Journal of Clinical Investigation No. · 2014. 1. 29. · Journal of Clinical Investigation Vol. 41, No. 11, 1962 EVALUATION OFVAGINAL SMEAR, SERUMGONADOTROPIN, PROTEIN-BOUND IODINE,

Journal of Clinical InvestigationVol. 41, No. 11, 1962

EVALUATION OF VAGINAL SMEAR, SERUMGONADOTROPIN,PROTEIN-BOUNDIODINE, ANDTHYROXINE-BINDING

AS MEASURESOF PLACENTALADEQUACY*

By J. T. NICOLOFFt R. NICOLOFF,4 AND J. T. DOWLING§

(From the Departments of Medicine, Veterans Administration Center and University of Cali-fornia Medical Center, Los Angeles, and the Department of Obstetrics and

Gynecology, Kaiser Hospital, Hollywood, Calif.)

(Submitted for publication March 8, 1962; accepted July 19, 1962)

During pregnancy the characteristic increasein the thyroxine-binding capacity of the circu-lating globulin, TBG (1-3), is associated with aretarded rate of thyroxine turnover (4) and anincrease in the concentration of hormonal iodine(PBI) in the plasma (1-4). These alterationsin iodine metabolism have not been found inwomen with choriocarcinomas who excrete largequantities of chorionic gonadotropin (5), but dooccur in nonpregnant individuals given large quan-tities of estrogens (6-8). It has been concluded,therefore, that during pregnancy these phenomenaresult from systemic effects of placental estrogens.Moreover, the sera of aborting women often failto show the characteristic increase in serum PBI(9) and in thyroxine-binding by TBG (2).

In the past, attempts have been made to use theconcentration of PBI as an indicator of fetalviability (9, 10); however, as appears to be trueof studies of urinary or plasma chorionic gonado-tropin determinations, pregnandiol and estrogenlevels, and vaginal smears (11-15), the PBI hasnot proved satisfactory for routine use. It ap-peared possible, however, that assessment of TBGmight prove of value as an indicator of placentalfunction. For that reason a technique was de-veloped for rapidly estimating serum thyroxine-binding, and its value in predicting the outcome

* Work supported by grant A-3000 from the NationalInstitute of Arthritis and Metabolic Diseases, U. S.Public Health Service, grant T-154 from the AmericanCancer Society, and a grant-in-aid from the PermanenteFoundation. Presented in part at the 1960 meeting of theEndocrine Society.

tPresent address: Wadsworth General Hospital, VACenter, Los Angeles, California.

$ Present address: Kaiser Hospital, Hollywood, Cali-fornia.

§ Present address: The King County Hospital System,Seattle, Washington. I

of pregnancy was compared with that of the PBI,serum chorionic gonadotropin, and the vaginalsmear.

It was found that serum gonadotropin levels andassessment of the vaginal epithelium were poorprognostic indicators in early pregnancy. On theother hand, the parameter of serum thyroxine-binding, by virtue of its reproducibility and tech-nical ease, appears to allow prediction of the out-come of most threatened abortions.1

MATERIALS ANDMETHODS

Measurement of serum TBG. Since a variety of ion-exchange resins have been found to have a high affinityfor thyroxine (16, 17), previously described methods forassessing serum thyroxine-binding (18-21) were adaptedfor quantitative assessment of TBG by using the anionicexchange resin, Dowex 2, A-G 2 X 7, 200 to 400 mesh,obtained from a commercial supplier in the chloride form.The resin was converted to the maleate form in from4 to 7 days by repeated washing with 0.2 M Tris maleatebuffer (pH 8.6) until no precipitate appeared in the su-pernatant fluid upon addition of a few drops of 0.1 Msilvernitrate. After repeated washing with distilled water andfinally with acetone, the resin was dried in room air ona filter-paper pad in a Buchner funnel.

To establish optimal characteristics for the procedure,the effects of the quantity of resin employed, amount ofserum, concentration of PBI, pH, and time of incubationon the resin uptake of labeled thyroxine were observed.The most useful preparation proved to be 100 mg of dryresin incubated with 0.5 ml of serum which contained0.1 ,tc of 1`31-labeled thyroxine and 0.19 ,ug of stable thy-roxine. The latter was added to reduce variationsdue to differences in endogenous thyroxine content.The resin was weighed into tared counting tubes witha semi-automatic balance, and the labeled serum wasadded directly to the tube. The resin serum mix-

1 The more general term, serum thyroxine-binding,has been used here interchangeably with TBG for con-venience without implying a necessary identity with thy-roxine-binding by whole serum or by specific thyroxine-binding proteins determined by other methods.

1998

TBG IN PATHOLOGICPREGNANCY

ture was incubated at room temperature with a shak-ing rate of 80 strokes per minute. Since it was foundthat pH values ranging from 7.0 to 8.6 had no ap-

preciable effect on the resin uptake of thyroxine fromserum, subsequent experiments did not require adjust-ment of serum pH. Time curves were performed underthe foregoing conditions, which indicated that the maxi-mum difference between the uptake of thyroxine frompregnancy sera and from normal sera occurred after fiveto ten minutes of incubation. Nearly complete uptake ofthe hormone from either pregnancy or normal sera oc-

curred after from 4 to 6 hours. A ten-minute period ofincubation was selected as the optimal time for demon-strating differences in serum thyroxine-binding. Forpurposes of control, resin uptakes of thyroxine frompooled, normal sera and pooled, normal, pregnancy sera

were determined during all studies of experimental sera.

