The prevalence of maternal F cells in a pregnant population and potential overestimation of foeto-maternal haemorrhage as a consequence

Abstract

Background

Acid elution (AE) is used to estimate foeto-maternal haemorrhage (FMH). However AE cannot differentiate between cells containing foetal or adult haemoglobin F (F cells), potentially leading to false positive results or an overestimate of the amount of FMH. The prevalence of F cells in pregnant populations remains poorly characterised. The purpose of this study was to ascertain the incidence of HbF-containing red cells in our pregnant population using anti-HbF-fluorescein isothiocyanate flow cytometry (anti-HbF FC) and to assess whether its presence leads to a significant overestimate of FMH.

Material and methods

Eighty-eight pregnant patients were assessed for the presence of F cells and foetal red cells by AE and anti-HbF FC. The “FMH equivalent”, estimated by AE and anti-HbF FC, was calculated.

Results

Thirty-six percent of the pregnant population had F-cell populations detectable by anti-HbF FC while AE detected F cells in 48% of the population. The mean estimated FMH equivalent determined by AE and anti-HbF FC was 0.59 mL (0–23.93 mL) and 0.41 (0 to 2.19 mL), respectively (p=0.012). In 3% of our population, AE overestimated the FMH by >3 mL due to the presence of an F-cell population of at least 16%.

Discussion

Thirty-six percent of a prospectively evaluated group of consecutive pregnant women were found to have F-cell populations. In some patients, these findings were clinically significant as AE overestimated the degree of FMH as a consequence.

Introduction

Rh haemolytic disease of the foetus and newborn (HDFN) was a major cause of perinatal morbidity and mortality in the past¹. Since the advent of anti-D prophylaxis to RhD-negative women, the incidence has plummeted². It is critical that sufficient anti-D is administered when indicated, to prevent RhD alloimmunisation². It is, therefore, important to quantify the amount of the foeto-maternal haemorrhage (FMH) in order to administer the correct amount of anti-D³. The Kleihauer-Betke, or acid elution (AE), test is a widely used method of estimating FMH⁴. This test is based on the principle that adult haemoglobin (HbA) is eluted from red cells in an acidic solution, while foetal haemoglobin (HbF) is resistant to this technique⁵. Rarely, red cells of maternal origin containing unusually high levels of HbF can persist. These may be detected using a flow cytometry technique which differentiates between HbF of maternal and foetal origin by employing a staining technique to detect intracellular HbF⁶. This finding is of no clinical consequence to the mother but can result in a false positive AE test. A false positive AE result can lead to unnecessary administration of prophylactic anti-D^6–8. The prevalence of elevated HbF-expressing red cells of maternal origin in a pregnant population is unknown and reports of such cases resulting in excessive anti-D administration are scarce^6–9. In order to determine the true prevalence of false positive AE tests due to elevated maternal HbF, we aimed to assess the prevalence in a population of pregnant or recently-postpartum women.

Materials and methods

Prior to its commencement, the research study was approved by the Research and Ethics Committee at the Rotunda Hospital and the University of Ulster at Coleraine.

All antenatal patients who had experienced a potentially sensitising antenatal event associated with the risk of FMH, as defined by British Committee for Standards in Haematology (BCSH) guidelines³ between June 2012 and August 2012 were included in the study. Furthermore during the same period all immediate postnatal RhD-negative mothers who delivered a RhD-positive baby were included. Testing was performed by AE and fluorescein isothiocyanate anti-HbF flow cytometry (FC).

Five millilitres of adult blood samples were taken by venepuncture into EDTA-anticoagulated tubes (Vacuette, Greiner, Kreimsmünster, Austria) and were stored at 2–8 ºC until testing. Samples from postnatal RhD-negative cases were taken 30 minutes after delivery while antenatal samples were taken at admission for suspected sensitising events. All samples were tested within 72 hours.

To determine foetal red blood cell volume by AE, whole blood was mixed with phosphate-buffered saline (Baker, Deventer, The Netherlands) in a 1:1 (v/v) ratio. Cells were spread on slides, air-dried, fixed then stained using reagents from a commercial Kleihauer kit (Immucor, Rödermark, Germany). Smears were analysed according to current BCSH guidelines³. The degree of smear positivity was graded as follows: very dark pink (3+), moderately pink (2+) and pale pink (1+). At least 10,000 red cells were counted and results are reported as percentages of positive cells. The presence of moderately pink or stronger staining cells was considered to represent the presence of a significant HbF-containing red blood cell population.

