patients had an initial baseline echocardiogram or appropriate follow-up echocardiograms based on current guidelines. This is perhaps not surprising given the variation in guideline advice and the paucity of clinical data. Furthermore, within our clinical service there is no formal system to monitor echocardiogram surveillance in these patients. Currently, echocardiograms are booked by the responsible clinician as part of the patient’s outpatient clinic review on an ad hoc basis. However, our service appears to be similar to other areas, with none of the 14 health boards in NHS Scotland having protocols to guide decision making regarding echo surveillance. Having surveyed the responsible clinicians it is clear that the growing body of literature on this subject is influencing practice. It may be the case that the very low rate of baseline echocardiograms reflects the fact that clinicians are risk stratifying patients and, as patients are generally started at a lower dose titrated up over several months, it is difficult to identify high risk patients prior to treatment initiation.

Whilst most echocardiograms showed no valvular abnormalities some were reported in a small number of studies, which in the most part were non-progressive and not of any clinical significance. This study was not designed to comment on whether these changes were above the baseline incidence for this population.

The literature review revealed inconsistent results. Of the 25 articles identified, 12 showed no clinically significant change in valvular function following treatment with cabergoline. These were a mixture of case-controlled, cross-sectional, prospective multicentre control matched and single arm prospective studies. They predominantly looked at valve regurgitation, valve thickening and mitral valve tenting area (area between the mitral annulus and the leaflets body and is a marker of regurgitation, subvalvular remodelling and leaflet thickening).

The remaining 13 articles were a mixture of case-controlled, cross-sectional and case-record retrospective studies. They also predominantly looked at valve regurgitation, valve thickening and mitral valve tenting area.

Only one study²⁴ demonstrated a case of severe mitral regurgitation that had the associated valve thickening due to cabergoline. It is important when discussing the above papers that to meet the definition of cabergoline-associated valvulopathy you must determine there is ‘the triad of moderate or severe regurgitation, with a restricted and thickened valve’.⁷

One study,¹ showed no significantly increased risk of clinically relevant valvular regurgitation or leaflet thickening but did report a higher mitral valve tenting area in patients receiving cabergoline. Another study²⁵ showed an increase in mitral valve tenting area and an increase in mild tricuspid regurgitation (7.8% versus 0.0%) in the cabergoline group compared with controls.

Four studies reported an increase in mild to moderate tricuspid valve regurgitation and mild pulmonary valve regurgitation at high cumulative doses. Of these, one study¹⁵ showed a statistically significant increase in mild tricuspid and mild pulmonary regurgitation in the cabergoline group at a mean cumulative dose of 311 mg. One study¹⁷ reported a statistically significant increase in moderate tricuspid regurgitation in the cabergoline group at cumulative doses greater than 280 mg. One study³⁰ reported an increase in mild tricuspid regurgitation in the cabergoline group taking cumulative doses of cabergoline greater than 180 mg (p = 0.024). One study³¹ reported a single case of moderate mitral regurgitation in the cabergoline group at a cumulative dose of 130 mg.

Two studies, a cross-sectional study¹⁹ and its 2-year single-arm follow up,²⁶ pooled all patients taking a dopamine agonist and reported a statistically significant increase in aortic valve calcifications and mild tricuspid regurgitation.

One study²¹ showed no significant differences in valvular function in the cabergoline group compared with controls but found that female patients had a higher prevalence of mild tricuspid regurgitation than female controls and male patients had a higher prevalence of trace tricuspid regurgitation than male controls. These valvular regurgitations were not clinically significant and showed no correlation to dose or duration. One study²⁹ showed no statistically significant difference in clinically relevant valve regurgitation. However, subclinical valve fibrosis occurred more often in the cabergoline-treated group compared with controls.

One study⁹ showed 2.6% had a moderate regurgitation or stenosis of any valve. However, the investigators concluded the triad of cabergoline-associated valvulopathy was not present.

One study⁷ showed five cases of valvular thickening but none of them met the definition of cabergoline-associated valve disease.

Therefore, it can be seen that the risk of clinically significant valvular dysfunction reported in the literature is very low. This brings into question the need for echocardiogram surveillance in these patients. Whilst further studies are required to quantify with certainty whether there is a clinically significant risk of valvular heart disease, the current evidence does not support this and suggests that the existing guidance for echocardiogram surveillance may be leading to many unnecessary echocardiograms in this group of patients. This is similar to the findings of previously reported literature reviews.

Based on these findings, we challenge the current echocardiogram recommendations and suggest that patients could be stratified into high-dose (≥2 mg weekly) and low-dose (<2 mg weekly) cohorts to determine the need for echocardiogram surveillance.

If appropriate 2 yearly follow-up echocardiogram surveillance was only performed on patients taking high-dose cabergoline (≥2 mg weekly), we estimate that 40 echocardiograms could be saved over a 2-year period in our patient cohort. This would equate to £4000 worth of activity in our health board (population 250,000). While this may seem modest, if our results were extrapolated to Scotland (population 5.5 million), this would be equivalent to approximately £100,000 of activity per annum. In addition to this, unnecessary screening could result in additional expense and inconvenience to patients and may increase anxiety. A reduction in unnecessary echocardiograms in an already pressured physiology service would seem desirable and is a key objective in the Scottish Government’s current Healthcare Science National Delivery Plan.

Limitations

This was a single-centre study and therefore there is a risk that our results are not generalizable. However, our hospital provides care for all patients of our area and apart from one patient whose area was provided solely in a tertiary centre, it is unlikely that we have missed any patients. Cases were identified from the PIS which covers all community pharmacy prescribing and dispensing in Scotland. This system will not cover private prescriptions but it is unlikely that these occur in this patient group and thus we are confident that we have captured all relevant patients. Some patients were started on cabergoline before (2007) the publication of the pivotal studies linking cabergoline to valvular dysfunction. It may have been the case that clinicians at this time were unaware of the need for echocardiogram surveillance. It is not clear whether this was included on the SPC at the time.

Author recommendations

We propose that clearer national guidelines are promoted and that individual echocardiography departments are empowered to challenge the use of inappropriate surveillance. We also suggest that a national survey be carried out to look at the variations in clinical practice for echocardiogram surveillance for these patients.

In our institution, we plan to review the current guidelines and discuss this at a local level with the aim to set recommendations for echocardiogram screening only in at-risk patients. We recommend that echocardiogram surveillance should be reserved for those with known valvular heart disease or an audible murmur and those on doses of at least 2 mg weekly of cabergoline or a high cumulative dose.

We recommend that high dose patients should receive follow up echocardiograms at 2 yearly intervals, in line with previous guidance. As patients are generally started at a lower dose, which is titrated up over several months, it may be a challenge to identify high dose patients and to allocate baseline echos prior to treatment initiation. Patients with abnormal cardiovascular examination or known cardiac valve pathology should receive a baseline echocardiogram within 6 months prior to starting cabergoline and then follow up echocardiograms at 2 yearly intervals.

Conclusion

In our evaluation of a clinical service covering around 250,000 patients we found significant variation in practice, with very few patients receiving appropriate initial or follow-up echocardiograms. However, a review of the literature suggests that there is little evidence to support the link between low-dose cabergoline treatment for hyperprolactinaemia disorders and clinically significant valvular heart disease and we are aware of only one case where clinically important valvular heart disease developed. We suggest that the value of echocardiographic screening in patients receiving low-dose cabergoline (<2 mg/week) is limited and should be discontinued. It is also apparent that organizations and departments managing patients on high-dose cabergoline should establish robust recall systems to ensure appropriate monitoring during treatment.