Fetal OK‐432 pleurodesis: complete or incomplete?

Congenital chylothorax occurs in 1 in every 12 000–15 000 pregnancies and is associated with fetal pulmonary hypoplasia due to compression of the fetal lungs. Therapeutic measures include maternal dietary modification1, repeat thoracocentesis2, thoracoamniotic shunting3, intrapleural injection of maternal blood4 and pleurodesis with OK-4325-9. We read with interest the articles (two published in 20015, 6 and one in 20037) in this journal on pleurodesis using OK-432. The first two cases of our own experience with OK-432 as a therapeutic agent for hydrops fetalis (in one of which it was used for pleurodesis) were published elsewhere9. Only four cases up to 2005—according to a MEDLINE search using the keywords ‘fetal chylothorax’ and ‘OK-432’—claimed to have achieved success in managing fetal chylothorax with this novel treatment modality5-9 (Table 1). We have applied the same strategy in four further patients since our earlier publication9 and report on our findings here. As the last two cases had successful outcomes (defined as survival beyond 1 month of age), we feel we are closer to understanding what factors may be involved in, or may contribute to, successful treatment, and we would like to share our experience with the readers of this journal. All four fetuses (Cases 6–9 in Table 1) proved to have no in-utero infection such as cytomegalovirus, herpes simplex virus, toxoplasmosis, rubella virus or parvovirus B19, nor were there any rhesus incompatibilities. Their chylothoraces were all confirmed by demonstrating the percentage of lymphocyte count to be more than 95% in the aspirated pleural fluids. 0.1KE (in a concentration of 1KE in 10 mL normal saline) OK-432 (Picibanil®, Chugai Pharmaceutical, Tokyo, Japan) was injected into the pleural cavity with or without aspirating dry the pleural effusion in each case we treated. A difficult orientation, for example when the fetal heart was just beneath the probe or the placenta was thick, was the main reason for failure to fully aspirate the pleural fluid before OK-432 instillation. In our two successful cases (Cases 8 and 9) and, indeed, in the four successful cases reported by others—i.e. Cases 1–4 in Table 1—the pleural fluid was aspirated dry and adhesions were visible on ultrasonography after a certain period of time. In our opinion, complete aspiration of pleural fluid prior to OK-432 instillation, and subsequent ultrasound demonstration of adhesions, would thus seem to be two requirements for a successful outcome. Case 9, we feel, is particularly noteworthy. The fetus was found to have bilateral chylothorax as early as 21 weeks' gestation and the fetal heart was just beneath the ultrasound probe, which made aspiration difficult and dangerous. However, we then decided to approach the pleural effusion through the fetal liver (Figure 1), and successfully performed OK-432 pleurodesis with the awareness that ‘completeness’ is vitally important in achieving success. The fetal lungs expanded immediately and remained so 2 weeks after the pleurodesis, at which stage adhesions were obvious on ultrasound imaging (Figure 2). No adverse effects were observed on the fetus during the pregnancy and the neonate was well at the time of writing (1 month). Ultrasound demonstration at the low thoracic level shows the needle (N) used in the pleurodesis procedure passing through the fetal liver (Li) into the pleural cavity (Pl). Fetal ultrasound image of Case 9 2 weeks after pleurodesis, showing a bright adhesion band (arrow). The mode of action of OK-432 is thought to be cellular and cytokine-mediated. The elevated cytokine levels are likely to increase the permeability of the vascular endothelium within the cystic hygroma, and leukocytes are recruited. The local inflammatory reaction further increases the permeability of the lining endothelium and accelerates the drainage of accumulated lymph10. The maternal death in Case 7 was obviously of considerable concern. However, the time interval between the administration of OK-432 and amniotic fluid embolism was 8 weeks (OK-432 pleurodesis was performed at 16 weeks' gestation and maternal death occurred at 24 weeks), implying that the two incidents were unrelated. It is difficult to conclude that amniotic fluid embolism may be directly associated with OK-432 use because the disorder itself is poorly understood and its incidence is extremely low (1 in 30 000 to 1 in 8000)11. We acknowledge that it is controversial to provide therapy to aneuploid fetuses12, as occurred in Cases 7 (Turner syndrome) and 8 (Down syndrome). However, such action is the result of the autonomous choice of the prospective parents. Furthermore, fetal therapy was not continued when the response was poor in Case 7. In conclusion, we consider that the two requirements for successful OK-432 pleurodesis are (1) complete aspiration of the pleural cavity so that it is as dry as possible and (2) demonstration of adhesions by ultrasonography following the procedure. Technical difficulties encountered that may hinder the achievement of both requirements should be tackled as thoroughly as possible, and may even involve apparently radical steps such as piercing the fetal liver. This study was supported by research grants of the Changhua Christian Hospital, Changhua, Taiwan (CCH-9282), and of the National Science Council, Taiwan (94-2314-B-371-007) to M. Chen. M. Chen* ?, J. C. Shih?, B. T. Wang , C. P. Chen?, C. L. Yu , * Department of Obstetrics and Gynaecology, Changhua Christian Hospital, Changhua, Taipei, Taiwan, Centre for Medical Genetics, Changhua Christian Hospital, Changhua, Taipei, Taiwan, Department of Medical Genetics, National Taiwan University, and Hospital, Taipei, Taiwan, ? Department of Obstetrics and Gynaecology, National Taiwan University, and Hospital, Taipei, Taiwan, ?