Abstract

Objective

Recurrent hydatidiform mole (RHM) is a rare disorder which is characterized by the presence of at least two molar pregnancies. The mutations in the NLRP7 and KHDC3L genes are responsible for the majority of recurrent molar pregnancies. This study aimed to demonstrate the diversity and frequency of NLRP7 and KHDC3L gene mutations in our Turkish cohort with recurrent molar pregnancies, and to establish genotype-phenotype correlation.

Materials and Methods

It was aimed to represent the detected NLRP7 and KHDC3L gene variants and reproductive history of 32 recurrent mole hydatidiform patients. We analysed the retrospective clinical and sequence data of 32 patients, who were referred to the laboratory for NLRP7 and KHDC3L sequencing.

Results

Among the detected 32 patients with recurrent molar pregnancy, 18 of 32 patients had no mutation in these two genes; we found 7 cases of homozygous NLRP7 variant, 1 case of heterozygous NLRP7 variant, 3 cases of homozygous KHDC3L gene variant, and 1 case of heterozygous KHDC3L gene variant. Among the detected NLRP7 variants, 3 of 11 variants were classified as pathogenic, 7 of 11 variants were classified as likely pathogenic, and 1 of 11 variants was classified as variant of unknown significance (VUS). Among the detected KHDC3L variants, 1 of 4 was classified as pathogenic, 2 of 4 were classified as likely pathogenic, and 1 of 4 was classified as VUS. Seven unpublished NLRP7 gene variants and two unpublished KHDC3L gene variants were first reported in this study.

Conclusion

Here we report new RHM patients with NLRP7 and KHDC3L mutations. The current study highlights the importance of defining new cases and novel mutations in the pathogenesis and clinical management of RHM. Understanding genotype-phenotype correlations in RHM patients will also contribute to the selection of treatment methods and patient management.

Introduction

Molar pregnancy or mole hydatidiform is a gestational trophoblastic disease characterized by abnormal embryonic development and excessive proliferation of trophoblasts. Most cases of molar pregnancy are sporadic, occurring in approximately 1 in 600 pregnancies^{(1, 2)}. A hydatidiform mole is histopathologically classified into two main categories based on parental origin: complete hydatidiform mole and partial hydatidiform mole. Complete mole hydatidiform is characterized by excessive trophoblastic proliferation and the absence of extra-embryonic membranes. In contrast, partial molar hydatidiform shows mild trophoblastic proliferation and may contain extra-embryonic membranes and embryonic tissue. Most complete mole hydatidiform cases are diploid androgenetic, while most partial mole hydatidiform cases are dispermic triploid. Recurrent mole hydatidiform (RHM) is defined as the occurrence of ≥2 molar pregnancies^{(1, 2)}. Although the frequency varies among the different ethnicities, RHM accounts for approximately 1-10% of all molar hydatidiform cases⁽³⁾. The most common genetic cause of recurrent molar hydatidiform is a homozygous mutation of the NLRP7 gene (OMIM 231090), which accounts for approximately 55% of cases^(4-6). The second most common cause is homozygous mutations of the KHDC3L gene (OMIM 611687), accounting for 5% of cases. In cases with mutations in NLRP7 or KHDC3L, molar hydatidiform tissues are found to be diploid biparental^(6-9). Both genes regulate gene expression during oocyte and embryo development through genomic imprinting or epigenetic mechanisms. The NLRP7 and KHDC3L genes are maternal-effect genes and components of the subcortical maternal complex, playing roles in the epigenetic reprogramming of the oocyte and the activation of embryonic development^{(6, 10)}. Mutations in the NLRP7 and KHDC3L genes contribute to the pathogenesis of molar pregnancy by disrupting cytokine secretion and the implantation process. In addition to playing a role in oocyte and embryo development, NLRP7 also regulates the release of interleukin-1 beta, contributing to inflammation and immune responses⁽⁸⁾. Mutations in the NLRP7 or KHDC3L genes lead to the inactivation of the maternal allele and the expression of only the paternal allele. Monoallelic paternal expression of these genes causes defective placenta-specific imprinting and recurrent molar pregnancies^{(10, 11)}. Several studies have demonstrated that mutations in the NLRP7 gene negatively impact oocyte quality and lead to arrest in embryonic development. NLRP7 plays a role in sustaining genomic stability by regulating the alternative splicing of genes involved in homologous recombination repair⁽¹²⁾. Similarly, the KHDC3L mutation causes genomic instability in embryonic cells, which can result in increased DNA damage and subsequent embryonic developmental arrest⁽¹³⁾. Mutations in both NLRP7 and KHDC3L disrupt DNA repair mechanisms, thereby contributing to impaired embryonic development⁽⁹⁾. Mutations of the NLRP7 and KHDC3L genes have been reported to be associated with recurrent pregnancy losses and recurrent molar pregnancies.

It is not easy to clinically and histologicallydistinguish recurrent from spontaneous molar pregnancies. Therefore, the diagnosis of RHMs is often delayed. Although the mechanisms leading to molar pregnancy are not fully known, screening for known genetic factors is necessary to establish the genetic diagnosis of recurrent molar pregnancies. Late diagnosis in these patients may lead to recurrent molar pregnancies, unnecessary medications, and in vitro fertilization (IVF) treatments. Early diagnosis of recurrent molar pregnancies is essential for patients to receive appropriate treatment. The aim was to highlight the importance of the genetic diagnosis of RHM in clinical decision-making by reporting new cases. Here, we present the reproductive history and genomic variants, specifically NLPR7 and KHDC3L mutations, in Turkish patients with RHMs, including nine previously unpublished variants.

