Article Text

Persistent pupillary membranes
  1. B J L BURTON,
  2. G G W ADAMS
  1. Moorfields Eye Hospital, City Road, London EC1V 2PD
  1. Miss Adams.

Statistics from

Editor,—Persistent pupillary membranes (PPMs) which are large enough to interrupt the visual axis are uncommon. If left untreated they may cause stimulus deprivation amblyopia.1There is evidence that early treatment to remove obstructions of the visual axis in infants may minimise the risk of amblyopia.2 We report a case of bilateral persistent pupillary membranes which were successfully treated surgically before the patient was 6 weeks old.


A healthy baby induced at 41 weeks was born by normal vaginal delivery following a normal pregnancy. He weighed 8 lb 8 oz (3.8 kg) and was noted to have an abnormal red reflex from both eyes on routine ophthalmic check by a paediatrician before discharge from hospital after birth. He appeared to fix briefly on a light but did not really look at toys. The red reflexes were significantly reduced centrally and there were peripheral spoke-like opacities. Even with dilatation only a poor fundal view was obtained. The view was particularly poor on the right eye. An ultrasound scan of the posterior pole was normal for each eye. Further examination under anaesthetic showed that the child had hyperplastic persistent pupillary membranes in both eyes (Fig 1). Blood could be seen flowing through the blood vessels within the membranes under the operating microscope. There were no other abnormal examination findings. Screening tests for toxoplasmosis, cytomegalovirus, rubella, and treponemal antibodies were all negative. There was no known history of substance abuse by the mother. The visual axis especially on the right side was significantly obstructed even when the pupils were dilated and consequently we thought that the risk of stimulus deprivation amblyopia was high if they were left in place. The membranes were removed from the right eye when the child was 4 weeks old and from the left eye when he was 5 weeks old.

Figure 1

Persistent pupillary membranes shown in the right eye (A) and the left eye (B) of a 4 week old child after dilating drops have been given.

The same operative technique was used for each eye. A small superior corneal section was made and a further side incision was made. Provisc (Healon) was injected between the membrane and the lens in order to protect the lens from coming into contact with the surgical instruments. Intraocular diathermy of the membrane vessels was attempted but this failed to have any effect. The membrane strands were cut with vitreous scissors and the freed membrane was removed with intraocular forceps. There was a small amount of bleeding from one of the more vascularised strands of membrane but the other smaller vessels retracted without bleeding. The Provisc was removed with a Simco irrigation/aspiration canula. The corneal wounds each required a single 10/0 nylon suture which were removed under anaesthetic 3 months postoperatively.

The visual axis was clear following surgery and there was no evidence of cataract. Histology showed a thin membrane of spindle-shaped cells some of which were pigmented. A fine network of blood vessels extended through the membrane.

Follow up of the child 5 months after the operation revealed no evidence of cataract formation in either eye. The eyes were straight and visual assessment showed that he objected to occlusion of each eye equally with a binocular acuity of 6/19 using Cardiff cards.


Remnants of pupillary membranes are very common occurring in 95% of normal newborn babies.3 Early development of the crystalline lens at about the sixth week coincides with the growth of the tunica vasculosa lentis which nourishes it. The anterior portion of this is continuous with the pupillary membrane which draws its blood supply from the capillaries which arise from the branches of the long posterior ciliary arteries and the major arterial circle.4The pupillary membrane is fully developed by 9 weeks’ gestation. These blood vessels begin to undergo remodelling and regression in a process involving phagocytosis by macrophages at around 5 months’ gestation. Dysfunction of macrophage invasion and phagocytosis is thought to play a role in the pathogenesis of PPMs.5

PPMs are thought to be derived from the tunica vasculosa lentis (TVL) and have normally disappeared by the 34th week of gestation.6 These membranes are often observed by neonatologists as they are frequently seen in premature babies. They can be used to identify the gestational age of a newborn infant (between 27 and 34 weeks)7 and they continue to regress at the same rate as if the child had not been born prematurely.8 There is some more recent evidence that intrauterine stress, particularly from chronic maternal hypertension, may accelerate the disappearance of these membranes.9

PPMs should be distinguished from congenital idiopathic microcoria which results from underdevelopment of the dilator pupillae.10

Although not commonly associated with any other pathology there has been a previous report of positive toxoplasma serology occurring in a 30 week premature baby with PPMs.3 There has also been a report of PPMs occurring in an infant with congenital rubella virus.6 We believe that any child with a congenital abnormality of the anterior chamber should be screened for possible intrauterine infections. A series of five cases of persistent membranes associated with congenital dystrophia myotonica have been described.11 There have also been reports of PPMs occurring with a variety of other abnormalities of the anterior chamber which in some cases seem to be familial with autosomal dominant inheritance but in our case there was no family history.12 13 There have been case reports of spontaneous haemorrhage occurring form PPMs but these are unusual.14

When present PPMs are usually so small as to need no treatment since they shrink significantly over the first year of life. There is some evidence that not all PPMs behave in this way. Some membranes appear to become more hyperplastic after birth rather than regressing.15 Some authors believe that hyperplastic membranes should be considered as a distinct type of PPM16and that they may be less likely to regress than thinner membranes that appear simply to be persistent rather than hyperplastic although no formal study has looked at this.17

Smaller membranes than the ones we report have been successfully treated with mydriatic therapy until spontaneous involution occurred.18 However, the smallest useful pinhole size (and presumably pupil aperture) is thought to be 1.5 mm19without which normal visual development is unlikely to take place. In our patient the visual axis, especially on the right side, was significantly obstructed even when the pupils were dilated and consequently we thought that the risk of amblyopia using this treatment strategy was high. Early surgery has been described in this condition at 5 days of age with good fixation and a clear visual axis a year later.17 If there is clear evidence that the membranes are shrinking over the first few weeks of life then it may be reasonable to wait for spontaneous resolution. In our case there was no evidence of resolution over a period of 3 weeks and this contributed towards the decision to operate. There have been reports of neodymium:YAG laser treatment to remove similar membranes from teenage children for cosmetic reasons.20 21 The long term risk of cataracts from this technique is likely to be high particularly if it were to be used on an infant. There are no published studies of long term follow up of these patients. We believe that a conventional surgical technique is more appropriate than laser treatment in an infant although the operator must be aware of the risk of inducing cataract from unnecessary contact with the lens.

There is growing evidence that the severity of deprivation amblyopia is minimised by earlier treatment to remove any obstruction to the visual axis particularly in the case of unilateral congenital cataracts.2 It is likely that a similar critical period applies to bilateral visual obstruction. Some authors feel that there is no visual advantage to be gained by removing these membranes after the sensitive period for visual development is over16although modest improvements have been reported20 21 in teenagers.

If bilateral membranes are large enough to put a child at significant risk of amblyopia then it is our view that a surgical approach to this condition should be taken within the first 6 weeks of life as after this time the amount of irreversible visual dysfunction from amblyopia may well increase. Patients treated surgically for this condition should be followed up carefully because of the long term risk of developing cataract,15 amblyopia requiring patching, and glaucoma which may accompany congenital abnormalities of the anterior segment.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.