All studies were performed in duplicate. After exactlyten minutes of incubation, 5 ml of distilled water was

promptly added to the tube from a wash bottle and theresin was sedimented by centrifugation at 3,000 rpm for2 to 3 minutes. The supernatant fluid was removed by as-

piration, and the washing procedure was repeated 3 more

times with 5 ml of water. The tube was then countedin a well-type scintillation counter. Thyroxine-bind-ing was arbitrarily expressed as the decimal ratio ob-tained by dividing the net counts per minute accumulatedfrom the control serum by the net counts accumulatedfrom the experimental serum. Thus, the normal valuewas 1.0, increased thyroxine-binding was indicated byvalues greater than 1.0, and reduced binding by valuesless than 1.0. These values are referred to herein as the"resin uptake ratio."

To establish the validity of the resin uptake ratio as a

measure of thyroxine-binding by TBG, replicate studieswere made of the resin uptake and of the electrophoreti-cally determined saturation capacities of TBG in 71 sub-jects who exhibited wide variations in thyroxine-bindingby TBG (Figure 1). The electrophoretic method em-

ployed was as described elsewhere (4, 8, 22, 23). Thisconsisted briefly of the addition of tracer thyroxine toserum, enrichment of samples of the labeled serum withstable thyroxine to six concentrations (10, 40, 120, 230,340, and 450 ,lg per cent),2 and zonal electrophoresis of0.020 ml samples of each portion on filter-paper (What-man no. 3) in Tris-maleate buffer, pH 8.6, ionic strength0.10 at 140 v for 16 hours. The quantities of total ra-

dioactivity associated with TBG and TBPA (thyroxine-binding prealbumin) were then assessed at each concen-

tration of added thyroxine. Saturation capacities of theproteins for thyroxine were determined from these data.In the study group was one euthyroid subject with re-

markably subnormal serum thyroxine-binding and con-

centration of PBI (24), and members of a family in

2 The present concentrations of added thyroxine provedto be more suitable than other suggested concentrations(23) in comparing sera that exhibited an extreme range

of saturation capacity of TBG (4 to 62 lug per cent thy-roxine).

60

50

0

1?

30

hi

10 -0

0.80

na71

0

1.00 1.20RESIN UPTAKE (CONTROL/SUBJECT)

140

FIG. 1. COMPARISONOF THE DETERMINATION OF SE-

RUMTHYROXINE-BINDING BY ZONAL ELECTROPHORETICAND

ION-EXCHANGE RESIN TECHNIQUES. The crosses repre-

sent members of a family, four of whom exhibited idio-pathic increases in thyroxine-binding by TBG. The pointenclosed by a circle represents an isolated instance of re-

duced thyroxine-binding by TBG. Ordinate: saturationcapacity, Aug per cent thyroxine. Abscissa: resin uptakeratios, net counts per minute from control serum/netcounts per minute from experimental serum.

whom representatives from three generations exhibitedidiopathic increases in TBG (25, 26). Many indi-viduals were selected who exhibited abnormally lowthyroxine-binding capacities of TBPA (20, 22), or

moderate hypoalbuminemia. The correlation coefficientbetween these methods was 0.92 (p < 0.001). Alteredbinding by TBPA, or moderate hypoalbuminemia, didnot affect the estimates. The linear regression equationderived from these data was x = 72.224y - 51.3. If we as-

sume continued linearity, the absence of thyroxine-bindingby TBG would thus be represented by a theoretical resinuptake ration of 0.71.

Sera were prepared from blood obtained from 103normally pregnant women in various stages of pregnancy

in order to establish normal ranges for subsequent com-

parisons of experimental data. Experimental sera were

obtained from 78 patients during the pathologic preg-

nancies described in Table I and in the subsequentfigures. These women were grouped according to ac-

cepted clinical criteria without regard to similarities or

differences in the etiology or pathogenesis of their dis-orders.

Other measures of pregnancy status. Quantitative se-

rum gonadotropin levels were assayed by the male frogtechnique (27).3 To assure accuracy and reproducibility,sera were stored in the frozen state and tested in largenumbers when seasonal variations in the frogs were mini-

3 Kindly performed by Ronald Edwards, GregoryWright, and Laura Austin of the Clinical Laboratory ofthe Kaiser Hospital, Los Angeles.

1999

.

J. T. NICOLOFF, R. NICOLOFF, AND J. T. DOWLING

0

o 0,, 0, o

0 0 0 0

.> C4

o0 oCU0

t C;o¢ m ; U o0u¢U m O Q z e0~~~~~~~~~'

bo0~e S^Ch + er -ce S~hO Ch+D\0X~bo

CU CC0 -a n 0o 0~NN N - N

Ov V.00 0> o2CM$c (U

> co* V. Cd4ic.. .

oooo .0 tO be WbeCd to C:

CUctCU co Ul 0 t0 ro c ao w.

cu .0 ci0s0CU0

Z ~0~ 5c '2coC: COT 1-0 &.

0 U-C UO U

No~- (C CC4~ C47

00000r

00 0000 0000O00 X b 00n

0! b00 >.0

-4 0 0 5

V0 C

t- f-4 U

m' 0 C4

t) tC

-4 U-) r

C4

u C CO ° O4°

CC 00CU) C-)

)4 C

m

1or

c

03

CZ

0

I Cd

44

0

'0

.;3lv

_4 so oomm

_; _;

,CCC

0._

. 4.0

CC0,

3

0

=

o

.0

000

CUz

O ,

COa

0 b0.0

a.0

'. H t-

t-

00)00E' C00 --_ - C -- N"

°-. 0

C

CCO ,4. 0 C'q

0 C0 U)

0 -4 C')

-4-1

2000

0

0

an

CUv

, B0

be0

aB

'0100be =

-00-

m

coo

64 OC -O'x __

Cd10*

OCU

C.,

5t

?0

'Cw

0

0.0

C.,

0o

0

U

.zo

00

0.

COr 0

0 0

=be

CU

4d I..

CU

V

-°e E

3>

3dc

>oo.

C)

C0)

Cq

eq

-4

CU

Cd

Cd

r

0

Cd

0

r0

0

.0

>0

C.)

