Assessing Gestational Age From Histology of Fetal Skin: An Autopsy Study of 379 Fetuses

JORG ERSCH, MD, AND THOMAS STALLMACH, MD

Objective: To develop guidelines for recognizing stages of fetal skin development to allow objective assessment of gestational age. Methods: Three hundred seventy-nine human fetuses with known gestational ages of 1230 weeks were examined at autopsy. Skin samples were taken from the abdomen. Two hundred fty samples constituted a reference to establish histologic criteria from the development of skin appendages, hair, and apocrine and eccrine glands. One hundred twenty-nine additional skin samples were evaluated as a test set by the same investigators without knowledge of the condition of pregnancy or gestational age. Results: The assessment of 250 skin samples allowed a description of those features that are suddenly discernible and can be used to determine gestational age as having reached or passed weeks 14, 16, 18, 20, 23, 25, and 29 or 30 (after the last menstrual period). Two examiners assessed these histologic features in a random and masked manner in the 129 skin samples of the test set. A 97% agreement with the actual gestational age of the fetus was achieved. Conclusion: Skin development is a continuous process, but some discrete patterns are strongly related to fetal age, are easy to recognize, and are insensitive to artifacts. In our institution, they have become indispensable in evaluating fetal and neonatal pathology. (Obstet Gynecol 1999;94:7537. 1999 by The American College of Obstetricians and Gynecologists.)

is a sequence of easily recognizable architectural patterns within the skin of the abdomen.13 Identication of these age-related morphologic patterns allows a denition of fetal age and offers background information for detecting congenital skin diseases, such as ectodermal dysplasias, blistering diseases, and ichthyosis.4 6

Materials and Methods

We studied 379 human fetuses at 1230 weeks gestational age. Of these, 250 constituted a reference to establish histologic criteria (Table 1). Gestational age derived from the last menstrual period (LMP) coincided with early ultrasound measurements and developmental indices (especially foot length) at postmortem examination. The causes of abortion or fetal death included intradecidual hemorrhage with abruptio placentae, premature rupture of membranes with chorioamnionitis, and pregnancy termination for localized fetal malformations. At autopsy, skin samples were excised parallel to the longitudinal opening of the abdomen below the umbilicus. Skin from other parts of the body differs in the quantity of skin appendages and the time sequence of developmental events. Tissues were xed in buffered 4% formalin and embedded in parafn. Sections 3 m thick were stained with hematoxylin and eosin. After several weeks, 129 additional skin samples were evaluated as a test set by the same investigators, but without knowledge of the condition of pregnancy or gestational age, using the previously established criteria. Causes of fetal death in this group included termination of pregnancy for chromosomal aberrations such as trisomies 21, 18, and 13 and tetrasomy 9p, as well as severe fetal growth restriction (FGR) that had been observed for several weeks.

When postmortem examination of a fetus is performed, the gestational age is usually given by the clinician. However, in many cases this information cannot be provided or the accuracy of the age given is doubted by clinicians and pathologists because of severe growth retardation or organ hypoplasia. Growth of a fetus is accompanied by developmental changes in all organs. However, only a few maturational events are easy to recognize on routinely prepared tissue slides. Beginning at 14 weeks and continuing until 30 weeks, the development of skin appendages
From the Division of Pediatric and Developmental Pathology, Department of Pathology, University Hospital, Zurich, Switzerland.

Results
Our evaluation used criteria that were easily recognized and were not sensitive to tilting of the specimens or

VOL. 94, NO. 5, PART 1, NOVEMBER 1999

0029-7844/99/$20.00 PII S0029-7844(99)00379-8

753

Table 1. Fetal Skin Samples Examined for Each Gestational Age in the Reference Group
Gestational age (wk) Result No. of skin samples 12/13 5 14/15 20 16 20 17 20 18 20 19 20 20 30 21 15 22 20 23 20 24 15 25 10 26 10 27 5 28 5 29 5 30 10

putrefaction. Embryos up to 9 weeks gestation (after the LMP) are covered by a single layer of epidermal cells. From 9 to 13 weeks, two-layered skin is seen with a second, supercial layer called periderm. From 13 weeks on, an intermediate layer is added and fetal skin is stratied, but still does not show appendages. Starting at 14 weeks, the skin of the abdomen shows budding of cells of the basal layer into the underlying mesenchyme. At 16 weeks, a corresponding proliferation of mesenchymal cells is seen just below the epidermal cell bud. Hair follicles grow rapidly, and second and third generations of buds arise from the basal skin layers. In the 18th week, hair shafts become clearly visible and sebaceous glands are recognizable. Fetal skin at 20 weeks is characterized again by budding of cells from the basal cell layer, resulting in the anlage of eccrine sweat glands. In contrast to the budding hair follicles, cells of the eccrine sweat glands are more eosinophilic and slightly cylindrical, and budding is not accompanied by a corresponding proliferation and condensation of the underlying mesenchymal tissue. At 23 weeks, the eccrine gland ducts have elongated and reached the level of transition of dermis to fatty tissue, and a lumen is hardly visible. At 24 weeks, enlargement and kinking of the duct end occur within the depth of the dermis, with the duct ends beginning to coil at 25 weeks. At 29 or 30 weeks, eccrine sweat glands are coiled to such an extent that multiple transverse sections can be seen within the depth of the dermis (Figure 1). Based on the histologic sections, the stages of the gradual development of epidermis and skin appendages are illustrated to clarify the structures and their nomenclature (Figure 2). Patterns of skin development are very easy to identify as soon as one recognizes the morphologic differences between budding of basal cells leading to hair follicles and budding leading to eccrine glands. The latter process is characterized by slightly different cell morphology within the buds and the lack

