Cystic hygroma: Comparison between prenatal and postnatal diagnosis

Cystic Hygroma: Comparison Between Prenatal and Postnatal Diagnosis By Ross Fisher, Ann Partington, and Evelyn London, England 0 Antenatal diagnosis of congenital malformation often leads to opportunities for intervention in the pregnancy that may have a profound effect on the outcome of the fetus. The accuracy of the diagnostic technique is a critical factor in judging the appropriateness of such interventions. This article reports the current accuracy of prenatal ultrasound diagnosis of cystic hygroma (CH) in a region with a population of 3.5 million and an annual birth rate of 52,000. During a 32-month period (to December 1994). 113 diagnoses of CH were reported to the South East Thames Regional Congenital Malformation Registry. Eighty-nine diagnoses were made prenatally and 24 were made postnatally (at delivery or postmortem). The number of confirmed cases of CH was 52 (46% total diagnoses). The prenatal diagnoses were established at a mean gestational age of 19 weeks. Fifty-six (63%) pregnancies with a prenatal diagnosis of fetal CH were terminated; in only 25% of these was the presence of CH confirmed after termination. Overall, of the 89 prenatal diagnoses, 28 (32%) were confirmed, 45 (50%) were incorrect, and 16 (18%) could not be confirmed or refuted. Only 24% of fetuses with a prenatal diagnosis of CH were livi born. The survival rate at 1 year for the live-born infants with CH in this series was 88%. The diagnosis of fetal CH has a recognised association with chromosomal abnormalities, and consequently may lead to termination of the pregnancy. A potential error rate of almost 70% in the prenatal ultrasound diagnosis of this condition requires clinicians to adopt a more cautious approach to the management of the fetus with “cystic hygroma.” Copyright o 1996 by W.B. Saunders Company INDEX WORDS: Cystic hygroma, tal diagnosis, ultrasonography.

nuchal

translucency,

prena-

T

HE WIDESPREAD availability of prenatal diagnosis leads not only to planned postnatal treatment but to opportunities for fetal manipulation and intervention in the pregnancy. Before embarking on any interventional course, it is essential that the accuracy of the prenatal diagnostic technique be established and that the natural history of the abnormality be well understood. There is a recognised association between fetal cystic hygroma (CH) and chromosomal abnormalities,‘” but the precise definitions and terminology used to describe CH remain confusing. An inaccurate diagnosis inevitably will affect parental counseling and the subsequent management of the pregnancy as well as the ultimate outcome for the fetus diagnosed to have CH.

JournalofPediafric

Surgery,

Vol31,

No 4 (April),

1996: pp 473-476

Dykes

The purpose of this study was to review the current accuracy of prenatal ultrasound diagnoses of CH and the ultimate outcome of fetuses with this condition from an established regional database. MATERIALS

AND METHODS

The South East Thames Congenital Malformation Registry (CMR) was instituted in April 1992 as the result of an initiative supported financially by the Regional Medical Audit Committee. The region has a population of approximately 3.5 million and an annual live birth rate of approximately 52,000. Notification of all identified cases of major congenital structural abnormalities was requested from all obstetric, fetal medicine, genetic, paediatric and pathology units in the region. This large reporting group offers extensive coverage of the region and allows identification of patients who may have been missed at a previous stage. Once notified of a case, the CMR coordinator actively collects data on the time and method of diagnosis, any prenatal interventions and investigations, time and place of delivery, and ultimate outcome (up to 1 year after birth). The initial ultrasound scans are performed as part of routine antenatal care in the majority of cases. All data entered in the Registry are taken directly from the clinical records and are maintained according to the regulations stipulated by the Data Protection Act (1994). For the purpose of this study, all complete data sets in which a diagnosis of CH had been made at any time (prenatally, or postnatally to 1 year of age) were reviewed to establish (1) the total number of diagnoses, (2) the correlation between prenatal and postnatal diagnoses, (3) the incidence of associated abnormalities, and (4) the outcome of the pregnancies involved. RESULTS

Between April 1992 and December 1994 (32 months), 113 diagnoses of “cystic hygroma” were From the South East Thames Congenital Malformation Registry, Children’s Hospital, Lewisham, London, England. Presented at the 42nd Annual International Congress of the British Association of Paediati’c Surgeons, Shefield, England, July 25-28, 1995. The South East Thames Regional Congenital Malformation Regishy isfinded b the Regional Medical Audit Committee. The opinions expressed in this paper are those of the authors and do not necessatily represent those of the Steering Group or of the Regional Health Authority. Registy Steering Group: Miss E. Dykes (Director), DrD. Garvie, MrP. Knott, Dr C. Beny, DrM. Driver, DrM. Newbold, Dr C. Wolfe, Mrs A. Partington (Coordinator). Address reprint requests to Miss Evelyn Dykes, Congenital Malformation Registry, Children’s Hospital Lewisham, London SE13 6LH, England. Copyright o I996 by W.B. Saunders Company 0022-3468/96/3104-0002$03.00l0