Flow cytometry to detect HbF was performed using the QuikQuant kit (Trillium Diagnostics, Groningen, The Netherlands). Cells were labelled with anti-HbF- fluorescein isothiocyanate according to the manufacturer’s instructions and then analysed on an FC500 Cytomics flow cytometer (Beckman Coulter, Miami, FL, USA) with a minimum of 50,000 events considered. A tri-level FETAtrol control (IQ Products, Groningen, The Netherlands) was run with each batch. In order to exclude debris and agglutinated cells, red cells were gated on log/log forward scatter (FS) versus side scatter (SS). Autofluorescence was excluded on SS vs FL2 dot plots. Compensation was adjusted using FL1 vs FL2 dot plots. A red cell histogram was generated with marker regions representing a HbA peak (which exhibited no anti-HbF staining), an F-cell peak (if present; defined as cells exhibiting intermediate anti-HbF staining and arising as a “shoulder” from the negatively staining HbA peak) and a foetal HbF peak (cells exhibiting very strong anti-HbF staining)⁶. The percentage of gated cells was determined by software (CXP Cytometer, version 2, Beckman Coulter) and foetal red cells were quantified and expressed as the percentage of foetal cells detected. The percentages and mean fluorescence of these gated cells under the markers were determined automatically using computer software. For both methods the volume of FMH was calculated as % foetal cells×21.96=FMH (mL packed red cells)¹⁰.

Statistical analysis

The volumes of FMH, calculated by AE and FC, were tested for normality of distribution by the Shapiro-Wilk test. As the population was found not to be normally distributed (p=0.205 and p=0.775, for AE and FC respectively), the non-parametric Wilcoxon signed rank test was used to compare differences between groups, using SPSS version 20 (IBM, Armonk, NY, USA).

Results

A total of 88 patients were recruited. Thirty-two (6 RhD-positive and 26 RhD-negative) were antenatal patients who underwent FMH determination due to a potentially sensitising event at a median of 39 (11–40+) weeks gestation and 56 were RhD-negative patients recruited following delivery of a RhD-positive infant. The presence of an F-cell population by anti-HbF FC was defined as the visual detection of a distinct peak in the adult HbF region of the FC histogram or as >5% cells in the adult HbF gate. For AE, the presence of F cells was defined by moderate to strong pink staining, as the majority of F-cell populations are heterocellular (due to varying cellular HbF concentration). F-cell populations were detected in 36% and 45% of the pregnant population by FC and AE respectively (Table I). As anti-HbF FC is considered more effective in discriminating maternal F cells and foetal cells⁶, our final reported F-cell incidence (36%) was determined from the results obtained using this technique.

Table I.

Prevalence of maternal F-cell populations, as characterised by anti-haemoglobin F flow cytometry (anti-HbF FC) and acid elution (AE).

Gestation	N.	Detection of F cells by anti-HbF FC		Detection of F cells by AE
		No	Yes	Negative-weak staining	Moderate-strong staining	Strong staining only
<24 weeks	5	1	4	2	3	3
24–<28 weeks	10	5	5	5	5	3
28–<32 weeks	7	4	3	4	3	3
32–<36 weeks	4	2	2	1	3	3
≥36 weeks	62	44	18	36	26	22
Total	88	56 (64%)	32 (36%)	48 (55%)	40 (45%)	34 (36%)

In order to assess whether FMH was being overestimated by AE due to the presence of maternal F cells, the FMH equivalent was calculated by AE and anti-HbF FC (Table II). The mean FMH equivalent determined by AE was significantly greater than that determined by anti-HbF FC, being 0.59 (0–23.93) mL and 0.41 (0–2.19) mL, respectively (p=0.012).

Table II.

Comparison of FMH equivalent estimation using anti-haemoglobin F flow cytometry (anti-HbF FC) and acid elution (AE).

Gestation	Mean FMH (mL) (range)
	Equivalent FMH by anti- HbF FC	Equivalent FMH by AE
<24 weeks	0.31 (0.22–0.66)	0.18 (0–0.44)
24–<28 weeks	0.44 (0–1.32)	0.15 (0–0.66)
28–<32 weeks	0.55 (0–1.10)	0.16 (0–0.44)
32–<36 weeks	0.38 (0.22–0.66)	0.33 (0–23.94)
≥36 weeks	0.40 (0–2.20)	0.33 (0–3.95)
Total	0.41 (0–2.19)	0.59 (0–23.93)

The potential clinical impact of maternal F cells in the overestimation of FMH by AE was characterised by analysis of a subgroup of nine patients found to have a large FMH equivalent (>1 mL) by AE (Table III). These patients’ FMH equivalents were 1.10–23.94 mL by AE and 0.22–1.98 mL by anti-HbF FC, the latter quantifications being overall much lower. The differences in FMH equivalent between the two methods ranged from 0.34 mL to a remarkable 23.28 mL. In six patients the FMH equivalent was overestimated by >1 mL and four of these patients had detectable maternal HbF levels ranging from 8.88 to 32.3% (Table III). Overall, it was observed that the larger the maternal F-cell population detected by FC, the greater the overestimation of FMH equivalent by AE. In particular, in the three patients who had maternal F-cell populations >16%, the FMH equivalent was overestimated by >3 mL (Table III).

Table III.

Comparison of volumes of foeto-maternal haemorrhage (FMH) determined by acid elution (AE) and anti-haemoglobin F flow cytometry (anti-HbF FC) in patients with 0.88 mL FMH by AE.

