SEARCH

SEARCH BY CITATION

Introduction

  1. Top of page
  2. Introduction
  3. Case histories
  4. Materials and methods
  5. Cytological and immunocytochemical findings
  6. Discussion
  7. References

Two cases of pleural effusion in neonates with fetal hydrops are reported. In both, the cytomorphology mimicked acute lymphoblastic leukaemia (ALL). However, immunocytochemistry and the examination of cord blood revealed reactive lymphocytes. Cytopathologists should be aware of this pitfall, which highlights the need to immunophenotype cells in such fluids.

Case histories

  1. Top of page
  2. Introduction
  3. Case histories
  4. Materials and methods
  5. Cytological and immunocytochemical findings
  6. Discussion
  7. References

Case 1

A 26-year-old woman (gravida 1, para 1) was referred at 20 weeks’ gestation with a diagnosis of fetal hydrops, with bilateral pleural effusion and ascites detected by ultrasonography and magnetic resonance imaging (MRI). In the amniotic fluid, the chromosomal pattern was normal, and DNA for cytomegalovirus or parvovirus B19 was negative by polymerase chain reaction (PCR) analysis. Ultrasound-guided aspiration was performed for the fetal bilateral pleural effusion (left, 2.0 ml; right, 7.7 ml) and ascites (6.1 ml) at 21 weeks’ gestation, and the bilateral pleural effusion (left, 69 ml; right, 71 ml) at 30 weeks’ gestation. These fluids contained a large number of blast-like cells. Percutaneous umbilical cord blood was obtained 2 days after the re-aspiration performed at 30 weeks’ gestation because of a suspected diagnosis of ALL. However, no atypical cells characteristic of ALL were identified in the blood film.

Case 2

A 35-year-old woman (gravida 2, para 2) was referred at 35 weeks’ gestation with a diagnosis of fetal hydrops, with bilateral pleural effusion and skin oedema having been detected by ultrasonography and MRI. An antibody screen was negative. At 37 weeks’ gestation, a 2.4-kg female infant was born by caesarean delivery after ultrasound-guided aspiration for fetal bilateral effusions (total volume, 130 ml). These effusions were infiltrated by numerous immature cells. At 1 hour after the birth, re-aspiration was performed (total volume, 40 ml). A few erythroblasts were found in a peripheral blood smear, but no lymphoblasts. On days 7, 29 and 32 after the birth, re-aspiration was performed (total volumes of 40, 76 and 15 ml, respectively). Pleural effusion did not recur later than day 32 after the birth.

Materials and methods

  1. Top of page
  2. Introduction
  3. Case histories
  4. Materials and methods
  5. Cytological and immunocytochemical findings
  6. Discussion
  7. References

Pleural and ascitic fluids were centrifuged at 630 g for 5 minutes. One of the smears was air dried and stained by Wright–Giemsa, whereas the others were fixed in 95% alcohol for immunocytochemistry.

For immunocytochemistry, we employed the polymer-peroxidase method [Histofine® Simple Stain MAX PO (MULTI); Nichirei Biosciences (Nichirei), Tokyo, Japan]. We used commercially available polyclonal antibodies against CD3 [DAKO Cytomation (DAKO), Glostrup, Denmark; diluted 1 : 300], CD117 (c-KIT) (Nichirei; diluted 1 : 1) and terminal deoxynucleotidyl transferase (TdT) (DAKO; diluted 1 : 50). In addition, we used monoclonal antibodies against CD4 (4B12; Nichirei; diluted 1 : 80), CD8 (C8/144B; Nichirei; diluted 1 : 80), CD10 (56C6; Leica Microsystems, Newcastle, UK; diluted 1 : 50), CD20 (L26; DAKO; diluted 1 : 1000), CD34 (NU-4A1; Nichirei; diluted 1 : 200) and CD79acy (HM57; DAKO; diluted 1 : 100). For TdT, sections were autoclaved in 0.05 M citrate buffer, pH 6.5, for 13 minutes. The percentage of immunoreactive cells was determined in at least 400 cells.

Cytological and immunocytochemical findings

  1. Top of page
  2. Introduction
  3. Case histories
  4. Materials and methods
  5. Cytological and immunocytochemical findings
  6. Discussion
  7. References

In both cases, the smears revealed a large proportion of monotonous blast-like lymphoid cells: 95.3 ± 1.8% (mean ± standard deviation) in Case 1 (bilateral pleural effusions and ascites at 21 weeks’ gestation, and bilateral pleural effusions at 30 weeks’ gestation) and 94.8 ± 1.5% in Case 2 (bilateral pleural effusions at 37 weeks’ gestation and bilateral pleural effusions at 1 hour after the birth), together with a few histiocytes and eosinophils. The blast-like lymphoid cells displayed a high nuclear to cytoplasmic ratio, irregular nuclear outlines, fine dispersed chromatin and visible nucleoli (Figure 1a,b). In addition, a few erythroblast-like cells were observed in Case 2 (Figure 1b). In Case 2, in the right pleural fluid taken on day 7 after the birth, the percentage of blast-like lymphoid cells had decreased to 76.1% of total cells. Pleural fluids obtained on days 29 and 32 after the birth contained a few blast-like lymphoid cells.

image

Figure 1.  (a, b) Cytology of ascitic fluid smears at 21 weeks’ gestation in Case 1 (a) and of pleural fluid smears at 37 weeks’ gestation in Case 2 (b); a large number of monotonous blast-like lymphoid cells showing a high N/C ratio, irregular nuclear outline, fine dispersed chromatin and visible nucleoli (a, b), together with a few erythroblast-like cells (arrow) in Case 2 (b) (Wright–Giemsa). (c–f) Immunocytochemical staining of Case 1: (c) CD3; (d) CD4; (e) CD8; (f) CD20. Bars, 20 μm. Original magnification, ×100.

