Studies of Peripheral Thyroxine Distribution

in Thyrotoxicosis and Hypothyroidism

JOHNT. NICOLOFF and J. THOMASDOWLING

From the Departments of Medicine of the King County Hospital System andthe University of Washington, Seattle, Washington 98104; WadsworthVeterans Administration Hospital and UCLA School of Medicine,Los Angeles, California 90033

A B STRACT Compartmental analysis of theperipheral distribution of labeled thyroxine wasapplied to various groups of subjects with thyro-toxicosis and hypothyroidism. It was observed thatthe hepatic incorporation of thyroxine was aug-mented in subjects with Graves' disease when com-pared to non-Graves' disease control groups at alllevels of thyroid function. Decreased values ofhepatic incorporation occurred in primary hypo-thyroid subjects. These lowered values were notacutely corrected by elevation of the serum thy-roxine level, but were observed to be rectified af-ter several months' therapy with exogenous thy-roid hormone. These alterations of the hepaticthyroxine-3113 incorporation were independentlyverified by direct quantitative liver scintiscandeterminations.

Employing a dual thyroxine tracer system, wewere able to demonstrate that during the earlyphases of equilibration of a tracer dose of thy-roxine, alterations in the rate of deiodination wereobserved to be present in the various thyroid dis-ease states. Increased deiodination rates werefound in subjects with Graves' disease and thereverse was noted in patients with primary hypo-thyroidism. Kinetic analysis of thyroxine com-partmental distribution during this early phase ofequilibration of a labeled thyroxine tracer indi-cated that the primary tissue uptake occurred inthe liver. These findings supported the contention

Address requests for reprints to Dr. John T. Nicoloff,University of Southern California School of Medicine,2025 Zonal Avenue, Los Angeles, Calif. 90033.

Received for publication 28 January 1968 and in revisedform 8 May 1968.

that the amount of labeled thyroxine incorporatedin the liver may be directly related to the deiodina-tion rate of thyroxine by that organ. The patho-genetic basis of these alterations is presentlyunknown.

INTRODUCTION

Evidence that disturbances in the peripheral me-tabolism of thyroxine may occur in some formsof thyroid disease in man became apparent fromthe investigations of Lennon, Engbring, and Eng-strom (1), and Ingbar and Freinkel (2). Len-non and coworkers studied the pattern of theacute disappearance of an injected tracer dose ofthyroxine-"3"I in various thyroid disease states.They found that during the 20-50 min postin-jection period there was a relative acceleration inthe thyroxine-"3"I removal rates in thyrotoxic,eumetabolic, and hypothyroid Graves' disease pa-tients. Interestingly, a reduced disappearance ratewas noted in patients with primary hypothyroid-ism, and a normal or reduced rate was present insubjects with toxic nodular goiter and iatro-genic thyrotoxicosis. The investigations of Ingbarand Freinkel also supported the concept that pa-tients with Graves' disease have alterations in thy-roxine kinetics (2). Their studies focused on al-terations in the fractional turnover rates of anequilibrated tracer dose of thyroxine-""I both insubjects with thyrotoxic and eumetabolic Graves'disease. They observed an acceleration of thethyroxine turnover rates in both eumetabolic andtoxic Graves' disease subjects and the reverse inprimary hypothyroid patients. Ingbar, Freinkel,

2000 The Journal of Clinical Investigation Volume 47 1968

Dowling, and Kumagai extended these observa-tions to relatives of Graves' disease subjects inwhom they also found an increased turnover ratepresent in some cases. These latter findings sug-gest that the alterations in thyroxine turnovermight be an expression of an underlying inheritedtrait that antedates the onset of clinically detecta-ble Graves' disease (3, 4). Collectively, these stud-ies (1-4) appear to indicate that alterations in themanner of thyroxine distribution and disposal mayplay an integral part in the pathophysiology ofGraves' disease as well as other forms of thyroiddisease.

We have recently described several techniquesfor the measurement of the compartmental distri-bution and deiodination of extrathyroidal thyrox-ine in normal human subjects (5). These tech-niques allow the quantitative estimation of thethyroxine distribution in the plasma, extracellularfluid, hepatic, and extrahepatic tissue compart-ments, as well as a semiquantitative estimation ofthe relative deiodinating activities of the hepaticvs. the extrahepatic compartments. The purposeof this investigation was to apply these new tech-niques to the evaluation of peripheral thyroxinemetabolism in subjects with Graves' disease andother thyroid disease states. Additionally, therewill be described a newly devised method for thedirect quantitative approximation of the hepaticincorporation of labeled thyroxine.

METHODSThe subjects employed for this study were obtained fromthe inpatient and outpatient services of the King CountyHospital System in Seattle, Wash. and the WadsworthVeterans Administration Hospital in Los Angeles, Calif.The patients' diagnoses were established by a combina-tion of appropriate clinical and laboratory observationsby one or both of the authors. Subjects with a history orphysical findings of liver disease or obesity were ex-cluded except where specifically stated. The protein-bound iodine (PBI) determinations were performed bythe method described by Barker, Humphrey, and Soley(6). The triiodothyronine resin uptake tests (RU) weredetermined by a standard commercial method (Trisorb,Abbott Laboratories, North Chicago, Ill.). Thyroxine-'Iand thyroxine-'I equally labeled in the 3', 5' positionswere obtained from a commercial source (Abbott Labora-tories) and stored in a sterile lo serum albumin-salinesolution at 4'C. Before initiating the study, the contami-nating iodide was removed from the labeled thyroxinesolutions by dialysis against a strong anion exchangeresin (5, 7). An approximation method for the com-partmental distribution of labeled thyroxine in plasma,

extracellular fluid, hepatic, and extrahepatic tissuespaces was applied to all study subjects (5). Thismethod employed the measurement of the plasma disap-pearance rate of thyroxine&I concomitantly with itsuptake as observed in the liver and forearm. Plasma thy-roxine distribution space was determined by extrapola-tion of the early plasma curve to the time of injection.The expansion of the labeled thyroxine distribution spaceduring the first 3 hr postinjection was partitioned be-tween thyroxine distribution into the extracellular fluidand hepatic thyroxine spaces. The proportion of the la-beled thyroxine distribution entering the extracellularfluid space was assumed to be equal to the observedtranscapillary passage of the labeled thyroxine into theforearm by 3 hr. The remainder was then considered torepresent hepatic space. The further expansion of thethyroxine distribution space after the 3 hr period repre-sented the entrance of labeled thyroxine into the ex-trahepatic tissue spaces. Justification for these assump-tions was described in detail elsewhere (5). Additionally,a four compartmental mathematical model systent de-vised by Berman (5, 8) was also used in some studysubjects for comparison with the approximation method.

