Journal of Clinical InvestigationVol. 43, No. 9, 1964

Muscle Blood Flow in Normal Man and in Patients withIntermittent Claudication Evaluated by Simul-

taneous Xe'33 and Na2' Clearances *N. A. LASSEN

(From the Department of Clinical Physiology, Bispebjerg Hospital, Copenhagen, Denmark)

Intramuscular injection of a small amount ofradioactive sodium (Na24) was used by Kety asan index of local blood flow in skeletal muscle (1,2). The hydrophilic sodium ions are known toexchange fairly slowly over many biological mem-branes, in contrast to the lipophilic inert gases towhich such membranes offer no diffusion barriers(3). Hence a radioactive inert gas such asxenon'33 is probably a better indicator of bloodflow than Na24. Recent clinical studies with intra-muscular injection of Xe'33 dissolved in salinehave tended to support this conclusion (4, 5).However, the possible superiority of employingXe'33 instead of Na24 is best evaluated by usingthe two tracers simultaneously. Forty such simul-taneous observations in ten normal subjects and inten patients with intermittent claudication are re-ported in the present study.

Materials and Methods

Ten normal subjects, ages 41 to 73 years (average, 54years), were studied as the control group. These nor-mal subjects had no symptoms suggestive of intermittentclaudication, and all peripheral pulses were normal.None had clinically manifest heart or lung disease.Six of these control cases were hospitalized for minordisorders not affecting the cardiovascular system, andfour were healthy hospital employees.

Ten patients with intermittent claudication, ages 43 to73 years (average, 56 years), were studied as the patho-logical group. These patients all had typical symptomsafter walking a distance of 3 to 600 feet at normal speedand had, in addition, markedly subnormal oscillometricpulsation and absence of peripheral pulse. Three of thepatients (Cases 3, 6, and 9 of this group) were hos-pitalized and confined to bed because of ischemic ulcera-tions of the skin of the foot. The other seven patientswere all in good general condition and were able tomanage some degree of work despite the handicap pre-sented by the claudication. Of the ten patients, three

* Submitted for publication March 16, 1964; acceptedMay 15, 1964.

(Cases 1, 3, and 5 of this group) had only symptomsand signs of unilateral arterial insufficiency; in all theothers bilateral disease was manifest.

One-tenth ml of normal saline containing 50 /Ac ofXe'2' in solution and 20 ,uc of Na2' were injected intothe thickest part of the tibialis anterior muscle. A sharpneedle with an o.d. of 0.4 mmwas inserted 1.5 cm at anangle of about 450 with the surface (i.e., the tip of theneedle was about 1 cm below the skin). The injectionlasted about 15 seconds; the needle was first withdrawn30 seconds after the injection to reduce possible back-flow of the injected material along the needle track.

The local clearances of Xe'2' and Na' were followedwith one scintillation detector coupled via two spectrome-ters to two rate meters. The scintillation detector had aNaI (Th) crystal 5 cm in diameter and about 5 cmthick. A tubular heavy lead collimator 5 cm in diameter(wall thickness, about 2.5 cm; length, 5 cm) was usedwith the aperture about 2 to 3 cm above the injectionsite. Both rate meters had a time constant of 3 seconds,and their output was recorded on linear writing potenti-ometers.

During the study one spectrometer was set so thatthe corresponding rate meter received impulses fromonly the strong primary gamma radiation from Na2'.The other spectrometer was set to record only the peakintensity of the low energy primary gamma radiation(81 kev) of Xe'. For both rate meters the initialcounting rate was about 1,000 cps. Approximately a 15%correction for the Na2' Compton scatter, of energy simi-lar to the Xe'2 primary radiation, was applied by usingan artificial Na' source and noting the deflection onboth potentiometers. This artificial source was a solu-tion of NaCl containing Na' in a glass vial givingabout the same amount of scatter in the 81 kev regionas Na' injected intramuscularly at a depth of 1 cm. Asemilogarithmic plot of each isotope was made afteralso correcting for background radioactivity (Figure 1).