0

.0

00

CU

z

0

be

0

CU

0

.0

>0.0CU-00

0e0CU

.0

.0

*

wa

CU

>0

oQC0-

,0 Zco000.aCUO

C0 .j

1.0mcd


4), 6

0 .

0~~ ~ ~~~~.

.0 V .o 005co 0Ccd 0 0

4) 00 U>>v0 4)0

00 C .0.0 a) 0

a.

c Ei. .0 U) 2..0 ~~~~~~~~~0 cis~~~~~44a.0 M

.0. 0 . 0 4 vw0Vco 00 0

4)4) - 0.04 ) : '-'0 U0 .2-4) 4) .0 abe~ Oto 06 .04

0 0C c C

cn (n~~ ~CZ . ...: z 04 at At

U)U)~~ 0' C~~ U) U) v) 0

2 Z~~~~~~~~

00 ~ ad '0 a.. ..'z.. 'o

.001 .0 .0 cl

A4 "C < !t0

N0CU)

_- _4 " C

r- -W m

01 b1OA4¢

2001

V0

.0.4

00

0.0-cd

5 00

4) Y

c oe

i.n 0

0* t,

-o V

U)0 0'

oo z

Vo 4) .2

0

4)4)c

0oOD . *^0 g. .

G) Vd E 4)

4)

Xo X o X os~coCbo. 3 r.

_ d _0

OU.

co.>a OZ

.0 .0 ~H Cl

.0

~ ~ 0

.0 co 0o

o ao seU

00 C') _ U) C) C4 1 U) _

'0 U) U) C'a U)

X C

_ C' " ui Vo Cr 00 0

0 "-

N0 *, OC C

C')

t- 00 0'

_- _.

rCdW0 3

0)

E.UV 4j

Ucts

0.

(-)

H

9:0a

)aEE

0m

000be2

04)

>a.0

'd a. c

4)

06

o

C)

