Context: In patients with Cushing disease (CD) and a typical image of adenoma at MRI, transsphenoidal surgery is consensual. However, when MRI is inconclusive or normal, some authors now advocate medical treatment instead. The implicit assumption is that modern MRI should miss only very small microadenomas that are too difficult to visualize at surgery.
Objective: To analyze the evolution with time of the performances of MRI and the outcomes of surgery in patients with CD with a typical image of adenoma vs an inconclusive or normal MRI.
Design and patients: Retrospective single center study of 195 patients with CD treated by transsphenoidal surgery between 1992 and 2018, using first a translabial microscopic and then a transnasal endoscopic approach. Patients with inconclusive or normal MRI were explored by bilateral inferior petrosal sinus sampling. Four MRI groups were defined: microadenomas (n = 89), macroadenomas (n = 18), or MRI either inconclusive (n = 44) or normal (n = 44).
Results: The proportion of inconclusive/normal MRI decreased with time, from 60% (21/35) in 1992 to 1996 to 27% (14/51) in 2012 to 2018 (P = 0.037). In the four MRI groups, the per-operatory adenoma visualization rate was only slightly lower when MRI was normal (95%, 100%, 86%, 79%; P = 0.012) and postoperative remission rates were not different (85%, 94%, 73%, 75%; P = 0.11).
Conclusion: The diagnostic performances of MRI have improved but remain inferior to the eye of an expert neurosurgeon, best assisted by endoscopy. We propose that patients with CD and an inconclusive/normal MRI be addressed by an expert neurosurgeon for transsphenoidal surgery rather than being treated medically.
Copyright © 2019 Endocrine Society.