COMPARISONOF HEMATOPOIESISIN THE FETUS ANDDURINGRECOVERYFROMPERNICIOUS ANEMIA

TOGETHERWITH A CONSIDERATIONOF THE RELATIONSHIP OF FETALHEMATOPOIESIS TO MACROCYTICANEMIA OF PREGNANCY

AND ANEMIA IN INFANTS 1, 2

By M. M. WINTROBEAND H. B. SHUMACKER,JR.3

(From the Department of Medicine and the Surgical Hunterian Laboratory,the Johns Hopkins University, Baltimore)

(Received for publication July 3, 1935)

It is well known that the red corpuscles of thenewborn are larger than those of the normal adult.Little study has been made, however, of the sizeand number of the corpuscles of the fetus. Thepurpose of this communication is to describe cer-tain observations on the size and number of thered cells in the blood of the fetus and newborn ofseveral species of mammal; namely, man, rabbit,pig, rat, cat and dog. In the species examined,it has been found that low erythrocyte counts andlarge red corpuscles are consistently found in thefetus, and that the younger the fetus the lower isthe red cell count and the larger the red corpuscles.As compared with the blood of the adult of thesame species, there is in the fetus what may fordescriptive purposes be spoken of as " anemia "of the macrocytic type. As the fetus develops,the erythrocyte count rises and the mean size ofthe red corpuscles diminishes in a manner whichreminds one of the changes which take place inthe blood of patients with pernicious anemia dur-ing the response to liver therapy.

REVIEW OF LITERATURE

Malassez (1) in 1875 and Cohnstein and Zuntz (2) in1884 observed that the erythrocyte count in the earlystages of development of the fetus (rabbit, dog, sheep) islow, and that it gradually increases. In 1889, the formermeasured the diameters of the red corpuscles (3) of ahuman fetus of four and one-half months and found thecells to be larger than those of adult blood. The excep-tionally large size of the earliest nucleated red corpusclesof the embryos of mammals was noted much earlier thani889, however, for Milne-Edwards (1857) (4) quotesPrevost, Wagner, Gulliver and Bischoff in this connec.-tion.

1 Aided in part by a grant from the Committee onScientific Research of the American Medical Association.

2 Read at the meeting of the American Society forClinical Investigation, Atlantic City, N. J., May 6, 1935.

8 Stephen C. Clark Fellow in Surgery.

Jolly (5, 6) studied the blood of rat fetuses and new-born, and pointed out that in this species the red cellcount even at birth is only about one-fourth of that of theadult whereas in the rabbit and human newborn theerythrocyte count is essentially the same as in the adult.Nicholas and Bosworth (7) observed an increase ofhemoglobin from 30 to 65 per cent in rat fetuses rangingfrom the twelfth day of the fetal period to the newbornstage. Kindred and Corey (8) found increasing erythro-cyte counts in rat fetuses 15.4 to 43.0 mm. in length (six-teenth to twenty-second day). Smith (9) reported di-ameter measurements as well as red cell counts in ratsfrom two days before birth until the adult stage wasreached. She found a gradual increase in the number ofcorpuscles until the twenty-third day of life when theirnumber began to increase rapidly, attaining the adultvalues at about three months of age. Throughout thiswhole period the mean diameter of the red cells, theircolor index and variability, and the proportion of reticulo-cytes, gradually decreased.

Knoll (10, 11) has recorded successively increasing redcell counts in human fetuses one to six months of age.

Zeidberg (12) studied rabbit fetuses and found an in-crease in hemoglobin from 8 to 12 grams during the lastthird of pregnancy (twenty-second to thirty-second day)and an increase in packed red corpuscles from 28 to 42per cent. He noted, moreover, that the proportion ofbasophilic red corpuscles decreased during this periodfrom 28 per cent to 4 per cent. von Deseo (13) seems tohave been the first to measure the mean volume andhemoglobin content of the red corpuscles in fetal blood.In 25 beef fetuses ranging from 22 to 9 months of agehe observed, as the age of the fetuses advanced, an in-crease in the erythrocyte count from 3.74 to 7.80 million,in hemoglobin from 7.65 to 10.83 grams and in volume ofpacked red corpuscles from 34.0 to 42.3 cc., whereas themean corpuscular volume decreased from 90.9 to 54.2 c.iA.,mean corpuscular hemoglobin fell from 20.5 to 14.0 micro-micrograms and mean corpuscular hemoglobin concentra-tion fluctuated between 20.9 and 25.6 per cent.

MATERIAL AND METHODS

Through the courtesy of the Department of Obstetrics,blood was obtained from 12 obviously non-viable humanfetuses removed by hysterectomy. Blood was secured bycardiac puncture immediately following removal of the

837

M. M. WINTROBEAND H. B. SHUMACKER,JR.

fetus. The approximate age of the fetuses was estimatedfrom the crown-rump length and the entire length on thebasis of the charts of Keibel and Mall (14) and Streeter(15). Blood was also obtained, by jugular puncture,from 3 premature infants at the Harriet Lane Home.The values for the blood of newborn infants which areused for Figure 4 represent the average of 31 determina-tions made immediately after birth, and 18 determinationsmade at intervals until the twenty-fifth day of life (16).