of mesenchymal condensation below the buds (Figure 1, middle row left versus right; Figure 2, upper row middle versus left). Thus, six easily recognized patterns can be used as a reference when estimating gestational age (Figure 2). To test the reliability of estimating fetal age from histologic examination of skin, skin samples were read with the examining investigators uninformed as to fetal age and condition. The test set consisted of 129 skin samples, including cases of severe fetal growth disturbance and different types of genetic diseases, excluding primary skin disorders. The investigators were able to assign each sample to a pattern delineated in the gures with 97% accuracy (Table 2).

Discussion
Development of body shape and organ architecture in human fetuses has been completed largely by 12 weeks gestation. However, four organs show major changes in developmental patterns throughout further stages of fetal development. The migration and differentiation of neurons in the brain follow precise timing, but denition of distinctive stages is very difcult from routine examination. The development of the lungs, with branching of the bronchial tree, and development of the kidneys, with the generation of multiple layers of glomeruli, are affected by the intrauterine environment and by functional needs, exemplied by the many causes of lung hypoplasia and the severe differences in kidney development between the donor and the recipient in twintwin transfusion. In contrast, skin development is constant and is not affected by intrauterine malnutrition. Even with chromosomal abnormalities, no changes were seen in the follow-up of developmental stages. The few mistakes in the test did not involve skin samples derived from fetuses with severe FGR or trisomy. Instead, they hap-

Table 2. Skin Samples Examined for Each Gestational Age in the Test Set and Number Correctly Assessed by Two Perinatal Pathologists
Gestational age (wk) Result No. of skin samples Correct assessment by JE/TS 14 17 17/17 16 22 22/21 18 19 19/19 20 24 24/22 23 16 16/16 25 14 14/14 29/30 17 16/17

754 Ersch and Stallmach

Histology of Fetal Skin

Obstetrics & Gynecology

Figure 1. Abdominal skin of human fetuses at different gestational ages. Hematoxylin and eosin, original magnication 50 (bottom row).

125 (top and middle rows),

VOL. 94, NO. 5, PART 1, NOVEMBER 1999

Ersch and Stallmach

Histology of Fetal Skin

755

Figure 2. Details of skin development, nomenclature of constituents, and patterns reached at dened gestational ages in a schematic drawing. b basal layer; e eccrine coil; h hair bud; l lanugo hair; m mesenchyme; mc Merkel cell; p periderm; s sebaceous gland; sw sweat gland bud.

pened in very small skin samples or in those with odd cutting angles. Of course, congenital primary skin diseases have a great impact on developing skin morphology4 6; conditions such as restrictive dermopathy, ectodermal dysplasia, and different forms of epidermolysis bullosa show abnormalities already in the early stages of fetal development. Recognition of the early manifestations of congenital skin diseases is possible only with exact knowledge of normal histology, as provided in this study. It must be kept in mind that the maturation of fetal skin varies somewhat by site. Thus, the described stages of skin development would be 12 weeks ahead of schedule if skin from the neck were examined and 2 weeks behind if skin from the thigh were used. This phenomenon reects the overall development of mammalian embryos, which proceeds from the head pole to the caudal pole. Regardless, registration of skin devel-

opment in our institution has become indispensable for evaluating fetal and neonatal pathology.

References
1. Holbrook KA. The biology of human fetal skin at ages related to prenatal diagnosis. Pediatr Dermatol 1983;1:97111. 2. Muller M, Jasmin JR, Monteil RA, Loubiere R. Embryology of the hair follicle. Early Hum Dev 1991;26:159 66. 3. Urmacher CD. Normal skin. In: Sternberg SS, ed. Histology for pathologists. 2nd ed. Philadelphia: Lippincott-Raven, 1997:25 45. 4. Rodeck CH, Eady RAJ, Gosden CM. Prenatal diagnosis of epidermolysis bullosa lethalis. Lancet 1980;i:949 52. 5. Golbus MS, Sagebiel RW, Filly RA, Gindhart TD, Hall JG. Prenatal diagnosis of congenital bullous ichthyosiform erythroderma (epidermolytic hyperkeratosis) by fetal skin biopsy. N Engl J Med 1980;302:935. 6. Elias S, Mazur M, Sabbagha R, Esterly NB, Simpson JL. Prenatal diagnosis of harlequin ichthyosis. Clin Genet 1980;17:275 80.

756 Ersch and Stallmach

Histology of Fetal Skin

Obstetrics & Gynecology

Address reprint requests to:

Thomas Stallmach, MD Department of Pathology University Hospital Schmelzbergstrasse 12 CH 8091 Zurich Switzerland E-mail: thomas.stallmach@pty.usz.ch

Received December 17, 1998. Received in revised form March 29, 1999. Accepted April 8, 1999.

Copyright 1999 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.

VOL. 94, NO. 5, PART 1, NOVEMBER 1999

Ersch and Stallmach

Histology of Fetal Skin

757