.e4 t_ I_U 0'- 00 0

-U

- C' - C4

0' 00 C U) U 0'U a.)C4 C) U

_

- C' U LU IC

I

J.T. NICOLOFF, R. NICOLOFF, AND J. T. DOWLING

co 0

0 0 d (U

0 .0 (U 0)0 ~ ) 0

co C O C (U

~~~~ ~~~~aS 0 '. . o

0 Cd(U (U 06 -4

0 U (U M ) )

0 too toto((U ( U (Ug C.) 0

.0 .ok C6 u ta(= S' iO 0

" 4 z: lzMH

oo

08oo o c

bo . ,d

0 0

06

>U o-cr

as S @QZ =

eqA 040oUwmif 0)w ' 0 SO 0eq

-__a_eqU" _leq eq eq

-__

000- N

eq

t-

0 0 00 0eq - -_ U)

m0 b0: 1 lU)

6:" U)04

.0.0be Cd Cd bO eo oo be WbG08(U(U 08Pba ,

N 00 eq

U) -I.0 .0 .0

-4 0 eq

eq -4 U

AMU)n_)emUow NeCmmm - NN(m mm

"i _~ - oo o. l- ol o_ .;_ _ __;

O 0 be 0° b0.: -..z :;z;

*- U - * - *0 In *

(- U) 0) V0

U)

000 ov0 ~ 00.0)..

o0 toqb

v C v VU

z z

.0 0.0.0 be co be be¢ 124 04 P4(20

N N N N N0 0 0 0

- - -

_

-a

.0 .0.0 .0 .0 00 00Wt-

U U) -W U) : D

o o0 eq me 0000

- U - m V0 U)- 0%

-U 0 U

"U)

114eq 0) 0o -

d-q __

Wa E4)*~p 0

eq 00 0

U)-4

eq -

(U)'

-4 U) k(0) -4

V o

m4-

eq V U O - 0 U)O o - - ia V

% 0 U)U)m -I eq eq eq eq eq U eq eq eq eq eq

SC.)

m N 00 (

" m w0 t

co Vo eq4 %Ot)

(4 N " mr

0) 00 0o

eq

N- eWqUo t -) 0 0) 0 --4

Ci0v

la0(U

(.la

000 0 --

0la

0

._0

0-.0(UlaN8U

M

P.0

2002

0EU

.C (U

Cg Cd

E E

..dY

ocn

L-m9

be(U

z

a0(

.0w0

oo4

N0

U)

U)

in

AC.)

000,O.

be

oo0

.bU0

.

0(UdC

0C

00

co0000~

0beco

u

t

I

.4III

4, 4U >4

)

w° -.0 o00

0d4

000 0

~.0 4,C

-.id CXCO 4o o_oeaCa0:C tk be, , *6

_ H

0

.0 bll 0

< P) Z

CS

0 CU e

09 CU 0

U) %n 0_9 _)

09 CUm

S00

U1) mU ~o

U) 00

0) 0

in so I c

u.4 r4


CU 0

4,~ ~ ~ ~ c 0H d4) 0.-0 4,

4, _0Oc 0

0Y 4,)°)0

o0 o 0_ oo0.0 c' 0 o04)4) 0.04, U) ~~~4) 4)0 ~ ~ , ) 4

uouou ,e.. .0 o°

0

0 ;a 4u0 0E

W ) 4) .40 CU0..0 ~ .)d

v.d0. ~010 ~ ~ 0 0

4,-.- 0 0 b1

V cd 00 0 000W.

c)0 40 4) 4,4, -

0.0~~~~~~0

0

Im w

U) 4).04

CU

04) -

4) 4

U00

boo4)inZU).

CU).0

00E

.0

-tke

00E.0

00

4..

w 4,

4)0.-.:

00

;0.0.

0 _O

2003

.00

0 .

i5 v

obe

0

.a .04, 5.1

.4c U

MU 0) 00

0 0)

b00. .0 be0U:: < a4

00)

.0

_9 *0 -

cU CU 0

I" 4C . 04

0 CU 0 0 0

-0 .* . U)

0 C C9 C 00._

-4 CU) CU 09

C9 CU)

N0

E -0. .0

H CU)

U)

CU)00e 0U)0\00CU) 0)In

09 09-4 4

CU) (7

SH

0CU)inN

10

co

.0

4A0

4s

4).0

I.0

P4

0

S

Go

.0 0

.-4)

cd I.£0 cis

CdC

0o~

04

404ico

%)09N -4 W)-4 -O


o CU

0O ._ 0

0) CW.0

co c- I Edwon 0,E'04i

.~

ME0O E-r00 C

0a0~ -00°O

OC __~0

~~ E-oo, U H

gL

'Cb

C

_.

0

0 o:

00

C4 C4 NC

CU

win 0 t 0._ C oC :1 .

_ CU

mU 0

-ot--~o0U) 0'.

.60O.:B0 1Go .,'ItE := c

,.Ebe t94..26.

; cvc. r.

c)O

5 C,4

iin

Go-.dC).

I

.= r0

CX

CC

0 ..

c o

.C 0

._

_ U) V-(

000t- (9t.

00wtObta, A4 a a, a4 a4 Z a-

ct 4t

0 0 0

C4 C4 tC4

0 0-Io t--N

(21 t-_ ES(

'IC

_ N M M

00

.000

al

CU

00

CrU

b03

- 0m)(N

0toU) OC

(-00 _

U' U)C4

I....1

-01-

.00CUw

-Cv

Cd

r.;CU

00

00

_ 00 (NOOeC

0

1-

0._!3

coCU00w

400

0tD

03

b0tH

.0tD

tD

0 OsUito " N NS W)

0 0 tO 0 bC,... ...t_4 _ _; _41

be o b0e .0

0 )a,0

N be0 0

._.0 .0

C

1- 0

( N4

'o o-N'o X _ t"00I-4 -(9(9(91 -1 " N M M . .

-N - - 0

(N

0% 0 -

0

00

00 0 s tsO-_N(sN (Ns

woCm-

0

0'm

N0

D-

e4 t t, "fC( N U) "

U)to4

m

2004

0

.U

0

-

CU

0

w

00

._

be be be

~Z;z

In U) '0 U"000'. ) t-.

.2v

C.)

-.4

0

0U

.2

.00

r. 0 )0,0

0900.

CUI.

~CU

CU.

o; Cd

0)

m

la

0

CU

00

0be

CU

Uc5k

-1

1)t

.1

I

OI11.4

lk


mal. Results were expressed as the greatest dilutionof serum at which a positive result was obtained.

Vaginal epithelial changes were appraised by techniquesdescribed by Benson and Traut (28) for estrogenic,progestational, and abortion effects. These were read as

unknowns by two examiners independently and classifiedinto the following categories: not pregnant, pregnant,possible abortion, abortion, infected, and unreadable. Afinal reading was agreed upon by the examiners, and theevaluation was assigned to the proper patient and date.

Sera were accumulated in the frozen state for subse-quent simultaneous determination of concentrations ofserum PBI with the same reagents used.3

RESULTS

Control group. The resin uptake ratios of the103 sera obtained at various intervals during thecourse of normal pregnancy are depicted in Figure2. The inclosed area encompasses one standarddeviation from the interval mean obtained fromvalues accumulated over 5-week periods. Al-though significant alterations in the resin uptakeratio occurred as early as the fifth week, it was notuntil ten weeks that consistent elevations were

noted. Thereafter there was a gradual increase inthe resin uptake ratio to the time of delivery.

Hz, ht1u4n.1 nhtwtinn Th;Qi rrn 1n nf 19IlaVktifull WUart bUVtE. 1 XS1b gIUU

with a history of primary (three ctive abortions without previous,cies) or secondary (three or n

abortions following the last viahabitual abortions was studied i

1.6

.5-

Az 1.4-

4k 1 3-

k

t 1.2-

1.1 I1-

SD roNON-

PREGNANTCONTROLS 5

Of111 S.D.n= 103

*. S

:SO00

0.

* *0 *

0 0

S.

,~~~~~~~~-

10 15 20 25

WEEKSGESTATI

FIG. 2. STANDARDDEVIATION FROM IT

OF TBG LEVELS DURING NORMAL PRERESIN METHOD. Individual values derivdeterminations performed on sera obta6 to 40 of pregnancy.

1 .2

1. 0

5 10 15 20 25 30 35 40

WEEKS GESTATION

FIG. 3. PRIMARY AND SECONDARYHABITUAL ABORTION

GROUP. The cross indicates the time when the conceptuswas delivered. Of the 12 patients studied, 8 carried tonormal term deliveries; of the 4 who aborted, one didso after loss of normal pregnancy changes, one failed todevelop pregnancy changes, and the remaining two aborteddespite the finding of normal pregnancy changes in thy-roxine-binding. Mechanical factors were suspected inthese cases (induced abortion and submucous fibroids,respectively).