Patient	AE		Anti-HbF FC			Over-estimation
	% foetal-equivalent staining cells	FMH equivalent (mL)	% foetal cells	% F cells	FMH equivalent (mL)	mL
1	1.09	23.94	0.03	32.3	0.66	23.28
2	0.18	3.95	0.03	28.56	0.66	3.29
3	0.15	3.29	0.01	16.26	0.22	3.07
4	0.10	2.20	0.09	1.69	1.98	0.02
5	0.09	1.98	0.04	8.88	0.88	1.10
6	0.08	1.76	0.02	3.31	0.44	1.32
7	0.08	1.76	0.06	7.39	1.32	0.44
8	0.07	1.54	0.01	1.79	0.22	1.32
9	0.05	1.10	0.03	1.8	0.66	0.34

Discussion

FMH in a RhD-negative woman carrying a RhD-positive foetus can result in RhD alloimmunisation. To prevent this complication, it is recommended that RhD-negative women undergo AE testing in addition to receiving a standard dose of anti-D immunoglobulin. This is carried out in the setting of a potentially sensitising antenatal event after 20 weeks’ gestation or upon delivery of a RhD-positive infant³. The purpose of AE testing is to determine whether the volume of foetal red cells present in the maternal circulation exceeds that neutralised by a standard dose of anti-D³. Generally, AE-resistant red cells are interpreted to represent foetal red cells. The prevalence of false positive and potentially, clinically significant AE tests due to a population of maternal red cells expressing a higher than usual level of HbF is unknown. Recent retrospective studies found that the misinterpretation of maternal F cells as FMH was common^11,12. Approximately one third of 69 RhD-negative pregnant women with a positive AE test were found to have >4% F cells¹². Subsequently, 10% received an overdose of anti-D¹². Concordantly, 36% of our broader pregnant population were found to have >5% F cells. We included RhD-positive and -negative pregnant patients with gestation from 11 weeks to term regardless of their AE result. Furthermore, we demonstrated in a prospectively evaluated population of antenatal and postnatal patients, that patients expressing higher red cell maternal HbF are likely to have a falsely elevated FMH equivalent determined by AE and, of critical importance, that this overestimation may reach clinical significance with an associated potential for inappropriate anti-D administration.

HbF is the major haemoglobin synthesised during foetal and early neonatal life, and consists of two α and two γ globin chains¹³. In contrast to adult haemoglobin, HbF has a very high oxygen affinity and therefore facilitates the transfer of oxygen across the placenta from mother to foetus. A shift from γ to β globin synthesis occurs from gestational week 30 (resulting in approximately 50–70% HbF at term), but is not complete, with the result that most adults have <1–2% HbF^13,14. There is, however, significant variation in adult HbF levels. Residual low level HbF continues to be synthesised during adult life but unusually high HbF levels are occasionally present¹³. This elevation in HbF can be pancellular (occurring to the same extent in all red cells) or present in only a proportion of red cells (F cells)¹⁵. Using specialised techniques, F cells can be detected in up to 7% of normal adults¹⁵. We demonstrated a higher incidence of F cells, which were detectable in 36% of our pregnant population.

Inter-individual variation in HbF expression may be genetic or acquired¹³. Elevated HbF levels can occur in medical conditions including sickle cell anaemia and β thalassaemia and have been described in 17% of pregnancies, reaching a peak at 18–22 weeks gestation^15–18. Concordantly with these resports, we found a 5-fold increase in F cells during pregnancy compared to the levels in non-pregnant populations. Interestingly, this pregnancy-related increase in HbF may be associated with an increase in the number of maternal F cells rather than an overall increase in HbF in all maternal red cells¹⁸.

The incidence of FMH is relatively low and the volume of haemorrhage was estimated to be greater than 1 mL and 30 mL in 4% and 0.03% of deliveries, respectively¹⁹. However, these estimates were based on the AE test. As we have demonstrated the AE test is falsely elevated when maternal F cells are present and a third of the pregnant population expresses such cells, so the incidence of FMH may be overestimated. Anti-HbF FC has been shown to be accurate in the estimation of foetal red cells in a maternal sample and it is capable of distinguishing these cells from maternal F cells. The incidence of FMH may, therefore, be better ascertained using anti-HbF FC techniques.

The AE test is insufficiently sensitive to detect low numbers of circulating F cells in most pregnant women¹⁸. However, in the setting of a high-normal baseline F-cell number, it is possible that physiological changes occurring during pregnancy may promote HbF expression in F cells and increase F-cell numbers to levels detectable by the AE method, leading to the erroneous conclusion that foetal red cells are present in the maternal blood. This interpretation has the potential to lead to excessive anti-D administration¹. Early awareness of the possibility of elevated maternal F cells ensures that samples can be sent to appropriate reference laboratories at the earliest possible opportunity to limit unnecessary Anti-D administration.

In summary, we reveal a previously poorly recognised limitation of the AE test: false positivity due to elevated maternal HbF in maternal red blood cells. Future collaborative studies are necessary to further characterise this limitation, which has the potential to translate into unnecessary anti-D administration.