OBJECTIVES: Troglitazone is a thiazolidinedione and peroxisome proliferator
-activated receptor gamma (PPAR gamma) ligand used to treat diabetes mellit
us type II. Because hyper-insulinemia may be a factor in nonalcoholic steat
ohepatitis (NASH), we postulated that troglitazone could have beneficial ef
fects in this disorder. Our study was initiated before reports of idiosyncr
atic hepatitis induced by this agent and was completed before its recent wi
thdrawal from the market.
METHODS: We studied 10 female patients (age 44 +/- 16) with histological NA
SH. All but two were obese (mean body mass index, BMI = 38 +/- 6). One had
type 2 diabetes, and three had well-compensated cirrhosis with NASH. Trogli
tazone was given at a dose of 400 mg/day for less than or equal to6 months.
Responders (defined as normal ALT at the end of treatment) were rebiopsied
. Paired specimens were compared in blinded fashion. Mitochondria were quan
titated using ultrathin electron microscopy.
RESULTS: Seven of ten patients responded with normal ALT at the end of trea
tment. One of three nonresponders initially normalized ALT but returned to
pretreatment level at 3 months. In this patient, therapy was stopped, and t
he ALT has remained at the baseline level with no other clinical or laborat
ory findings. In the responders, ALT fell from 87 +/- 38 before to 39 +/- 9
at the end of treatment (p = 0.01), and AST decreased from 77 +/- 23 to 30
+/- 8 (p = 0.002). Biopsy comparisons before and after therapy showed pers
istent steatohepatitis in all cases, although four of seven showed a one-po
int improvement in the necroinflammatory grade. Electron microscopy reveale
d elongation of the mitochondria after therapy.
CONCLUSIONS: Normal ALT was seen in 70% of NASI-I patients at the end of tr
eatment, but this biochemical response was associated with only mild histol
ogical improvement, and all follow-up biopsies had evidence of NASH. Normal
ization of the liver enzymes in patients with NASH who are treated with thi
azolidinediones should be viewed with reservation. Follow-up biopsy is esse
ntial to evaluate the efficacy of these agents, which, at the histological
level, appears to be relatively modest. (C) 2001 by Am. Cell. of Gastroente
rology.