The subjects were studied after about 15 minutes ofrest at a room temperature of approximately 200 C. Af-ter the injection of the mixture of the two isotopes, theirclearance was followed in the resting muscle for about5 minutes. Then a cuff placed just above the knee wassuddenly inflated to a pressure of 250 to 300 mmHg, andthe subject was asked to move the ankle joint by doingat full force dorsiflexion and plantarflexion movements.After about 60 to 100 such movements, muscle fatigue

1805

N. A. LASSEN

and some degree of ischemic pain developed; furthermovements could not be carried out. At this point thecuff pressure was released, and the isotope clearance dur-ing the reactive hyperemia was followed for 10 to 15minutes.

The extra radiation dose involved in adding Xe"3 tothe conventional Na" clearance method is exceedinglysmall, since Xe"' is rapidly cleared through the lung.The two intramuscular inj ections of 50 Ac Xe"' havebeen calculated to result in a gonadal exposure of only0.03 millirads (6).

Calculations. For both Na" and Xe"' the rate of de-crease of the concentration per gram of tissue, dC/dt =

C', can be expressed by applying the Fick principle to100 g of muscle tissue:

where f is the blood flow in milliliters per 100 g per min-ute in the injected area. In the current studies f was notconstant. Cv is the concentration of the venous bloodleaving the area. Equation 1 does not on its right-hand side contain a term for the rate of supply of tracerby arterial blood (f -Ca), since recirculation is negligible(Ca = 0) due to the dilution (and exhalation in thecase of Xe"').

The clearance rate of an indicator, Cl, in millilitersper 100 g per minute, can as suggested by Renkin (7),be defined in analogy with the clearance concept in kid-ney physiology: Cl is that imaginary volume of blood inmilliliters per 100 g per minute that would have containedthe amount of tracer actually leaving 100 g tissue (f .Cv)if complete diffusion equilibrium had occurred. Atequilibrium CVequii = C/X, where X is the tissue-bloodpartition coefficient defined as (C/Cv) equl 1. Accordingto these definitions,

f * Cv = Cl *Cvequ,, = Cl C/X. [2]

Inserting Equation 2 in 1, one obtains

Cl (ml/100 g/minute) = 100 1 (- C'/C). [3]

XNa is the ratio of the tissue sodium concentration (in-cluding sodium in its contained blood) to the whole bloodsodium concentration. These two concentrations may beestimated at about 40 ,uEq per g (8) and 80 ,uEq per ml,respectively (sodium in red blood cells and in muscle cellshas not been taken into account because of the knownslowness of Na exchange in and out of cells). Thus,

40XN. 0.5 ml/g. [4]

