OBJECTIVE: To evaluate serum human chorionic gonadotropin (hCG) ratio, progesterone and inhibin A as single parameters and in combination for the prediction of spontaneous resolution of pregnancies of unknown location (PUL). STUDY DESIGN: Prospective observational study of 105 consecutive patients with a diagnosis of PUL. Serum levels of hCG, progesterone and inhibin A were determined at the first visit and after 2 days. Patients were followed clinically until a final diagnosis of spontaneously resolving PUL, viable or non-viable intrauterine pregnancy, or ectopic pregnancy with need of laparoscopic intervention had been reached. Different combinations of hCG ratio (hCG at 48 h/hCG at 0 h), s-progesterone and s-inhibin A were investigated to find the best predictor for successful expectant management. RESULTS: The final pregnancy outcomes were: 52 spontaneously resolving PUL (49.5%), 37 viable intrauterine pregnancies (35.2%), 8 non-viable intrauterine pregnancies (7.6%), 7 ectopic pregnancies (6.7%), and one molar pregnancy (1.0%). An hCG ratio<0.80 predicted spontaneously resolving PUL with positive and negative predictive values (PPV and NPV), sensitivity, and specificity of 0.98, 0.78, 0.72, and 0.99, respectively. In patients with hCG ratio ≥ 0.80, a combination of s-progesterone < 20 nmol/l and s-inhibin A < 30 pg/ml predicted spontaneously resolving PUL with PPV, NPV, sensitivity and specificity of 0.92, 0.96, 0.85, and 0.98 respectively. CONCLUSION: Our results suggest that patients with PUL and hCG ratio < 0.80 display a high probability of spontaneously resolving PUL with minimum need of follow-up. In cases of hCG ratio ≥ 0.80, a combination of s-progesterone < 20 nmol/l and s-inhibin A < 30 pg/ml, may be a reliable predictor of spontaneously resolving PUL. The safety of this approach should be tested in large prospective studies.
OBJECTIVE: To evaluate serum human chorionic gonadotropin (hCG) ratio, progesterone and inhibin A as single parameters and in combination for the prediction of spontaneous resolution of pregnancies of unknown location (PUL). STUDY DESIGN: Prospective observational study of 105 consecutive patients with a diagnosis of PUL. Serum levels of hCG, progesterone and inhibin A were determined at the first visit and after 2 days. Patients were followed clinically until a final diagnosis of spontaneously resolving PUL, viable or non-viable intrauterine pregnancy, or ectopic pregnancy with need of laparoscopic intervention had been reached. Different combinations of hCG ratio (hCG at 48 h/hCG at 0 h), s-progesterone and s-inhibin A were investigated to find the best predictor for successful expectant management. RESULTS: The final pregnancy outcomes were: 52 spontaneously resolving PUL (49.5%), 37 viable intrauterine pregnancies (35.2%), 8 non-viable intrauterine pregnancies (7.6%), 7 ectopic pregnancies (6.7%), and one molar pregnancy (1.0%). An hCG ratio<0.80 predicted spontaneously resolving PUL with positive and negative predictive values (PPV and NPV), sensitivity, and specificity of 0.98, 0.78, 0.72, and 0.99, respectively. In patients with hCG ratio ≥ 0.80, a combination of s-progesterone < 20 nmol/l and s-inhibin A < 30 pg/ml predicted spontaneously resolving PUL with PPV, NPV, sensitivity and specificity of 0.92, 0.96, 0.85, and 0.98 respectively. CONCLUSION: Our results suggest that patients with PUL and hCG ratio < 0.80 display a high probability of spontaneously resolving PUL with minimum need of follow-up. In cases of hCG ratio ≥ 0.80, a combination of s-progesterone < 20 nmol/l and s-inhibin A < 30 pg/ml, may be a reliable predictor of spontaneously resolving PUL. The safety of this approach should be tested in large prospective studies.