Impact of transsphenoidal surgery on asymptomatic cardiomyopathy in patients with acromegaly. A single-blinded study
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.217956
Source of Support: None, Conflict of Interest: None
Background: Patients with acromegaly have 2–3 times the expected mortality rates primarily due to cardiovascular risks. Echocardiographic studies showing improvement of cardiac function following transsphenoidal surgery (TSS) are limited.
Keywords: Acromegaly, cardiomyopathy, echocardiogram, transsphenoidal surgery
There is considerable literature that suggests a specific cardiomyopathy in acromegaly, resulting in structural and functional abnormalities that may be partially reversed by effective reduction in growth hormone (GH) or insulin-like growth factor type 1 (IGF-1) levels. The ventricular walls are concentrically thickened due to a relative increase in myocyte size without enlargement of cardiac chambers. Moreover, biopsy and autopsy studies indicate that interstitial fibrosis is the main histological feature that gradually impairs architecture and function. Most echocardiographic studies show a normal left ventricular systolic function but an impaired diastolic function as an early finding in acromegalic cardiomyopathy., The prevalence of left ventricular diastolic dysfunction is approximately 30% in untreated patients. In our previously published reports, the prevalence of systolic dysfunction was 5% at the time of presentation., The consequences of these findings are not so much at rest as during exercise, during which the ejection fraction is reduced, leading to reduced exercise tolerance.,
Patients with acromegaly have 2–3 times the expected mortality rates primarily due to the cardiovascular and cerebrovascular risks.,,, However; the cardiovascular risk factors play a major impact on the survival of patients of acromegaly. Before the introduction of effective therapy for patients with acromegaly, approximately 80% of the patients died before reaching the age of 60 due to the presence of cardiovascular disease.
The aim of the present study was to evaluate the effect of transsphenoidal surgery (TSS) of somatotropinomas on the echocardiographic parameters; as well as, to evaluate the possible impact of postoperative normalization of GH/IGF-I on the echocardiographic parameters of these patients.
The study was conducted from July 2013 to December 2014 and was approved by the Institutional Ethics Committee. Acromegaly was diagnosed on the basis of standard screening test in the form of IGF-1, estimated by electrochemiluminiscence, (Diasorin, Liason, USA). Minimal detectable IGF-1 concentration by this assay was 25 ng/ml. Age specific normal ranges for IGF-1 values were taken as provided by manufacturer of the kit. GH-OGTT was done for the confirmation of diagnosis. GH was estimated by electrochemiluminescence immunometric assay (Cobas 800, Roche-Hitachi, Germany), with a minimal detection limit of ≤0.05 ng/ml. The inter- and intra-assay coefficients of variation for GH and IGF-1 were < 5%. Pituitary adenoma was diagnosed by radiological imaging including computed tomography (CT) and magnetic resonance imaging (MRI) of the hypothalamopituitary area. In the prospective analysis, all patients with acromegaly underwent preoperative echocardiography for the variables, as mentioned in the appendix. Echocardiography and tissue Doppler (Philips i33, 3D ECHO) for cardiac indices were performed by observers blinded to the patients' information and they were supervised by the author, AB. In the prospective group, echocardiography was repeated after 6 months of surgery. A retrospective analysis of acromegalic patients who had undergone surgery for pituitary tumor was done for assessing their left ventricular diastolic dysfunction and ejection fraction. The echocardiographic parameters which were observed were systolic ejection fraction (EF), and the early diastolic filling time/atrial systole time (E/A), which reflects LV diastolic function (E/A) and LV mass index. The parameters were defined as normal or abnormal according to the American Society of Echocardiography guidelines. We also analyzed the isovolumetric relaxation time (IVRT). All patients underwent TSS for resection of the pituitary tumor.
Biochemical cure was confirmed at least after 6 months by the glucose-suppressed plasma GH concentrations (GH-OGTT) of less than 0.4 ng/ml, the random GH levels of less than 1 ng/ml, and the normal age-corrected IGF-1 values, as recommended. Clinical evaluation, GH-OGTT with glucose, IGF-1, and echocardiographic studies were obtained in all patients at baseline and at least 6 months after the transsphenoidal pituitary surgery. The following patients were excluded: patients who were on somatostatin analog or cabergoline for optimization of treatment, patients who did not give consent or withdrew consent from the study, or patients who died before completion of the study, were lost to follow-up, or had less than 6 months of follow-up; patients suffering from secondary cardiomyopathy, such as ischemic or valvular disease, or cardiac failure were also excluded.
The statistical analysis was carried out using the Statistical Package for the Social Sciences software package version 21 (SPSS Inc., Chicago, IL, USA). Normality of the data was checked using one-sample Kolmogorov–Smirnov test. Continuous data were presented as mean, median, and standard deviation, and categorical data were presented as percentages. Power of 80% and alpha error (P) of <0.05 was considered to be significant. Differences between parameters in the different patient groups were evaluated using Fisher exact test and Student's t-test where appropriate. Differences between proportions were evaluated using the chi-square test. We conducted univariate and multivariate analyses to assess whether associated comorbidities contributed to the increase in left ventricular mass (LVM). The univariate analysis for categorical factors was performed with the use of Fisher's exact test. For continuous factors, we used a single-variable logistic- regression model. A multivariate logistic-regression analysis was performed to assess factors considered to be significant (P ≤ 0.10) by univariate analysis or considered clinically important. The assessed risk factors were pre-specified in the protocol, and the statistical methods were pre-planned. Analysis of variance was applied wherever appropriate.