U1 I.-, Wullicilmanner during the course of pregnancy (Figure 3

:)r more consecu- and Table I). Despite a high expectation of fre-viable pregnan- quent abortions in this group, only 4 of the 12

iiore consecutive failed to carry to viable births. Moreover, two ofible pregnancy) the abortions may have resulted from mechanicalin a prospective factors (Patients 3 and 12, Table I, Category I).

Patient 2 (Table I) probably terminated herpregnancy at 16 weeks at the time of uterine bleed-ing. Subsequently, her resin uptake ratio de-creased over the next four weeks and she aborted

:at 20½ weeks. Patient 5 had a similar sequence

* of clinical events except that her serum never re-

*. ': vealed pregnancy changes in resin uptake... ' * In contrast, Patient 3, in whom there was evi-

.d . dence of self-induction, had normal pregnancy

. changes at 16 weeks after an initial low level at 13weeks (at which time she was threatening toabort). At 19 weeks her resin ratio was still

30 35 4O normal for pregnancy, although it had decreased

ONsharply. She aborted at 20 weeks. In a similarmanner, Patient 12, in whom submucous fibroids

NTERVAL AVERAGES were present, also aborted when her resin uptake;GNANcY BY THE

ed from duplicate ratio was at pregnancy levels. Of the eight womenlined from weeks who carried, seven noticed lower abdominal cramp-

ing or bleeding, or both, despite their displaying

2005


* INCOMPETENTOSAFIBROIDSaJPROBABLE INDUCED ABORTION

a OTHERCAUSES

n=19

I-

GOI520 5 30 35+

l'0 IS 210 2.5 30 A 4i0

1.5'

1.4-

Z 1.3-kIQ0. 1.2-

01'

5 10 15 20 25 30

WEEKSGESTATION

WEEKSGESTATION

FIG. 4. INCOMPETENT CERVICAL OS AND OTHER ME-

CHANICAL FACTORS IN PREGNANCY. A cross indicateswhen the conceptus was delivered. Where there is no

cross, delivery occurred shortly after the sample was ob-tained. Five of these 19 women carried their pregnan-

cies to term, and these were patients with incompetentcervical os. Of the remaining women, all but 3 aborteddespite the finding of normal pregnancy changes in serum

thyroxine-binding. Of the 3 who exhibited abnormalthyroxine-binding, one had earlier shown normal preg-

nancy changes.

resin uptake ratios characteristic of normal preg-

nancy.

Incompetent cervical os and other mechanicalfactors in pregnancy. These 19 women exhibiteda variety of abnormal clinical states in which therewas the common denominator of an intra-uterineabnormality which was either present before or

acquired during pregnancy (Figure 4; Table I,

Category II). These consisted of submucousfibroids (Patients 2, 4, 8), abruptio placenta (Pa-tient 5), probable traumatic expulsion of the con-

ceptus (Patient 1), incompetent cervical os (Pa-tients 9 through 19) and probable self-inducedabortions (Patients 3, 6, 7). Five of them, allwith a diagnosis of incompetent os, carried toterm. All but one of the remainder showed nor-

mal increases in thyroxine-binding. This excep-

tion was Patient 14, in whom a circlage of theuterine cervix (Shirodkar procedure) was per-

formed which probably delayed clinical evidence ofan otherwise immediate termination of her preg-

nancy.

Placental insufficiency. These 11 women were

patients in late pregnancy in whomgross pathologyof the placenta, and often, secondary alterationsof the fetus were found (Figure 5; Table I, Cate-

FIG. 5. PLACENTAL INSUFFICIENCY. Delivery of theconceptus occurred shortly after the last sample was ob-tained. Relatively normal values were found, even in thepresence of infarction of two-thirds of the placenta (Pa-tient 8, Table I).

gory III). Hypertension, vascular disease, dia-betes, and preeclampsia constituted the predomi-nant underlying disorders.

Despite as much as two-thirds infarction of theplacenta, pregnancy values of resin uptake were

noted in these women. When the data were com-

pared with control values during this period,however, resin uptake values were statisticallylower than expected (p < 0.05).

Threatened abortions in early pregnancy.

Thirty-six cases of threatened abortion were stud-

~1.0- _ a , +

5 10 15 20 25 30 35 40 45

WEEKSGESTATION

FIG. 6. THREATENEDABORTIONS THAT CARRIED TO A

LIVE FETUS. Cross denotes the time at which the fetuswas delivered. Despite the clinical history of threatenedabortion and in some cases a history of copious bleedingthroughout pregnancy, only one woman is known tohave delivered prematurely. Two women with low ini-tial values for serum thyroxine-binding reassuringly ex-

hibited progressive increases into the normal pregnancyrange at follow-up.

I.4-

.3-

1.27

kIM0..

0)I

n=lI

0*

1.0

35 40 45

2006

I1.51


ied in whom 13 carried to a viable pregnancy

(Figure 6; Table I, Category V), and 23 spon-

taneously aborted from unknown causes (Figure7; Table I, Category IV). In the former group,

normal pregnancy changes in resin uptake ratiowere noted throughout pregnancy in those pa-

tients who were followed. In the latter group,

initial values were in the nonpregnant range or

rapidly became so during follow-up in all but two

instances. Differences between either the initialor the final values of the aborting group and thepregnant control values were highly significant(p < 0.001).

Results of ancillary methods for evaluating thestatus of pregnancy. Assessment of the vaginalepithelium and titers of serum chorionic gonado-tropin poorly predicted the outcome of pregnancy

except in the case of known mechanical causes ofgestational failure or in placental insufficiency(Table II). Low or absent titers correctly pre-

dicted failures in 10 of 12 instances, whereas se-

rum thyroxine-binding was abnormal in only 2 of19 patients studied. On the other hand, gonado-tropin titers predicted failures in 3 of 10 women

hi

(It

1.5-

I.4-

I.3-

1.2-

I. I-

0.9,

n= 23

