Substances That Can Cross the Dialyzer Membrane
During Hemodialysis

Includes normal solutes, uremic toxins, and trace/suspected contaminants from water or membrane. Dialyzer membranes are semipermeable; pore size, membrane type, and water purity determine what actually crosses.

1. Intended solutes & uremic toxins — cross by design

Category	Examples	Molecular Weight	Clinical relevance
Small water‑soluble	Urea, creatinine, uric acid, K⁺, Na⁺, Ca⁺⁺, Mg⁺⁺, phosphate, glucose	<500 Da	Routinely removed by diffusion
Small middle‑molecules	β2‑microglobulin, cystatin C, complement factor D, cytokines IL‑6, TNF‑α	500 Da – 25 kDa	Removed by high‑flux/MCO membranes; accumulation causes amyloidosis, inflammation
Large middle‑molecules	α1‑microglobulin, YKL‑40, free light chains	≥25 kDa	Partially removed by MCO; poor removal by standard high‑flux
Protein‑bound uremic toxins	Indoxyl sulfate, p‑cresyl sulfate	~200 Da but 90% albumin‑bound	Poor removal even with HDF; RR ~48‑53%

2. Bacterial products & fragments — cross from contaminated dialysate

These are not intended but can cross, especially with high‑flux/MCO membranes and poor water quality.

Substance	Source	Size / Notes	Evidence of crossing	Clinical effect
Endotoxin (LPS)	Gram‑negative bacteria in water/biofilm	Intact LPS ~10‑20 kDa; fragments smaller	Anti‑endotoxin antibodies in HD patients prove crossing	Pyrogenic reactions, chronic inflammation, CV disease, ↑CRP/IL‑6
Bacterial DNA fragments (bDNAF)	Dead bacteria in dialysate	Low‑MW oligonucleotide	Can pass MCO membranes; 500 ng/mL induces IL‑6	↑ CRP/IL‑6; strong predictor of CV disease in PD patients
Peptidoglycan, muramyl peptides	Gram‑positive cell wall fragments	<10 kDa	Suspected to cross like LPS fragments	Cytokine‑inducing, micro‑inflammation
Exotoxins, bacterial metabolites	Pseudomonas, other Gram‑negatives	Variable, often <20 kDa	Suspected in outbreaks with bacteria+endotoxin	Fever, hypotension, inflammation

3. Chemical contaminants from water — cross if water treatment fails

These caused 217 cases and 14 deaths 1960‑2007. They cross freely when present.

Contaminant	Source	Effect if crosses	Notes
Aluminum	Exhausted DI tanks	Seizures, dialysis dementia, osteomalacia	3 deaths
Chloramine	Carbon filter failure	Hemolytic anemia	41 patients affected
Copper	Low pH water + copper pipes	Hemolytic syndrome	4 fatalities
Fluoride	Municipal spill or exhausted DI	Fluoride intoxication	4 deaths
Formaldehyde, hydrogen peroxide	Inadequate rinse after disinfection	Patient intoxication, hemolysis	Multiple cases
Silicates, carbonate scales	Hard water, biofilm	Deposited on membranes	May leach trace elements
PFAS (per‑ and polyfluoroalkyl substances)	Environmental, membrane manufacturing	PFHpA, PFNA, PFDA, PFUnDA detected; levels vary by membrane type	Persistent pollutants; associated with kidney disease

4. Disinfection byproducts & organics

Substance	Source	Notes
Trihalomethanes, haloacetic acids	Chlorinated municipal water + organics	Not routinely tested; small MW may cross
N‑nitrosodimethylamine (NDMA)	Chloramine disinfection byproduct	Suspected carcinogen; can cross
Ozone byproducts	If ozone used in water treatment	Aldehydes, ketones; trace levels possible

5. Membrane‑related or leachables

Substance	Source	Notes
BPA, phthalates	Polysulfone/polyethersulfone membranes, plasticizers	Suspected endocrine disruptors; trace leaching possible
PVP (polyvinylpyrrolidone)	Hydrophilic agent in membranes	Can elute; some allergic reactions reported
Acetate	Acetate‑based dialysate	Crosses freely; replaced by bicarbonate in most units but still used

Key factors that increase crossing

High‑flux & MCO membranes

Larger pores → more middle‑molecules, LPS fragments, bDNAF pass

No ultrafilter on dialysate

ISO/AAMI recommend ETRFs; without them, pyrogens cross

Biofilm in water system

Constant source of LPS, bDNA, peptidoglycan

Backfiltration

If TMP reverses, dialysate contaminants are pushed into blood

Membrane age / reprocessing

Pores enlarge after cleaning, ↑ ET leakage

Bottom line for everyone:

The dialyzer is not a perfect barrier. Even with ultrapure dialysis fluid <0.03 EU/mL, endotoxin fragments, bDNAF, and trace chemicals can still cross, especially with high‑flux/MCO membranes. That’s why AAMI, ISO, and JSDT recommend ultrapure fluid for routine HD and why chronic inflammation remains common in HD patients.

— based on membrane science, water quality standards, and clinical evidence —

Substances That Can Cross the Dialyzer MembraneDuring Hemodialysis