473

474

FISHER,

Table

1. Outcome

for 113 Patients

Diagnosed

to Have Cystic

Confirmed Cystic Hygroma Time of Diagnosis

Clinical outcome*

Plus Other Anomaly

PARTINGTON,

AND

DYKES

Hygroma

No Cystic Hygroma

No Other Anomaly

All W)

Plus Other Anomaly

No Other Anomaly

All (%I

Postmortem Not Performed (%)

Prenatal TOP

56

13

14 (25)

26

2

28 (50)

SPA

4

0

2 (50)

IUD

8

3

4 (50)

0 3

0 1

0 4 (50)

LIB

21

3

89

19

8 (38) 28 (32)

TOP

10

IO

10 (100)

0

SPA

3

3

3 (100)

0

IUD

2

LIB

9

Total prenatal

7

6

13 (62)

36

9

45 (50)

14* (25) 2 (50) 0 NA 16 (18)

Postnatal

Total

0

2

9 (100)

NA

24

16

24 (100)

113

35

52 (46)

postnatal

All diagnoses Abbreviations: *One

2 (100)

TOP, termination

fetus was found

of pregnancy;

to have an abnormal

SPA, spontaneous

karyotype

pre-TOP

abortion;

for 89 Prenatally Cystic

Karyotype

Cases

No. (%) 10 (11)

Normal Failed test Abnormal Trisomy

Diagnosed

Hygroma

2 32 (36) 21

12

Turner’s Trisomy

(45,X0) 18

11

Trisomy

13

Other

5 1 3

Not examined

45 (51)

Total

89

IUD, in utero death;

9 LIB, live birth;

45 (40)

16 (14)

NA, not applicable.

but did not have a postmortem.

reported to the Regional Congenital Malformation Registry. All diagnoses referred to lesions of the cervical or nuchal region. Eighty-nine cases (79%) were diagnosed prenatally and 24 were diagnosed postnatalIy (ie, after delivery, including stillbirths and terminations). Subsequent follow-up showed only 52 confirmed cases of CH (46% of all diagnoses; 28 prenatal and 24 postnatal) (Table 1). The mean gestational age at the time of prenatal diagnosis was 19 weeks. Fetal karyotyping was undertaken in 44 of the 89 cases (49%); in 32, the result was abnormal (Table 2). Nonimmune hydrops fetalis was identified in 15 cases of fetal CH (17%), 8 of which had additional chromosomal or structural malformations. In 56 of the 89 cases (63%) with a prenatal diagnosis of CH, the pregnancy was terminated. Only 14 (25%) terminations were confirmed to have CH postmortem; in 26 of the 28 without CH, other significant abnormalities were found (chromosomal Table 2. Fetal Karyotype

0 36

of

and/or structural). For the remaining 14 terminations, no confirmatory examinations or investigations were done. It appeared that three pregnancies were terminated because of isolated fetal CH, only one of which was confirmed. Additionally 12 of the 89 cases (14%) of fetal CH died in utero (8 in utero deaths, 4 spontaneous abortions). In only six of these was the prenatal diagnosis confirmed. Overall, 61 (68%) of the prenatal diagnoses of fetal CH were not confirmed subsequently. This group comprised 13 live-born infants and 48 dead fetuses (either terminated or demising in utero), including 16 cases without postmortems. Of the 52 confirmed cases of CH, 24 (46%) were not diagnosed prenatally; 10 were identified only after termination because of other abnormalities, and 14 were diagnosed at the time of delivery. Additional abnormalities occurred in 35 of the 52 confirmed cases of CH (67%) (Table 3). Only 17 infants with CH were live born (8 prenatally diagnosed, 9 postnatally diagnosed). Five of them had additional structural malformations, but none had chromosomal abnormalities. The survival rate at 1 year of age for the live-born infants with CH in this series was 88% (15 of 17). DISCUSSION

The fetal lymphatic system develops around the fifth week of gestation as outgrowths of the venous system, with the major connections occurring in the neck at the internal jugular vein. Congenital malformations of lymphatic drainage result in pathological accumulations of fluid, often first observed in the

CYSTIC

HYGROMA:

Table

PRENATAL

3. Additional

Abnormalities in 52 Confirmed Cystic Hygroma

Type of Abnormality

10 (19) 1 7 1

(45,X0) 18

1

Partial translocation Cardiovascular

17 (33) 9 (17)

Pulmonary Musculoskeletal

9 (17) 7 (14)

Renal/urologic Multiple

(> 1) additional

Total with additional

Cases of

No. of Cases (%)