Hepatic thyroxine-'I uptake employing a quantitativescintiscan method. It was apparent from our previousstudies in normal subjects (5) and those of other in-vestigators in man (9-16) and laboratory animals (17-19) that the hepatic uptake of thyroxine may be an im-portant factor in determining the form of the initialportion of the thyroxine-'I plasma disappearance curveand that the liver constituted a major site for tissuethyroxine sequestration. A need seemed apparent for thedevelopment of a more direct quantitative technique formeasurement of thyroxine-'I hepatic incorporation. Theobservation by Van Middlesworth, Turner, and Lipscomb(9), that the liver could be scintiscanned after the in-jection of a thyroxine-'I tracer, seemed to offer anexcellent opportunity for achieving this goal. The methoddevised for this study was to perform a liver scintiscan3 hr after the injection of 20-60 *c of thyroxine-'Itracer. This time period was selected because the peakliver uptake of labeled thyroxine was observed to plateauon or before the 3 hr period in all subjects we have thusfar investigated, regardless of hepatic or thyroid status(5). The instrument employed for the scanning was aPicker Magnascanner III. A 0.340-0.390 Mev win-dow and a zero background subtraction setting wasused. The probe employed had a 19 hexagonal port col-limator with a 4 inch focal length. This is the standardcollimator employed for qualitative liver scintiscans. Theliver outline was determined by inspection of the dot scanpicture. The total liver counts were calculated by sum-ming all of the dots incorporated within the area of theliver outline and multiplying them by the dot factor em-ployed. The dot factor represents the number of countsaccumulated in producing one dot on the scan picture.The amount of thyroxine-'I administered varied de-pending on the anticipated fraction of the injected la-beled thyroxine tracer which, judging from previous ex-perience, would be incorporated into the liver. In all

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidism 2001

cases, the accumulated gross counts over the liver weregreater than 2500. The vascular background correctionwas approximated by counting a portion of the dot scanpicture corresponding to the left lower portion of theabdomen which was equal in area to the hepatic scinti-scan image. Standardization of the in vivo scintiscan wasperformed by scintiscanning a molded plastic liver phan-tom containing the appropriate fraction (15-40%o) of theadministered thyroxine dose. The liver phantom formwas produced by casting in plaster of Paris a normal1400-g liver at autopsy. The plaster casting was thenused to construct a hollow acrylic plastic liver phantom.The walls of the phantom measured approximately '/2inch in thickness and the volume of the phantom con-tained 1400 ml. The phantom was filled with a 1% hu-man serum albumin-normal saline solution which con-tained an appropriate percentage of the injected tracerdose. The phantom was positioned in a manner to con-form as closely as possible to the in vivo geometry of theliver. A typical example of a scintiscan of the liver invivo and the liver phantom are portrayed for comparisonin Fig. 1. Scanning was performed at a 3 inch distancefrom the anterior surface of the phantom, while in thestudy subjects, scanning distance was 1 inch from theanterior chest wall.

Evaluation of the accuracy of the scintiscan method.The influence of scintiscan probe distance on the effi-ciency of the detection was determined by repetitive scinti-scan measurements performed at varying distances fromthe liver phantom source. Normalizing the total countsproduced by the phantom at 3 inches to be 100, we ob-served the relative alteration in counts at 2 inches to be107, at 4 inches 94, and at 5 inches 99. The estimated

* * -V . . * .

. ..

FIGURE 1 Comparison of dot scintiscan obtained from a

liver phantom (upper scan) and an in vivo scan of a

subject with thyrotoxic Graves' disease (lower scan) per-formed 3 hr after injection of a thyroxine-'I tracer.

in vivo counting distance from the anterior surface ofthe liver was easily encompassed by this 2-5 inch span.Thus, a maximum error of 11% of the net liver uptakecounts could be produced by variation in probe dis-tance. The attenuation of counts produced by the shield-ing effect of the anterior chest wall was appraised bythe placement of an autopsy specimen of an anteriorchest wall over the liver phantom. The chest wall em-ployed was composed of the rib cage, associated muscle,and subcutaneous tissue structures. The average over-allthickness of the specimen was %inch. Its placement overthe liver phantom produced a net 6% reduction in liverphantom counts. The X inch acrylic plastic walls of theliver phantom itself approximated the same shielding ef-fect. Therefore, no correction was used for the chest wallshielding effect. It therefore appeared that neither thechest wall nor the scan distance variations should resultin major errors in the hepatic scintiscan measurements.

No attempts were made to correct for the errors pro-duced from variations in the in vivo liver size and form,but in general, the in vivo scintiscan image correspondedfairly closely to that produced by the phantom itself, asillustrated in Fig. 1. The justification for the hepatic vas-cular background correction employed was based on theobservation that a rapid scintiscan performed during theinitial 10 min postinjection period resulted in a net he-patic uptake of 2% in a subject who eventually provedto have an uptake of 30% at 3 hr. Thus, the near zerotime hepatic uptake value corresponded closely to theactivity observed in the left lower quadrant of the ab-domen. Additionally, a hepatic scintiscan performed usingalbumin-"sI instead of thyroxine-'lI revealed almost thesame activity in the hepatic area as was observed in asimilar area of the left lower quadrant.

Incorporated in this study was a previously describedtechnique for the assessment of the relative deiodinationrates of a labeled thyroxine tracer during the equilibrationphase (5). This method involved the measurement ofradioactive iodide activity in timed urine samples afterthe acute injection of a "pure" thyroxine-'I tracer tosubjects who had previously been equilibrated with athyroxine-"3'I tracer dose. The observed change in theratio of 'I to "sI activities in the hourly urine samplesallowed a qualitative analysis of the relative deiodinationrates which occurred during the equilibration period.This method was applied to various forms of thyroiddisease states in an attempt to ascertain if the alteredmanner of thyroxine incorporation into the tissue com-partments might be reflected by changes in the rate ofdeiodination of the labeled thyroxine tracer by thesecompartments.

RESULTS

Compartmental distribution of thyroxine in hy-pothyroidism. 13 subjects with untreated hypo-thyroidism and 5 subjects with treated hypothy-roidism were evaluated for their peripheral com-partmental distribution of thyroxine (5). The

2002 J. T. Nicoloff and J. T. Dowling

method employed allows an approximation of theplasma thyroxine distribution space, extracellularfluid distribution space, as well as hepatic andextrahepatic thyroxine compartments at equilib-rium. The results are listed in detail in Table I.In the untreated hypothyroid subjects, the hepaticcompartment contained only 9 + 10 uig I/M2 + SDvs. a normal value of 51 ± 7 ug I/M2, while theextrahepatic compartment contained 52 + 26 ugI/M2 vs. a nomal value of 122 + 34 ,mg I/M2. Thisreduction in the hepatic compartment was mostprominent in those subjects with severe long-stand-ing hypothyroidism. In some cases, there was nodetectable hepatic incorporation of labeled thyrox-ine. Additionally, when an external scintillationprobe was concomitantly placed coer the liver, nodetectable hepatic uptake was present in these se-verely hypothyroid subjects. When a measurablehepatic uptake was noted in subjects with milderforms of hypothyroidism, no apparent delay in therate at which the liver achieved its peak was notedeither by external liver probe analysis or indirectlyfrom the early rapid drop in the forearm countrate. As noted in the treated hypothyroid group,subjects B.W. and L.N. had a repetition of theircompartmental studies performed very shortly af-ter the initiation of exogenous thyroid hormonetherapy. In both cases a single large dose of stablethyroxine was administered intravenously (500,ug for B.W. and 1500 ug for L.N.). It is evidentfrom the results of these repetitive studies, asshown in Table I, that the altered distribution ofthyroxine was only minimally rectified by theacute elevation of the circulating thyroxine levels.In contrast, A.B. achieved a near normal com-partmental distribution pattern after 6 months oftherapy for hypothyroidism. It is of further in-terest that, in general, those patients with the mostsevere reduction in hepatic incorporation also dis-played the lowest rate of extrathyroidal thyroxineturnover. This was most evident in subject A.B.when her turnover rate went from 5.87%o beforetherapy to 10.5%o after thyroid therapy.

