Prediction of response to antiretroviral therapy by human experts and by the EuResist data‐driven expert system (the EVE study)

Objectives

The EuResist expert system is a novel data‐driven online system for computing the probability of 8‐week success for any given pair of HIV‐1 genotype and combination antiretroviral therapy regimen plus optional patient information. The objective of this study was to compare the EuResist system vs. human experts (EVE) for the ability to predict response to treatment.

Methods

The EuResist system was compared with 10 HIV‐1 drug resistance experts for the ability to predict 8‐week response to 25 treatment cases derived from the EuResist database validation data set. All current and past patient data were made available to simulate clinical practice. The experts were asked to provide a qualitative and quantitative estimate of the probability of treatment success.

Results

There were 15 treatment successes and 10 treatment failures. In the classification task, the number of mislabelled cases was six for EuResist and 6–13 for the human experts [mean±standard deviation (SD) 9.1±1.9]. The accuracy of EuResist was higher than the average for the experts (0.76 vs. 0.64, respectively). The quantitative estimates computed by EuResist were significantly correlated (Pearson r=0.695, P<0.0001) with the mean quantitative estimates provided by the experts. However, the agreement among experts was only moderate (for the classification task, inter‐rater κ=0.355; for the quantitative estimation, mean±SD coefficient of variation=55.9±22.4%).

Conclusions

With this limited data set, the EuResist engine performed comparably to or better than human experts. The system warrants further investigation as a treatment‐decision support tool in clinical practice.     

Membrane expression of platelet calpain

Platelet calpain has many platelet substrates, including external membrane proteins. We thus investigated whether platelet calpain II was associated with platelet membranes in unstimulated and thrombin- activated platelets. A monospecific, goat polyclonal antibody was reared to purified platelet calpain II. Sixteen whole platelet lysates were found to contain 4.5 +/- 0.7 micrograms calpain antigen II per 10(8) platelets (mean +/- SEM) as determined by a competitive enzyme- linked immunosorbent assay. Using the dipeptide fluorogenic substrate, Suc-Leu-Tyr-MCA, 17 human platelet lysates contained 3.6 +/- 0.4 micrograms calpain activity per 10(8) platelets. Platelet calpain II was associated with the Triton X-100 insoluble platelet cytoskeletons from both unstimulated and thrombin-activated platelets. When compared with the total cell content of platelet calpain II, calpain antigen (10% to 13%) and calpain activity (24% to 28%) was associated with platelet cytoskeletons in unstimulated and thrombin-activated platelets, respectively. On immunoblot, the heavy chain (80 Kd) of calpain II was detected in platelet cytoskeletons. Subcellular fractionation studies on both unstimulated and thrombin-activated platelets, revealed that half of the total platelet calpain II antigen was associated with cytosol, and the other half was associated with the membrane fraction. Platelet calpain II was not seen on the surface of unstimulated, paraformaldehyde fixed platelets by immunofluorescence. However, on thrombin-activated platelets, rim immunofluorescence was seen, indicating that activated platelets externalize their calpain. This observation was confirmed by the finding that about 2,000 molecules per platelet of an 125I-anti-calpain II Fab' specifically bound to thrombin-activated but not unstimulated platelets. Both dibucaine (1 mmol/L) and platelet activating factor (1.86 mumol/L) in the absence of external Ca++, but not collagen (5 micrograms/mL) or ionophore A23187 (2.5 mumol/L) in the absence of external Ca++, were also able to externalize platelet calpain II antigen, as indicated by a similar level of specific 125I-anti-calpain II Fab'-platelet binding. These combined studies indicate that platelet calpain II is a major protein, comprising 2% of total platelet protein, a substantial portion of which is membrane-associated. When platelets are activated by thrombin and platelet activating factor, calpain II antigen also becomes present on the external platelet surface.     

Parathyroid hormone resistance and B cell lymphopenia in propionic acidemia

The mechanisms of hypocalcemia, recurrent infections and hypogammaglobulinemia associated with metabolic decompensation of propionic acidemia due to propionyl-CoA carboxylase deficiency have not been defined. A 7-week-old infant with this disorder presented with severe hypocalcemia and B cell lymphopenia during an episode of metabolic acidosis and hyperammonemia. Hypocalcemia (1.1 mmoll ¹) was associated with elevated serum intact parathyroid hormone (122 ng 1 ¹), hyperphosphatemia, hypophosphaturia and hypercalcuria, indicating parathyroid hormone resistance. B cell lymphopenia (20 cells μl^-1) was associated with transient neutropenia, anemia and subsequent hypogammaglobulinemia (IgG < 294mgdl^-1, IgM < 8mgdl^-1, IgA < 8mgdl ¹), while T cells were normal. Parathyroid hormone resistance and B cell lymphopenia resolved following treatment with hemodialysis, diet and carnitine. These complications may be due to interference with parathyroid hormone renal tubular action and B cell maturation/proliferation by accumulated organic acids.     

Infant growth and aorta total lipid fatty acids

Abnormal fetal and infant growth have increasingly been correlated with adult onset cardiovascular disease. To date, there is little known about the lipid fatty acid profiles in infant cardiovascular tissue. Therefore, we analysed total lipid fatty acids from thoracic and abdominal aorta intima and media from 24 normally grown sudden infant death syndrome cases. Aorta from small for gestational age (n = 2), failure to thrive from birth (n = 3), and premature (n = 1) infants were also examined. Dihomo-gamma-linolenic acid (C20:3n-6) and oleic acid (C18:1n-9) concentrations were significantly lower in the thoracic than in the abdominal aorta. Similar dietary related differences were found in the subgroup (n = 15) of infants fed on formula milks. Both abdominal and thoracic intimal arachidonic (C20:4n-6) to dihomo-gamma-linolenic acid ratios were greater in the infants with retarded growth after birth than in their normally grown counterparts. Growth restriction in infancy might disrupt the normal accretion of vascular endothelial polyunsaturated fatty acids.