TABLE I

Clearance rates in skeletal muscle for Na24 and Xe"" in ten normal subjects

At rest During hyperemia

Case no. Age Leg CINa Clx, Cla/ClXe CINa ClX, CINa/ClXe

ml/100 g/minute ml/100 g/minute1 41 Left 4.8 2.8 1.71 12.6 41.8 0.30

Right 5.8 3.9 1.49 16.8 49.0 0.34

2 45 Left 3.5 6.1 0.57 12.6 70.4 0.18Right 0.8 1.3 0.62 8.1 53.4 0.15

3 47 Left 1.4 1.5 0.93 11.4 47.5 0.24Right 2.1 2.7 0.78 7.8 66.5 0.12

4 51 Left 2.4 3.1 0.77 12.9 61.4 0.21Right 2.6 2.9 0.90 12.5 50.6 0.25

5 53 Left 1.0 0.7 1.43 9.3 35.9 0.26Right 2.0 2.0 1.00 7.1 50.6 0.14

6 53 Left 1.6 1.7 0.94 14.3 79.5 0.18Right 2.6 2.5 1.04 13.5 71.4 0.19

7 54 Left 3.8 4.8 0.79 11.3 53.9 0.21Right 4.3 5.2 0.83 12.1 55.2 0.22

8 61 Left 2.3 2.1 1.10 11.2 63.0 0.18Right 1.4 1.8 0.78 11.0 55.2 0.20

9 63 Left 1.6 2.0 0.80 17.5 49.5 0.35Right 2.0 2.1 0.95 16.4 49.0 0.33

10 73 Left 3.9 5.5 0.71 19.4 53.9 0.36Right 1.4 2.1 0.67 20.7 58.4 0.35

Mean 54 2.6 2.8 0.94 12.9 55.8 0.24SD 9 1.4 1.5 0.30 3.7 10.4 0.08

1806

XEl33 AND NA24 CLEARANCEIN MUSCLE

TABLE II

Clearance rates in skeletal muscle for Na24 and Xe13; in ten patients with intermittent claudication

At rest During hyperemia

Case no. Age Leg CINa CIXe CiNa/CIXe CiNa Clx. CINa/CIxe

ml/100 g/minute ml/100 g/minzue11 43 Left 1.6 2.2 0.73 9.8 13.0 0.75

Right* (2.6) (3.9) (0.67) (11.9) (38.5) (0.31)

12 44 Left 1.0 0.8 1.25 13.8 23.9 0.58Right 1.3 1.0 1.30 12.7 26.7 0.48

13 49 Left 3.2 5.3 0.60 8.9 21.3 0.42Right* (2.2) (3.1) (0.71) (10.7) (54.5) (0.20)

14 55 Left 2.4 3.2 0.75 7.6 20.6 0.37Right 1.2 1.9 0.63 5.3 21.3 0.21

15 56 Left* (1.3) (1.1) (1.18) (14.1) (53.9) (0.26)Right 1.5 1.2 1.25 6.0 13.5 0.44

16 59 Left 2.0 1.1 1.82 7.5 7.2 1.04Right 2.0 2.6 1.70 9.5 13.5 0.70

17 63 Left 1.7 1.9 0.89 11.2 24.4 0.46Right 1.6 2.0 0.80 9.8 17.8 0.55

18 64 Left 1.1 1.5 0.73 12.3 23.2 0.53Right 0.7 0.9 0.77 10.1 31.0 0.33

19 68 Left 3.4 2.3 1.48 14.2 21.1 0.33Right 4.2 2.9 1.45 9.5 10.7 0.89

20 73 Left 2.2 2.7 0.81 8.4 15.9 0.52Right 2.6 2.9 0.90 8.9 18.4 0.48

Mean 57 2.0 2.1 1.05 9.7t 17.4t 0.521SD 10 0.9 1.1 0.39 2.5 6.4 0.20

* Values for three asymptomatic legs with normal peripheral pulsations are set in parentheses and are not includedin the mean values given below the columns.

t Significantly different from mean value in the normal group (p < 0.01).t Highly significantly different from mean value in the normal group (p < 0.001).

Xx. has been determined experimentally (9). It variesslightly with the hematocrit of the blood, but this vari-ation has been neglected, since all subjects studied hada normal hematocrit value. The in vAvo value of Xx.corrected for the specific gravity of blood to obtain it inthe units employed here was, for blood with 15 g hemo-globin per 100 ml, 0.73/1.05, i.e.,

Xx. c 0.7 ml/g. [5]

- C'/C is the relative rate of decrease of tissue concen-

tration. It can be obtained from the observed curve atany time. This curve gives the tissue concentrationmultiplied by a factor related to counting efficiency andgeometry. But, as this factor affects both numeratorand denominator, it cancels out. - C'/C is equal to d ln

C/dt, which is the slope of the tangent of In C. By con-

ventional semilogarithmic plotting, the slope of the tangentis 0.693 per tj, tj being the half-time in minutes of thetangent.