Download figure to PowerPoint

In the immunocytochemistry, CD3-positive and CD4-positive cells predominated in each case (Figure 1c–f), and no cells were positive for CD10, CD34, c-KIT or TdT (Table 1).

Table 1. Percentage of immunocytochemically positive cells in serous effusions in the two cases
 Case 1 (n = 5*)Case 2 (n = 4)
Mean ± standard deviation
  1. *Five samples of bilateral pleural effusions and ascites at 21 weeks’ gestation and bilateral pleural effusions at 30 weeks’ gestation.

  2. Four samples of bilateral pleural effusions at 37 weeks’ gestation and bilateral pleural effusions at 1 hour after birth.

  3. TdT, terminal deoxynucleotidyl transferase.

CD383.9 ± 1.282.7 ± 1.7
CD473.8 ± 2.775.7 ± 2.9
CD820.8 ± 2.616.4 ± 3.4
CD2010.1 ± 4.78.9 ± 3.5
CD79acy7.2 ± 2.76.8 ± 2.4
CD100 ± 00 ± 0
CD340 ± 00 ± 0
c-KIT0 ± 00 ± 0
TdT0 ± 00 ± 0

Discussion

  1. Top of page
  2. Introduction
  3. Case histories
  4. Materials and methods
  5. Cytological and immunocytochemical findings
  6. Discussion
  7. References

Fetal hydrops is a rare condition defined as an abnormal accumulation of fluid in two or more fetal compartments.1,2 Although several previous reports have revealed similarities between the lymphocytes present in body fluids obtained from a chylothorax and the cellular morphological findings typical of ALL,3–5 few cytopathologists have so far had the opportunity to diagnose using such fluids. However, developments in perinatal medicine have led to an increase in such cytological materials. In the present study, we reveal the morphological findings relating to the lymphoid cells present in body fluids obtained from two cases of fetal hydrops, and demonstrate that these lymphoid cells could not be distinguished morphologically from the lymphoblasts found in ALL or from benign stem cells. However, we were able to diagnose them as mature lymphocytes by immunocytochemistry of body fluids and cytological analysis of peripheral blood or PUB. For this reason, we wish to emphasize that immunocytochemistry and flow cytometry of body fluids, and/or cytological analysis of either peripheral blood or percutaneous umbilical cord blood, are helpful in avoiding a diagnostic pitfall.

In the immunocytochemistry in the present cases, the majority of the lymphoid cells had a mature T-cell phenotype (CD3 positive, CD4 and CD8 variably positive, and TdT negative). These are the usual immunohistochemical findings for mature lymphocytes. Furthermore, in the present two cases, the mean proportions obtained for CD4 and CD8 cells were 73.8–75.7% and 16.4–20.8%, respectively. Thus, in our cases, CD4 was predominant. However, when Horn and Penchansky3 examined six cases in which infants had chylous pleural effusions, they found that the ratio between CD4 and CD8 was variable, as befitting a peripheral population consisting of a mixture of helper and suppressor T phenotypes. In ALL, the immunohistological features usually include positivity for TdT. Furthermore, in T-ALL, the T cells are double negative for CD4 and CD8, whereas, in cortical T stage, they are double positive for CD4 and CD8; the medullary stage may express either CD4 or CD8.

In conclusion, we report cytological findings relating to lymphoid cells in body fluids in fetal hydrops. These findings emphasize the existence of a potential diagnostic pitfall, and thereby highlight the need to immunophenotype the cells present in such fluids.

References

  1. Top of page
  2. Introduction
  3. Case histories
  4. Materials and methods
  5. Cytological and immunocytochemical findings
  6. Discussion
  7. References
  • 1
    Abrams ME, Meredith KS, Kinnard P, Clark RH. Hydrops fetalis: a retrospective review of cases reported to a large national database and identification of risk factors associated with death. Pediatrics2007;120:849.
  • 2
    Rocha G. Pleural effusions in the neonate. Curr Opin Pulm Med2007;13:30511.
  • 3
    Horn KD, Penchansky L. Chylous pleural effusions simulating infiltrate associated with thoracoabdominal disease and surgery in infants. Am J Clin Pathol1999;111:99104.
  • 4
    Kallanagowdar C, Craver RD. Neonatal pleural effusion. Spontaneous chylothorax in a newborn with trisomy 21. Arch Pathol Lab Med2006;130:e223.
  • 5
    Kren L, Rotterova P, Hermanova M et al. Chylothorax as a possible diagnostic pitfall: a report of 2 cases with cytologic findings. Acta Cytol2005;49:4414.