Compartmental distribution of thyroxine inGraves' disease. This portion of the study per-tains to the evaluation of 25 subjects with the es-tablished diagnosis of Graves' disease. Thesecases were divided into subgroups of 11 toxic, 10eumetabolic, and 4 hypothyroid subjects on thebasis of clinical and laboratory evaluation. The

clinical status of each patient was relatively stablefor 3 months or longer before the performance ofthe compartmental studies. Four of the eumeta-bolic subjects (J.L., V.R., H.B., and A.L.) hadreceived no antithyroid therapy. The remainder ofthe eumetabolic and hypothyroid group had beentreated with radioactive iodine except for subjectM.F. who was treated by subtotal thyroidectomy.All thyrotoxic subjects were studied before theadministration of any therapy. In the eumetabolicsubjects, eumetabolism had been achieved for aperiod of 1 yr or longer, the longest period being7 yr for subject M.S. and 8 yr for subject V.R.

The most notable alteration observed in thecompartmental analysis in the subjects withGraves' disease was the relative augmentation inthe quantity of thyroxine estimated to be incorpo-rated into the liver compartment. This was 250 ±81 MCg I/M2 + 1 SD in the toxic subjects and 117 ±38 Mbg I/M2 in the eumetabolic subjects; these val-ues were significantly greater than the 51 + 7 ugI/M2 observed in the control subjects. This relativeincrease of hepatic incorporation seemed to be ex-tended even to the hypothyroid Graves' diseasegroup. The 32 + 4 Mug I/M2 observed to be in thehepatic compartment of subjects with hypothyroidGraves' disease was significantly greater than 9 +

10 Mg I/M2 found in the hepatic compartment ofsubjects with primary hypothyroidism. Althoughthe hepatic thyroxine compartment of the Graves'disease groups appeared to be grossly enlarged,and these groups statistically different fromtheir non-Graves' disease counterpart (P < 0.01),there is great difficulty in properly matchingthese patients with non-Graves' disease subjects.Possibly the most satisfactory standard for com-parison would be to relate the hepatic compart-ment size relative to the PBI level. This compari-son is graphically represented in Fig. 2. It is evi-dent that the hepatic compartment of subjects withGraves' disease generally exceeds that of the non-Graves' disease subjects at all PBI levels with thedifference being less evident at the lower PBIranges. In contrast, the extrahepatic pool size didnot appear to be greatly enlarged in the toxicGraves' disease groups, as listed in Table I, withthe exception of subject C.K. The toxic and eu-metabolic Graves' disease groups displayed an in-creased fractional rate of thyroxine disposal of17.34 +4.56% ± 1 SD for the toxic group and

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidism 2003

TABLE IEstimation of Thyroxine Turnover Kinetics and

Category Subject Sex Age m2 PBI K TDS

Primary hypothyroidism

Primary hypothyroidism,

treated

Graves' disease-toxic

Graves' disease,eumetabolic

A. B.B. W.V. T.B. D.L. N.M. G.H. L.A. P.S. Y.J. C.

MeanSD=-

B. W.L. N.A. B.G. D.L. W.

C. K.A. P.L. R.F. G.C. F.J. K.K. J.R. S.E. V.S. R.A. C.

MeanSD4

C. F.L. H.A. L.H. B.S. K.A. S.M. S.J. L.V. R.J. L.

MeanSD±

F 72 1.24F 85 1.31F 80 1.31M 45 1.94M 30 2.20F 50 1.93M 78 1.67M 73 1.83F 56 1.99M 58 1.94

63 1.7418 0.34

F 72 1.34M 30 1.96F 72 1.17F 69 1.94F 64 1.85

M 33 1.87F 40 1.84F 69 1.27F 66 1.37M 47 1.81M 52 1.75M 43 2.15M 40 1.80F 51 1.50F 28 1.67M 60 2.02

48 1.7013 0.27

M 61 1.87M 51 1.83M 32 1.60M 65 1.72F 27 1.71F 50 1.94F 24 1.73F 28 1.74M 52 1.97F 40 1.68

43 1.7815 0.12

pg/lut) ml

1.00.40.70.60.81.81.71.32.91.9

vo

5.055.874.955.776.877.854.405.777.879.24

1.3 6.36.8 1.55

4.0 6.309.9 7.953.4 10.504.1 9.909.4 8.45

17.512.510.310.210.1

9.99.69.39.29.18.7

10.62.5

23.9023.1019.3021.0015.0019.80

9.1114.7013.1017.3314.40

17.344.56

7.8 15.757.5 11.727.4 15.407.2 10.986.6 13.865.8 10.205.1 10.405.0 13.804.1 9.003.4 12.16

6.0 12.321.5 2.29

Abbreviations used are as follows: PBI, protein-bound iodine; K, daily fractional rate of thyroxine turnover; TDS,thyroxine distribution space; PTDS, plasma thyroxine distribution space; ETDS, extracellular fluid thyroxine iodinedistribution space; HC/m2, hepatic thyroxine iodine content per square meter of body surface; EHC/m2, extrahepaticthyroxine iodine content per square meter of body surface; D/m2, daily turnover of thyroxine iodine per square meter ofbody surface.* Triosorb, Abbott Laboratories, North Chicago, Ill.

2004 1. T. Nicoloff and T. Dowling

liters12.312.314.719.614.012.716.112.710.815.9

14.12.6

12.013.010.511.2

9.9

13.79.6

10.512.512.214.713.914.112.110.811.0

12.31.7

10.013.911.612.79.4

12.79.8

13.914.512.2

12.11.8

Distribution in Thyroid Disease States

ResinPTDS ETDS HC/m2 EHC/mS D/m2 uptake* Comment

liters liters pAg I pg I pg/dajy2.20 2.56 0 61 5.0 - Precoma2.20 2.30 0 21 2.2 Precoma2.58 3.07 1 47 3.9 19.5 Precoma3.85 4.55 3 32 3.5 18.3 Severe4.00 2.63 4 23 3.5 Severe3.00 2.20 20 50 9.3 21.02.73 2.08 5 110 7.2 15.42.90 2.70 5 45 5.2 18.72.64 1.83 22 70 12.4 22.53.89 2.88 26 63 14.4 27.0

3.00 2.68 8.6 52 6.70.68 0.76 10 26 4.3

2.20 2.20 15 212 22.6 - 2 days treated4.10 3.16 56 234 52.2 - 7 days treated2.20 1.90 32 154 32.0 - 4 months treated3.40 3.00 32 70 23.4 23.5 1 yr treated2.63 2.53 92 212 42.5 40.5 1.5 yr treated3.80 3.31 395 222 306 50.3 Atrial fibrillation