Relation of clearance rates to blood flow. The rela-tion of clearance rates to blood flow is given by Equa-tion 2. If diffusion equilibrium is maintained, then Cv=

Cvequii, i.e., C1 equals f, and the CINa/ClX. ratio is unity.If equilibrium is not maintained, then Cv < Cvequii andCl < f. In this situation the ClNa/Clx. ratio is not likelyto remain equal to unity, as an equal degree of disequi-librium is rather improbable. A ClN./Clx. ratio belowunity shows predominant exchange limitation for Na,whereas a ratio above unity shows predominant exchangelimitation for Xe.

Results

During rest the clearance rates were low in nor-mal as well as diseased legs, and there were nosignificant differences between the two groups(Tables I and II). The average values for all40 legs were as follows: ClNa = 2.3 ml per 100 gper minute (SD 1.2), and Clxe = 2.5 ml per 100g per minute (SD 1.4). The average ClNa/ClXewas 0.98 (SD of mean 0.05), a value suggestingthat diffusion equilibrium is essentially reachedfor both tracers in the resting muscle, i.e., that

1807

N. A. LASSEN

the clearance values equal resting capillary bloodflow in the anterior tibial muscle.

After the release of the tourniquet the two clear-ance rates increased (Figures 1 and 2). Theinitial hyperemic response was similar for bothclearances until about 5 to 10 ml per 100 g hadbeen cleared [the cumulative amount of bloodcleared, ft2 Cl(t)dt, equals 10O-X(In C[t] - InC [t]), and it can consequently be obtained di-rectly from the semilogarithmic plot of the ob-served radioactivity curves]. This initial phasewhere ClNa/ClXe 1 lasted a fraction of a minutein normal legs and somewhat longer in diseasedlegs. It would seem to represent mainly the wash-out of blood having reached diffusion equilibriumbefore the release of the cuff.

After this brief initial phase the two clearancesdid not follow the same course. Clxe rose to amaximal hyperemic value (Figures 1 and 2). Inall cases an extended period (0.5 to several min-utes) of maximal Clxe was apparent by an almoststraight line of steepest slope on the semiloga-rithmic plot of the observed curves. In the 20

Case no.6,left leg,normal

0 1 2 3 4 5 6 7 8 9 10 11 12time in minutes

FIG. 1. THE WASHOUTCURVES OF NA' AND XEi,THEIR CLEARANCES [Cl (ml/100 g/minute) = X- 1OOdlnconcentration (C)/dt] AND CLEARANCERATIOS IN A NOR-

MAL LEG. The evaluation of C1 depends on assessingthe slope of the washout curve; hence it is subject tosome error when this slope changes rapidly.

Ca

E

0T

0

0

c

E

9

EEcin

0c

40I

E0.o

CLtnc

0u

6

4

2

.8.8

.c 4

2 80

c 60

LA 40u

c

Tv 20wa,

X 1.0

Z 0.5

Case no.14,left leg,arterial obstruction

0 1 2 3 4 5 6 7 8 9 10 11 12

time in minutes

FIG. 2. THE WASHOUTCURVES OF NA2' AND XE3,THEIR CLEARANCES,AND CLEARANCERATIOS IN A LEG WITH

OBSTRUCTIVE ARTERIAL DISEASE. See legend to Figure 1.

normal legs maximal Clxe averaged 56 ml per 100g per minute (SD 10), and in 17 legs with inter-mittent claudication maximal ClOe averaged 17 mlper 100 g per minute (SD 6). This difference isstatistically highly significant (p < 0.001), andthere was no overlapping between the two groups.

In most of the legs with claudication the maximalClxe first came over 1 minute after release of thetourniquet (cf. Figure 2).

