Eye (1990) 4, 845-849

Ophthalmic Complications of Amniocentesis

G. NAYLOR, J. P. ROPER, H. E. WILLSHAW Birmingham

Summary Reports of ocular damage from amniocentesis needles are rare. We report four cases of ocular perforation in which an amniocentesis needle may have been the cause and a proven case of central nervous system perforation during amniocentesis which resulted in hemianopia and gaze palsy. Ocular damage during amniocentesis may be commoner than the paucity of reports would suggest, and should be considered in all cases of congenital ocular conditions.

Amniocentesis was introduced in 1930' as part which to perform the amniocentesis. The of the technique of amniography to determine increased risks of the procedure were placental position in cases of painless bleeding explained to the mother who elected to pro­ in the third trimester of pregnancy. Since then ceed. Amniocentesis proved difficult, only the technique has been mainly used in the 2ml of blood stained liquor being aspirated management of haemolytic disease· of new­ following two needle insertions. Subsequent born, assessment of fetal maturity, and pre­ analysis of the fluid revealed an alphafetopro­ natal genetic diagnosis. tein (AFP ) of 240mg/1 (normal 36mg/l at 16 Most amniocenteses now take place in mid­ weeks gestation). Acetylcholinesterase trimester, and although some studies point to (ACH) assay showed two bands to be present. a slightly increased risk of fetal loss,2,3 the The blood in the sample was 100% fetal. NICHD group in 19764 concluded that amnio­ Cytogenetic results of the cultured cells centesis was a 'highly accurate and safe pro­ showed a normal male karyotype-46XY. The cedure that does not increase the risk of fetal raised AFP was thought to be due to the pres­ loss or injury.' ence of fetal red blood cells and that the two There are, however, numerous reports in bands of ACH could be the result of a cerebral the literature of fetal damage caused by puncture. The diagnosis of fetal puncture, needle puncture during amniocentesis, and probably cerebral, was made and consequent we describe four cases of presumed ocular ultrasonography suggested a minor dilatation injury and one proven case of optic tract dam­ of one ventricle. The mother elected to con­ age following midtrimester amniocentesis. tinue the pregnancy. Following delivery by Caesarean section at 38 weeks, a small scar Case 1 was noted in front of the right fronto-parietal The mother of this child underwent amnio­ suture, but the child appeared otherwise centesis at 16 weeks gestation because of normal. Early ultrasonography suggested a maternal age. Preceding ultrasonography cystic lesion communicating with the right lat­ showed an anterior placenta with conse­ eral ventricle, and the possibility ofdamage to quently only a small area of uterustrn-ough the right optic tract by the amniocentesis

Correspondence to: Mr H. E. Willshaw FRCS, FCOphth , Eye Department, Birmingham Childrens Hospital, Ladywood Middleway, Birmingham B16 SET. 846 G. NAYLOR. 1 P. ROPER. H. E. WILLSHAW

At the age of 14 months, he still had a left homonymous hemianopia and a marked gaze defect to the left, although there had been some improvement as he did generate occasional saccades to the left .

Case 2 The mother of this child underwent amnio­ centesis/chorionic villus sampling at 30 weeks gestation (by dates) for chromosomal analysis (subsequently normal) and because of intra­ uterine growth retardation and oligohydram­ nios. On ultrasound, the placenta was found Fig. 1. CT scan of case I showing dilated right lateral to be lying posteriorly and laterally with the ventricle, prominent Sylvian fissure and right sided fetus in a cephalic presentation. Despite two cortical atrophy. attempts under ultrasound guidance, no villus fragments were aspirated . Following delivery needle was considered. Ocular examination at 31 weeks because of fetal distress, weighing was normal. 715 grams, the baby required special care for The child appeared to progress satisfacto­ hypoglycaemia, thrombocythaemia, hypo­ rily for seven months but although he albuminaemia , chest infection, and choles­ responded to sounds on the left by turning his tatic jaundice following an episode of head, he did not turn his eyes . Left sided necrotising enterocolitis. His respiratory motor development was slow, with poor use problems, however, did not necessitate more of the left hand and reduced movement of the than three days of 30% oxygen. left leg . Clinically, he had a left homonymous Following discharge his parents noted his hemianopia, although visual evoked potential left pupil was abnormal and he was sub­ on pattern reversal showed no consistent sequently seen in our eye department at nine hemisphere asymmetry. onths of age. On examination, visual acuity At seven months, fitting began with an epi­ � m both eyes was within normal limits (by sode of status epilepticus, controlled by medi­ forced choice preferential looking). The right cation. CT scans showed moderate dilatation upil was round , central and normally react­ of the right lateral ventricle and lesser dilata­ p mg. The left pupil was distorted towards the 3 tion of the left occipital horn. The left frontal o'clock position with a small tag of iris drawn horn appeared to arise like a diverticulum up towards a full thickness corneal scar. There from the right. Prominent interhemisphere was no cataract, and fundus was normal. (Fig. fissure and right Sylvian fissureindicated atro­ 2). phy rather than compression by the dilated right anterior horn. (Fig. 1). Case 3 This child was noted to have a left esotropia from birth, the parents also noting that the left eye appeared smaller than the right . The mother underwent amniocentesis at 16 weeks gestation because of maternal age , with sub­ sequent delivery of an otherwise normal girl at 38 weeks (labour being induced because of maternal hypertension). The child was seen by us at 10 months of age. On examination, the left eye was microphthalmic, with reduced visual acuity, esotropia and limitation of Fig. 2. Case 2 with pupil peaked to corneal abduction . Anterior segments appeared perforation. normal, but fundus examination revealed a OPHTHALMIC COMPLICATIONS OF AMNIOCENTESIS 847