Rabbits of blue-black Dutch stock and of mixed lab-oratory stocks were mated in our own laboratory and,since the buck and doe were together for three or fourhours only, the age of the rabbit fetuses studied could beestimated with accuracy. The fetuses were removed byhysterectomy and the blood immediately obtained. When-ever possible, blood was secured by cardiac puncture. Inthe case of very small fetuses the heart was cut open andthe blood was drawn into a capillary pipet as it welledout. In order to obtain sufficient amounts of blood, in anumber of instances the blood of several fetuses from thesame uterus was mixed.

The blood of rat fetuses was obtained in the same wayas that of the rabbit fetuses. The rats were not deliber-

KDUM

6.0

ately mated, however, and the age of the fetuses hadtherefore to be calculated from weight and crown-rumplength on the basis of data published by Donaldson (17).The age of newborn rats was determined from accuratelitter records. The rats were of an inbred laboratorystock on a standard diet (18).

Pig fetuses were obtained in a neighboring abbatoir.The uteruses of pregnant sows were brought to the lab-oratory immediately after the animals had been slaugh-tered and the fetal blood was there collected without delay.In no instances had clotting commenced, and in some ofthe pig fetuses the heart was still beating when the bloodwas withdrawn. Blood was frequently collected fromthe umbilical cord, it having been found that the erythro-cytic content of blood obtained in this way was the sameas that of blood taken from the heart. In the case of thesmaller pig fetuses, blood was taken directly from theheart in the manner already described. The age of thepig fetuses was estimated from their weight and crown-rump length on the basis of Warwick's data (19). Bloodof newborn pigs was obtained by cardiac puncture on ani-mals at the U. S. Research Center, Beltsville, Maryland,through the courtesy of Dr. Hugh McPhee.

360

306270

13.0J 210

160

$Si 120

R.SLC 6.* 60.

6*

4O 48 -KG56 4 2 sq.S3 96 04 12FETAL PERIOD WtY 16 24 3k 40 4NEWBORN

838

1l 24

FIG. 1. ERYTHROCYTECOUNTS (E), MILLIONS PER C.MM.), MIEAN CORPUSCULARVOLUME (U, CUBIC MI-CRONS), MEANDIAMETERIN WETPREPARATIONS(X, MICRONS), PROPORTIONOF NUCLEATEDREDCORPUSCLES(A)AND OF RETICULOCYTES (A) IN THE BLOODOF 98 PIG FETUSES, 22 NEWBORNPIGS, AND 5 ADULT PIGS.

66ADULT

"2

I

I

I

.1

-1

HEMATOPOIESIS IN FETUS AND PERNICIOUS ANEMIA

22 24 2b 2b 3b 2 4 6 8 2iliNEWBORNDAYS I

FIG. 2. ERYTHROCYTECOUNTS ((D, MILLIONS PER C.MM.), MEAN CORPUSCULARVOLUME (*, CUBIC MI-CRONS), MEANDIAMETER IN WETPREPARATIONS (X, MICRONS) AND PROPORTIONOF NUCLEATEDRED CORPUSCLES(A) IN THE BLOODOF 34 RABBIT FETUSES AND 8 NEWBORNRABBITS, COMPAREDWITH AVERAGEVALUES FOR THEADULT RABBIT.

Blood of newborn dogs and cats was obtained by cardiacpuncture on animals born in the laboratory. The age offetuses was estimated from weight and crown-rumplength (20, 21).

The blood was collected either in heparin or without ananticoagulant, it having been found that coagulation isextremely slow and imperfect in the blood of youngfetuses. Hemolyzed specimens were discarded.

Fresh blood preparations and blood smears were made,and two diameters of 25 to 100 unselected red corpusclesin the wet and in the dried, stained films (Wright's stain)were measured by means of a calibrated ocular microm-eter. The proportion of nucleated corpuscles was deter-mined in the stained preparations. Reticulocyte countswere made in wet preparations to which a small quantityof brilliant cresyl blue (1 per cent in normal saline) hadbeen added. Erythrocyte counts (two in each instance),hemoglobin and hematocrit determinations, and calcula-tions of the mean volume and hemoglobin content of thered corpuscles were made as described elsewhere (22, 23).

Two chief sources of error must be kept in mind in in-

terpreting the results of these studies: (1) except in thecase of the rabbit fetuses, the age of fetuses has beenestimated from data for length and weight and is there-fore only approximate; (2) when the fetuses were verysmall, such minute amounts of blood were available thatvalues in the smaller fetuses for hemoglobin, volume ofpacked red cells, and mean volume and hemoglobin con-tent of the red corpuscles cannot be considered as repre-senting the same degree of accuracy as is possible whenstudying the blood of adults (23). Erythrocyte counts,however, may be cons