Growth and Resorption of Chronic Subdural Hematomas: Gardner, Weir, and the Osmotic Hypothesis Revisited.

Background: To explain why some chronic subdural hematomas (CSDHs) grow and/or resorb, a physically decreasing outer membrane (OM) surface area (SA) to CSDH volume (V) ratio has been reexplored, and a critical CSDH size inferred (OM SA ≈ V). Gardner showed that since CSDH protein exceeded cerebrospinal fluid (CSF) protein, CSF→CSDH osmosis occurred across a semipermeable inner membrane (n = 1). By contrast, Zollinger and Gross demonstrated that serum→CSDH osmosis could also occur across the OM (n = 1). Notably, Weir refuted Zollinger and Gross by finding equal CSDH and serum total protein (n = 20); however, Weir did not refute Gardner. Although all extant mechanisms, especially rehemorrhages, explain CSDH growth, only OM SA ≥ V simultaneously permits resorption. We aimed to reevaluate the osmotic hypothesis.

Methods: Paired serum and CSDH samples were measured in a prospective cohort.

Results: Results were consecutively obtained in 116 patients (87 men; mean age, 73 ± 13 years). Serum osmolality and CSDH osmolality were similar (285.70 ± 7.99 vs. 283.85 ± 7.52 mmol/kg, respectively; P = 0.11) and significantly correlated (r = 0.75, P < 0.0001). Serum total protein significantly exceeded CSDH total protein (66.6 ± 6.8 vs. 43.68 ± 20.24 g/L, P < 0.0001) as did serum albumin (35.62 ± 4.46 vs. 30.85 ± 8.5 g/L, P < 0.0001) and serum total globulins (31.5 ± 6 vs. 18.6 ± 11.4 g/L, P < 0.0001). Serum and CSDH proteins were not correlated (total protein: r = 0.003; albumin: r = 0.08; globulins: r = 0.21).

Conclusions: Only crystalloids equilibrated. CSDH colloids were significantly decreased. CSDH dilution or colloidal flocculation is implied. CSDH dilution (by CSF→CSDH inner membrane [IM] osmosis or OM transudation/exudation) could favor CSDH growth, as would repeated OM hemorrhages. Contrariwise, isolated colloidal flocculation could favor CSDH shrinkage by OM CSDH→serum osmosis. The latter may result in OM SA ≥ V favorable for ultimate resolution. Our results refute Weir and Zollinger and Gross, but not Gardner. Osmotic gradients simultaneously exist for both CSDH growth and resorption. Each equilibrium could depend on each gradient relative to each IM/OM semipermeability.