Kidney International, Vol. 35 (1989), pp. 95—98

Plasma 5-hydroxyindoleacetic acid as an endogenous index ofrenal plasma flow

THIERRY HANNEDOUCHE, DOMINIQUE LAUDE, MICHELE DECHAUX, JEAN-PIERRE GRUNFELD,and JEAN-LUC ELGHOZI

Department of Nephrology, Hépital Necker-Enfants Malades, /61 rue de Sèvres; and Departments of Pharmacology and Physiology,INSERM U7, Faculté de Médecine Necker-Enfants Malades, 156 rue de Vaugirard, Paris, France

Plasma 5-hydroxyindoleacetic acid as an endogenous index of renalplasma flow. 5-Hydroxyindoleacetic acid (5-HIAA), the major serotoninmetabolite, was measured in human plasma and urine using liquidchromatography and electrochemical detection. The renal extraction of5-HIAA was 78%, a value compatible with an almost complete extrac-tion during a single passage of the blood through the kidney. Inaddition, plasma levels of 5-HIAA measured in patients with a widerange of renal function was inversely correlated (r = 0.85) with theclearance of para-aminohippuric acid (PAH). The results indicate that5-HIAA determinations can be used as an alternative procedure to thePAH clearance method for the estimation of renal plasma flow inclinical practice.

Levels of 5-hydroxyindoleacetic acid (5-HIAA), the majorserotonin metabolite, have been measured in the urine toinvestigate the effects of drugs on serotonin (5-hydroxytrypta-mine, 5-HT) metabolism [1—3] or to help in diagnosing patho-logical states such as malignant carcinoid, phenylketonuria ormigraine [4—8]. Renal elimination of 5-HIAA, like that ofp-aminohippuric acid (PAH) is achieved by glomerular filtrationand active tubular secretion in the proximal tubule [9, 10].Tubular secretion concerns the elimination of weak acids and isdue to an active transport mechanism which is inhibited byprobenecid [9, 10]. Interestingly, the excretion of 5-HIAA is notaltered by change of urinary pH or oral administration of fluid[11, 12].

Since the 5-HIAA elimination is analogous to the PAHsystem, it is expected that 5-HIAA clearance (or plasma level of5-HIAA) parallels PAH clearance (or l/PAH clearance) classi-cally used as an index of renal plasma flow [13, 14]. Indeed,Despopoulos and Weissbach [9] demonstrated in a patient withmalignant carcinoid that the renal clearance of 5-HIAA wasmuch greater than the glomerular filtration rate. The recentimprovement in the electrochemical procedure used to detect5-HIAA permitted the development of an assay of endogenous5-HIAA in human plasma. In this study, we analyzed the renalextraction coefficient of 5-HIAA, and compared the renalclearance of 5-HIAA to the renal clearance of PAH to test

whether endogenous 5-HIAA could be used as an index of renalplasma flow in humans.

5-HIAA determinations

Plasma and urine collections. A 2 ml sample of venous bloodwas collected in heparinized tubes and centrifuged for fiveminutes at 500 g. Plasma was stored at —18°C until the assay.Table I shows the influence of the collection procedure on the5-HIAA level. Platelet 5-HT stores do not affect plasma 5-HIAA even after leaving the heparinized blood for two hours atambient temperature. However, when blood was collectedwithout heparin and clotted, the serum 5-HIAA was slightlyelevated when compared to plasma 5-HIAA, reflecting themetabolism of 5-HT leaking from lysed platelets. On the day ofanalysis, plasma was deproteinized by adding 10% HCIO4 4 Nand centrifuged for four minutes at 15000 g. Fifty d of thesupernatant was then injected into the chromatographic system.

Aliquots of urine (10 to 20 ml) were stored at —18°C until theassay. On the day of the analysis, urine was diluted 10 times and25 itl was injected into the chrornatographic system.

Chromatographic conditions. A coupling of liquid chroma-tography with on-line amperornetric detection was used forassaying plasma and urinary 5-HIAA. The general procedurehas been described elsewhere [15]. The mobile phase was a 3vol!2 vol citric acid-monohydrogen phosphate buffer, pH 3.3,containing 15% methanol. The glassy carbon working electrodewas set to a potential of + 500 mV against a silver-silverchloride reference electrode. The eluting flow rate was 1.5 ml!mm. In these conditions mean retention time for 5-HIAA was10.2 minutes (Fig. 1). Fresh standard was injected the day of thedosage (50 /11 of l0 mol/liter for plasma and 10 p1 of 10—6 moltliter for urine). 5-HIAA was identified and quantified by com-paring retention time and peak height to standards. The repro-ducibility of the method was calculated from nine determina-tions by the formula (1 — SD/rn) x 100%. Mean intra-assaycoefficients of reproducibility for plasmatic and urinary 5-HIAAwere 95.2% and 97.6%, respectively. Mean inter-assay coeffi-cients of reproducibility for plasmatic and urinary 5-HIAA were91.2% and 95.9%, respectively.

