The Kell protein is a zinc endopeptidase that converts endothelin-3 to an active vasoconstrictor. The Kell antigen is present on immature erythroid progenitor cells in early foetal life at 10–11 weeks of gestation.4 It is believed that Kell proteins are believed to play a role in RBC growth and differentiation.5

HDFN associated with maternal anti-K1 may result in severe foetal anaemia, foetal hydrops, asphyxia, and perinatal death. Hyperbilirubinemia can lead to kernicterus and neurological disabilities in newborns. Unfortunately, a foetus with anti-K1 can develop hydrops during early pregnancy, hydrops can develop at <20 weeks of gestation.6 Severe foetal anaemia should be detected before the development of hydrops.

A study by Lindenburg et al. revealed that the occurrence of anti-K1 was significantly higher (24 %) in pregnancies requiring intrauterine transfusions (IUTs) prior to reaching 20 weeks compared to transfusions after 20 weeks (11 %).7

Studies have shown that the survival rate of RhD blood group incompatibility is 89 %, which is higher than the HDFN caused by anti-K1 (58 %); the difference is due to delayed diagnosis. Administration of anti-D antibodies during pregnancy also decreased the RhD incidence of Rh D alloimmunisation.

The current management of pregnancies complicated by maternal Kell (K1) antibodies includes antibody screening. Antibodies are identified during the indirect antiglobulin test phase. Early antibody screening can help improve the pregnancy outcome,8 with the next step after confirming the presence of an antibody being to identify the fatherʼs zygosity. If the father is heterozygous (K, k), foetal Kell typing is performed through amniocentesis. In rare cases, if the father is homozygous (K, K), antibody titration is performed without foetal Kell typing. Heterozygous K-positive cells (single doses) are used for the antibody titration. Two registered technologists typically determine the titres to ensure agreement.9 There are many observations to determine the threshold values for anti-K1 titres. Since the disease occurs with low titres, many studies have concluded that there is no correlation between titres and pregnancy outcome.8 The amniotic fluid bilirubin test is not very helpful and does not correlate with the severity of foetal anaemia. In one study, a threshold of 1:2 was used, which was monitored from 17 weeks onwards; when the tires were higher than the threshold, a middle cerebral artery (MCA) ultrasound was performed to measure the peak systolic volume to evaluate the PSV.4 If the multiples of the median (MoM) is <1.5, Doppler ultrasound of the MCA is repeated after 1–2 weeks. If MoM >1.5 or hydrops is present, foetal blood sampling was performed. Haematological parameters help clinicians decide the need for IUT. Antibody titres and Doppler ultrasound help determine the necessity of foetal blood sampling and the need for intervention.

The risk of developing severe anaemia and hydrops in pregnancies with anti-K1 can be predicted by non-invasive investigations.

Foetal blood sampling is the gold standard test to detect foetal anaemia, but it carries a foetal loss rate of 1.4 % per procedure.10

This study aimed to explore the effectiveness of non-invasive investigation, anti-K1 titre, and MCA ultrasound before cordocentesis in detecting high-risk pregnancies and improving perinatal outcomes.

The meta-analysis followed the Preferred Reporting Items for Systemic Reviews and Meta-Analysis (PRISMA) protocol to explore and identify studies investigating the role of anti-K1 titres and peak systolic velocity (PSV) of the foetal MCA in predicting the severity of foetal anaemia and hydrops, and the need for IUT in Kell alloimmunised pregnancies.11 The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were used to ensure the quality of the included studies.12

For this meta-analysis, relevant literature was obtained using an electronic database and manual search. PubMed, Embase, and Science Direct databases were systematically searched from January 2000 to June 2022. Google Scholar was manually searched. The same search terms, phrases, and test descriptions were used for both manual and electronic searches. The keywords, phrases and test descriptions included: “haemolytic disease of the foetus and newborn”, “anti-K1 titre”, “Kell alloimmunised pregnancies”, “Maternal Kell alloantibody”, “hydrops”, “intrauterine transfusion”, “MCA Doppler ultrasound”, “middle cerebral artery peak systolic velocity”, “hyporegenerative anaemia”, “hyperbilirubinemia” and “Decreased reticulocytes in Kell alloimmunised pregnancies”. Furthermore, some references in the articles were identified as appropriate studies.