~~~ ~ ~ ~ ~ ~ ~ ~ ~ :

~~~~~~~~~~~~~~~~~~ . ..........

d-"e-S

10 1'5 20 25 30 35 40

WEEKSGESTATION

FIG. 7. THREATENED ABORTIONS THAT TERMINATED

IN ABORTION. Patient aborted shortly after the last valuewas obtained (see Table I) unless later termination is in-dicated by a dashed line and cross. These latter sixwomen had pregnancies classified as missed abortion.One of these terminated 14 weeks after two consecutivevalues for serum thyroxine-binding were in the non-

pregnant range. Only 2 of these 23 women who had no

known cause for fetal wastage aborted after exhibitingnormal pregnancy changes in serum thyroxine-binding(Patients 5 and 14 of this category, Table I).

TABLE II

Predictive value of vaginal smears, chorionic gonadotropin titers, and serum thyroxine-bindingin the prognosis of pregnancy

Vaginal smear Gonadotropin Resin TBG

No. No. No.correctly Total no. correctly Total no. correctly Total no.

Category Outcome predicted* of studies predicted* of studies predicted* of studies

I. Primary and sec- Aborted 1 2 1 3 3 3ondary habitualabortion Carried 3 5 3 4 9 9

II. Incompetent os Aborted 4 5 8 10 2 16and mechanicalfactors Carried 0 1 2 3 3 3

III. Placental Aborted (or 0 1 2 2 0 3insufficiency stillborn)

Carried 3 6 5 7 8 8IV. Threatened Aborted 4 10 10 13 21 23

abortionV. Threatened Carried 6 10 7 11 13 13

abortion

Total aborted 9 18 21 28 26 45Total carried 12 22 17 25 33 33

Total 21 40(52%) 38 53(72%) 59 78(76%)Total for categoriesI, IV, and V 14 27(52%) 21 31(68%) 46 48(96%)

* Based on values obtained nearest the date of abortion.

2007


who carried their pregnancies and were exhibitingnormal pregnancy changes in serum thyroxine-binding. The over-all correctness of predictionby the vaginal smear was 52 per cent, by serumgonadotropin titers, 75 per cent, and by thyroxine-binding, 76 per cent. Moreover, it will be notedfrom Table I that serial determinations of smearsand gonadotropin titers were quite variable in thesame patient during a period when the status ofpregnancy was clearly indicated by assessment ofserum thyroxine-binding (e.g., Patients 8 and 10,Category V).

More germane to the comparison of these meth-ods to the resin technique were the results obtainedin the case of women who were threatening toabort who had no known causes for abortion andwho either aborted or carried (Table I, CategoriesI, IV, and V). In these groups, comprising 48patients, vaginal smears were assessed in 27, andgonadotropin titers were measured in 31. A cor-rect prediction was made by the vaginal smear inonly half the cases and by the gonadotropin titersin about two-thirds of the cases. The resin tech-nique correctly predicted the outcome in 46 of 48pregnancies, or 96 per cent.

The concentration of serum PBJ, as noted byothers, correlated moderately well with clinicalstatus (9, 10). After 10 weeks of pregnancy,concentrations of serum PBI in women whothreatened to abort and shortly did so were some-what lower than concentrations in women whothreatened to abort, but carried to delivery of vi-able infants (p < 0.02). When the same groupswere compared on the basis of their resin uptakeratios, however, the difference proved to be ofmuch greater significance (p < 0.001).

DISCUSSION

Measurement of the accumulation of labeledthyroxine from serum by an anionic ion-exchangeresin is a simple and rapid technique for apprais-ing thyroxine-binding by the serum thyroxine-binding globulin (TBG) (18-21). When ap-plied to a specific clinical problem, it proved to bea practical method for predicting the eventualoutcome of pregnancy in women who were threat-ening to abort. In the present study it was a morereliable indicator of the prognosis of threatenedabortion than previously described techniques, in-

cluding the evaluation of the vaginal smear andmeasurement of serum gonadotropin titers. Inlate pregnancy, however, the finding of low or ab-sent serum gonadotropin titers despite normalpregnancy changes in TBG appears to carry anominous prognosis for pregnancy. Since the con-centration of serum PBI is a function of severalvariables (29-33), it correlated only moderatelywell with the status of pregnancy (9, 10).

Beyond its clinical application, measurement ofthe resin uptake of serum thyroxine clarifies cer-tain physiologic events that occur in pathologicpregnancies. Previous work has shown that boththe concentration of PBI and thyroxine-bindingby TBG return slowly to normal after pregnancyor after discontinuing the administration of largeamounts of estrogen, and achieve nonpregnant val-ues only after from 3 to 6 weeks (2, 6). Thisgradual decrease was observed in one woman inthe habitual abortion group .(Patient 2, Table I,Category I) and in two patients in the threatenedabortion group (Patients 11 and 18, Table I,Category IV). Thus, it is likely that placentalfailure to elaborate normal amounts of estrogenwould not be reflected by a decreased resin uptakeratio for at least several days. Accordingly, mostwomen who were threatening to abort and wholater did so may have incurred placental failure aslong as several weeks before the event of abortion.This interpretation agrees with the conclusion ofa number of workers who have found evidencethat fetal death occurs approximately six weeksprior to abortion (34-36). On the other hand, anumber of women with known mechanical causesof gestational failure, such as incompetent cervicalos, submucous leiomyomata, abruptio placenta,and induced abortion may exhibit initially ade-quate elaboration of estrogen, as judged by normalpregnancy levels of serum thyroxine-binding.This finding indicates that death of the conceptusis a late event in these types of abortion. More-over, it follows that the finding of normal preg-nancy levels of TBG in patients who soon abortsuggests search for such mechanical factors eventhough they are not clinically evident. If thesepatients are excluded from the group of womenwho are threatening to abort, then failure to findnormal pregnancy TBG changes or loss of thesechanges confirms gestational failure and inevitableabortion regardless of the patient's apparent clini-

