Amnion nodosum is commonly regarded as a placental hallmark of severe and prolonged oligohydramnios. It consists of nodules of amorphous granular material present on the surface of the amnion. We reviewed all 45 cases of amnion nodosum from our placental database from 1994 through 2003 (study group). The control group consisted of 45 cases from the same database matched for gestational age but without amnion nodosum. Oligohydramnios, multiple pregnancy, perinatal mortality, macerated stillbirths, and chronic twin-twin transfusion were the most common clinical features encountered in the study group. Luminal vascular abnormalities of chorionic villi, fibrosis of chorionic villi, and placental edema were the placental features that were more frequently present in the study group.

Amnion nodosum (AN) is a placental lesion consisting of numerous small nodules on the amnion of the chorionic disc, placental membranes, or the umbilical cord. Prior to 1950 when Landing1 first used the term amnion nodosum, the characteristic plaques containing squamous cells on the fetal surface of the amnion were referred to as amniotic nodules. Grossly, the lesion consists of multiple, firm, circumscribed, round to ovoid, raised, shiny, yellow nodules, ranging from 1 to 5 mm in diameter visible on the amniotic surface1–3 (Figure 1). Microscopically, they are composed of varying proportions of squamous cells (occasionally keratinized) embedded in degenerative amorphous acidophilic debris1,4,5 (Figure 2). These lesions either are found on the amniotic surface or occasionally embedded in amniotic mesoderm or project through the amnion into the spongy layer cleft between it and the chorion.5 The amniotic epithelium may be present either as a complete or as an incomplete cell layer between the basal side of the nodule and the basement membrane. The epithelium is identical to that of the adjacent uninvolved amnion between nodules and occasionally appears continuous with it, often showing hyperplastic changes.6 

Figure 1.

Gross appearance of the amniotic surface of the fetal membranes showing multiple, firm, circumscribed, round to ovoid, yellow nodules, typical of amnion nodosum. Figure 2. Microscopic appearance of amnion nodosum nodules showing varying proportions of squamous cells embedded in degenerative amorphous acidophilic debris, attached to the amniotic epithelium (hematoxylin-eosin, original magnification ×100). Figure 3. Firm white subchorionic nodules of subchorionic fibrin deposition. Figure 4. Squamous metaplasia of amnion. A, Irregular white hydrophobic plaques. B, Keratinizing squamous epithelium on the chorionic plate of squamous metaplasia (hematoxylin-eosin, original magnification ×100). Figure 5. Microscopic appearance of chorion nodosum showing flat nodules containing squamous debris buried in the chorionic mesenchyme (hematoxylin-eosin, original magnification ×100).

Figure 1.

Gross appearance of the amniotic surface of the fetal membranes showing multiple, firm, circumscribed, round to ovoid, yellow nodules, typical of amnion nodosum. Figure 2. Microscopic appearance of amnion nodosum nodules showing varying proportions of squamous cells embedded in degenerative amorphous acidophilic debris, attached to the amniotic epithelium (hematoxylin-eosin, original magnification ×100). Figure 3. Firm white subchorionic nodules of subchorionic fibrin deposition. Figure 4. Squamous metaplasia of amnion. A, Irregular white hydrophobic plaques. B, Keratinizing squamous epithelium on the chorionic plate of squamous metaplasia (hematoxylin-eosin, original magnification ×100). Figure 5. Microscopic appearance of chorion nodosum showing flat nodules containing squamous debris buried in the chorionic mesenchyme (hematoxylin-eosin, original magnification ×100).

Close modal

Ultrastructural analysis shows that the nodules are composed mostly of closely packed bundles of fibrillary material of high electron density and cellular elements of various kinds, irregularly dispersed within the fibrillary mass. The basement membrane of the amnion epithelium is usually present under the nodules and sometimes appears multilaminated.6 This indicates that the process is largely superficial and the underlying stroma does not participate in the formation of the lesion. The microscopic and ultrastructural features lend credence to the belief that AN represents deposits of cellular elements from the fetal skin accumulating and organizing on the surface of the amniotic epithelium and undergoing secondary degenerative changes, with subsequent invasion of the squamous cell mass by connective tissue.