ClNa, in contrast, typically decreased again af-ter having reached its maximal value at a timebefore Clxe had become maximal (Figures 1 and2). This maximal ClNa was reached in a singlepoint of the curve and was difficult to evaluate ac-

curately. After its maximum, ClNa found a stablelower level. These values of ClNa, obtained fromthe curves in the same interval of time as that in

which Clxe was at its maximal level, are listed in

Tables I and II. This hyperemic value of ClNaaveraged 12.9 ml per 100 g per minute (SD 3.7)in 20 normal legs and 9.7 ml per 100 g per minute(SD 2.5) in the 17 legs with claudication. Thedifference was statistically significant (p < 0.001),but there was considerable overlapping between

the two groups.

t he observations H~~~~~eischemia

andwork

the clearances

aXe,Na

the clearance ratios

1808

XE'33 AND NA24 CLEARANCEIN MUSCLE

Corresponding to the above variations of theindividual clearances, the ratio CINa/Clxe de-creased to values much below unity during hypere-mia (Figures 1 and 2). With the values corre-sponding to the period of maximal Clxe, this ratioaveraged 0.24 (SD 0.08) in the 20 normal legsand 0.52 (SD 0.20) in the 17 diseased legs. Thisdifference is highly significant (p < 0.001), a re-lation also apparent from Figure 3, where ClNahas been plotted against Clxe. The figure showsthat ClNa does not rise over a level of about 12 mlper 100 g per minute irrespective of the Clx,.Only in a single case, the diseased leg with the low-est Clxe, was (ClNa/ClXe) - 1 during maximalClse.

According to the theoretical consideration givenabove, this result indicates that Na24 does notdiffuse freely from the tissue to the capillaryblood during maximal hyperemia. For the Xe133maintenance, diffusion equilibrium during hypere-mia cannot be evaluated from the present studiesper se. If it is essentially maintained, then Clx,3is a measure of capillary muscle blood flow evenduring maximal hyperemia.

Regardless of whether this is so or not, thepresent series demonstrated clearly that Xe'33gave a more clearcut separation between normaland diseased legs than Na24 (cf. Figure 3).

Discussion

The Na24 clearance and muscle blood flow.The transport of electrolytes from blood to tissuein isolated skeletal muscles has been investigatedin a series of studies by Renkin (7, 10-12). In-fusing K42 and Rb8V at a constant concentrationvia the arterial supply, he found that only at verylow blood flow levels of about 2.0 ml per 100 gper minute did the initial extraction ratio (Cartery- Cvein)/Cartery approach the value close to unitytheoretically expected for complete equilibrium.At higher flow rates a progressive decrease of thisextraction ratio was found, values about 0.5 be-ing reached at a flow level of about 10 ml per100 g per minute. These findings could be ac-counted for by shunting of arterial blood past thecapillaries. However, to postulate gross shuntingreaching 50%o at so moderate a flow level as 10ml per 100 g per minute was on general groundsconsidered most unreasonable, as such shunted

blood would serve no nutritive purpose. There-fore, Renkin believed that incomplete transcapil-lary exchange of electrolytes was causing the dropin extraction ratio with increasing flow.

This conclusion is confirmed by the presentfindings. Both Na24 and Xe'33 are cleared locallyvia the capillaries and are thus unable to measureshunt flow, i.e., the sharp decrease of the ClNa/Clxe ratio during hyperemia points to increasingexchange limitation (lack of equilibrium) at thecapillary level for Na with increasing flow.

At high flow rates CINa is probably a measureof the capillary diffusion capacity. This capacityor "permeability-surface area product" is de-fined as the limiting clearance value (in millilitersper 100 g per minute) obtained when blood flowis so rapid (at unchanged capillary permeabilityand surface area) that the capillary tracer con-centration remains very low in the capillary blood(7). Thus it is a measure of the unidirectionalNa flux from tissue to capillary blood.

It follows from Renkin's studies and from thedata obtained in the present study that for muscleblood flows between a low level of about 2 ml anda high level of 15 to 20 ml per 100 g per minute,ClNa cannot readily be interpreted. In this rangeof blood flow, Prentice, Stahl, Dial, and Ponteriostudied the total flow and the Na24 clearance inthe isolated biceps muscle of the dog (13). Onlya fairly gross correlation of the two parameterswas found, which may be considered as a corollaryof the complexity of the factors determining ClNain this flow range.