Chromosomal Trisomy 21 Turner’s Trisomy

475

DIAGNOSIS

anomalies

abnormality

15 (43) 35 (67)

nuchal region. Excess fluid may manifest as skin or subcutaneous oedema or as discrete fluid collections. The extreme condition of hydrops fetalis occurs when there is absolute failure of lymphatic drainage to the venous system, resulting in generalised oedema, ascites, pleural effusion and, usually, fetal death. Ideally, the term “cystic hygroma” should be used only to describe large, localised accumulations of lymphatic fluid, which often are multiloculated or septated, but the term is often Ioosely applied to the entire spectrum of lymphatic abnormalities4 Excess fluid in the nuchal region is considered to be physiological up to the 11th week,5a6 but with the increasing availability of high-resolution prenatal ultrasonography, fetal lymphatic abnormalities are being detected more frequently, often at very early stages of gestation. It has been suggested that the observation of fetal nuchal fluid or translucency greater than 3.0 mm in depth at the end of the first trimester is a strong marker for chromosomal abnormality, justifying further intervention to establish the fetal karyotype and exclude other anomalies.4,6 Although the association between persistent and obvious fetal lymphatic abnormality and other anomalies may be valid, the use of such fine measurements as a discriminator in early pregnancy is bound to be partly subjective. Errors in measurement can occur because of differences in skin thickness (which has been found to increase proportionately with fetal age and size), the amount of fetal hair, and the proximity of the amniotic sac, umbilical cord, or fetal limbs to the back of the head.’ There is aIso evidence lhat many lymphatic drainage problems are transient, and resolve either partiahy or completely during intrauterine life, presumably by the formation of collaterals or by recanalisation of obstructed channels.3,6J,9 This study has shown that the antenatal ultrasound diagnosis of CH must be viewed with caution. In this

series, only one third of all such diagnoses were confirmed postmortem or at the time of delivery. Whether the remaining unconfirmed cases were the result of true diagnostic errors, resolution of the lesion in utero, or misuse of terminology is not possible to determine with certainty from the information available. Nevertheless, a consistently high proportion of prenatal ultrasound diagnoses could not be confirmed postnatally at various stages of gestation (50% terminations, 50% in utero deaths, 62% live births), which suggests that much more precise definitions of the observed fetal abnormalities and the terminology used to describe them are required. Also of concern is the observation that in the present series at least two pregnancies (4% terminations) appear to have been terminated simply because of an ultrasound diagnosis of isolated CH, which was not confirmed at postmortem, nor was any other anomaly detected. Conversely, only 53% of the confirmed cases of CH had a prenatal diagnosis; this demonstrates the apparently poor sensitivity of prenatal ultrasonography in the detection of this condition. It may be argued that some of the missed cases developed later in gestation, after routine ultrasound scans had been completed, but because 20% of these cases were terminations for other sonographically diagnosed abnormalities, continuing unease about the diagnostic accuracy must persist. There can be little doubt about the association between true fetal lymphatic malformation and other anomalies,r”JO although in the present series, structural defects were more common than chromosomal abnormalities. This observation tends to support the need for some concern if abnormal nuchal fluid is demonstrated in the fetus at any stage, and may justify further investigation, in particular detailed anomaly scanning for other structural markers of chromosomal abnormality. However, the clinician’s understandable desire to provide an early (and inevitably invasive) diagnosis of abnormal fetal karyotype must be balanced against the risks to the fetus, together with the low confirmation rate for ultrasonographic diagnosis of CH in early pregnancy. Aggressive intervention on the basis of a single ultrasound demonstration of isolated fetal lymphatic abnormality should be actively discouraged. ACKNOWLEDGMENT The authors Thames Region

thank all the reporting specialists who contributed to the Registry.

in South

East

476

FISHER,

PARTINGTON,

AND

DYKES

REFERENCES 1. Azar GB, Snijders RJM, Gosden C, et al: Fetal cystic significance of cystic hygroma in the first trimester. Prenat Diagn hygromata: Associated malformations and chromosomal defects. l&643-651,1990 Fetal Diagn Ther 6:46-57,199l 7. Van Zalen-Sprock RM, Van Vugt JMG, Van Geijn HP: First 2. Shulman LP, Emerson DS, Felker RE, et al: Frequency of trimester diagnosis of cystic hygroma-Course and outcome. Obcytogenetic abnormalities in fetuses with cystic hygroma diagnosed stet Gynecol167:94-98,199Z in the first trimester. Obstet Gynecol80:80-82, 1992 8. Bronshtein M, Bar-Hava I, Blumenfeld I, et al: The difference 3. Bernstein HS, Filly RA, Goldberg JD, et al: Prognosis of between septated and non-septated nuchal cystic hygroma in the fetuses with a cystic hygroma. Prenat Diagn 11:349-355,199l early second trimester. Obstet Gynecol81:683-687,1993 4. Nicolaides KH, Azar G, Byrne D, et al: Fetal nuchal translu9. Chevernack F, Isaacson G, Blakemore K, et al: Fetal cystic cency; Ultrasound screening for chromosomal defects in the first hygroma. Cause and natural history. N Engl J Med 309822-825, trimester of pregnancy. Br Med J 304:867-869, 1992 1983 5. Wilson RD, Venir N, Farqhaharson DF: Fetal nuch fluidPhysiological or pathological?-in pregnancies less than 17 men10. Abramovicz JS, Warsof SL, Doyle DL, et al: Congenital strual weeks. Prenat Diagn 12:755-763,1992 cystic hygroma of the neck diagnosed prenatally: Outcome with 6. Cullen MT, Sandro G, Green JJ, et al: Diagnosis and normal and abnormal karyotype. Prenat Diagn 9:321-327,1989