Two gross lesions commonly confused with AN are subchorionic fibrin deposits (Figure 3) and squamous metaplasia of the amnion (Figure 4, A and B).7 

The pathogenesis and mechanism of formation of the nodules in AN are largely unknown and have been the subject of controversy in the past, but several hypotheses have been formulated. Simpson2 suggested that AN was caused by hemorrhagic effusion from which the blood pigment goes in time, leaving sacs of serum with contracted coagula. That was in part because the clinical pictures preceding the findings were consistent with abortion or antepartum hemorrhage. The nodules were also previously suggested to be the result of amniotic metaplasia or transplantation of fetal epithelial cells. It is possible that some of the cells accumulated in the nodules are of amniotic origin, but most of the cellular elements embedded in the nodular masses are fetal epidermal cells in various stages of preservation and maturation.6,8 Another possible pathogenetic mechanism that has been suggested is inflammation9; however, there is no associated inflammatory or other tissue reaction.10 

Amnion nodosum does not represent degenerate areas of squamous metaplasia, as it consists of amorphous acidophilic material intermingled with squamous elements of varying age under the amniotic epithelium.4 It has been observed in the membranes of 9- and 16-week-old fetuses, thereby supporting the hypotheses that the amniotic lesions seen in late pregnancy are the end result of abnormal processes that begin early in development rather than recent events.11 One must be careful with diagnosis of AN in cases of prolonged intrauterine fetal demise as amniotic debris attached to the amnion as part of the maceration and autolysis process may simulate it. It is possible however to distinguish between keratin from fetal skin (cytokeratin 14 positive and cytokeratin 13 negative) and amnion (cytokeratin 13 positive and cytokeratin 14 negative).12 

It is generally believed that AN is associated with conditions that lead to significant prolonged oligohydramnios. The association with marked oligohydramnios was first reported in 1912, when it was thought that perforation of the amnion by fetal hair had resulted in proliferation of the epithelium forming the nodules.13 It is found in placentas from fetuses with renal agenesis, following premature rupture of membranes, in the donor twin of the twin transfusion syndrome, in diamniotic acardiac twins, and in sirenomelia.10 Relationship between AN and oligohydramnios have been explained in 2 ways: (1) squames from the hyperconcentrated amniotic fluid might adhere to the surface of the amnion, producing secondary degeneration of the amniotic epithelium, and (2) fetal movement could erode the amniotic epithelium, leading to the incorporation of fetal squames and further proliferation of the exposed amniotic mesoderm.4 Available evidence suggests that reduced liquor in oligohydramnios allows the fetus to come into direct contact with the amnion, thereby leading to transfer of fetal squames to the amnion by a detritic mechanism.7 

As our preliminary results suggested that AN is an important lesion associated with extremely high perinatal mortality and morbidity,14 the review of existing literature showed that the most recent original reports on AN date back decades ago. At the same time, even recent textbooks on placental pathology quote only well-known data on morphology and associations with oligohydramnios and predisposing conditions thereto. No data on incidence of AN or its association with other placental features are quoted, however. Therefore, we decided to devote more attention to this lesion by analyzing our 10-year experience in this case-controlled study.

Forty-five cases of AN were retrieved (study group [SG]) from our placenta and autopsy database at the University of Cincinnati Medical Center during a period of 10 years (1994–2003). Cases of chorion nodosum were excluded. Patients' medical records and placental reports were reviewed for clinicopathologic correlation. The hematoxylin-eosin slides for all the cases were also reviewed. In a similar way, 45 cases of a control group (CG) without AN were matched for gestational age but otherwise randomly selected from the placental database. The clinical and placental (gross and microscopic) features were statistically compared using the 1-tail Fisher probability test or analysis of variance when appropriate. The difference was considered significant when P ≤ .05. Amnion nodosum was diagnosed grossly by finding multiple, firm, circumscribed, round to ovoid, raised, shiny, yellow nodules, ranging from 1 to 5 mm in diameter on the amniotic surface of the membranes (Figure 1). Microscopically, these nodules were located on the surface of the amnion and occasionally embedded in the amniotic mesoderm and were made up of varying proportions of squamous cells embedded in degenerative amorphous acidophilic debris (Figure 2).