A total of 43 patients diagnosed with acromegaly gave consent for conduction of the study. Five patients were excluded from the study, 3 patients were lost to follow-up before 6 months, and 2 patients who had secondary cardiomyopathy due to ischemic heart disease were excluded. Thus, 38 patients were available for the final analysis [Figure 1]. All patients had a pituitary macroadenoma. No patient received preoperative/postoperative adjuvant treatment for acromegaly till the completion of the study. All patients underwent gross total excision of the adenoma. The strategy for the management was TSS only as the first-line therapy; if they had elevated GH and IGF-1 after 6 months, they received gamma knife/intensity modulated radiation therapy (IMRT) and octreotide long acting release (LAR) and/or cabergolin in the interim period. The preoperative clinical and hormonal characteristics of the studied patients are depicted in [Table 1]. At a follow-up of 6 months, biochemical cure was achieved in 13 patients (34%) [they achieved both target GH and IGF-1]. The remaining patients in whom biochemical cure was not achieved were divided into two groups, i.e., patients with GH-OGTT between 1 ng/ml and 5 ng/ml (20 patients) and GH-OGTT >5 ng/ml (5 patients). Hence, the subgroup analysis was done on the basis of GH-OGTT at 6 months only, not considering IGF1.
Pre- and postoperative echocardiographic parameters are shown in [Table 2]. Prior to surgery, 9 out of 38 (28%) patients met the criteria for left ventricular hypertrophy (LVH). A significant decrease in LVM and LVM index (LVMI) were observed (P < 0.001) in both the groups. There was significant decrease in LVM and LVMI in patients who were cured as well as in those with GH-OGTT 1–5 ng/ml [Figure 2]. In the prospective group, LVMI completely normalized in 2 patients, new onset deterioration occurred in 1 patient (who was not cured), and an improvement that was not below the defined cut-off occurred in 8 other patients. Left ventricular systolic function was abnormal at baseline in 18 (47.3%) patients, which normalized in 11 (61.1%) patients postoperatively; in 7 patients, it improved significantly, although it did not normalize completely. There was also a significant improvement in the left ventricular systolic function as reflected by improvement in the left ventricular ejection fraction (P = 0.01). Post TSS, in patients with GH-OGTT >5 ng/ml, there was no significant decrease in the LVM and LVMI and the EF also did not improve. In 1 patient, there was deterioration in the ejection fraction, which was probably due to uncontrolled disease.
In the retrospective group, 62 patients, with a mean follow up of 20.3 months, were analyzed for change in ejection fraction. There was a significant improvement in the left ventricular ejection fraction in patients who were cured (P = <0.001). In toto, 18 (29.0%) patients had an abnormal ejection fraction, and following TSS, it normalized in 12 (66.0%) patients, improved significantly but could not normalize in 6 patients, and 1 patient had a new onset deterioration of ejection fraction. Postoperative nadir GH in this patient was 5 ng/ml. It was directly proportional to the level of glucose suppressed GH value [Table 3].
Using multinomial logistic regression analysis with the state of disease as the dependent variable (6-month GH-OGTT levels: <1 ng/ml; 1–5 ng/ml, and >5 ng/ml) it was found that age, sex, presence of hypertension, diabetes mellitus, hypocortisolic status, and the hypothyroid status did not affect the echocardiographic parameters.
Fifty percent of the patients in the prospective group showed left ventricular diastolic dysfunction (LVDD) as indicated by inversion of the E/A ratio (ratio of the early [E] to late [A] ventricular filling velocities) and isovolemic relaxation time [IVRT]. However, all of them had decreased LVMI, which is an indirect evidence of left ventricular diastolic dysfunction [LVDD].
In the present study, we confirmed that successful TSS significantly improves and may normalize echocardiographic parameters (LVM, LVMI, EF, and LVDD) in a significant proportion of patients who are cured. It is important to note in our study that all patients were apparently asymptomatic and a large proportion of patients had normal baseline echocardiographic parameters with no previous treatment received for the same. Although several studies have been conducted for the evaluation of cardiovascular parameters in acromegaly, the methods of conducting them are different, along with varying end points.,,, There have been differences in the population selected, i.e., the study may have been a single hospital-based one or a multi-hospital based one; and, the type of treatment instituted may have varied amongst them. There is hardly any data on this aspect from our country.
We analyzed IVRT and early diastolic filling time/atrial systole time (the E/A ratio), which reflect left ventricular diastolic function; however, normal values of IVRT and E/A can occur both in case of normal left ventricular diastolic function as well as in severe diastolic dysfunction (pseudonormalization). Hence, we considered a decrease in the LVMI as an indirect indicator of left ventricular diastolic function.