"Xe and 2'Na Clearances during hyperemia

zNORMALLEGS °0

g 50 CLAUDICATIONe:

40z/Z

z 30

w

<~~~~~~~~~/20 00

'"XENON CLEARANCEIN ML/lOOg MIN

FIG. 3. SIMULTANEOUSCLN, AND CLxe VALUES IN THEINTERVAL WHENTHE LATTER HAD REACHEDITS MAXIMALLEVEL DURING REACTIVE HYPEREMIA.

1809

1N. A. LASSEN

As a consequence of the above analysis, itwould appear difficult to employ clearance data forNa24 or other ions to prove or disprove the exist-ence of a dual circulation (shunt) in skeletalmuscle.

The Xe'33 clearance and muscle blood flow.If Clxe is to be a measure of local blood flow, thenthis tracer must remain essentially in diffusionequilibrium regardless of the flow. Due to itslipoid structure the total capillary membrane sur-face is presumably available for inert gas trans-fer. In contrast, the exchange of ions in thecapillaries of the extremities appears to be limitedto pores in the wall occupying only about 1/1,000of the surface area (14). Disregarding mem-brane resistance, the diffusibility of the inert gasesin the tissues is high. It can be calculated thatthe inert gas concentration in capillary bloodleaving a tissue is likely to be within a few percents of the equilibrium value [Copperman asquoted by Kety (3)].

Experimental evidence supports this conclusion.In the myocardium both heavy water and 1131-

antipyrine have been shown to remain in diffusionequilibrium even under rapidly varying bloodconcentrations and at high blood flow levels (15,16). Presumably the lipophilic inert gases dif-fuse even better in muscle tissue than these twomore hydrophilic tracers.

If a shunt exists, then Clxe measures onlycapillary or nutritive flow, but even a less drasticinhomogeneity of flow would have the effect of ashunt. A progressive fall in Clxe would resultin this situation even if muscle blood flow wereabsolutely stable. I have on several occasionsseen a moderate decrease of Clxe in the restingmuscle studied over 30 minutes. Probably, how-ever, this reflects an unsteady state of decreasingblood flow. It has not been found in patients stud-ied after prolonged (for hours), complete im-mobilization of the limb when injecting, as de-scribed, at a slow rate through a thin needle only0.1 ml Xe133-saline containing no additives, e.g..no added bacteriostatic agent.

To critically evaluate ClOe as a measure of localcapillary blood flow in the muscle, this methodshould, ideally, be compared to a reliable inde-pendent method. Such a method does not exist.The best approach would therefore seem to be tocompare Clxe to total venous outflow from an iso-

lated muscle in the experimental animal. In mana comparison to the result of venous occlusionplethysmography of the calf is of interest, bear-ing in mind that the latter method also measuresblood flow in bone and skin. Only the resultsobtained in normal subjects will be comparedhere, since differences of patient groups arelikely to exist.

Plethysmographic studies of calf perfusion areusually expressed in milliliters per 100 ml calfper minute, but since the specific gravity of thecalf is close to 1, I have without correction changedthe unit to milliliters per 100 g calf per minute.The resting calf perfusion averaged 1.9 and 3.6ml per 100 g per minute in the two series of nor-mal subjects of all age groups (17, 18). The cor-responding resting value averaged 2.8 ml per 100g per minute in the present series (20 legs).The maximal calf perfusion after ischemic workaveraged 44 ml per 100 g per minute in a seriesof normal subjects of all age groups (19). Thecorresponding maximal Clxe value averaged 56ml per 100 g per minute.

Since the maximal flow rate in the musclesalone can be expected to exceed that of the calfas a whole, the observed agreement is consideredsatisfactory and tends to validate the Xe133 method.