95

Methods

Received for publication February 22, 1988and in revised form June 30, 1988

© 1989 by the International Society of Nephrology

2

5HIA

A

01...

.

o - -

-

ciiiI

iiii T

ITH

TT

T)

O.ln

A1

96 Hannedouche et a!: 5-HIAA measures plasma flow

Table 1. Effect of the collection procedure on plasma 5-HIAA level

Fig. 1. Chromatogram of a solution ofauthentic 5-HIAA, (10—i mol/liter, 50 pi)(right), and deproteinized plasma (50 pJ) froma volunteer (middle) and a patient with arenal failure (left).

A B C

Heparin 75 lU/mi 250 lU/mi 0 (serum)Ascorbic acid l0— mol/liter 0 0Centrifugation 1000 g for 10

mm

7000 g for 10 mmafter 2 hourslatency

7000 g for 10 mmafter 2 hourslatency

5-HIAA 3.73 0.39 3.87 0.41 4.5 0.32a10 mol/literMean SF.M, N = 6

P < 0.05 for comparison with A, paired t-test

Venous sampling for extraction ratio determinationsMultiple venous blood samplings were performed in seven

hypertensive patients (2 females and 5 males, 12 to 64 years and34.9 6.1 years, respectively, mean SEM) in order tocompare renin secretion of both kidneys. Creatinine clearancewas in the normal range in all cases. We used inferior venouscava (IVC) blood (sampled below renal veins) as an equivalentof peripheral (arterial) blood. We also obtained an arterialsample in one case and arterial and IVC 5-HIAA concentrationswere equal. Blood sample from the renal vein (RV) of thecontrol side was used to estimate 5-HIAA renal extraction ratiousing the formula:

Extraction ratio = [(IVC — RV)/IVC] x 100%.

Clearance determinationsFour healthy volunteers and 23 patients (20 males and 7

females, l9to 83 years and 36.7 3.1 years, respectively, meanSEM) were investigated. We included patients with varying

Fig. 2. Direct correlation between P5-HIAA and CrAg (N = 27).

degrees of renal failure seconoary to nypertension, nephroli-thiasis, chronic glomerulonephritis or interstitial nephropathy.All subjects were studied during the morning after fastingovernight. None was receiving any drug for the day of investi-gation. Each subject drank 500 ml water 30 minutes beforeemptying the bladder, and water loading was maintained by anintake of 300 ml of water every 30 minutes for the duration ofthe study. A bolus of 10 ml of a saline solution containing inulin(40 mg x kg body wt; mutest®, Laevosan Gmbh, Linz, Austria)and PAH (8 mg x kg body wt; Nephrotest®, Ba Gmbh Lich,FRG) was injected intravenously, followed by an infusion of asaline solution containing inulin (100 mg x estimated ml/minGFR per liter) and PAH (50 mg x estimated GFR per liter) at aconstant rate of 2 mI/mm for 150 minutes. Estimated GFR was

P5-

HIA

A, —

8 m

o//li

ter

9 • -

- •_

•

- ••

•

SI

• ••

• • •

• •

I ••

•

•

E

Ia-

calculated according to Cockroft's formula. With this protocol,plasma concentration plateaus of 200 mg/liter for inulin and 10mg/liter for PAH were obtained. After 90 minutes of equilibra-tion, two consecutive clearance periods of 30 minutes eachwere evaluated. Urine was collected every 30 minutes byvoluntary voiding, Blood specimen was drawn at the beginningand the end of each collection period. A specimen of theinfusion solution was collected for assaying inulin and PAH.Plasma and infusion solution concentrations of inulin and PAHwere determined by a colorimetric reaction using an autoana-lyzer Technicon I (Technicon Instruments, Tarrytown, NewYork, USA). 5-HIAA and creatinine were measured on the firstsample of each period. Plasma and urine creatinine wereassayed by the Jaffe's kinetic method with an Astra 8 BeckmanAnalyzer (Beckman Instruments, Fullerton, California, USA).Clearances were calculated using the standard clearance for-mula: C = UV/P, where V is the urine flow, U the urineconcentration of the substance and P its plasma level. The meanof the two determinations were used for comparisons.