cho rioretinal scar in the temporal periphery in 100 cases, and Karp et al8 a 2.1% occur­ with obvious overlying vitreous condensa­ rence in 190 cases. tions. The view was not good enough to iden­ The majority of needle punctures present as tify any incarceration (Fig. 3). non-pigmented dimple like marks on the chest, back, and limbs and are often over­ Case 4 looked at birth, becoming more apparent A three and a half year old boy came to an during infancy due to tethering caused by sub­ ophthalmology de partment with a mild con­ cutaneous scar tissue.3,7,9 junctivitis. Visual acuity was 6/5 in both eyes Creasman5 points out that there is often with no refractive error. He had no past very little evidence of abdominal puncture history of eye problems or head injury. Slit following intrauterine transfusion a fe w days lamp examination revealed a small adherent prior to birth, and that fetal skin healing is leukoma near the limbus at 7 o'clock. No lens very rapid. opacity was seen and fundus examination was Ocular injuries are rare, only five isolated normal. His 34 year old mother had under­ cases being described in the American liter­ gone amniocentesis at 15 weeks gestation. ature,I4-18 with no ocular damage being The appearance of the child's eye was consis­ re ported in the larger series. tent with an earlier perfo rating injury, and in Of the cases describe d here case 1 has a the absence of any other history of injury gaze palsy and he mianopia which must be due could have been caused by the amniocentesis to the proven amniocentesis puncture. needle. In cases 2 and 3, the ocular abnormalities were no ted by the parents at birth. The find­ Case 5 ings in case 2 were compatible with a limbal A 5 year old girl attended an eye casualty with perforating injury, the amniocentesis needle a small chalazion on her right upper lid. being the most likely cause. In case 3, micro­ Examination showe d a small full-thickness phthalmos, a chorioretinal scar and restricted scar near the limbus at 9 o'clock. Visual acuity abduction could, we feel be explained by a was no rmal, the lens clear and fundus normal. posterior perfo ration with lateral rectus He r mother could not recall any previous damage. trauma, but had amniocentesis performed Cases 4 and 5 had signs of previous anterior during the mid trimester, suggesting a poss­ segment perforations. These could obviously ible cause for the corneal scar. relate to a previous injury occurring at any time between birth and presentation, but the Details of the amniocentesis are not avail­ lack of any history of trauma or other eye able for the last three cases. problems leads us to suggest that amniocen­ tesis needle perforation is a possible cause. Discussion The eye is still at an active stage of develop­ Re ports of fetal damage during amniocentesis ment during the mid-trimester, and the lids include abdominal perforation,5 pneumo­ still fuse d. One might therefore expect more thorax,5,6, central nervous system perfor­ evidence of ocular disruption from such per­ ation,5 and limb and trunk puncture.5,7-12 forating injuries than found in our cases, with Several large studies have specifically evidence of lid damage. examined newborns for signs of needle punc­ Of the five previously reported cases, three ture , with varying results. The NICHD group4 describe anterior segment perforation,14.15,16 found no evidence of skin puncture in 1040 two following amniocentesis near term (3414 children following amniocentesis, either as and 3715 weeks gestation) and one mid trimes­ newborns or at one year, and Ta borI3 no scars ter (19 weeksI6). None developed any vision in 2302 newborns. although two cases were initially noted to However, Epley et at found a 9% fI:e­ have clear media and normal posterior poles. quency of fetal injury following mid trimester A furthe r case describes a posterior segment amniocentesis in 107 infants, Finegan et aP a perforation following midtrimester amnio­ 5.5% incidence in 91 cases, Broome et all! 4% centesis at 17 weeks with leukocoria and total 848 G. NAYLOR. 1. P. ROPER. H. E. WILLSHAW

