Pre-filter haemodiafiltration cleared middle molecules better in children — and raised diastolic pressure

Design

randomised, multicentre, crossover trial with a four-week washout and two four-week periods

Population

24 children on maintenance dialysis, 21 completing; median age 15.1 years, 71% girls

Primary outcome

reduction in alpha-1-microglobulin, a large-middle molecular weight uraemic retention solute

Effect

convection volume 36.2 vs 14.3 l/m2 per session; pre-dialysis alpha-1-microglobulin median difference -17.0 ng/ml (95% CI -27.0 to -3.5), beta-2-microglobulin -5.19 mg/l (-9.34 to -1.76), myoglobin -76.0 ng/ml (-100 to -44.5); diastolic blood pressure higher on pre-filter

Twenty-four children on maintenance dialysis were randomised in a crossover design to four weeks of pre-filter haemodiafiltration then four weeks of post-filter, or the reverse, with every other dialysis parameter held constant. Twenty-one completed. Median age was 15.1 years and 71% were girls.

The convection volumes achieved were not marginally different: 36.2 l/m2 per session with pre-filter against 14.3 with post-filter, roughly two and a half times as much. Pre-dialysis concentrations of alpha-1-microglobulin, beta-2-microglobulin and myoglobin were all significantly lower on pre-filter, reduction ratios higher, and interleukin-10, tumour necrosis factor alpha and high-sensitivity CRP lower. Removal of protein-bound uraemic toxins — indoxyl sulfate and p-cresyl sulfate — was also better. Small-molecule clearance, Kt/V and albumin did not differ.

The cost sits in the blood pressure. Diastolic pressure was significantly higher on pre-filter, and it correlated with the replacement volume and with post-dialysis serum sodium — a coherent mechanism rather than an incidental finding. Higher replacement volumes deliver a higher sodium load, and in children on dialysis sodium loading is not a small matter.

The trial measured solutes and inflammation, not outcomes. What it establishes is that pre-filter haemodiafiltration removes what it is meant to remove, at a price that needs managing rather than ignoring.

• Consider pre-filter haemodiafiltration where middle-molecule clearance is the goal and convection volume is limiting
• Check the replacement fluid sodium concentration and post-dialysis sodium when moving to higher convection volumes
• Monitor blood pressure through the switch rather than at the next routine review
• Do not expect better small-molecule clearance; Kt/V and small solute removal were unchanged
• Watch albumin, which did not fall here but is the standard concern with high convection volumes

The statistics, in plain English

A crossover design in 21 children is efficient — each child is their own control, which removes between-patient variation and is why differences reach significance in a sample this small. The confidence intervals for the middle-molecule differences all exclude zero, but the outcomes are laboratory measures and inflammatory markers, not clinical events. Whether better removal of alpha-1-microglobulin translates into better growth, cardiovascular outcomes or survival is exactly what this trial could not test, and the existing evidence for post-filter haemodiafiltration rests on trials that did.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.