2.26 2.48 186 137 151 52.1 Exophthalmos2.90 2.60 279 127 164 44.72.40 2.65 372 183 195 50.33.02 3.00 270 75 102 43.3 Exophthalmos3.28 3.28 260 200 165 46.1 Atrial fibrillation3.82 3.40 181 117 56.5 33.03.08 3.12 268 141 107 42.12.88 2.48 231 170 97.2 44.7 Exophthalmos3.00 2.30 168 128 102 43.53.30 2.30 141 91 68 -

3.07 2.81 250 145 137.6 45.0.49 .42 81 45 148 6.8

2.80 2.27 100 105 65.7 37.7 Exophthalmos3.08 3.02 160 160 66.8 36.0 Exophthalmos2.78 2.72 160 123 82.6 39.22.75 3.00 170 126 58.4 32.5 Exophthalmos2.60 3.32 78 58 50.3 33.0 Exophthalmos3.13 3.62 91 91 38.7 - Exophthalmos2.81 1.49 83 83 30.0 26.53.40 3.00 141 63 55.1 41.53.05 2.05 118 88 27.2 28.1 Exophthalmos2.78 2.93 66 65 30.0 29.9

2.92 2.74 117 96 50.5 33.80.24 0.63 38 33 18.7 4.8

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidism 2005

TABLE I-(Continued)

Category Subject Sex Age M2 PBI K T I)S

pg/100 ml % liters

Graves' disease- M. F. F 37 1.52 4.0 10.23 7.2hypothyroid J. S. F 40 1.56 2.5 9.76 11.4

J. N. F 70 1.65 2.5 9.49 12.1G. G. M 19 2.35 1.8 9.65 15.3

Mean 42 1.77 2.7 9.78 11.5SD±] 21 0.39 0.9 .32 3.3

Toxic nodular goiter I. C. F 80 1.56 10.9 8.45 11.0I. B. F 50 1.65 8.0 10.50 9.6E. G. F 77 1.57 8.0 13.60 8.0

Mean 69 1.59 9.0 10.85 9.5SD4 1 7 0.04 1.7 2.60 1.5

I iscellaneous P. M. F 40 1.60 11.2 8.67 14.0A. D. M 59 2.04 4.6 12.20 11.8G. H. M 66 1.71 5.2 13.30 9.8R. L. M 59 1.85 2.5 23.20 13.5

Control (n = 13) Mean 47 1.69 5.5 10.66 10.4SDi ; 20 0.20 1.1 1.66 1.4

12.32 + 2.29%o for the eumetabolic Graves' dis-ease group as compared to the control group val-ues of 10.66 ± 1.66%. The hypothyroid Graves'disease group also displayed a surprisingly highfractional disposal rate of 9.78 + 0.32%o. This isin marked contrast with the 6.36 ± 1.55%o valueobserved in the non-Graves' disease hypothyroidgroup. These results again support the conceptthat there may be a direct relationship in these con-

300

200.

100.

HEPATICTHYROXINE 0*CONTENT

jAg I/m2200.

100.

0

GRAVES' DISEASE

n rfl-fl 1 --lfl DI -1-<

CONTROL

n nn~ltfl MRw p 1I .__ I .

2 4 6 8 10 12PROTEIN - BOUND IODINE jig /IOOml

FIGURE 2 Comparison of the hepatic thyroxine contentand protein-bound iodine levels for Graves' disease andnon-Graves' disease subjects.

ditions between hepatic incorporation and thefractional disposal rate of thyroxine. Indeed, acorrelation coefficient of 0.839 was obtained forthese two parameters in the control, hypothyroid,and Graves' disease groups, but, an even highercorrelation coefficient value of 0.925 was ob-served between the hepatic uptake and the net dis-posal rate as illustrated in Fig. 3.

Thyroxine distribution in toxic nodular goiter.In contrast to the Graves' disease subjects, noapparent alteration in the relative distribution ofhepatic vs. extrahepatic thyroxine pools was ap-parent in this group, as shown in Table I and Fig.2 in the non-Graves' disease group. The fractionalturnover rate of thyroxine in these subjects wasalso noted not to be markedly increased. PatientL.W., a subject with treated primary hypothy-roidism, should probably be included in this por-tion of the result section since she appeared to bechronically over-treated with exogenous desic-cated thyroid which resulted in an elevated PBIlevel of 9.4 ug/100 ml as seen in Table I. It isof interest that her compartmental distribution ofthyroxine was quite similar to that found in atoxic nodular goiter group.

Thyroxine distribution in subjects with liverdiseases and low thyroxine-binding globulin lev-

2006 J. T. Nicoloff 'and J. T. Dotbling

_ 1. .. .1- I u

ResinPTDS ETDS HC/m2 EHC/m2 D/m' uptake* Comment

liters liters ;gg I jsg I ;,g/day

2.38 1.28 22 45 19.4 21.5 14 yr posttherapy2.25 1.90 35 81 17.8 9 yr posttherapy3.00 2.05 47 60 17.4 25 yr posttherapy3.50 3.10 23 43 11.3 8 months posttherapy

2.78 2.08 32 57 16.50.58 .76 12 18 3.6

2.27 3.00 83 320 64.9 39.52.56 1.87 50 200 48.9 -2.35 1.99 72 115 55.4 31.0

2.39 2.29 68 212 56.4.15 .62 17 103 8.0

2.95 2.70 0 585 85.0 - Acute hepatitis3.28 2.72 18 103 32.5 - Laennec's cirrhosis2.50 1.93 40 123 39.6 Laennec's cirrhosis3.20 3.20 73 23 42.3 56.6 Low-thyroxine-binding globulin

2.72 2.70 51 122 34.7 31.70.47 .69 7 34 3.6 2.85

NE

so

a

z

x

0-z0

z

0IxI..U

40.'Uz

404A

300k

2001

tool

4k AA

A

AA AA

A AAA

A

'I

^M

AA- BRAVES'OISEASE0- HYPOTHYROID

UNORMALS

loo 200

DISPOSAL RATE pg irn2

FIGURE 3 Comparison of thyroxine disposal rates andhepatic thyroxine content for normal, Graves' disease, andhypothyroid subjects. Shaded area encloses the two stand-ard deviation limits for the control group.

els. Three subjects with liver disease were in-cluded in this present study for the purpose ofdemonstrating that liver disease will cause amarked decrease in the hepatic thyroxine uptake.In subject P.M., with acute infectious hepatitis,

virtually no hepatic uptake of thyroxine could bedemonstrated. As in the hypothyroid subject A.B.,this lack of hepatic uptake was qualitatively veri-fied by external counting over the area of the liver.It is of interest that the disposal rate of 85 Fsg I/m?was in the range that was observed in mildly thy-rotoxic subjects. In mild alcoholic cirrhosis, sub-ject G.H. was found to have a minimal alterationin compartmental distribution, but in subjectA.D., with severely decompensated cirrhosis,there was a marked decrease in hepatic thyroxineuptake. In subject R.L., who has a near absenceof thyroxine-binding globulin on an idiopathicbasis, there was a calculated elevation of 73 MugI/M2 in the hepatic incorporation of thyroxine anda reduction to 23 Mg I/mi2 in his extrahepatic thy-roxin compartment. The thyroxine-binding glob-ulin saturation capacity for this subject was 7 pgT4/100 ml which compares with the normal rangein our laboratory of 15-25 ug T4/100 ml.