Forty-five cases of AN were found among 4342 placentas analyzed (1%). The clinicopathologic and placental features are summarized in Tables 1 and 2, respectively. The mean gestational age in both groups was 28.6 weeks. The mean fetal weight was 1349 g in the SG, whereas it was 1534 g in the CG. In the SG, oligohydramnios was diagnosed clinically in 10 (22.2%) cases, whereas it was diagnosed in 2 (4.4%) cases in the CG. Congenital malformations were found in 12 (26.7%) cases in the SG, and none in the CG. In the SG, associated multiple gestations, perinatal mortality, prolonged preterm premature rupture of membranes, and chronic twin-twin transfusion syndrome (chronic TTTS) were reported in 13 (28.9%), 16 (35.6%), 11 (24.4%), and 4 (8.9%) cases, respectively. In the CG, multiple gestations, perinatal mortality, and prolonged preterm premature rupture of membranes were present in 12 (26.7%), 10 (22.2%), and 8 (17.8%) cases, respectively. There was no reported case of chronic TTTS. The mean placental weight was 314 g in the SG, whereas it was 325 g in the CG. In the SG, the most common placental findings included acute chorioamnionitis (14 cases, 31.1%), fibrosis of chorionic villi (9 cases, 20%), chorangiosis (6 cases, 13.3%), meconium staining (5 cases, 11.1%), villous edema (5 cases, 11.1%), decidual arteriolopathy (4 cases, 8.9%), and chronic villitis (3 cases, 6.7%). The most common placental findings in the CG included acute chorioamnionitis (22 cases, 48.9%), chorangiosis (10 cases, 22.2%), decidual arteriolopathy (7 cases, 15.6%), chronic villitis (4 cases, 8.9%), meconium staining (4 cases, 8.9%), and fibrosis of chorionic villi (2 cases, 4.4%). There was no associated villous edema (Table 2).

Table 1. 

Clinicopathologic Correlation of Amnion Nodosum

Clinicopathologic Correlation of Amnion Nodosum
Clinicopathologic Correlation of Amnion Nodosum
Table 2. 

Placental Correlations of Amnion Nodosum

Placental Correlations of Amnion Nodosum
Placental Correlations of Amnion Nodosum

Four cases of chorion nodosum associated with the severe early amnion rupture sequence were identified. In these cases, the mean gestational age was 22 weeks, the mean fetal weight was 499 g, and the mean placental weight was 165 g.15 These were not included in the study because they did not meet the diagnostic criteria for AN.

Amnion nodosum is a lesion that merits more attention and reevaluation as our analysis has shown that, probably unlike any other single placental feature, it is associated with an extremely high risk of fetal and perinatal mortality (35%). This outcome is mainly because of lethal congenital malformations, predominantly of the genitourinary system and oligohydramnios-associated fetal pulmonary hypoplasia. The perinatal mortality in the CG was also very high, hence absence of statistically significant differences, but this was because of other causes, predominantly the amniotic sac infection syndrome and complications of prematurity (average gestational age was 28.6 weeks in both the SG and the CG). It is probably impossible to select a true gestational age–matched CG, because (1) only placentas from abnormal pregnancies are sent for examination in our department at the discretion of referring obstetricians and (2) the SG is composed predominantly of premature deliveries (average gestational age 28.6 weeks). It is theoretically possible that because of significantly higher proportion of macerated stillbirths in the SG (Table 1), AN may be the result of oligohydramnios secondary to amniotic fluid resorption after fetal death. However, this mechanism is unlikely and is not raised in the literature.

Half of our AN cases occurred in the second trimester, hence this is at variance with the statement that “AN develops only late in fetal life because earlier in gestation there is insufficient vernix and that when oligohydramnios is present in the second trimester, AN does not develop.” 13(p7) An opinion that placentas from gestations of 26 weeks or less rarely show AN despite an accompanying history of marked oligohydramnios16 is more acceptable. Although AN is generally regarded as a hallmark of prolonged oligohydramnios, surprisingly, this clinical condition was diagnosed in only 10 (22%) cases of AN. This may mean that AN may not be a reliable sign of oligohydramnios, and mechanisms other than a simple increased concentration of vernix in the amniotic fluid such as adhesion of squames to traumatic erosion of the amnion because of fetal movements4,6 might be operational. Another explanation could be that clinically significant oligohydramnios is much more prevalent than that diagnosed antenatally by sonography or that antenatal sonography was not performed in most cases. However, the patient population of our hospital is composed predominantly of poor socioeconomic groups that do not have benefits of the whole spectrum of antenatal care.17 

Although the typical model association of AN are anomalies of the fetal urinary system, our analysis has shown that AN secondary to prolonged premature rupture of membranes is almost as common as AN associated with congenital malformations and more common than AN secondary to genitourinary malformations alone. This, however, may be because prolonged premature rupture of membranes is a very common condition, much more common than fetal genitourinary malformations.

