Endogenous Formate Elimination and Total Body Clearance During Hemodialysis

To the Editor:

Kerns et al. question the utility of hemodialysis for the case of a late-presenting acidemic methanol-poisoned patient because their data indicate no statistical difference between endogenous formate elimination and total body clearance during hemodialysis [1]. If this data could be substantiated, it may change the paradigm of how patients with methanol toxicity will be managed.

Hepatic methanol metabolism produces formate, and its endogenous clearance is relatively slow with an elimination half-life between 2.5–20 hours [2][3][4][5], whereas the elimination half-life during hemodialysis is 1.1–2.8 hours [3][6][7]. The molecular weight of formate is 46, it is not protein bound, and its volume of distribution (V_d) is 0.5 L/kg [6], which theoretically makes it easily dialyzable. The mean dialysance rate for formate has been reported to be 148 mL/min (range, 128–161 mL/min) [6] and 223 mL/min (range, 196–254 mL/min) [1].

Kerns et al. compared serum formate half-lives before and during hemodialysis [1]. Quality data for kinetic calculations of formate half-life should include at least three to five data points depending on the importance of the conclusion, an observation period of at least three half-lives, and a constant blood flow during hemodialysis. In addition, variables such as the presence of ethanol affecting formate kinetics should be considered. The patient data presented for analysis by Kerns et al. is summarized in Table 1. There were less than three data points for calculating endogenous formate half-life for patients one and two. The out-layer value of 79 min in patient 2 further makes this data point suspect. Assuming this patient to be 75 kg, the V_d (75×0.5) would be 37.5 L [6]. Applying the equation for first order kinetics [(37,500 mL×0.693)/79 min], the endogenous total body clearance (TBC) of formate would be 328 mL/min. During hemodialysis the TBC would be 328+231=559 mL/min. The contribution of hemodialysis to TBC is 41%, which is generally accepted as significant enough to recommend the procedure. However, this calculation is not recommended on this poor set of data and is shown only as an example.

Table

			FA	HD	NonHD T1/2		HD T1/2		HDBF	HDCL
	Age (Sex)	BAL	(mmol/L)	(hrs)	(min;	n≥3)	(min;	n≥3)	(mL/min)	(mL/min;	n≥3)
1.	42 (M)	?	20.6	5	210;	—	158;	+	100–400a	196;	+
2.	61 (M)	?	21.0	7	79;	—	94;	+	250	231;	—
3.	18 (M)	?	5.3	—	234;	+	—		—	—
4.	53 (M)	?	9.9	5	210;		196;	+	—	—
5.	34 (F)	?	0.5	—	—		—		—	—
6.	35 (M)	?	0.6	—	—		—		—	—
7.	25 (F)	?	27.7	4	149;	+	162;	+	400	254;	—
8.	45 (M)	?	34.8	16	347;	+	146;	+	200–400a	213;	+

BAL, blood alcohol level; FA, initial formate; HD, hemodialysis; T1/2, half-life; HDBF, hemodialysis blood flow; HDCL, hemodialysis clearance; n, number of data points.

a. Variable blood flow during dialysis.

Hemodialysis was not performed on patients 3, 5, and 6. There was no hemodialysis blood flow or clearance data available on patient 4. Surprisingly, in patients 2 and 7 the formate half-lives during hemodialysis were actually longer than endogenous elimination. The variability in blood flow during hemodialysis in patients 1 and 8 undermines kinetic calculations. The only possible quality data suitable for analysis comes from patient 8, and his plasma formate half-life was significantly shorter during hemodialysis. Although the variable blood flow during hemodialysis could theoretically counteract an otherwise first-order elimination during hemodialysis, Fig. 1 in Kerns' paper showed otherwise. The conclusion in Kerns' paper appears to contradict the best data points available for analysis.

Understandably, perfect data may be difficult to obtain even from a well-designed prospective multicenter study, and it would be unethical to delay medical therapy for the purpose of increasing data sampling. However, the same scientific rigor should be applied to interpreting all data and the conclusions should reflect critical interpretation of such data. Conclusions derived from imperfect data should be viewed with skepticism and an impetus for further research. Clinical toxicologists and nephrologists need to be provided with convincing data before accepting the concept that formate elimination is not enhanced by hemodialysis and considering a change in clinical practice.

Luke Yip

Rocky Mountain Poison and Drug Center

Denver, Colorado, USA

Dag Jacobsen

Ullevaal University Hospital

Division of Medicine MICU

Oslo, Norway