Quantitative liver thyroxine uptake employinga scintiscan method. The results of the compart-

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidism 2007

sos .

TABLE I IEstimation of the 3 Hr Hepatic Uptake of Labeled Thyroxine by External Liver Scintiscan Method

Subject Age Sex

C. H. 42 MD. A. 45 MJ. G. 42 MH. H 44 MA. C. 46 MM. F. 30 MR. M. 44 MR. T. 33 MD. F. 44 MD. G. 53 MR. B. 43 FT. T. 46 MP. S. 18 F

MeanSD±

Resin /PBI uptake*

5.5 30.45.0 29.45.3 31.65.8 27.66.0 35.34.5 29.97.1 30.26.0 32.75.0 34.74.6 26.76.1 29.26.2 31.24.9 32.0

to Hepaticuptake Comment

26.7 Peptic ulcer15.8 Peptic ulcer15.5 Chronic pyleonephritis13.8 Arterioscl. heart disease19.5 Psychoneurosis16.1 Peptic ulcer22.2 Chronic pyelonephritis20.0 Normal18.8 Functional bowel disease20.8 Mild diabetes18.6 Normal23.4 Normal25.0 Normal

5.54 30.84 19.710.747 2.48 3.88

Primary hypothyroidism

Toxic Graves' disease

Eumetabolic Graves'

disease

H. W. 55 M 2.5M. S. 70 F 2.0B. B. 80 M 2.0A. D. 70 M 1.5R. L. 46 M 1.5

MeanSDA

W. B. 48 MT. F. 47 MH. M. 32 ML. G. 35 MA. T. 38 MJ. H. 19 MP. S. 48 MJ. C. 59 MC. K. 33 MN. O. 36 MR. E. 42 M

MeanSD4

T. L. 24 M 7.8

L. H. 51 MK.-J. 49 ML. L. 49 FL. H. 65 MC. P. 58 MK. O. 39 MR. L. 45 ML. S. 49 ML. B. 33 FF. V. 44 M

L. A. 68 MT. N. 69 M

MeanSD-

1.90.418

16.116.014.214.214.113.113.012.412.212.110.0

18.821.418.713.220.8

18.63.24

~50.052.448.049.842.548.242.937.240.348.842.0

8.615.410.413.214.8

12.482.91

23.628.838.337.638.827.831.334.025.022.027.4

1 gr of thyroid/day

13.4 45.65 30.421.78 4.84 6.02

6.04.04.55.06.66.65.67.25.94.24.17.0

5.731.28

38.9

30.129.329.130.128.032.427.331.127.231.026.234.8

30.423.45

33.3 Stable 6 months, no

treatment39.0 Stable 1 yr post 'l~I25.1 Stable 1 yr, post 131121.8 Stable yr, post 131131.0 Stable 1 yr, post 131I23.9 Stable 3 yr, post 'l1I24.8 Stable 5 yr, post 13'l24.6 Stable 6 yr, post 131125.3 Stable 6 yr, post 13'I25.5 Stable 6 yr, post 1311

25.7 Stable 9 yr, post 131j25.9 Stable 10 yr, post 131138.3 Stable 15 yr, post 13l1

28.025.57

PBI, protein-bound iodine.* Trisorb, Abbott Laboratories, North Chicago, Ill.

2008 J. T. Nicoloff and J. T. Dowling

Category

Control

TABLE I I-(Continued)

Category

Hypothyroid Graves'disease

Subject

E. H.J. G.R. B.H. H.

MeanSD4

Age Sex PBI

42 M 1.647 M 2.038 M 2.548 M 2.0

Resinuptake*

17.626.921.318.9

%Hepaticuptake

28.421.118.614.3

Comment

9 yr post 13112 yr post 13116 months post 13116 yr post 131I

2.025 21.175 20.580.369 4.11 5.91

R. D. 65 M 12.0J. P. 70 M 10.0

MeanSD

R. L.Q. C.R. B.

37.241.8

11 39.51.41 3.25

59 M 2.5 56.652 M 2.0 41.738 M 5.9 32.1

J. M. 41 M 17.6R. B. 42 M 15.1

41.0

38.2

24.622.8

23.71.27

42.0 Idiopathic low TBG30.8 Idiopathic low TBG

9.4 Severe Laennec'scirrhosis

16.3 Thyrotoxicosis factitia18.3 Thyrotoxicosis factitia

mental analysis of thyroxine distribution revealedthat the hepatic uptake of thyroxine could be sub-stantially altered in a variety of thyroid diseasestates. It appeared desirable to establish a directtechnique for the quantitative measurement of thehepatic incorporation of labeled thyroxine whichwas independent of the indirect methods previouslyemployed. To accomplish this purpose, a secondgroup of subjects were evaluated for their hepaticthyroxine-131I uptakes using a quantitative ex-

ternal scintiscan technique. While patients were,

with few exceptions, different individuals fromthose employed for the previous compartmentalanalysis study, the clinical classification of thesesubjects was the same as previously employed.The compartmental analysis results revealed thatthe hepatic uptake values at 3 hr should be on theaverage 22.5% for control and 16.0% for primaryhypothyroid subjects, while in Graves' diseasesubjects they would be 28.6% for toxic, 25.6%for eumetabolic, and 19.7% for hypothyroid sub-jects. The corresponding values employing the 3hr hepatic scintiscan method were found to be19.7% for control and 12.5% for primary hypo-thyroid subjects, while in Graves' disease sub-jects they were 30.4% for toxic, 28.0%o for eu-

metabolic, and 20.6% for hypothyroid subjects as

listed in Table II. In those subjects, the concomi-tant measurement of hepatic thyroxine uptakevalues and compartmental analysis by the approxi-mation method were performed. Close correlationbetween these methods regarding the 3 hr hepatic

uptake values were observed as recorded in TableIII. No alteration in the 3 hr hepatic thyroxine up-

take percentages were observed in subjects withtoxic nodular goiter and factitial thyrotoxicosis,while an increased hepatic uptake was noted intwo subjects with an idiopathic decrease in serum

thyroxine-binding globulin. A marked decreasewas found in one patient with severe Laennec'scirrhosis.

Mathematical model analysis of thyroxine dis-tribution. In a previous study employing normalsubjects, a hypothetical system was developed forthe compartmental distribution of thyroxine (5).The model system employed was that describedby Berman and Schoenfeld (20), and was basedon a set of differential equations which portrayedthe size and kinetic interplay of a theoretical fourcompartmental model. The restrictions placed on

the model system were the same as previously em-

ployed in normal subjects with the exception thatthe estimation of the 3 hr hepatic uptake value was

TABLE II IComparison of the 3 Hr Hepatic Uptake by the

Compartmentaland Liver Scan Methods

Sub-jects

R. L.C. K.L. H.

Diagnosis

Low TBGToxic Graves'Eumetabolic Graves'

Per cent 3-hr hepaticuptake

Compart-mental Liver

analysis scan

45.8 42.032.0 25.039.0 39.0

Thyroxine DiLtrihutinn in Thyrotoxicosis and Hypothyroidism

Toxic nodular goiter

Miscellaneous

2009

100.x OBSERVED

. CALCULATED 80.