ClNa and Clxe during reactive hyperemnia as apractical test for diagnosing arterial obstruction.The marked difference in the clearance of twotracers during the main part of the hyperemic re-sponse has been commented on. Only initially(Figures 1 and 2) was the ClNa fairly consistentlysubnormal in the patients with arterial obstruc-tion, and, as mentioned, reliable quantitation ofthis early difference of ClNa proved difficult, sinceit had its maximal value at only a single point(time) after release of the cuff. Apparently Xe133was superior to Na24 with respect to separatingthe two clinical groups.

In addition, Xel33 has several other advantagesover Na24. The half-life of Xe'33 is longer (5.3days) than for Na24 (14 hours). The softnessof the gamma radiation from Xe"33 (81 kev)renders this isotope considerably easier to handlein the laboratory than Na24 (1,370 and 2,760 kev),and being a gas, Xe"33 does not contaminate count-ing equipment and laboratory space. The lowradiation energy and especially the very rapidelimination from the body via the lungs render

1810

XE'S' AND NA24 CLEARANCEIN MUSCLE

the Xe'33 radiation doses more than 1,000 timessmaller than with the same amount of Na24 (6).

The only clinically useful measure of muscleblood flow in patients with intermittent claudica-tion has so far been obtained by venous occlusionplethysmography. The classical studies of Shep-herd (20) established that resting blood flow isnormal in such patients. A subnormal and de-layed flow response after ischemic work was foundin some patients with intermittent claudication(type B of Shepherd). He also found a group ofpatients with intermittent claudication who hadonly subnormal hyperemic flow but no delayed on-

set of maximal blood flow after release of the cuff(type A of Shepherd). We have found bothtypes in the present study: of 17 legs with claudi-cation, 12 had a delay of the onset of maximalClxe of more than 1.0 minute after release of thetourniquet, this delay representing the upper limitin the normal series. (A more detailed discussionof these clinical aspects of the Xel33 clearancemethod has been published elsewhere [5].)

Summary

Muscle blood flow was evaluated by simultane-ously injecting Na24 and Xe'33 into the anteriortibial muscle in ten normal subjects (20 normallegs) and ten patients with intermittent claudica-tion (17 symptomatic legs). Studies were madeat rest and during maximal reactive hyperemiaafter ischemic work.

Comparison of the clearances of the two iso-topes indicated marked diffusion limitation ofNa24 at high blood flow levels and demonstratedthat Xe'33 allowed a much clearer separation be-tween normal and pathological cases than Na24.Calculated on the basis of the Xe'33 clearance stud-ies, the blood flow averaged 2.5 ml per 100 g per

minute in resting muscle, there being no signifi-cant difference between normal subjects and pa-

tients with intermittent claudication. Duringmaximal hyperemia the maximal blood flow aver-

aged 56 ml per 100 g per minute in normal legsas compared to 17 ml per 100 g per minute in thediseased legs, this difference being highly signifi-cant statistically (p < 0.001). These data are ingood agreement with the results of venous oc-

clusion plethysmography of the calf reported inthe literature, an agreement tending to validate

Xe'33 clearance as a method for measuring quan-titatively the capillary blood flow in skeletalmuscles.

The results obtained with Na24 relative to Xel'33would seem to invalidate clearance data for Na24as a measure of capillary flow except at very lowflow levels. During hyperemia the capillary dif-fusion capacity for Na is rate limiting to an in-creasing extent. These findings also invalidateattempts to employ Na24 or Il31 clearance meas-urements of a dual circulation in skeletal muscle.

Acknowledgments

The author wishes to express his gratitude for a veryfruitful discussion with Drs. E. M. Renkin and D. C.Tosteson of the data presented in this study.

References

1. Kety, S. S. Quantitative measurement of regionalcirculation by the clearance of radioactive sodium(abstract). Amer. J. med. Sci. 1948, 215, 352.

