Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Wearable and Implantable Renal Replacement Therapy
Introduction
Dialysis is the semi-selective removal of solutes from a solution by diffusion across a semipermeable membrane. It was first used in vitro by Thomas Graham in 1861 to separate ions from macromolecules in a colloidal solution. Early in the 20th century, John Jacob Abel isolated epinephrine, insulin, and other hormones from the blood of living dogs using dialysis, which he termed “vividiffusion.” Kolff is generally credited with the first successful dialytic treatment of kidney failure in a human patient in 1944. Alwall independently developed a hard-shell dialyzer, allowing for pressure-driven ultrafiltration to control extracellular fluid volume in addition to solute clearance. In 1960, Scribner, Quinton, and Dillard developed a reliable make-and-break connection to the circulation, the Scribner shunt, and maintenance hemodialysis became possible. The technology and medical, economic, and social infrastructures of dialysis have evolved so that by 2020, over 500,000 American patients with end-stage kidney failure are able to survive for months and years after the diagnosis of kidney failure. Dialysis is thus the first engineered replacement for a failed vital organ and has been successful in that a procedure that 50 years ago was rationed due to scarcity is now the standard of care. Recent literature has suggested that overenthusiastic prescription of dialysis to inappropriate patients is a concern, rather than appropriate use of a scarce resource.
A patient dependent on hemodialysis is medically distinct from a patient with normal kidneys or advanced chronic kidney insufficiency. Some of the most compelling data in this regard are the vastly prolonged survival of kidney failure patients who receive a transplant compared to those remaining on the waitlist for an organ. Time to first appropriate shock from an implantable defibrillator is much sooner in dialysis patients than in those with chronic kidney disease, and epicardial atherosclerosis appears much more resistant to therapy in dialysis patients compared to subjects without kidney failure. Patients dependent on dialysis have impaired vaccine responses and die of pulmonary infection at a rate 20 times the rate in the general population. Young women with kidney failure on dialysis often do not menstruate and rarely conceive or carry a pregnancy to term. The clinical presentation of a dialysis patient who is “well dialyzed” by present standards may be disappointingly similar to the presentation of a patient with kidney failure who has not been dialyzed at all.
Limits of Conventional Hemodialysis
Maintenance hemodialysis is an episodic extracorporeal membrane blood purification procedure, and in the United States, it is generally performed three times each week for 3 to 5 hours a session. Dialysis dose is in practice adjusted by setting the duration of the treatment. Treatment time is chosen in order to clear approximately one volume of distribution of urea per patient per treatment, using a membrane that has an area approximately equivalent to the patient's body surface area. Urea is, in many ways, a sensible choice for a biomarker of dialytic adequacy, as the nitrogenous wastes of protein metabolism are clearly linked to acute uremia, and diets that are sparing in protein can delay the onset of uremic symptoms when dialysis is not feasible. This episodic treatment pattern is essentially unlike the continuous, around-the-clock solute removal of healthy kidneys. In this section, we describe how the episodic treatment schedule of dialysis contributes to the morbid phenotype of dialyzed kidney failure.
Solute Removal
Urea
Urea is the end product of protein metabolism. Typical dietary protein intake in adults results in about 10–13 mg/min of urea nitrogen production. In the course of a dialysis session, small solutes are removed in proportion to their concentration. In the case of urea, a two-compartment first-order kinetic model describes urea removal well, and a single-compartment model is sufficient for most urea kinetic modeling.
Where K is the instantaneous clearance, V is the volume of distribution of urea, and t is treatment time. A Kt/V of 1.0 is approximately equivalent to a dialysis session clearance of one urea volume of distribution. In practice, there is some urea rebound from the (not modeled) peripheral compartment, and so the targeted single-pool Kt/V target is 20%–40% higher than 1.0. From this, we can estimate net urea removal in a single session:
Over 2 days between dialysis sessions, somewhere between 28 and 43 g of urea will be generated and need to be removed. As removal is concentration dependent, one might ask, how high must the predialysis urea nitrogen rise to achieve this removal? Hemodialysis urea clearance is usually limited by blood flow rather than dialysate flow or membrane efficiency. If we assume K is around 300–350 mL/min, we can evaluate the integral in Eq. 13.2 and determine that the predialysis urea nitrogen must rise to around 70 mg/dL or greater to remove urea generated at a rate of 10 mg/min. The time-averaged urea concentration is likely close to 40 mg/dL. The lesson is that in maintenance dialysis on a thrice-weekly schedule, azotemia is the price of nutrition. The dialysis patient with a low predialysis blood urea nitrogen (BUN) either has significant residual renal function or is not eating enough protein.
In contrast, the time-averaged concentration of any solute that is continuously cleared by a first-order mechanism tends toward the ratio of the generation rate to the clearance ( G / K ). If urea is generated at 10 mg/min and cleared at 15 mL/min, the long-term concentration will be about 0.66 mg/mL or 66 mg/dL. If we double the clearance to 30 mL/min, the time-averaged concentration will still be elevated: 33 mg/dL.
Phosphorus
Phosphorus is an unusual molecule in dialysis. Inorganic phosphate is a small molecule with a molecular weight (100 Da) less than that of creatinine (113 Da); it is negatively charged and passes through modern dialyzer membranes slightly less freely than creatinine. However, dialytic removal of phosphorus is inadequate to balance dietary intake, so dietary restriction and oral phosphorus binders are required to keep phosphorus levels low. Very few dialysis patients are able to attain target phosphorus levels through dialysis, binders, and diet. The basis for limited dialytic removal of phosphorus is that phosphorus is best described by a multicompartment kinetic model with nonlinear kinetics. Extracellular phosphate falls rapidly during treatment, decreasing the concentration gradient that drives removal, but refilling the extracellular space from intracellular stores is slow. Thus, hemodialysis can extract phosphorus from blood but not from all body tissues. Interestingly, phosphorus concentrations in blood fall precipitously; but then plateau as incompletely understood regulatory mechanisms mobilize phosphorus from reservoirs, possibly bone, to prevent more severe hypophosphatemia. The role of intradialytic hypotension in decreasing blood flow to phosphorus-rich tissues, such as striated muscle remains uncertain, but it is a plausible factor complicating dialytic phosphorus removal. In contrast, continuous renal replacement therapies (CRRTs), as practiced in some critical care units, have instantaneous clearances an order of magnitude lower than hemodialysis but can remove so much phosphorus over 1 to 2 days that supplementation is often needed. This is not surprising as one recalls that long-term steady-state serum concentrations will tend towards G / K . To balance dietary absorption of 800 mg/day or 0.5–0.6 mg/min, an instantaneous clearance of only 15 mL/min can maintain phosphorus levels of about 3 mg/dL. More than that, as in CRRT, will always drop levels lower.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here