Also, in our material, AN in multiple pregnancy was 3 times more common than the TTTS diagnosed clinically or at autopsy. Amnion nodosum may therefore be the first clinical indicator of the TTTS before other features thereof appear, or, alternatively, the oligohydramnios in monochorionic pregnancy may occur independently and not as a part of chronic TTTS or may occur in dichorionic twin pregnancy. The diagnosis of TTTS is usually made based on the findings of polyhydramnios-oligohydramnios sequence with weight discordance judged by obstetrical sonogram in monochorionic twins.18 Surprisingly, our CG and SG comprised similar numbers of monochorionic versus dichorionic placentas (Table 1).

Among other clinical conditions that were more common, although insignificantly, in the SG than in the CG were maternal preeclampsia and diabetes mellitus, both of which may be associated with oligohydramnios. The mechanism may be a marked uteroplacental underperfusion19 because of pathologic changes in uterine arteries providing blood to the placenta20 or decreased fetal glomerular filtration rate.21 Oligohydramnios in postterm pregnancies is associated with a reduction in renal artery end-diastolic velocity, suggesting that increased arterial impedance is an important factor.22 Our analysis also revealed a conspicuous paucity of other placental findings associated with AN. The high macerated stillbirth rate in the SG was most likely responsible for the placental regressive changes (stem vessel vascular luminal abnormalities and fibrosis of chorionic villi).23 Placental edema, more commonly seen in the SG, may be explained by decreased venous return to the heart because of compression of chorionic plate vessels and by congenital malformations—both known to be associated with hydrops, seen earlier in the placenta than in the fetus. Laminar necrosis of placental membranes was more commonly observed in the CG. It is a placental lesion diagnosed microscopically as a bandlike coagulative decidual and/or trophoblastic necrosis in the placental membranes in cases with perinatal acute hypoxia, particularly in cases complicated by hypertensive disorders in pregnancy. The lesion, unlike AN, is a feature of acute hypoxia24 therefore less commonly diagnosed in the SG. As this is an acute placental lesion, it was less commonly found in the SG. This is most likely the consequence of perinatal hypoxia as a cause of fetal death. Although AN has no relevance to management of pregnancy5 as it is diagnosed only postnatally, it may be important to identify AN at the time of birth, because this could rapidly suggest previously unrecognized oligohydramnios and its consequent, frequently serious conditions including congenital malformations of the neonate thus leading to rapid initiation of relevant diagnostic procedures or in cases of perinatal mortality to strengthen indications of perinatal autopsy.

Two gross lesions commonly confused with AN are subchorionic fibrin deposits (Figure 3) and squamous metaplasia of the amnion (Figure 4). In the former, the amnion can be made to slide over these deposits, whereas in the latter, there is a more diffuse thickening of the amnion and the plaque cannot be separated from the membrane. The term squamous metaplasia is a misnomer because the amnion is an immature squamous epithelium continuous with the fetal skin. The histologic appearance of squamous metaplasia is obvious, showing stratified squamous epithelium with focal keratinization.7 Histologically, AN should also be differentiated from chorion nodosum, a rare grossly inapparent lesion occurring in placentas from pregnancies complicated by severe early amnion rupture sequence associated with extra-amniotic pregnancy, particularly in the limb-body wall complex.15 Microscopically, chorion nodosum features flat nodules containing squamous debris, buried in the chorionic mesenchyme and without overlying amnion (Figure 5).

In conclusion, much remains to be learned about the pathogenesis of AN. Nevertheless, this placental condition portends poorly for pregnancy as it is associated with extremely high perinatal morbidity and mortality.

