Developing a Simple Recognition System of Acromegaly

Background: Acromegaly is a slowly progressive disease caused by growth hormone (GH) excess. The estimated prevalence of the disease is 40 cases/1000000 population with 3-4 new cases/1000000 population per year. Most patients have classic manifestations due to acral and soft tissue change, including thickening of the calvarium, increased size of frontal sinus, enlargement of the nose, prognathism, and widely spaced teeth. The hands and feet are large, thickened and bulky. Acromegaly was also associated with many comorbidities including cardiovascular complications (acromegalic cardiomyopathy and arterial hypertension), respiratory complications (sleep apnea), metabolic complication (impaired glucose tolerance, diabetes mellitus), joint and bone complications (carpal-tunnel syndrome, osteoarthritis).3 The outcome of growth hormone level suppression after transsphenoidal surgery is far better for microadenomas (80-90%) than for macroadenomas (less than 50%). Furthermore, higher growth hormone level at diagnosis was related to poor prognosis. An early detection of acromegaly with early intervention was desirable. However, due to the slowly progression of the disease, times from symptom onset to diagnosis were 5.2 to 5.9 year4.

Current method to diagnose patients with acromegaly was to check serum GH and insulin-like growth factor 1 (IGF-1). However, tests for growth hormone and IGF-1 are currently unavailable in most hospitals in Taiwan and were expensive. We want to find a better screening tool for acromegaly with less expensive and high efficacy. Due to specific facial appearance of acromegaly, there were studies to recognize acromegaly by computer program1,2,3. They tried to develop computer programs to separate facial photographs of patient with acromegaly and normal subjects. The accuracy could reach 71.9% to 86%. Due to enlarged hand with increased soft tissue in patients with acromegaly, we also want to compare the difference of face, hand and feet morphology and face, hand and feet X ray of acromegaly patients with those of normal population. We hope to set a computer model to early detect acromegaly patients in Chinese people.