Results

The 5-HIAA extraction ratio calculated from caval and renalvenous samplings was 78.4 2.5% (mean SEM, 72 to 88%,N = 7). This very high extraction ratio was also illustrated bythe CSHIAA/CIfl ratio calculated for each patient included in theclearance determination study. This ratio was 6.30 0.66(mean SEM, N = 27) calculated in the same patients. Themost striking result of this study is illustrated in Figures 2 and3. The correlation coefficient of the inverse relationship be-tween 5HIAA and CPAH was very high (r = 0.85, P < 0.001,N = 27). 5HIAA was stable during the investigation period(difference between the first and the last 5-HIAA determination:0.15 0.11 10-8 mol/liter, NS, N = 27). The direct correlationbetween C5HIAA and CPAH shown in Figure 4 was also highlysignificant (r = 0.76, P < 0.001, N = 27) with a slope of 1.15(CSHJAA = 1.15 CPAH — 16). 5-HIAA was inversely correlatedwith CSHIAA (CSHIAA = (l234/P51AA) — 3.7, r = 0.76, P <0.001, N = 27). This study also confirmed the very close

CPAH, mi/mm

Fig. 4. Direct correlation between C5HIAA and CPAH (C5HJAA = 1.15

CPAH — 16, N = 27, r = 0.76, P < 0.00/).

relationship between Cr and C1 (Car = 0.97 C1 + 13, r =0.93, P < 0.001, N = 27).

Discussion

The test using PAH clearance provides an accurate index ofrenal plasma flow because this compound is almost completelyextracted during a single passage of the blood through thekidney. However the extraction ratio in humans varies fromabout 0.7 to 0.9 at plasma concentrations well below thetransport maximum [13]. The main reason for the incompleteextraction of PAH is a failure of part of the blood supply to themedulla to perfuse pars recta segments of proximal tubules(sites of PAH secretion) [13]. This anatomical feature is appli-cable to the other compounds eliminated via the same mecha-nisms and implies that the extraction ratio of such compounds

Hannedouche et a!: 5-HIAA measures plasma flow 97

A B

15

10

5

.

0 40 60 801 OOO/C,,,, 1000/CPAH

Fig. 3. Inverse correlation between 5-HiAA and C1,, (left, P5HJAA = [1401C1,J + 1.04, N = 27, r = 0.91, P < 0.00/) and between 5-HIAA and CPAH(right, 15-H1AA = [770/CPAH] + 1.0, N 27, r = 0.85, P < 0.001; values identical to those represented in Fig. 2).

100 0 5 10 15

.

I/,0

1000

500

0

.

98 Hannedouche et a!: 5-HIAA measures plasma flow

fall within the range of the PAH extraction ratio. The presentlyobserved 5-HIAA extraction ratio of 0.78 is therefore a satis-factory result in agreement with the working hypothesis thatthis metabolite is almost completely extracted during a singlepassage of the blood through the kidney. An incomplete corticalextraction, evidenced with PAH [161, could also occur for5-HIAA and account for its reduced extraction ratio. Thissimilarity, together with the complexity of the PAH clearancedetermination which restricts the use of the method to clinicalresearch led us to carefully compare plasma levels and renalclearance of 5-HIAA to the reference PAH method. The recentdevelopment of LCEC helped us to assay plasma levels of5-HIAA, which has the considerable advantage over the PAHof being an endogenous compound and therefore obviating theneed of an infusion.

The glomerular filtration of the 27 patients estimated from theinulin clearance ranged from 14 to 152 mi/mm. The PAHclearance used as the reference method for renal plasma flowwas largely in excess of inulin clearance, ranging from 134 to788 ml/min, with a mean CPAH/CIfl of 5.63 (1/filtration fraction).

The most striking result of this study was the strong inversecorrelation (r = 0.85, P <0.001) between 5-H1AA and CPAH.This relation indicates that plasma 5-HIAA reflects renalplasma flow estimated from CPAH. The formula of a hyperbolaand the results, as may be seen in Figures 2 and 3, agree withthe theoretical assumption. Such an hyperbolic (inverse) rela-tionship has been repeatedly observed between endogenouscreatinine level and glomerular filtration rate [171. Brod [17]pointed out that deviations from this relationship may occur ifthe investigation is not carried out in a metabolic steady state.Plasma 5-HIAA, like plasma creatinine, depends of both itsrenal elimination and total body production. This latter factorbeing not controlled, one should therefore assume that 5-HIAAformation was similar in all the tested patients, whatever theirrenal status. Actually the main influence over 5-HIAA produc-tion is the dietary intake of serotonin (converted to 5-HIAA) orthe intake of ethanol, reducing the 5-HT metabolic conversionto 5-HIAA [6]. Since the blood was drawn in the morning whilepatients were fasting, these dietary factors were minimized andonly the metabolism of endogenously released 5-HT was re-sponsible for 5-HIAA production. This probably reduced thedeviation from the inverse relationship between 5HIAA andCPAH which was therefore mainly dependent upon the renalstatus. The upper plasma 5-HIAA level which was observedwith a renal plasma flow within normal limits was 4.l0_8 M.When the 5-HIAA level exceeded 4.108 M practically all thesubjects had reduced rates of renal plasma flow. The hyperbolicrelationship makes uneasy the estimation of high plasma flows,but within low renal plasma flow ranges plasma 5-HIAA levelsare markedly affected (Fig. 2).