EH

p

ECF

H

20 30HOURS

FIGuRE 4 Computer analysis of the temporal pattern oflabeled thyroxine distribution in the major body fluidand tissue compartments in a subject with primary hy-pothyroidsm. EH, P, ECF, and H represent the ex-trahepatic, plasma, extracellular fluid, and hepatic com-partments, respectively.

obtained by the approximation method rather thanfrom hepatic biopsy data as previously employed(5). Close agreement was observed between theapproximation method estimates and the mathe-matical model system in all thyroid disease statesstudied. Typical examples of this close correlationare graphically demonstrated in Figs. 4 and 5, inwhich the results from a hypothyroid and a toxicGraves' disease subject are portrayed.

The change in rate of deiodination of an equili-brating tracer dose of thyroxine in thyroid diseasestates. The results from both compartmentalanalysis and hepatic thyroxine scintiscan studiesrevealed an increase in hepatic incorporation ofthyroxine in Graves' disease and a decrease in pri-mary hypothyroidism. The question then arose, ifthe hepatic thyroxine uptake is variable, wouldthe contribution of the liver to the total thyroxinedisposal rate also be variable. In other words, ifthe hepatic uptake were elevated, would the hepaticthyroxine disposal rate be elevated and vice versa.Because the hepatic uptake of a tracer dose of thy-roxine is accomplished long before that of the ex-trahepatic tissues (5), any major alteration in thedisposal rate attributable to the liver would bemost apparent during the first few hours after theinjection of the thyroxine tracer. Wehave previ-ously described a method for measurement of therelative deiodination rate for labeled thyroxinethroughout the time course of its equilibration (5).The method employed dual tracers of thyroxine,

DDOSE

.Ix OBSERVED: CALCULATED

II -0HI ECF

O ;tq EH3 10 20 30 40 5O

HOURSFIGURE 5 Computer analysis of the temporal pattern oflabeled thyroxine distribution in the major body fluidand tissue compartments in a subject with thyrotoxicGraves' disease. H, P., ECF, and EH represent hepatic,plasma, extracellular fluid, and extrahepatic compart-ments, respectively.

one of which was in an equilibrated state and theother in the process of achieving equilibration.The relative rates of deiodination of these twotracers were assessed by measuring the ratio ofthe two labeled iodide isotopes as they were ex-creted in timed urine samples after the adminis-tration of the second (unequilibrated) tracer. The

'-a

x so- A_>,< _ ,

0 p

0 col GRAVES' DISEASES 40 ST,0 / * * HYPOTHYROIDISMO a40&. * * CONTROLW A&7 20 F,,

0 10 20 30 40 50 60 70 soHOURSPOST THYROXINE-12I INJECTION

FIGURE 6 The change in the rate of deiodination of la-beled thyroxine during the equilibration period as mea-sured in serial timed urine samples for normals and sub-jects with primary hypothyroidism and Graves' disease.The vertical bar represents the value for one standarderror of the mean for the control group. The open andthe solid squares represent the values for two patientswith thyrotoxic Graves' disease. After equilibration ofthe two tracers has been achieved, the established ratiowas normalized as 100%o.

2010 1. T. Nicoloff and J. T. Dowling

I

-9-

so

2

TABLE IV3-Hr Thyroxine Deiodination Values in Thyroid Disease

3-hrdeiodination

Category Subject Age Sex PBI rate HC/m2

Control

Primary hypothyroidism

Eumetabolic Graves'disease

lThyrotoxicdisease

L. N.A. S.G. H.H. G.M. B.E. A.J. A.J.S.J. G.J. N.

MeanSD±

B. D.L. N.A. P.H. L.S. Y.

MeanSD±

J. L.M. F.V. R.K. J.

MeanSD4

A. P.E. V.L. H.N. 0.A. T,C. K.

MeanSD=

24506640193936464645

41.113.2

4530727856

56.219.6

40455243

455.1

404251354033

40.26.3

FMMMFMMFMF

MMMMF

FFMM

p.g/loo ml

5.77.65.25.03.45.26.05.45.54.6

5.41.1

39.840.046.048.236.947.549.047.550.648.0

45.44.7

0.6 11.90.8 13.9.1.3 28.51.7 22.32.9 37.8

1.5 20.90.9 14.9

3.5 76.54.0 59.74.1 38.77.8 65.5

4.9 60.12.0 15.9

F 12.5 63.0F 9.2 56.3M 8.2 79.0M 8.0 66.0M 11.9 80.0M 17.5 58.5

11.2 67.13.6 10.2

results of such an approach are illustrated in Fig.6 for normal, hypothyroid, and Graves' diseasesubjects. When the two isotopes achieve fullequilibration, the ratio of the iodide tracers in theurine maintains a fixed ratio. The value of thisratio was normalized as 100%o. Values obtainedbefore achieving the ultimate ratio were then ex-

pressed as a percentage. In subjects with Graves'disease, there was an augmentation of the earlydeiodination rate of thyroxine, and in subjectswith primary hypothyroidism a marked reduction

was noted. These findings correlate with the aug-

mented proportion of labeled thyroxine found inthe liver of Graves' disease patients and the di-minished values found in patients with primaryhypothyroidism. The maximum differences in thedeiodination rate appeared to occur during the 3-4hr urine collection period. The differences betweenthe two groups became less apparent as equilibra-tion was approached. For this reason, a compari-son of Graves' disease, normal, and hypothyroidgroups was made employing the 3 hr urinary col-

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidism

485445

506045395245

48.676.16

3452

22

7.87.98

6622

118117

80.846.1

186231200

20623

2011

lection values. The results of the comparison ofthe 3 hr deiodination rates are listed in Table IV.The observed 3-hr value was expressed as a per-centage of the ultimate ratio value. This valuewas then multiplied by the ratio of the 3 hr thy-roxine distribution volume/ultimate thyroxine dis-tribution volume in order to compensate for therelative differences in specific activities of the un-equilibrated and equilibrated thyroxine tracers inthese various clinical disease states. In other words,after 3 hr of equilibration in subjects with Graves'disease, the thyroxine tracer may have achieved75% of its ultimate distribution volume, while in asubject with primary hypothyroidism, this mayonly be 50%. These differences are illustrated inFig. 7. Using this method, the 3-hr deiodinationrate in normal subjects was found to be 45.5 ±4.7% + 1 SD which was in contrast to the pri-mary hypothyroid group which demonstrated a20.9 ± 14.9%o value. Elevated values were ob-served for the eumetabolic (60.1 + 15.9%) andthe thyrotoxic Graves' disease subjects (67.1 +

10.2%). A gross correlation between hepatic thy-roxine incorporation and the 3 hr deiodination ratein these groups was observed (r = 0.73). Al-though the relative 3-hr deiodination rates of thesethree groups were significantly different (P <0.01), there was a wide variability noted withineach group, especially in the Graves' diseasesubjects.

