Professor J. J. M. van Dongen, Department of Immunology, Erasmus University Rotterdam/University Hospital Rotterdam, PO Box 1738, 3000 DR Rotterdam, The Netherlands. E-mail: firstname.lastname@example.org
A 36-week pregnant woman was diagnosed with acute lymphoblastic leukaemia. Delivery was initiated prematurely, and a healthy child was born. Cord blood and peripheral blood samples from the neonate (obtained at 6 weeks, 3 months and 6 months) were analysed for the presence of minimal residual disease by polymerase chain reaction analysis of a leukaemia-specific IGH gene rearrangement and the E2A–PBX1 fusion gene transcript. In the cord blood sample, a tumour load of ≈ 4 × 10−4 was found, whereas all later blood samples were negative. Our data indicate that the maternal leukaemic cells did not engraft in the neonate.
It has become widely accepted that the placenta is not an absolute barrier, but that single cells can pass from mother to child and vice versa (Geha & Reinherz, 1983; Maloney et al, 1999; Zhou et al, 2000). Recently, Catlin et al (1999) reported transplacental transmission of a maternal natural killer cell lymphoma and subsequent engraftment of these cells in the fetus, with fatal consequences for the infant. In addition, direct evidence has been obtained for intraplacental metastasis of (pre)leukaemic cells between monozygotic twins who developed concordant acute lymphoblastic leukaemia (ALL) several years after birth (Ford et al, 1998). The clonal origin of the twin leukaemias was proved by the presence of exactly the same clonotypic fusion sequence of a chromosome aberration and an identically rearranged IGH allele.
Reports of patients with ALL during pregnancy are rare. In most cases reported so far, chemotherapy was given during pregnancy, and a healthy child was born (Bergstrom & Altman, 1998; Tewari et al, 1999). We report the presence of maternal ALL cells in cord blood (CB) of a neonate, born to a mother who was diagnosed with ALL during late pregnancy.
Materials and methods
Case report A previously healthy pregnant woman at 36 weeks' gestation presented with symptoms of abdominal pain, extensive night sweating and nose bleeding. She was referred to the St Antonius Hospital (Nieuwegein, The Netherlands) and diagnosed with ALL of pre-B-ALL immunophenotype (TdT+, CD10+, CD19+, CD22+, cytoplasmic CD79+, HLA-DR+ and cytoplasmic Igµ+) with 90% blast cells in her peripheral blood (PB). Before the start of the cytotoxic treatment, delivery was initiated prematurely, and a healthy female child (2·72 kg) was born. No abnormalities were found upon physical examination of the newborn, and her PB was normal by routine biochemical and cytomorphological analyses. The child developed normally, was released from the hospital 3 weeks after birth and has continued to do well for 30 months.
A CB sample was taken, and PB samples were obtained from the newborn at 6 weeks, 3 months and 6 months after birth. Mononuclear cells (MNCs) were obtained by Ficoll density centrifugation and used for molecular investigations.
Molecular analysis of blood samples from the newborn using Ig/TCR gene rearrangements as targets To detect clonal immunoglobulin (Ig) and T-cell receptor (TCR) gene rearrangements in the maternal ALL cells, DNA was extracted from MNCs obtained at diagnosis. Polymerase chain reaction (PCR) analysis was performed using 26 different primer combinations to detect rearrangements of the IGH, IGK, TCRD and TCRG gene loci (Pongers-Willemse et al, 1999). PCR products were denatured and renatured (60 min at 4°C) for heteroduplex analysis to confirm the clonality of the detected rearrangements. Monoclonal rearrangements were subjected to direct sequence analysis of the junctional regions (Pongers-Willemse et al, 1999).
The tumour load in the CB and PB samples from the child were analysed by dot blotting followed by hybridization with a patient-specific junctional region probe (van Dongen et al, 1998; Pongers-Willemse et al, 1999) and by real-time quantitative PCR (RQ-PCR) analysis as described previously (Pongers-Willemse et al, 1998). Tenfold serial dilutions of ALL DNA into DNA from normal MNCs (10−1−10−6) were made to define the sensitivity of the two PCR techniques and for assessment of the tumour load.
Molecular analysis of blood samples from the newborn using fusion gene transcripts as PCR target RNA was extracted from maternal MNCs at diagnosis, and cDNA was prepared. Reverse transcription (RT)–PCR was performed to detect the most common chromosome aberrations in ALL (van Dongen et al, 1999). RNA samples from CB and PB MNCs from the child were analysed by a nested RT–PCR approach using E2A and PBX1 primers as described previously (van Dongen et al, 1999). To check for the integrity of the RNA samples, an ABL RT–PCR was performed. In addition, RQ-PCR analysis was performed for the E2A–PBX1 fusion transcript. The forward primer (5′-GAC TCC TAC AGT GCT TCC CTG TTT AT-3′) and TaqMan probe (AGC CCA GGA GGA GGA ACC CAC AGA) were positioned in exon 13 of the E2A gene, and the reverse primer (5′-CGC TAA CAG CAT GTT GTC CAG-3′) was positioned in exon 2 of the PBX1 gene. Tenfold serial dilutions of ALL RNA into HL-60 RNA (10−1−10−6) were made to define the sensitivity of the two PCR techniques and for assessment of the tumour load.
PCR heteroduplex analysis and subsequent sequencing revealed two clonal IGH gene rearrangements, of which a VH3–11–JH6c rearrangement was selected for this study. A junctional region-specific probe was developed for PCR dot-blot analysis reaching a sensitivity of 10−5(Fig 1). Analysis of the CB sample showed the presence of maternal leukaemic cells, with an estimated tumour load of 10−3. The PB samples taken at 6 weeks and 3 months were negative (Fig 1). RQ-PCR analysis of the VH3–11–JH6c rearrangement resulted in a sensitivity of 10−4. The calculated tumour load of the CB sample was 5 × 10−4, whereas the later neonatal PB samples were negative.
RT–PCR analysis of the diagnosis sample identified the E2A–PBX1 fusion transcript, characteristic for t(1;19) and in line with the pre-B-ALL immunophenotype. Via nested RT–PCR analysis, E2A–PBX1 fusion transcripts were found in the CB sample, whereas all neonatal PB samples were negative. RQ-PCR analysis of the E2A–PBX1 fusion gene transcript resulted in a sensitivity of 5 × 10−5. The relative tumour load of the CB sample was 3 × 10−4, whereas all PB samples were negative.
This is the first report on the presence of maternal ALL cells in the CB sample of a neonate. The occurrence of maternal leukaemic cells in the CB sample and in the child's PB samples was determined by analysis of a clonal IGH rearrangement and E2A–PBX1 fusion gene transcripts via classical PCR approaches (van Dongen et al, 1998, 1999; Pongers-Willemse et al, 1999) as well as via the recently developed RQ-PCR technique (Pongers-Willemse et al, 1998). Fully concordant results were obtained with the four different minimal residual disease (MRD) PCR techniques, and accurate quantification of the leukaemic load (≈ 4 × 10−4) was achieved by the RQ-PCR analyses.
Formally, we cannot exclude the possibility that the CB sample has been contaminated with a minor amount of maternal PB during sampling, but we took all necessary precautions to obtain uncontaminated CB. However, it is fair to assume that the maternal leukaemic cells passed the placental barrier, like normal maternal leucocytes (Geha & Reinherz, 1983; Maloney et al, 1999; Zhou et al, 2000) and natural killer cell lymphoma cells (Catlin et al, 1999). Consequently, we conclude that these maternal leukaemic cells did not engraft in the fetus: within 6 weeks after birth, the leukaemic cells became undetectable and were probably eliminated by the immune system.