2008


cal course. Alternatively, in the absence of theforegoing disorders, women threatening to abort,regardless of their clinical course, who continu-ously exhibit normal pregnancy changes in TBGare likely to continue pregnancy at least to theo-retical fetal viability and in most cases to term.

In no way should it be construed that this workindicates that placental failure to elaborate normalamounts of estrogen or abnormalities of thyroidfunction are causes of gestational failure. Thesedata suggest that abortion results from prior deathof the conceptus and that in most, if not all casesof threatened abortion, therapy with exogenousestrogen cannot be of benefit (34, 35). It is alsoapparent that estrogenic treatment would vitiateassessment of serum TBG as a method for fol-lowing the course of pregnancy.

Many of the present women, particularly thosewith a diagnosis of habitual abortion, receivedrelaxin or progestational steroids throughout thestudy-norethindrone (Norlutin) and hydroxy-progesterone caproate (Delalutin). These prepa-rations appeared to have no effect on serum thy-roxine-binding and therefore did not vitiate thepredictive value of the assessment of serum TBGby the resin method. It is parenthetically note-worthy that despite previous suggestions (37,38), the present data support the view that thearbitrary diagnosis of habitual abortion is arti-ficial and misleading (39). In the prospectivestudy of 12 women with this diagnosis, only 4failed to carry to term and mechanical causes ofabortion were suspected in two of these . Notablealso was the fact that these women did respondto the stimulus of pregnancy with normal changesin thyroxine transport.

It is likely that techniques for assessing serumthyroxine-binding can be considerably improved.Such an attempt, with the use of immunochemicaltechniques, is currently under study. Measure-ment of changes in other circulating transport pro-teins may also be of value in following the statusof pregnancy. It has been shown that plasmacortisol levels, the cortisol-binding protein (trans-cortin), ceruloplasmin, iron-binding globulin, andhaptoglobin are all increased during the adminis-tration of large amounts of estrogen or duringnormal pregnancy (35-44). It is likely that acombined assessment of one or more of thesetransport proteins and of other parameters, such

as gonadotropin titers, will allow improved pre-diction of the outcome of disturbed pregnancies.Moreover, the profile of results obtained may di-rect diagnostic attention toward specific pathogene-ses of gestational failure.

SUMMARY

Studies of serum thyroxine-binding employingzonal electrophoresis and a technically simple ion-exchange resin method correlated favorably andthereby allowed clinical evaluation of serum thy-roxine-binding as an indicator of fetal viability.

Comparison of the method with assessments ofprotein-bound iodine, serum chorionic gonado-tropin, and vaginal smears indicated its greaterprecision in predicting the outcome of threatenedabortions.

Studies of 103 normally pregnant women and78 patients during pathologic pregnancies sug-gested that normal pregnancy changes, in the ab-sence of mechanical disturbances to gestation, wereprognostically favorable in women who werethreatening to abort.

These data confirm the view that fetal death con-siderably antidates clinical abortion in womenwhose pregnancies terminate after threatening toabort.

ACKNOWLEDGMENT

The authors wish to express their gratitude for thetechnical assistance of Mrs. Dzidra Razevska.

ADDENDUM

The similar findings of Brody and Carlstr6m, derivedfrom studies of pathologic pregnancies by an immunologicassay for human chorionic gonadotropin, have been re-cently published (45).

REFERENCES

1. Dowling, J. T., Freinkel, N., and Ingbar, S. H.Thyroxine-binding by sera of pregnant women.J. clin. Endocr. 1956, 16, 280.

2. Dowling, J. T., Freinkel, N., and Ingbar, S. H.Thyroxine-binding by sera of pregnant women,new-born infants, and women with spontaneousabortion. J. clin. Invest. 1956, 35, 1263.

3. Robbins, J., and Nelson, J. H. Thyroxine-bindingby serum protein in pregnancy and in the new-born. J. clin. Invest. 1958, 37, 153.

4. Dowling, J. T., Hutchinson, D. L., Hindle, W. R.,and Kleeman, C. R. Effects of pregnancy on io-dine metabolism in the primate. J. clin. Endocr.1961, 21, 779,

2009


5. Dowling, J. T., Freinkel, N., and Ingbar, S. H.Iodine metabolism in hydatidiform mole and chorio-carcinoma. J. clin. Endocr. 1960, 20, 1.

6. Dowling, J. T., Freinkel, N., and Ingbar, S. H.Effect of diethylstilbestrol on the binding of thy-roxine in serum. J. clin. Endocr. 1956, 16, 1491.

7. Engbring, N. H., and Engstrom, W. W. Effects ofestrogen and testosterone on circulating thyroidhormone. J. clin. Endocr. 1959, 19, 783.

8. Dowling, J. T., Freinkel, N., and Ingbar, S. H.The effect of estrogens upon the peripheral metabo-lism of thyroxine. J. clin. Invest. 1960, 39, 1119.

9. Man, E. B., Heineman, M., Johnson, C. E., Leary,D. C., and Peters, J. P. The precipitable iodine ofserum in normal pregnancy and its relation to abor-tions. J. clin. Invest. 1951, 30, 137.

