LP were detected in a significant proportion of dialysis patients, probably related to underlying CAD with left ventricular dysfunction. Prolongation of fQRS after dialysis could be explained by the acute reduction in serum potassium levels.

Several comparative studies have claimed that procedures based substantially or exclusively on pressure-driven water-solute transport, such as haemodiafiltration or haemofiltration, afford better protection of the cardiovascular tolerance to fluid removal than conventional haemodialysis. During each depurative modality, several factors are set in motion that might impact, each in its own right, upon the haemodynamic response to fluid withdrawal. To explore the haemodynamic effect of each of them singularly, one needs to keep all other components unvaried. However, this is very difficult to accomplish. For instance, to confirm the alleged greater protection of cardiovascular stability by pure convection vs diffusion, one needs to keep unvaried all the other factors potentially affecting haemodynamic tolerance, i.e. the rate of body fluid removal, the membrane, the buffer, the blood temperature in the extracorporeal circuit, depuration efficiency, the sodium balance, the fluid sterility and so on. Such studies are still awaited. However, clinical trials published to date have not resolved the question of whether haemofiltration and haemodiafiltration provide a better haemodynamic tolerance to fluid removal. If we limit our consideration to controlled trials only, most prospective studies have adopted a cross-over design implemented on very small patient samples and for very short periods. Such an approach is liable to generate misleading results because the incidence of dialysis hypotension often fluctuates from time to time. Owing to such fluctuations, results can be strongly affected by the 'order effect' of the cross-over from one technique to the other. The negative results provided by parallel comparisons of procedures should be taken with caution because patients samples did not include a suitable proportion of unstable patients.

The purpose of this study was to evaluate the autonomic response to standard haemodialysis and the changes associated with the onset of intradialytic hypotension in 12 normotensive patients with uraemia. Power spectra of R-R interval and of blood pressure fluctuations were obtained during a standard dialysis session and estimated in the low-frequency (LF, 30-150 mHz) and high-frequency (HF, 150-400 mHz) range. The absolute power of the LF component of blood pressure variations and the LF/HF ratio of R-R interval were assumed as indexes of sympathetic activity. Standard haemodialysis induced hypotension in six patients (unstable) while a minor pressure decline was present in the other six (stable). Normalized blood volume before dialysis and percentage volume reduction were similar in the two groups. Tachycardia in response to pressure and volume decrease was more pronounced in stable than in unstable patients, as evidenced by a higher slope of the relation between R-R interval and systolic blood pressure (7.9 versus 0.9 ms/mmHg, P<0.01). Sympathetic tone was enhanced during early dialysis in all patients (+2+/-1 for R-R LF/HF ratio, +2.4+/-0.6 mmHg2 and +7.2+/-2 mmHg2 for absolute LF power of diastolic and of systolic blood pressure respectively, P<0.05), compared with baseline predialysis values. During late dialysis, unstable patients showed an impairment of sympathetic activation which preceded hypotension and was maximal during the crisis (-2.9+/-1.4 for R-R LF/HF ratio, -2.7+/-1.4 mmHg2 and -8.6+/-4.0 mmHg2 for absolute LF power of diastolic and of systolic blood pressure respectively, P<0.05). On the contrary, stable patients showed constantly elevated indexes (+3.7+/-1.4 for R-R LF/HF ratio, +5.9+/-2.7 mmHg2 and +13.3+/-6.2 mmHg2 for LF of diastolic and of systolic blood pressure, P<0.05). Values returned to predialysis levels after the end of the dialysis session in all patients. We conclude that standard haemodialysis activates a marked and reversible sympathetic response in both stable and unstable uraemic patients. However, in unstable patients, such activation is impaired in late dialysis, therefore contributing to the onset of the hypotensive crisis.

Background. In patients on on-line convective treatments, given the considerable quantity of dialysis fluid re-infused, the small amount of acetate present in bicarbonate dialysis fluid as a pH stabilizing factor may allow a significant transfer of that anion to the patient, possibly inducing cytokine activation. Methods. To verify this hypothesis, we performed on-line haemodiafiltration (OL-HDF) with (3 mmol/l) and without acetate in dialysis fluid in a cross-over randomized order on 12 prevalent patients. Results. In comparison with the pre-treatment values, plasma acetate levels were unchanged during and after acetate-free OL-HDF, while they were 5-6 times higher in the course of OL-HDF containing acetate in dialysis fluid; plasma acetate levels returned to basal values 2 h after the end of the procedure. The total increase of bases in the patient attributable to acetate was 36%.Plasma bicarbonate values at the end of treatment were significantly lower in treatments without acetate, as compared to those with acetate. Interleukin-6 plasma levels were super-imposable at the beginning and in the course of the two methods compared, but there was a tendency towards a greater increase at an interval of 2 h following OL-HDF with acetate. Conclusions. Our preliminary results confirm the assumption that body gain of acetate is particularly high in convective treatments, while acetate-free OL-HDF slows down acetate burden. Clinical advantages due to these effects should be evaluated in properly designed prospective studies.

35 dialysis patients underwent anti-HBV vaccination. We classified patients in responders or non-responders using an anti-HBs titer of 50 UI/l as the discriminating serum level and tried to assess whether the antibody response bears any relationship with the nutritional status. 26 patients (74%) reached the target atb titer, which was maintained during follow-up (average 360 UI/l). The weak response in the other 9, with values never exceeding 20 UI/l, was short-lived. Anthropometric and impedenziometric parameters were higher in responders than in nonresponders, but the difference did not reach statistical significance. We conclude that the atb titer which discriminates uremics in responders or not must be > 50 UI/l and that the nutritional status may interfere with the seroconversion rate, but this conclusion needs to be validated in a wider population.