-6.c4)u

0.0

Ix

GRAVESE UTHYROID

HYPOTHYROID

3 10 20 30

Hours Postinjection

FIGURE 7 The relative percentage of the ultimate dis-tribution volume for a thyroxine-"lI tracer was mea-sured during the period of isotonic equilibration in anormal subject and in patients with primary hypothyroid-ism and Graves' disease.

DISCUSSION

The results of these compartmental studies in thevarious thyroid disease states are consistent as tothe direction of alterations in hepatic incorpora-tion of thyroxine. In subjects with Graves' dis-ease, it appeared that an increased proportion oflabeled thyroxine was taken up and deiodinatedby the peripheral tissues, and that this occurredprimarily in the liver. In subjects with primaryhypothyroidism the reverse was observed. Theincrease in hepatic thyroxine uptake found in sub-jects with thyrotoxic Graves' disease was alsoobserved in the eumetabolic and hypothyroid pa-tients with Graves' disease when compared totheir non-Graves' disease counterparts. Resultsof urinary deiodination studies employing dualthyroxine tracers indicated that the labeled thyrox-ine incorporated into the liver promptly becameavailable for deiodination. It might, therefore, beexpected that the amount of thyroxine incorpo-rated into the liver should then be directly pro-portional to the net hormonal disposal rate. Theclose correlation observed between the hepatic in-corporation values and the net hormonal disposalrate in primary hypothyroid and Graves' diseasesubjects would seem to verify this relationship.This is graphically illustrated in Fig. 3. This rela-tionship was not so apparent in subjects with toxicnodular goiter and factitial thyrotoxicosis, whenthe increased tissue thyroxine incorporation wasapproximately equally partitioned between thehepatic and extrahepatic tissue compartments.

The question of why subjects with Graves' dis-ease persist in maintaining a relatively largehepatic thyroxine pool in the eumetabolic or evenhypothyroid state is one of the most puzzling prob-lems which has evolved from this study. Thisfinding would appear to be consistent with theobservations of Ingbar and Freinkel (2), whichindicated a persistence of elevated thyroxine turn-over rates in such subjects. In addition, theseresults would also support the observations ofLennon and coworkers (1) as to the presence ofrapid early thyroxine disappearance rates in pa-tients with Graves' disease. This would be theexpected result since the liver is the major tis-sue site for the early uptake of thyroxine. It isdoubtful that alterations in total serum thyroxine-binding protein properties could explain thesechanges in distribution and disposal since the T:,

2012 J. T. Nicoloff and J. T. Dowling

resin uptake values obtained in the eumetabolicand hypothyroid Graves' disease groups revealedno major alteration to be present. Additionally,restoration of normal thyroxine-binding capacitiesfor thyroxine-binding globulin and thyroxine-binding prealbumin in treated Graves' diseasesubjects has recently been observed by Braverman,Foster, and Ingbar (21). At present, the mostlikely explanation would appear to be that an in-crease in the number of intracellular hepatic thy-roxine-binding sites occurs in subjects withGraves' disease which persists after therapeuticmeasures to lower the serum thyroxine levels tonormal or to below normal ranges. The reversewould also seem to be true in the early treatmentphase of subjects with primary hypothyroidism.The studies on hypothyroid subjects B.W. andL.N., revealed that a rapid elevation of thyroxinelevels to normal concentrations did not result inrestoration of a normal hepatic thyroxine poolsize. These findings support the contention thathepatic thyroxine binding sites were decreased inpatients with primary hypothyroidism and that aprolonged period of eumetabolism was necessaryto restore the hepatic thyroxine pool size to nor-mal as was observed in subjects A.B., G.D., andL.W. The factors which regulate these intracel-lular thyroxine-binding sites are presently un-known.

The investigations of Lennon and coworkers(1), Ingbar and Freinkel (2), and this presentstudy, are relatively consistent in that they de-scribe distinct qualitative alterations in the mannerin which Graves' disease subjects handle labeledthyroxine. Not all investigations have been able toconfirm these qualitative differences. Webster,Britton, Volpe, and Ezrin (22), employing es-sentially the same methodology as Lennon and co-workers (1), found that the acute disappearanceof labeled thyroxine was directly related to thecirculating levels of thyroxine and indirectly tothe concentration of thyroxine-binding globulin,no matter which thyroid disease state was con-sidered. They concluded from their findings thatthere was no unique qualitative difference in themanner in which Graves' disease or primary hy-pothyroid subjects handled thyroxine. There is noapparent explanation for these disparate results.These apparently conflicting results indicate thecomplexity of the factors which affect peripheral

thyroxine metabolism and only serve to point Upthe need for the development of more direct in-vestigative approaches.

The hepatic thyroxine content and hepatic de-iodination rates were not as closely correlated asthe hepatic thyroxine content and thyroxine dis-posal rates. One reason for this finding may bethat the technique employed for the estimation ofhepatic thyroxine deiodination rates was one whichrequired that the equilibrating thyroxine tracer beof the highest purity. Contamination of labeledthyroxine from commercial sources with a va-riety of other labeled thyronines has previouslybeen described (23) and was noted to be presentin some of our shipments. Even in cases where theapparent purity of the delivered isotope allowedthis technique to be employed, it is quite possiblethat minor degrees of contamination were stillpresent and could account for some of the varia-tions observed in the groups studied. Despite allof these shortcomings, there was no difficulty inclearly distinguishing differences between the ma-jor thyroid disease groups.

Our observations of an increase in the size ofthe hepatic thyroxine compartment in subjectswith idiopathic low thyroxine-binding globulin,and a decreased hepatic thyroxine compartment inpatients with hepatic disease, confirms the recentinvestigations of Cavalieri and Searle (10, 24).Although our calculated values differ somewhatfrom theirs, primarily because of the differentmathematical models employed, the qualitative al-terations are similar. The findings that the liverconcentrates thyroxine quite well in subjects withidiopathic low thyroxine-binding globulin, infersthat hepatic tissue-binding sites are not involvedin this disorder.

The compartmental analysis method employedin this study revealed that the extracellular fluiddistribution of thyroxine was on the average equalto that of the plasma compartment in virtually allconditions studied. The early extravascular uptakepattern of thyroxine in the forearm was employedin determining the size of the extracellular fluidthyroxine space (5). The assumption was madethat the other extracellular fluid compartmentsof the body possessed approximately the samekinetic pattern as that observed in the forearm.Also, all extrahepatic tissues were grouped to-gether for purposes of simplifying the analysis.

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidism 2013

It is quite obvious that these various compart-ments of the body are not homogenous and thatthese estimates should be considered only as roughapproximations. Since direct measurements ofthe extracellular fluid and extrahepatic compart-ments were not obtained in this study, the solu-tion obtained from either the approximation orthe mathematical model systems cannot be con-sidered entirely unique. In spite of these limita-tions, these model systems do appear to be con-ceptually useful for the future design of experi-ments which may establish more exactly the kineticcharacteristics of the extracellular fluid and ex-trahepatic thyroxine compartments. Although theprecision of the measurement of the extracellularfluid and extrahepatic tissue thyroxine pools maybe in question, this would not appear to invalidatethe observations of the differences found for theliver pool size in these various disease states.

