Exploring the morbidity and mortality in acromegaly

Introduction: Acromegaly is a rare clinical condition characterized by chronic growth hormone (GH) excess, resulting into body disfigurement and systemic complications. Despite the therapeutic advancements in the management of patients with acromegaly, there still remains significant morbidity and mortality. Of the different complications associated with acromegaly, arthropathy has the highest prevalence and is one of the main determinants of quality of life in these patients, while cardiovascular disease remains one of the leading causes of mortality in acromegaly patients. Aims: The aim of this thesis is to explore further the unresolved issues of impaired quality of life, arthropathy and cardiovascular mortality associated with acromegaly and in particular to: i) establish whether quality of life (QoL) changes over time in patients previously treated for acromegaly and who have biochemically stable or improved disease and identify positive and negative predictors of quality of life in patients with acromegaly; ii) characterise the joint alterations in patients with acromegaly, by evaluating subchonadral knee bone shape (a surrogate marker of osteoarthritis) using a novel imaging technique based on automated segmentation of MR images of knee bones and calculation of bone area using active appearance models and identify factors, which may be associated with altered bone shape in patients with acromegaly; and iii) explore whether acromegaly is associated with increased thrombotic potential, by evaluating the effects of GH/IGF-1 excess on clot formation and lysis, as a potential mechanism for the increased cardiovascular mortality observed in patients with acromegaly. Results: For the QoL evaluation a two-arm study was conducted, consisting of a cross-sectional arm to compare QoL between patients with treated and controlled acromegaly and healthy controls; and a longitudinal arm to assess QoL changes in patients with biochemically stable disease during 5.7±0.6 years of follow-up. Fifty eight (58) patients and 116 matched controls were recruited for the cross-sectional arm; 28 patients completed the longitudinal arm. Three generic questionnaires [Psychological General Well-Being Schedule (PGWBS), 36-item Short-Form (SF-36), EuroQoL (EQ-5D)] and the disease-specific acromegaly QoL questionnaire (AcroQoL) were applied for QoL evaluation. QoL assessment was performed 11.6±8.2 years following diagnosis and treatment of acromegaly. Patients with treated acromegaly had lower QoL scores compared with controls in all questionnaires with the exception of the PGWBS “Anxiety” subscale. The “Appearance” subscale of the AcroQoL and the “physical function” subscales of the remaining questionnaires were the most underscored domains. No difference in the total and subscale scores of all questionnaires was observed between baseline and follow-up assessment, with the exception of the SF-36 “Physical Function”, where a decline was found between baseline and follow-up (58.5±24.7% vs. 43.1±31.1%; p=0.002). Duration of IGF-1/GH control was positively correlated with QoL scores in most questionnaires at baseline, whereas use of GH lowering therapy at the time of QoL assessment was a negative predictive factor of QoL. For the characterisation of the the arthropathy in acromegaly, a two-arm study was conducted. In the cross-sectional arm, 60 patients with acromegaly (29 males, mean age 54.8±12.9yrs) were compared with 300 age/gender-matched controls from the publicly available Osteoarthritis Initiative (OAI) database via propensity score matching. An additional control group of 886 individuals without any evidence of knee osteoarthritis on serial MRI scans, also referred as non-OA group, was selected from the OAI. In the longitudinal arm, bone shape data from 42 patients with acromegaly were compared at baseline and after at least 12 months of follow-up. Acromegaly patients attended for bilateral knee MRI scan at baseline and after at least 12 months of follow-up. Knee bone shape, joint space width (JSW) and cartilage thickness were measured based on automated segmentation of MR images of knee bones and calculation of bone area using active appearance models. Acromegaly patients had increased medial JSW compared with controls [6.21mm (95% CI 6.03-6.40) vs. 5.78mm (95% CI 5.70-5.87) respectively, p<0.001] and increased lateral and medial femorotibial cartilage thickness. Patella and medial tibia bone areas were also increased in acromegaly patients. B-score (a biomarker associated with severity and risk of progression of OA) was significantly higher in patients compared with controls [1.7 (95% CI 1.32-2.08) vs. 1.01 (0.84-1.18) respectively, p=0.001]. Thirty-five percent (n=21) of acromegaly patients had B-score ≥2, which is indicative of OA. These patient had higher GH levels at the time of diagnosis of acromegaly and required a higher number of therapeutic interventions during the disease management compared with patients with B-score <2 (n=39). Additionally, patients with B-score ≥2 had significantly larger femoral, tibial and patella bone areas, increased medial JSW and lateral and medial femorotibial cartilage thickness compared with the remaining patients. However, when patients with B-score <2 (within the B-score reference range) were compared with the non-OA control group, acromegaly patients still demonstrated differences in bone shape, with increased bone area at the lateral patella, medial patella and medial tibia; increased B-score [patients:0.69 (95% CI 0.36-1.02), non-OA controls: 0.018 (95% CI -0.05 – 0.08), p<0.001]; and increased lateral and medial femorotibial cartilage thickness. Comparison between baseline and follow-up scans for the 42 patients who completed the longitudinal arm of the study did not show any change in the bone size area, B-score or JSW. For the assessment of potential cardiovascular risk factors and clot dynamics in patients with acromegaly, a cross-sectional study with 40 patients with acromegaly (21 males, age 53±13yrs) and 40 age/gender-matched controls was conducted. Clot structure was analysed using a validated turbidimetric assay and fibrin networks were visualised by laser scanning confocal microscopy (LSCM). Metabolic profile parameters, body composition, plasma fibrinogen and PAI-1 were also assessed. Twenty-two patients had active acromegaly and 18 were in remission. There was no difference in qualitative patient characteristics between the two groups. Both groups had less favourable body composition and cardiovascular risk profile compared with controls. Despite no difference in clot formation and lysis parameters between the two patient groups, active disease patients had higher fibrinogen and clot maximum absorbance compared with controls, after adjusting for BMI (3.8±0.2 vs. 2.6±0.2mg/ml, p<0.001; and 0.39±0.02 vs. 0.33±0.01 arbitrary units, p=0.03, respectively). Patients in remission had higher fibrinogen compared with controls following adjustment for BMI (3.3±0.2 vs. 2.6±0.2mg/ml, p=0.02) but not clot maximum absorbance (0.35±0.03 vs. 0.33±0.02 arbitrary units, p=0.6). LSCM showed increased fibrin network density only in active disease patients, consistent with turbidimetric analysis. In addition to active disease, BMI, fat mass and skinfold thickness were associated with higher clot density and longer lysis time. Conclusions: Patients with biochemically controlled acromegaly demonstrate impaired quality of life, which persists despite long-term disease control, is primarily consisted of impaired physical function and secondly of impaired psycho-social well-being. Duration of biochemical disease control and current use of GH lowering therapy were the predominant factors determining patients’ QoL. Acromegaly patients despite higher B-score (a biomarker of osteoarthritis) and larger bone area particularly of the patella and medial tibia bones have preserved and/or increased joint space due to increased cartilage thickness, which distinguishes acromegalic arthropathy from osteoarthritis. The higher pre-treatment GH values and the higher number of therapeutic interventions seen in patients with B-score ≥2, indicate that overall exposure of peripheral tissues to excessive GH levels is a risk factor for more pronounced changes to the knee bone shape and potentially more severe arthropathy. Clot structure was analysed using a validated turbidimetric assay and fibrin networks were visualised by laser scanning confocal microscopy (LSCM). Metabolic profile parameters, body composition, plasma fibrinogen and PAI-1 were also assessed.