Nine out of 38 patients had LVH (23.8%). However, the LVH was normalized in only 1 out of 9 patients. There was a significant decrease in the LVM, LVMI, and LVEF in the prospective group (P < 0.001), which suggests improvement in the left ventricular systolic function. In a case–control observational study by Colao et al., LVH and diastolic and systolic dysfunctions were more common in acromegalic patients as compared to controls, and these findings in the patients correlated with disease duration and the patient's age. We did not find a similar correlation. However, 50% of patients with increased LVMI had associated hypertension. Long-term echocardiographic follow-up of acromegalic patients in a prospective study by Hradec et al., found results similar to our study. This study concluded that both hypertrophy and dilation of the left ventricle in acromegaly are reversible after successful surgery, but in a previous retrospective study by Terzolo et al., there was no difference in the left ventricular mass in patients suffering from an active disease and the cured acromegalics.
Jaffrain-Rea et al., in their study, confirmed significant improvement in the LVM and LVMI, IVRT, and E/A in the 31 patients of acromegaly following successful surgery. However, in their study, all 31 patients who had biochemical cure after surgery were selected. Thus, there could be a possibility of a selection bias due to the inclusion of only cured patients. Further, the cardiac variables were studied in only cured patients. However, in our study, all patients who underwent surgery for acromegaly satisfying the inclusion and exclusion criteria were included, irrespective of whether or not cure was achieved in them. Moreover, analysis of the IVRT and E/A parameters can give a false impression of an improving left ventricular function, especially in patients with severe diastolic dysfunction, which is common in acromegaly. However, there was no significant reduction of LVM, LVMI, and LVEF in patients with GH-OGTT >5 ng/ml in both the retrospective and prospective groups. Thus, our study further substantiates the fact that improvement in cardiac function is significantly reflected by the reduction in the circulating GH levels to the safe limit.
Minniti et al., in a similar study among 31 acromegalic patients, compared the impact of surgery on the cardiovascular factors in patients with postoperative 6-month GH levels <10 mU/L and >10 mU/L (4 ng/ml), and concluded that there was no improvement in the cardiac parameters in uncontrolled disease (GH >10 mU/L). However, subgroup comparison was done between the 'cured' and the 'not cured' cohorts. Thus, our study confirms the difference in echocardiographic parameters between the well- and poorly-controlled patients. Furthermore, it is noted that postoperative cardiac improvement can be achieved within a few months following successful surgery over a wide range of GH-OGTT values at follow up.
In a study by Colao et al., comparing the impact of first-line surgery and medical treatment with somatostatin analogs on the cardiac parameters in acromegaly, it was found that diastolic function improved significantly in both the groups; however, systolic function and ejection fraction improved significantly in the medically treated group and not in the surgery group. This result was attributed to the effect of somatostatin analogs on the cardiomyocytes or the effect of these drugs on the remaining pituitary function. However, this study was retrospective and nonrandomized. In our study, it was convincingly evident that there was a significant improvement in the left ventricular diastolic function as well as systolic function (ejection fraction) in 6 months of follow up. The improvement in ejection fraction (P < 0.001) was also confirmed in the retrospective analysis of 64 patients of acromegaly treated by surgery after a mean follow-up of 20 months. The improvement in the cardiac parameters following surgery in our study can be attributed to the decrease in the circulating levels of GH as well as the reversible changes of the cardiac myocytes. These findings need to be further validated with studies on a larger number of patients and a longer follow up.,
Analysis of IVRT and E/A can incorrectly predict left ventricular diastolic function in patients with severe diastolic dysfunction. Ventricular hypertrophy helps in preserving systolic function, but may be deleterious for diastolic function; alterations in diastolic relaxation have been reported to correlate well with the degree of ventricular hypertrophy. It should be noted, however, that significant improvement in systolic function at peak exercise can also be observed after normalization of GH/IGF-I.,,
The strength of our study is that the patients studied were apparently asymptomatic for any cardiac illness, thus, negating the influence of the confounding factors (cardiac dysfunction) on the results and a single-blinded observer (AB). This was not mentioned in any of the previous studies done so far. The other strength is the use of M-mode echocardiogram to measure cardiac dimensions and cardiac motion, which is supposed to be more accurate than two-dimensional echocardiography and tissue Doppler. The limitations of our study were the smaller number of patients recruited for the study; this being a single centre study; this study not evaluating all of the echocardiographic parameters; there being lack of a healthy control group; the relatively shorter duration of the study period in the prospective group; and, a lack of detailed echocardiographic parameters in the retrospective group.
We conclude that reduction in growth hormone levels and IGF-1 can decrease the LVM and LVMI, which directly or indirectly contribute to the improvement in diastolic function. Reduction of LVM by improving the left ventricular function might decrease the morbidity and mortality associated with acromegalic heart disease. A long-term follow-up study is necessary to assess for possible ongoing improvement in cardiovascular parameters in acromegalic patients with well-controlled blood hormonal levels, with special reference to their diastolic function.
We are thankful to Dr. Anand Srinivasan, DM Pharmacology for the statistical analysis.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]