Relationship of glutathione S-transferase genotypes with side-effects of pulsed cyclophsphamide therapy in patients with systemic lupus erythematosus

Aims
Cyclophosphamide (CTX) is an established treatment of severe systemic lupus erythematosus (SLE). Cytotoxic CTX metabolites are mainly detoxified by multiple glutathione S-transferases (GSTs). However, data are lacking on the relationship between the short-term side-effects of CTX therapy and GST genotypes. In the present study, the effects of common GSTM1, GSTT1, and GSTP1 genetic mutations on the severity of myelosuppression, gastrointestinal (GI) toxicity, and infection incidences induced by pulsed CTX therapy were evaluated in patients SLE.
Methods
DNA was extracted from peripheral leucocytes in patients with confirmed SLE diagnosis (n = 102). GSTM1 and GSTT1 null mutations were analyzed by a polymerase chain reaction (PCR)-multiplex procedure, whereas the GSTP1 codon 105 polymorphism (Ile→Val) was analyzed by a PCR-restriction fragment length polymorphism (RFLP) assay.
Results
Our study demonstrated that SLE patients carrying the genotypes with GSTP1 codon 105 mutation [GSTP1*-105I/V (heterozygote) and GSTP1*-105 V/V (homozygote)] had an increased risk of myelotoxicity when treated with pulsed high-dose CTX therapy (Odds ratio (OR) 5.00, 95% confidence interval (CI) 1.96, 12.76); especially in patients younger than 30 years (OR 7.50, 95% CI 2.14, 26.24), or in patients treated with a total CTX dose greater than 1.0 g (OR 12.88, 95% CI 3.16, 52.57). Similarly, patients with these genotypes (GSTP1*I/V and GSTP1*V/V) also had an increased risk of GI toxicity when treated with an initial pulsed high-dose CTX regimen (OR 3.33, 95% CI 1.03, 10.79). However, GSTM1 and GSTT1 null mutations did not significantly alter the risks of these short-term side-effects of pulsed high-dose CTX therapy in SLE patients.
Conclusions
The GSTP1 codon 105 polymorphism, but not GSTM1 or GSTT1 null mutations, significantly increased the risks of short-term side-effects of pulsed high-dose CTX therapy in SLE patients. Because of the lack of selective substrates for a GST enzyme phenotyping study, timely detection of this mutation on codon 105 may assist in optimizing pulsed high-dose CTX therapy in SLE patients.

Learning vector representation of medical objects via EMR-driven nonnegative restricted Boltzmann machines (eNRBM)

Electronic medical record (EMR) offers promises for novel analytics. However, manual feature engineering from EMR is labor intensive because EMR is complex - it contains temporal, mixed-type and multimodal data packed in irregular episodes. We present a computational framework to harness EMR with minimal human supervision via restricted Boltzmann machine (RBM). The framework derives a new representation of medical objects by embedding them in a low-dimensional vector space. This new representation facilitates algebraic and statistical manipulations such as projection onto 2D plane (thereby offering intuitive visualization), object grouping (hence enabling automated phenotyping), and risk stratification. To enhance model interpretability, we introduced two constraints into model parameters: (a) nonnegative coefficients, and (b) structural smoothness. These result in a novel model called eNRBM (EMR-driven nonnegative RBM). We demonstrate the capability of the eNRBM on a cohort of 7578 mental health patients under suicide risk assessment. The derived representation not only shows clinically meaningful feature grouping but also facilitates short-term risk stratification. The F-scores, 0.21 for moderate-risk and 0.36 for high-risk, are significantly higher than those obtained by clinicians and competitive with the results obtained by support vector machines.