ADDENDUMSince the submission of this investigation for publica-tion, Musa, Kumar, Ogilvie, and Dowling (25) havepublished in abstract form, that the hepatic space insubjects with thyrotoxic Graves' disease, toxic nodulargoiter, and hypothyroidism is increased, normal, and de-creased, respectively. They have employed the methodof Cavalieri and Searle (10) for estimation of the he-patic space. In addition, they observed no correlation ofthese alterations in hepatic thyroxine space with bindingcapacities of either thyroxine-binding globulin or thy-roxine-binding prealbumin levels.

ACKNOWLEDGMENTS

We wish to express our gratitude to Mr. Dwight W.Warren for his excellent technical assistance and to Mr.Hojat Rostami for assistance with the mathematicalanalysis. The computer analysis was performed at theHealth Science Computing Facility at the UCLA Medi-cal Center, Los Angeles, Calif.

This paper was supported in part by grants AM-05825-02 and 2A-5331 from the National Institute of Arthritisand Metabolic Diseases, National Institutes of Health,Bethesda, Md., and by the Clinical Investigatorship Pro-gram of the Veterans Administration, Los Angeles,Calif., and Army Contract DA 49-193-MO-2420, issuedby the Office of the Surgeon General, Department of theArmy, Washington, D. C.

REFERENCES1. Lennon, E. J., N. H. Engbring, and W. W. Engstrom.

1961. Studies of the rate of disappearance of labeledthyroxine from the intravascular compartment. J.Clin. Invest. 40: 996.

2. Ingbar, S. H., and N. Freinkel. 1958. Studies ofthyroid function and the peripheral metabolism

of I3'-labeled thyroxine in patients with treatedGraves' disease. J. Clin. Invest. 37: 1603.

3. Ingbar, S. H., N. Freinkel, J. T. Dowling, and L. F.Kumagai. 1956. Abnormalities of iodine metabolismin euthyroid relatives of patients with Graves' disease.J. Clin. Invest. 35: 714. (Abstr.)

4. Ingbar, S. H. 1960. Clinical and physiologic iniplica-tions of thyroxine turnover in man. In Clinical Endo-crinology. I. E. B. Astwood, editor. Grune & Stratton,Inc., New York. 91.

5. Nicoloff, J. T., and J. T. Dowling. 1968. Estimationof thyroxine distribution in man. J. Clin Invest. 47:26.

6. Barker, S. B., M. J. Humphrey, and M. H. Soley.1951. The clinical determination of protein-boundiodine. J. Clin. Invest. 30: 55.

7. McCall, M. S., and M. F. Camp. 1961. A simpletechnique for prolonging the storage life of IP1-labeled proteins and polymers. J. Lab. Clin. Med.58: 772.

8. Berman, M. 1960. Application of differential equationsto the study of the thyroid system. In Proceedings ofthe 4th Berkeley Symposium on Mathematical Statis-tics and Probability. Jerzy Neyman, editor. Univer-sity of California, Berkeley. 4: 87.

9. Van Middlesworth, L., J. A. Turner, and A. Lips-comb. 1963. Liver function related to thyroxinemetabolism. J. Nucl. Med. 4:132.

10. Cavalieri, R. R., and G. L. Searle. 1966. The kineticsof distribution between plasma and liver of l"I-labeled L-thyroxine in man: observations of subjectswith normal and decreased serum thyroxine-bindingglobulin. J. Clin. Invest. 45: 939.

11. Blomstedt, B., and L. 0. Plantin. 1965. The extra-thyroidal distribution of 'I thyroxine. Acta Endo-crinol. 48: 536.

12. Rall, J. E., J. Robbins, D. Becker, and R. W. Raw-son. 1953. The metabolism of labeled L-triiodothyro-nine, L-thyroxine, and D-thyroxine. J. Clin. Invest. 32:596. (Abstr.)

13. Pochin, E. E. 1964. Liver concentration of thyroidmetabolites. In Dynamic Clinical Studies with Radio-isotopes. R. M. Kniseley and W. N. Tauxe, editors.Atomic Energy Commission-Technical InformationDocument No. 7678, Oak Ridge. 413.

14. Inada, M., and K. Sterling. 1967. Thyroxine turnoverand transport in Laennec's cirrhosis of the liver.J. Clin. Invest. 46: 1275.

15. Vannotti, A., and T. Beraud. 1959. Functional rela-tionships between the liver, the thyroxine-bindingprotein of serum, and the thyroid. J. Clin. Endocrinol.Metab. 19: 466.

16. Roche, J., and R. Michel. 1960. On the peripheralmetabolism of thyroid hormone. Ann. N. Y. -Acad.Sci. 86: 454.

17. Ford, D. H., K. R. Corey, and J. Gross. 1957. Thelocalization of thyroid hormones in the organsand tissues of the guinea pig: an autoradiographic andchromatographic study. Endocrinology. 61: 426.

2014 J. T. Nicoloff and J. T. Dowling

18. Gorman, C. A., E. V. Flock, C. A. Owen, Jr., andJ. Paris. 1966. Factors affecting exchange of thyroidhormones between liver and blood. Endocrinology.79: 391.

19. Albert, A., and F. R. Keating, Jr. 1952. The role ofthe gastrointestinal tract, including the liver, in themetabolism of radiothyroxine. Endocrinology. 51: 427.

20. Berman, M., and R. Schoenfeld. 1956. Invariants inexperimental data on linear kinetics and the formula-tion of models. J. Appl. Phys. 27: 1361.

21. Braverman, L. E., A. A. Foster, and S. H. Ingbar.1967. Thyroid hormone transport in the serums ofpatients with treated thyrotoxicosis. In Annual Meet-ing of the American Thyroid Association, Inc. AnnArbor. 35. (Abstr.)

22. Webster, B. R., A. Britton, R. Volpe, and C. Ezrin.1967. Further studies on the rate of disappearance of

labeled thyroxine from the intravascular compart-ment of man, with reference to the role of thyroxinebinding proteins. Acta Endocrinol. 55: 497.

23. Volpert, E. M., M. Martinez, and J. H. Oppenheimer.1967. Radioiodinated impurities in commercial prepa-rations of 'I-thyroxine and their effect on the mea-surement of free thyroxine in human serum by equi-librium dialysis. J. Clin. Endocrinol. Metab. 27: 421.

24. Cavalieri, R. R., and G. L.- Searle. 1965. Role of theliver in the distribution of I' thyroxine in man.In The Fifth International Thyroid Conference, Rome.C. Cassano and M. Andreoli, editors. Academic Press,Inc., New York.

25. Musa, B. U., R. S. Kumar, J. T. Ogilvie, and J. T.Dowling. 1968. Hepatic thyroxine kinetics in man instates of altered thyroid function. Clin. Res. 16: 149.(Abstr.)

Thyroxine Distribution in Thyrotoxicosis and Hypothyroidis-m 2015