10. Singh, B. P., and Morton, D. G. Blood protein-bound iodine determinations as a measure ofthyroid function in normal pregnancy and threat-ened abortion. Amer. J. Obstet. Gynec. 1956, 72,607.

11. Hon, E. H., and Morris, J. M. Gonadotropin titersin disturbed pregnancies. Surg. Gynec. Obstet.1955, 101, 59.

12. Delfs, E., and Jones, G. E. S. Endocrine patterns inabortion. Obstet. gynec. Surv. 1948, 3, 680.

13. Russell, C. S., Paine, C. G., Coyle, M. G., and Dew-hurst, C. S. Pregnanediol excretion in normal andabnormal pregnancy. J. Obstet. Gynaec. Brit.Emp. 1957, 64, 649.

14. Hon, E. H., and Morris, J. M. Evaluation of maleBufo americanus and Rana pipiens for pregnancytesting. Amer. J. Obstet. Gynec. 1956, 71, 331.

15. Zondek, B. Hormonal diagnosis of placental dys-function leading to fetal death. Clin. Obstet.Gynec. 1960, 3, 1083.

16. Scott, K. G., and Reilly, W. A. Use of anionic ex-change resin for the determination of protein-bound131 in human plasma. Metabolism 1954, 3, 506.

17. Blanquet, P., Dunn, R. W., and Tobias, C. A. Iso-lation and quantitative estimation of iodide, thyrox-ine, and substances related to thyroxine by meansof an anion exchange resin. Arch. Biochem. 1955,58, 502.

18. Mitchell, M. L. Resin uptake of radiothyroxine insera from non-pregnant and pregnant women. J.clin. Endocr. 1958, 18, 1437.

19. Sterling, K. ,and Tabachnick, M. Resin uptake ofI"13-triiodothyronine as a test of thyroid function.J. clin. Endocr. 1961, 21, 456.

20. Nicoloff, R. L., Nicoloff, J. T., and Dowling, J. T.Evaluation of vaginal smear, serum gonadotropin,PBI and thyroxine-binding as measures of pla-cental adequacy. Clin. Res. 1960, 8, 111.

21. Braverman, L., and Ingbar, S. H. A rapid methodfor individually assessing hormonal binding by thy-roxine-binding proteins. Clin. Res. 1961, 9, 176.

22. Ingbar, S. H. Pre-albumin: a thyroxine-binding

protein of human plasma. Endocrinology 1958,63, 256.

23. Ingbar, S. H. Clinical and physiological observa-tions in a patient with an idiopathic decrease in thethyroxine-binding globulin of plasma. J. clin. In-vest. 1961, 40, 2053.

24. Nicoloff, J. T., and Dowling, J. T. Idiopathic de-ficiency of serum thyroxine-binding in a eumetabo-lic adult. Clin. Res. 1961, 9, 74.

25. Beierwaltes, W. H., and Robbins, J. Familial in-crease in the thyroxine-binding sites in serum alphaglobulin. J. clin. Invest. 1959, 38, 1683.

26. Florsheim, W. H., Dowling, J. T., Meister, L., andBodfish, R. E. Familial elevation of serum thy-roxine-binding globulins. J. clin. Endocr. 1962, 22,735.

27. Hodgson, J. E. Office use of the frog test for preg-nancy. J. Amer. med. Ass. 1953, 153, 271.

28. Benson, R. C., and Traut, H. F. The vaginal smearas a diagnostic and prognostic aid in abortion.J. clin. Endocr. 1950, 10, 675.

29. Berson, S. A., and Yalow, R. S. Quantitative as-pects of iodine metabolism. The exchangeableorganic iodine pool, and the rates of thyroidal se-cretion, peripheral degradation and fecal excre-tion of endogenously synthesized organically boundiodine. J. clin. Invest. 1954, 33, 1533.

30. Ingbar, S. H., and Freinkel, N. Simultaneous esti-mation of rates of thyroxine degradation and thy-roid hormone synthesis. J. clin. Invest. 1955, 34,808.

31. Sterling, K., and Chodos, R. B. Radiothyroxineturnover studies in myxedema, thyrotoxicosis, andhypermetabolism without endocrine disease. J.clin. Invest. 1956, 35, 806.

32. Robbins, J., and Rall, J. E. The interaction of thy-roid hormones and protein in biological fluids.Recent Progr. Hormone Res. 1957, 13, 161.

33. Ingbar, S. H., and Freinkel, N. Thyroid hormonesin Hormones in Human Plasma. Nature andTransport, H. N. Antoniades, Ed. Boston, Little,Brown, 1960, p. 515.

34. Simons, J. H. Statistical analysis of one thousandabortions. Amer. J. Obstet. Gynec. 1939, 37, 840.

35. Hertig, A. T., and Livingstone, R. G. Spontaneous,threatened and habitual abortions: their patho-genesis and treatment. New Engl. J. Med. 1944,230, 797.

36. Hamblen, E. C. Some clinical observations on theendocrinology of abortion. Amer. J. Obstet. Gynec.1941, 41, 664.

37. Malpas, P. A study of abortion sequences. J. Ob-stet. Gynaec. Brit. Emp. 1938, 45, 932.

38. Jones, G. E. S., and Delfs, E. Endocrine patternsin term pregnancies following abortion. J. Amer.med. Ass. 1951, 146, 1212.

39. Speert, H. Pregnancy prognosis following repeatedabortion. Amer. J. Obstet. Gynec. 1954, 68, 665.

2010


40. Peterson, R. E. The miscible pool and turnover rateof adrenocortical steroids in man. Recent Progr.Hormone Res. 1959, 15, 231.

41. Chen, P. E., and Mill, I. H. in discussion of Peterson,R. E. The miscible pool and turnover rate ofadrenocortical steroids in man. Recent Progr.Hormone Res. 1959, 15, 231.

42. Daughaday, W. H., Kozak, I., and Biederman, 0.

Corticosteroid binding globulin activity in bodyfluids and in fetal and maternal plasma (abstract).J. clin. Invest. 1959, 38, 998.

43. Russ, E. M., and Raymunt, J. Influence of estro-gens on total serum copper and caeruloplasmin.Proc. Soc. exp. Biol. (N. Y.) 1956, 92, 465.

44. Fay, J., Cartwright, G. E., and Wintrobe, M. M.Studies on free erythrocyte protoporphyrin, serum

iron, serum iron-binding capacity and plasma cop-

per during normal pregnancy. J. clin. Invest. 1949,28, 487.

45. Brody, S., and Carlstrom, G. Immuno-assay of hu-man chorionic gonadotropin in normal and patho-logic pregnancy. J. clin. Endocr. 1962, 22, 564.

2011

Documents

Journal of Clinical Investigation No. · 2014. 1. 29. · Journal of Clinical Investigation Vol. 41, No. 11, 1962 EVALUATION OFVAGINAL SMEAR, SERUMGONADOTROPIN, PROTEIN-BOUND IODINE,