Articles were identified using the search strategy and screened using the designated exclusion and inclusion criteria to obtain the eligible studies for the meta-analysis. After scrutinising the titles and abstracts, the full manuscript of potentially eligible articles was obtained. Some studies were manually selected based on predefined criteria. Articles with cell-alloimmunised pregnancies were selected. Studies were included if data on the anti-K1 cut-off titre, foetal anaemia, hydrops, intrauterine foetal death (IUFD) and MCA Doppler ultrasound for PSV could be extracted. The articles included were observational, retrospective, or prospective cohort studies.

Systematic reviews, case studies, literature reviews, and conference proceedings were excluded. Articles in languages other than English and those with unavailable translations were excluded.

Some studies were excluded because their full text was not available. Moreover, articles from which data related to Kell alloimmunised pregnancies could not be extracted were excluded.

The quality of all articles was assessed in terms of population, intervention, and outcomes. The eligible studies were evaluated for methodological quality and bias using published standards outlined by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.12

The information obtained from each study is summarised in the tables. The study characteristics and sample sizes tabulated are shown in Table 1. Table 2 illustrates the outcomes of anti-K1 sensitised pregnancies using anti-K1 titre cut-off values as a non-invasive test. The consequences of Kell-sensitised pregnancies using PSV of the foetal MCA as a non-invasive procedure above and equal to the cut-off value of 1.5 MoM are listed in Table 3.

A forest plot for the meta-analysis was constructed from data in Tables 3 & 4 by using Open Meta-Analyst software (version 4.06.15) downloaded from the Brown University website. Two-armed proportion analysis was used to express the Arcsine risk difference and 95 % confidence interval using binary random effect model with maximum likelihood. A forest plot was constructed with 95 % confidence intervals, p-values and heterogeneity. A p-value ≤ 0.05 was considered statistically significant.

Initially, 5143 potential articles were identified in PubMed, Science Direct, Scopus and Embase. In addition, 24 potential articles were identified through a manual search of Google Scholar. After removing duplicates, the titles of 3223 studies were screened. Most (2962 titles) were found to be irrelevant to the topic. Furthermore, 218 studies were excluded after screening the abstracts. The full texts of 42 articles were thoroughly explored. Twenty-five studies were excluded after reading the full texts. Ultimately, 18 studies were included in the analyser and were pooled for systematic review. As eight of them lacked sufficient data for quantitative analysis, they were excluded from the study and thus ten articles were found to have relevant data for the meta-analysis (Figure 1).

Figure 1.

PRISMA flow chart for the identification, screening, eligibility and inclusion of studies in the systematic review and meta-analysis of the role of anti-K1 titres and MCA-PSV in predicting the severity of anaemia, hydrops foetalis and requirement of intrauterine transfusion.

(0.53MB).

The included studies were retrospective cohort studies4,13–17 and prospective studies.10,18–20 The citations included studies from six different countries from 1999 to 2021. The sample included Kell-alloimmunised pregnant women and Kell-positive foetuses. In all studies, there were singleton foetuses except for one study that had one twin pregnancy.15Table 1 summarises the characteristics of the studies.

The roles of monitoring anti-K1 titres in six articles are summarised in Table 2.4,13,15–17,20 The titration methods are described below. In one study, this technique was not specified.17 In each eligible study, the lowest cut-off titre used to monitor pregnancy by foetal ultrasound of the middle cerebral artery was specified. Additionally, anti-K1 titres at the first IUT were outlined in each study. The number of Kell alloimmunised women who underwent IUT with the median gestational age at which they were transfused is also specified in Table 2. The median haemoglobin level at which the first IUT was performed was also extracted from the articles. In two studies, the median haemoglobin value was not determined.13,15 The outcomes of Kell-alloimmunised pregnancies are shown in Table 3. The number of cases with hydrops in each study, intrauterine deaths and live births in Kell-alloimmunised pregnancies are included.

