Impact of drug-eluting stent selection on long-term clinical outcomes in patients treated for unprotected left main coronary artery disease: the sirolimus vs paclitaxel drug-eluting stent for left main registry (SP-DELFT)

AIM: To compare the long-term relative efficacy and safety of SES and PES in patients undergoing percutaneous coronary intervention (PCI) for unprotected left main coronary artery (ULMCA) disease and to evaluate the role of lesion location and stenting technique in determining outcomes. METHODS AND RESULTS: From April 2002 to April 2004, 288 consecutive patients who underwent elective PCI with DES implantation for de novo lesions on ULMCA have been retrospectively selected and analyzed in seven European and US tertiary care centers. All patients had a minimum follow-up of 3 years. SES was used in 152 patients while 136 received PES. Isolated ostial-shaft disease was present in 27% of patients. Distal LM disease (73%) was treated with single and double stent approach in 29.5% and 43.4% of patients respectively. After 3 years, rates of survival free from any of the events investigated, were independent from lesion location and stenting approach and did not differ significantly between SES and PES groups. Freedom from MACE (SES vs. PES) was 76.3% vs. 83.1% in the ostial/shaft group, 80.3% vs. 72.8% in the distal-single stent group and 67.1% vs. 66.2% in the distal-double stent group. Definite stent thrombosis occurred only in 1(0.3%) patient at 439 days. CONCLUSIONS: In elective patients who underwent PCI for de novo lesions in the ostium, shaft or distal ULMCA, long-term clinical outcomes with SES and PES use were similar independently of lesion location and stenting technique.

Effects of HIV type-1 immune selection on susceptability to integrase inhibitor resistance

BACKGROUND: All site-specific interactions between HIV type-1 (HIV-1) subtype, human leukocyte antigen (HLA)-associated immune selection and integrase inhibitor resistance are not completely understood. We examined naturally occurring polymorphisms in HIV-1 integrase sequences from 342 antiretroviral-naive individuals from the Western Australian HIV Cohort Study and the Swiss HIV Cohort Study. METHODS: Standard bulk sequencing and sequence-based typing were used to generate integrase sequences and high-resolution HLA genotypes, respectively. Viral residues were examined with respect to drug resistance mutations and CD8(+) T-cell escape mutations. RESULTS: In both predominantly subtype B cohorts, 12 of 38 sites that mediate integrase inhibitor resistance mutations were absolutely conserved, and these included the primary resistance mutations. There were 18 codons with non-primary drug resistance-associated substitutions at rates of up to 58.8% and eight sites with alternative polymorphisms. Five viral residues were potentially subject to dual-drug and HLA-associated immune selection in which both selective pressures either drove the same amino acid substitution (codons 72, 157 and 163) or HLA alleles were associated with an alternative polymorphism that would alter the genetic barrier to resistance (codons 125 and 193). The common polymorphism T125A, which was characteristic of non-subtype B and was also associated with carriage of HLA-B*57/*5801, increased the mutational barrier to the resistance mutation T125K. CONCLUSIONS: Primary integrase inhibitor resistance mutations were not detected in the absence of drug exposure in keeping with sites of high constraint. Viral polymorphisms caused by immune selection and/or associated with non-subtype B might alter the genetic barrier to some non-primary resistance-associated mutations.

Code-specific learning rules improve action selection by populations of spiking neurons

Population coding is widely regarded as a key mechanism for achieving reliable behavioral decisions. We previously introduced reinforcement learning for population-based decision making by spiking neurons. Here we generalize population reinforcement learning to spike-based plasticity rules that take account of the postsynaptic neural code. We consider spike/no-spike, spike count and spike latency codes. The multi-valued and continuous-valued features in the postsynaptic code allow for a generalization of binary decision making to multi-valued decision making and continuous-valued action selection. We show that code-specific learning rules speed up learning both for the discrete classification and the continuous regression tasks. The suggested learning rules also speed up with increasing population size as opposed to standard reinforcement learning rules. Continuous action selection is further shown to explain realistic learning speeds in the Morris water maze. Finally, we introduce the concept of action perturbation as opposed to the classical weight- or node-perturbation as an exploration mechanism underlying reinforcement learning. Exploration in the action space greatly increases the speed of learning as compared to exploration in the neuron or weight space.