2. Kety, S. A. Measurement of regional circulationby the local clearance of radioactive sodium.Amer. Heart J. 1949, 38, 321.

3. Kety, S. S. The theory and applications of the ex-change of inert gas at the lungs and tissues.Pharmacol. Rev. 1951, 3, 1.

4. Lassen, N. A., K. Hoedt-Rasmussen, I. Lindbjerg, F.Peaersen, and 0. Munck. Muscle blood flow de-termined by use of xenon'33. Scand. J. clin. Lab.Invest. 1964, 15 (suppl. 76), 61.

5. Lassen, N. A., I. Lindbjerg, and 0. Munck. Meas-urement of blood-flow through skeletal muscle byintramuscular injection of xenon-133. Lancet1964, 1, 686.

6. Lassen, N. A. Assessment of tissue radiation dosein clinical use of radioactive inert gases, with ex-

amples of absorbed doses from 'H2, 'MKr and '33Xe.Sixth International Symposium on Radioactive Iso-topes in Clinical Medicine and in Research, BadGastein, Austria, January 1964. In press.

7. Renkin, E. M. Transport of potassium-42 fromblood to tissue in isolated mammalian skeletal mus-

cles. Amer. J. Physiol. 1959, 197, 1205.8. Bergstr6m, J. Muscle electrolytes in man. Scand.

J. clin. Lab. Invest. 1962, 14 (suppl. 68).9. Conn, H. L., Jr. Equilibrium distribution of radio-

xenon in tissue: xenon-hemoglobin associationcurve. J. appl. Physiol. 1961, 16, 1065.

10. Renkin, E. M. Exchangeability of tissue potassiumin skeletal muscle. Amer. J. Physiol. 1959, 197,1211.

11. Renkin, E. M., and S. Rosell. The influence of sym-pathetic adrenergic vasomotor nerves on transport

181

N. A. LASSEN

of diffusible solutes from blood to tissues in skele-tal muscles. Acta physiol. scand. 1962, 54, 223.

12. Renkin, E. M., and S. Rosell. Effects of differenttypes of vasodilator mechanisms on vascular tonusand transcapillary exchange of diffusible materialin skeletal muscle. Acta physiol. scand. 1962, 54,241.

13. Prentice, T. C., R. R. Stahl, N. A. Dial, and F. V.Ponterio. A study of the relationship betweenradioactive sodium clearance and directly meas-

ured blood flow in the biceps muscle of the dog.J. clin. Invest. 1955, 34, 545.

14. Pappenheimer, J. R., E. M. Renkin, and L. M. Bor-rero. Filtration, diffusion and molecular sievingthrough peripheral capillary membranes. A con-

tribution to the pore theory of capillary perme-

ability. Amer. J. Physiol. 1951, 167, 13.15. Thompson, A. M., H. M. Cavert, N. Lif son, and

R. L. Evans. Regional tissue uptake of D20 in

perfused organs: rat liver, dog heart and gastro-cnemius. Amer. J. Physiol. 1959, 197, 897.

16. Krasnow, N., H. J. Levine, R. J. Wagman, and R.Gorlin. Coronary blood flow measured by I1" iodo-antipyrine. Circulat. Res. 1963, 12, 58.

17. Allwood, M. J. Blood flow in the foot and calf inthe elderly; a comparison with that in young

adults. Clin. Sci. 1958, 17, 331.18. Hillestad, L. K. The peripheral blood flow in inter-

mittent claudication, V. Plethysmographic studies.The significance of the calf blood flow at restand in response to timed arrest of the circulation.Acta med. scand. 1963, 174, 23.

19. Carlsten, A., N.-J. Nilsson, and 0. Thulesius.Personal communication.

20. Shepherd, J. T. The blood flow through the calfafter exercise in subjects with arteriosclerosisand claudication. Clin. Sci. 1950, 9, 49.

1812