5-HIAA clearance largely exceeded inulin clearance, with amean C5HJAA/CIfl of 6.30, confirming that 5-HIAA is activelysecreted by the kidney. CSHIAA was correlated with CPAH (r0.76, P < 0.001) but the correlation was not as strong as theinverse correlation between 5.HIAA and CPAH (r =0.85). Thereason for this discrepancy between the quality of the two

correlations could depend upon several factors, such as theanalytical procedure used for urinary 5-HIAA and/or an intra-renal 5-HIAA metabolism.

In conclusion, the present study confirms that 5-HIAA isalmost completely extracted during a single passage of theblood through the kidney. Being an endogenous substanceeasily assayed in plasma by using liquid chromatography andelectrochemical detection, we recommend its use in clinicalpractice for a quick orientation on renal plasma flow.

Reprint requests to J.L. Elghozi, Department of Pharmacology,INSERM U7, Facu!té de Medecine Necker-Enfants Malades, 156 ruede Vaugirard, 75015 Paris, France.

References

I. ELGH0zI JL, LAUDE D, DUPRAT P. MIGN0T E: Pizotifen increases5-HIAA urinary excretion in male healthy volunteers. Eur J ClinPharmacol 27:191—196, 1984

2. GARFINKEL PE, WARSH JJ, STANCER HC, GODSE DD: CNSmonoamine metabolism in bipolar affective disorder. Arch GenPsychiatr 34:735—739, 1977

3. LAUERJW, IN5KIP WM, BERNSOHN J, ZEYER EA: Observations onschizophrenic patients after iproniazid and tryptophan. Arch Neu-rol Psychiatr 80:122—130, 1958

4. BoussER MG, ELGH0zI JL, LAUDE D, SolssoN T: Urinary 5-HIAA in migraine: Evidence of lowered excretion in young adultfemales. Cephalalgia 6:205—209, 1986

5. CURZON G, THEAKER P. PHILIPPS B: Excretion of 5-hydroxyin-dolyl acetic acid (5-HIAA) in migraine. J Neurol Neurosurg Psy-chiatr 29:85—90, 1966

6. ELOHOZI JL, LAUDE D, RaY F, LEDARD F, MIGNOT E, DUPRAT P,MEYER P: Le 5-HIAA urinaire: Signification et application auxantimigraineux, in Les Antimigraineux, edited by MEYER P, ALBINH, JAILLON P, LHOSTE F, Paris, Masson, 1985, pp. 212—228

7. GARNIERJP, BOUSQUET B, DREUX C: Determination of 5-hydroxy-indoleacetic acid in urine by high-performance liquid chromatogra-phy: A fully automated method. J Chromatogr 204:225—230, 1981

8. PARE CM, SANDLER M, STACEY RS: 5-hydroxytryptamine: defi-ciency in phenylketonuria. Lancet 1:551—553, 1957

9. Dspoi'ouios A, WEISSBACH H: Renal metabolism of 5-hydroxyin-doleacetic acid. Am J Physiol 189:548—550, 1957

10. HAKIM R, WATROUS WM, FuilMoTo JM: The renal tubular trans-port and metabolism of serotonin (5-HT) and 5-hydroxyindoleaceticacid (5-HIAA) in the chicken. J Pharmacol Exp Ther 175:749—762,1970

11. MILNE MD, CRAWFORD MA, GIRAO CB, LOUGHRIDOE L: Theexcretion of indolylacetic acid and related indolic acids in man andthe rat. Clin Sci 19:165—179, 1960

12. BERTACCINI G, BARONLO G: Urinary excretion of 5-hydroxyindole-acetic acid after administration of fluid in man. Arch liii Pharma-codyn 155:57—62, 1965

13. BRENNER BM, ZATZ R, IcHIKAwA I: The renal circulations, in TheKidney, vol. 1 (3rd ed.), edited by BRENNER BM, RECTOR FC JR,Philadelphia, Saunders, 1986, pp. 93—123

14. MORGAN T: Renal physiology, in Nephrology, edited by HAM-BURGER J, CROSNIER J, GRUNFELD JP, New-York-Paris, Wiley-Flammarion, 1979, pp. 59—100

15. LE QUAN-BUI KH, ELGH0ZI JL, DEVYNCK MA, MEYER P: Rapidliquid chromatographic determination of 5-hydroxyindoles anddihydroxyphenylacetic acid in cerebrospinal fluid of the rat. Eur JPharmacol 81:315—320, 1982

16. NI55EN 01: The extraction fraction of p-aminohippurate in thesuperficial and deep venous drainage area of the cat kidney. ActaPhysiol Scand 73:329—338, 1968

17. BROD J: The Kidney. London, Butterworth, 1973 p. 95