A prospective clinical crossover study comparing the functional perfor
mance and biocompatibility of three brands of polysulfone membranes (F
resenius Polysulfone(R) (Fresenius Ag, Bad Homburg, Germany), Polyphen
(R) (Minntech Corp., Minneapolis, MN), and Biosulfane(R) (WR Grace Inc
., Danvers, MA)) incorporated in ethylene oxide-sterilized dialyzers o
f comparable surface area (1.3 to 1.35 m(2)) was undertaken. The clear
ance of small molecules by each membrane was comparable, Plasma levels
of beta(2) microglobulin fell to 49.9% of pretreatment values by 210
min when using the Fresenius Polysulfone(R) membrane, 60.2% with the P
olyphen(R) membrane, and 63.1% with the Biosulfane(R) membrane. The re
duction achieved by the Fresenius Polysulfone(R) membrane was superior
(P = 0.003). The plasma reductions were associated with the recovery
of 195 mg beta(2) microglobulin from the dialysate for the Fresenius P
olysulfone(R) membrane and 158 mg for the Polyphen(R) membrane, but no
beta(2) microglobulin was recovered from the dialysate with the Biosu
lfane(R) membrane. The dialysate collected with the Fresenius Polysulf
one(R) membrane also contained a mean of 6853 mg of fetal protein, com
pared with 5490 mg with the Polyphen(R) membrane and 8422 mg with the
Biosulfane(R) (P = 0.04) membrane. The neutropenia was slight and inde
pendent of membrane brand, as were the changes in C3a des arg and SC5b
-9 complement components. The reduction in platelet counts was higher
for the Biosulfane(R) membrane than for the other brands (P = 0.003).
This study indicates that whereas the polymer base of the membrane is
the same, its production and subsequent handling during dialyzer produ
ction induce changes that attain statistical significance, most notabl
y in the way that the membrane removes beta(2) microglobulin and inter
acts with proteins. The differences observed are a consequence of the
different alloying polymers used during manufacture and, consequently,
the membranes cannot be considered equivalent.