occur despite preceding ultrasonography. The advent of real-time ultrasonography (Fig. 3) with guidance of a small bore needle (21122 G) may improve the safety of amniocentesis'9,2o and reduce the risk of fetal injury. In more recent reports, TaborD found no scars in 2302 newborns following mid-trimester amniocen­ tesis. and Tatayama21 none in 500 cases. Of our two cases where details of the amniocentesis are available (Cases 1 and 2), one was performed with real-time ultrasonog­ raphy and both were undoubtedly difficult Fig. 3. 3. Patient procedures. Real-time was unlikely to have been available when cases 4 and 5 underwent I� retinal detachment. The anterior segment amniocentesis although as with case 3, a pre­ was normal in this case and the retinal archi­ ceding scan was likely to have been tecture mature, the detachment being due to performed. traction from vitreous incarcerated in the per­ We therefore suggest that ocular damage foration. The final case describes a post-lim­ during amniocentesis may be commoner than bal perforation associated with a sealed originally thought. We would agree with contra-coup retinal tear with surrounding pig­ Merin and Beyth 16 that prenatal trauma mentation, no lenticular damage and useful should be considered in the differential diag­ vision, following amniocentesis at 20 weeks nosis of congenital ocular conditions, gestation. 17 This report therefore shows that minimal ocular disruption can follow ocular We would like to thank Mr G. Constantine and Ms M. perforation during fetal life. Bradley at the Birmingham Maternity Hospital for Of the three midtrimester reports. only one their assistance, and Paul England and Richard notes a lid notch.'6 Hildred for the photographs. It is possible that cases of ocular damage References following amniocentesis are commoner than I Menees TO. Miller JD. Holly FE: Amniography: the paucity of reports would suggest. but do Preliminary report. Am J Roentgen 1930.24: 363. not present if symptoms and signs are mini­ , Milunsky A: Risk of amniocentesis for prenatal diag­ mal. Although large reviews have failed to nosis. New Eng J Med 1975.293: 932-3. find any ocular damage following amniocen­ 'Finegan JAK. Quarrington B1, Hughes HE. Rudd NL. Stevens LJ, Weksherg R, Doran TA: Br J tesis the same can be said of skin perforation, Obstet Gvnaecol1985, 92: 1015-23. a complication which undoubtedly does 'The NICHD National Registry For Amniocentesis occur. Study Group. Midtrimester amniocentesis for Previously reported cases of fetal injury prenatal diagnosis. Safety and accuracy: J Am Med Ass 1976.236: 1471-6. 5 Creasman WT, Lawrence RA, Thiede HA: Fetal complications of amniocentesis. J Am Med Ass 1968,204: 949-52. h Cook L, Shott RJ, Andrews BF: Fetal complications of diagnostic amniocentesis: A review and report of a case of pneumothorax. Pediatrics 1974,53: 421-4. 7 Epley SL, Hanson JW, Cruikshank DP: Fetal injury with midtrimester diagnostic amniocentesis: Obstet Gvneco11979, 53: 77-80. H Karp LE a�d Hayden PW: Fetal puncture during midtrimester amniocentesis. Obstet Gynecol 1977,49: 115-7. 'Crandell BF. Howard J, Lebherz TB, Rubinstein L, Sample WF, Sarti 0: Follow up of 2000 second­ Fig. 4. Amniocentesis performed under real-time trimester amniocenteses. Obstet Gynecol 1980, ultrasonography. 56: 625-8. OPHTHALMIC COMPLICATIONS OF AMNIOCENTESIS 849

10 Fairweather DVI and Walker W: Obstetrical Con­ 15 Fortin JG and Lemire J: Une Complication Oculaire siderations in the routine use of Amniocentesis in de L'amniocentese. Can JOphthalmol 1975, 10: immunised rhesus negative women. JObs Gyn 511-13. 6 Brit Comm1968, 71: 48-53. 1 Merin S and Beyth Y: Uniocular congenital blind­ 11 Broome DL, Wilson MG, Wiess B, Kellogg B: ness as a complication of midtrimester amniocen­ Needle Puncture of Fetus. A complication of tesis. Am JOphthalmol1980, 89: 299-30l. 17 second-trimester amniocentesis. Am J Obstet Isenberg SJ, Heckenlively JR: Traumatized eye with Gynecol1976, 126: 247-52. retinal damage from amniocentesis. J Pediatr 2 Ophthalmol Strabismus 1985,22: 65-7. 1 Allen HH, Sergovich F, Stuart EM, Pozsonyi J, 18 Admoni MM and BenEzra D: Ocular trauma fol­ Murray B: Infants undergoing antenatal genetic lowing amniocentesis as the cause of leukocoria. J diagnosis: A preliminary report. Am J Obstet Pediatr Ophthalmol Strabismus 1988, 25: 196--7. GynecoI1974,118,310-3. 19 Williamson RA, Varner MW, Grant SS: Reduction 13 Tabor A, Philip J, Madsen M, Bang J, Obel EB, in amniocentesis risks using a real time needle Norgaard-Pederson B: Randomised controlled guide procedure. Obstet Gynecol1985,65: 751-5. trial of genetic amniocentesis in 4606 low risk 20 MacKenzie WE: Safety of amniocentesis. Lancet women. Lancet 1986,1 (8493 ) : 1287-93. 1986,2 (8500 ) : 225 (letter). 14 Cross HE and Maumenee AE: Ocular Trauma 21 Katayama KP and Roesler MR: 500 cases of amnio­ During Amniocentesis. Arch Ophthalmol 1973, centesis without bloody tap. Obstet Gynecol1986, 90: 303-4. 68: 70-3.