Landing
,
B. H.
Amnion nodosum: a lesion of the placenta apparently associated with deficient secretion of fetal urine.
Am J Obstet Gynecol
1950
.
60
:
1339
1342
.
Simpson
,
J. Y.
Pathological observations on diseases of placenta.
Edinb Med Surg J
1836
.
45
:
265
.
Bryans
,
A. M.
,
J. U.
Balis
, and
M. D.
Haust
.
Amnion nodosum: report of a case.
Am J Obstet Gynecol
1962
.
84
:
582
585
.
Bartman
,
J.
and
S. G.
Driscoll
.
Amnion nodosum and hypoplastic cystic kidneys: an electron microscopic and microdissection study.
Obstet Gynecol
1968
.
32
:
700
705
.
Thompson
,
V. M.
Amnion nodosum.
J Obstet Gynaecol Br Emp
1960
.
67
:
611
614
.
Salazar
,
H.
and
A. I.
Kanbour
.
Amnion nodosum: ultrastructure and histopathogenesis.
Arch Pathol
1974
.
98
:
39
46
.
Scott
,
J. S.
and
A. D.
Bain
.
Amnion nodosum.
Proc R Soc Med
1958
.
51
:
512
513
.
Bourne
,
G. L.
The anatomy of the human amnion and chorion.
Proc R Soc Med
1966
.
59
:
1127
1128
.
Rokitansky
,
C.
trans ed. Manual of Pathological Anatomy. Vol 2.
London, England: Sydenham Society; 1849:346
.
Baergen
,
R. N.
Manual of Benirschke and Kaufman's Pathology of the Human Placenta.
New York, NY: Springer; 2005
.
Bourne
,
G. L.
The Human Amnion and Chorion.
London, United Kingdom: Lloyd-Luke; 1962:196
.
Sarita-Reyes
,
C.
,
M.
Reyes-Múgica
, and
E.
Zambrano
.
Distinguishing amniotic bands from fetal skin by expression of cytokeratins in the diagnosis of amniotic band disruption sequence (ABS).
Presented at: Society for Pediatric Pathology Annual Meeting; February 11–17, 2006; Atlanta, Ga. Abstract 45
.
Sitzenfrey
,
A.
Ueber Amnionanomalien.
Beitr Geburtsch Gynaekial
1911
.
17
:
1
10
.
Adeniran
,
A.
and
J.
Stanek
.
Amnion nodosum revisited: clinicopathologic and placental features.
Presented at: 50th Annual Meeting of the Paediatric Pathology Society; April 22–24, 2004; Cape Town, South Africa
.
Stanek
,
J.
and
A.
Adeniran
.
Chorion nodosum: a placental feature of the severe early amnion rupture sequence.
Pediatr Dev Pathol
2006
.
9
:
353
360
.
Faye-Petersen
,
O. M.
,
D. S.
Heller
, and
V. V.
Joshi
.
Handbook of Placental Pathology. 2nd ed.
London, England: Taylor and Francis; 2006
.
Sundararajan
,
V.
,
L. S.
Levin
, and
J.
Stanek
.
High recurrence rates of acute and chronic placental inflammatory conditions.
9th International Federation of Placenta Associations Meeting; Mainz, Germany; September 24–27
.
Placenta
2003
.
24
:
A.31
.
Chiang
,
M. C.
,
R.
Lien
,
A. S.
Chao
,
Y. H.
Chou
, and
Y. J.
En Chen
.
Clinical consequences of twin-to-twin transfusion.
Eur J Pediatr
2003
.
162
:
68
71
.
Krous
,
F. T.
,
R. W.
Redline
, and
D. J.
Gersell
.
et al
.
Placental Pathology.
Washington, DC: American Registry of Pathology; 2004
.
Nessmann
,
C.
and
J. C.
Larroche
.
Atlas de Pathologie Placentaire.
Paris, France: Masson; 2001
.
Popek
,
E. J.
Normal anatomy and histology of the placenta.
In: Lewis SH, Perrin E, eds. Pathology of the Placenta. 2nd ed. New York, NY: Churchill Livingstone; 1999:49–88
.
Oz
,
Au
,
B.
Holub
, and
I.
Mendicioglu
.
et al
.
Renal artery Doppler investigation of the etiology of oligohydramnios in postterm pregnancy.
Obstet Gynecol
2002
.
100
:
715
718
.
Genest
,
D. R.
Estimating the time of death in stillborn fetuses, II: histologic evaluation of the placenta; a study of 71 stillborns.
Obstet Gynecol
1992
.
80
:
585
592
.
Stanek
,
J.
and
H.
Al-Ahmadie
.
Laminar necrosis of placental membranes: a histological sign of uteroplacental hypoxia.
Pediatr Dev Pathol
2005
.
8
:
34
42
.

The authors have no relevant financial interest in the products or companies described in this article.

Presented at the 10th Meeting of the International Federation of Placenta Associations, Asilomar, Calif, September 25–29, 2004.

Author notes

Reprints: Adebowale J. Adeniran, MD, Department of Pathology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10065 ([email protected])