The frequency of IUTs in Kell-sensitised pregnancies with median gestational age when PSV of the foetal MCA ≥1.5 multiple of medians (MoM) is summarised in Table 3. The results were obtained from five eligible studies.4,10,14,16,18 Median PSV (cm/s) was given in three studies.4,18,10 Two studies also mentioned the anti-K1-specific sensitivity and specificity of MCA Doppler ultrasound for peak systolic velocities.10,18

Strobe criteria were used to assess the quality of the eligible studies. The most relevant criteria are listed in Table 4. Most of the studies were high-quality studies and fulfilled the criteria for methodological quality assessment; however, in two articles, limitations were not discussed.13,19 One study failed to discuss all outcomes.18 In three articles, efforts to address potential bias were not defined.4,14,15 Most of the studies had high methodological quality, which suggests that the risk of bias through irrelevant articles is low (Table 5).

The proportion of IUT, hydrops foetalis and IUFD are summarised in forest plots in Figure 2A–C based on six included citations.4,13,15,17,20 Specific cut-off values of anti-K1 titres before monitoring by Doppler ultrasound and before IUT were determined in the studies, as shown in Table 3.

Figure 2.

A: Forest plot of meta-analysis of number of Kell alloimmunised pregnancies with and without intrauterine transfusions at different cutoff antibody titres. B: Forest plot of meta-analysis of number of Kell alloimmunised pregnancies of newborns with and without hydrops at different cutoff antibody titres. C: Forest plot of meta-analysis of number of Kell alloimmunised pregnancies with and without intrauterine foetal deaths at different cutoff antibody titres. D: Meta-analysis of intrauterine transfusion for Doppler ultrasound middle cerebral artery multiples of the median ≥1.5.

(0.16MB).

The Forest plots were interpreted using two-arm proportion analysis which was expressed in Arcsine risk difference 95 % confidence interval (95 % CI) using binary random effect model with maximum likelihood.

The numbers of pregnant women with and without IUT in Kell alloimmunised pregnancies at different anti-K1 cut-off titres and gestational weeks, as shown in Table 2, were analysed using a forest plot.

The result of the meta-analysis on IUT was found to be statistically significant (absolute risk difference - ARD: 0.351 95 % CI: −0.593 to −0.109; p-value = 0.004); however, high heterogeneity can be seen as the study has variable outcomes (p-value <0.001 and I2value = 79.928) as interpreted by the forest plot in Figure 2A.

The proportion of pregnancies with and without hydrops is summarised in a forest plot (Figure 2B). This result was statistically significant (ARD: 0.808; 95 %CI −1.145 to −0.472; p-value <0.001). Heterogeneity is high as there are variable outcomes. (p-value <0.001 and I2 = 92.802).

The alloimmunised pregnancies that resulted in live births and those that resulted in IUFD are summarized in a forest plot (Figure 2C). The results of the meta-analysis of IUFD proved its statistically significance (ARD: −0.938; 95 % CI: −1.344 to −0.533; p-value <0.001) but the heterogeneity was very high (p-value <0.001 and I2 = 92.976).

Meta-analysis was performed for the number of pregnancies with PSV >1.5 MoM and PSV <1.5 MoM. Table 3 shows the number of IUT in different studies of pregnancies with PSV >1.5 MoM.

The forest plot (Figure 2D) shows that the results were not statistically significant (ARD: 0.381; 95 % CI: −1.079 to −0.317; p-value = 0.285) with high heterogeneity (p-value <0.001 and I2 = 97.748).