Mediastinoscopy in Patients With Presumptive Stage I SarcoidosisA Risk/Benefit, Cost/Benefit Analysis
There is disagreement about the need for tissue confirmation in patients with a presumptive diagnosis of stage 1 sarcoidosis (SIS) when bilateral hilar adenopathy (BHL) is accompanied by erythema nodosum, uveitis, polyarthralgia or trivial chest symptoms, or when symptoms are lacking. Winter-bauer et al, in the seminal quantitative investigation of this issue, reviewed records of 100 patients with BHL. Sarcoidosis was confirmed in 29 persons who were asymptomatic and had normal results of physical examination and in 13 symptomatic persons with findings limited to uveitis or erythema nodosum, prompting the authors to advocate forgoing biopsy in favor of continued observation of such patients. To our knowledge, no published exceptions to this dictum have withstood careful scrutiny. Affirming that timely recognition of an alternative diagnosis (AD), ie, tuberculosis (TB) or lymphoma, might prove beneficial, some authors advocate histologic confirmation, either by mediastinoscopy or trans-bronchial lung biopsy, when more accessible tissue is unavailable. Advocates of continued observation maintain that biopsy is unnecessary because exceptions, if they exist, are so rare that a secure a priori diagnosis can be made on clinical grounds. The issue can be framed in quantitative terms: in a cohort of patients with BHL, presumed SIS (BHLps), what are the tangible benefits or disadvantages of mediastinoscopy; secondarily, is it cost-effective?
Materials and Methods
Incidence Estimates
We reviewed journal articles cited in texts and articles identified in a MEDLINE search to estimate the incidence with which each clinically relevant AD—ie, TB, Hodgkin’s disease (HD), and non-Hodgkin’s lymphoma (NHL)—presents as an isolated asymptomatic bilateral hilar lymphadenopathy (ABHL) as the product: proportion presenting with ABHLXannual incidence. This proportion was derived from a meta-analysis of the published literature on each AD and was computed as the product: percent asymptomatic X percent with isolated mediastinal disease X percent with isolated mediastinal involvement limited to BHL. Annual rate=incidenceXUS population (2.5X108).