Materials and Methods

Patient Selection

A total of 32 patients with recurrent hydatidiform mole who were referred to the Mikrogen Genetic Diagnosis Center for NLPR7 and KHD3L gene sequencing were retrospectively analyzed. Thirty-one patients had at least two molar pregnancies, and one patient had a history of choriocarcinoma. This study was approved by the Yüksek İhtisas University Medical School Ethical Committee (approval no: 295, date: 14.04.2025), and written consent was obtained. This study was conducted in accordance with the Declaration of Helsinki.

NLRP7 and KH3DL Gene Sequencing

Genomic DNA was isolated from peripheral blood samples using the QIAamp DNA Blood Kit (QIAGEN, Germany). Next-generation sequencing of the NLRP7 and KHD3L genes was performed on an Illumina MiSeq sequencing platform (Illumina Inc., San Diego, CA, USA) by following the manufacturer’s instructions. The exon/exon-intron junctions of NLRP7 and KHDC3L genes were sequenced to obtain a minimum read depth of 20x for >98% of the targeted bases. The Hg19 (GRCh37) sequence was used as a reference genome. FASTQ and VCF files were obtained. Bioinformatic solutions such as NextGENe (Version 2.4.2.3/SoftGenetics LLC-USA), Geneticist Assistant (Version 1.8.1.0/SoftGenetics LLC-USA), and Franklin Genoxx (Genoox, Israel) were used for VCF data analysis. All detected variants were evaluated according to their pathogenicity and classified according to the recommendations of international guidelines. Detected NLRP7 and KHDC3L variants have been classified into five categories based on international standards, which are determined by their pathogenic effects (ACMG 2015): pathogenic, likely pathogenic, variant with unknown clinical significance (VUS), likely benign, and benign. A pathogenic variant is defined as a genetic change for which there is strong and well-established evidence indicating a direct causative role in disease. A likely pathogenic (LP) variant has greater than 90% certainty, of being disease-causing, based on available evidence. Conversely, a likely benign variant also has greater than 90% certainty; however, in this case, the evidence suggests that the variant is not associated with disease. A benign variant is one for which there is conclusive evidence demonstrating that it does not cause disease. A VUS refers to a genetic alteration for which the current evidence is either insufficient or conflicting regarding its role in disease. Identification of a VUS does not confirm or exclude a diagnosis. The variant classification is made according to the mutation type, its functional effect, whether it is defined in relevant databases and literature, and whether it is compatible with the patient’s clinical findings.

Results

A total of 32 patients with RHMs were analyzed for mutations in NLRP7 and KHDC3L genes. 7 of 32 patients (21.8%) had a homozygous NLRP7 variant, 1 of 32 patients (3.1%) had compound heterozygous NLRP7 variants, 2 of 32 patients (6.2%) had a heterozygous NLRP7 variant, 3 of 32 patients (9.3%) had a homozygous KHDC3L variant and 1 of 32 patients (%) had a heterozygous KHDC3L variant (Figure 1). No mutation was found in 56.2% of patients (18/32). Among the detected NLRP7 variants, 3 of 11 variants were classified as pathogenic, 7 of 11 variants were classified as LP, and 1 of 11 variants was classified as VUS. Regarding the type of NLRP7 gene variants, one missense, five frameshift, two nonsense, and three splice variants were reported. Among the detected KHDC3L variants, 3 out of 4 were classified as LP, and 1 out of 4 was classified as VUS. Regarding the type of KHDC3L gene variants, one missense and three frameshift variants have been reported. Seven unpublished NLRP7 gene variants and two unpublished KHDC3L gene variants were first reported in this study (Table 1).

The clinical phenotypes of patients with NLRP7 or KHDC3L gene variants are presented in Table 1. Eighteen of 32 patients had no variant in the NLRP7 or KHDC3L genes. Seventeen of 18 patients (17/18), had at least two molar pregnancies, and one of them had a history of choriocarcinoma.

Discussion

Here, we have reported the retrospective data of 32 patients with RHM referred to the Mikrogen Genetic Diagnosis Center. Seven patients had a homozygous NLRP7 variant, one patient had a compound heterozygous NLRP7 variant, and two had a heterozygous NLRP7 variant. Individuals with NLRP7 variants exhibit a broad spectrum of reproductive outcomes, including stillbirth, spontaneous abortion, blighted ovum, partial hydatidiform mole, complete hydatidiform mole, and, in rare instances, live birth. The underlying mechanisms through which specific mutation types in NLRP7 contribute to the pathogenesis of molar pregnancies remain to be elucidated⁽¹⁴⁾. Biallelic NLRP7 variants cause recurrent hydatidiform mole disease. Still, female carriers have an increased risk of reproductive failure, recurrent pregnancy loss, or having offspring with aberrant methylation and imprinting disorders. Limited data are reported regarding the effect of heterozygous NLRP7 variants on phenotype. These studies suggested that female heterozygous carriers have a history of reproductive failure without molar pregnancy. Hayward et al.⁽¹⁵⁾ demonstrated aberrant methylation in embryonic tissues of NLRP7 heterozygous carriers, and they stated that the NLRP7 gene regulates oocyte growth or controls the transduction of signals to initiate imprinting. Some studies have demonstrated that reproductive failure in heterozygous carriers of NLRP7 variants is consistent with imprinting defects in placental tissues and increased maternal methylated transcripts^{(16, 17)}. Soellner et al.⁽¹⁸⁾ reported a case of reproductive failure and fetal aberrant methylation. They stated that heterozygous variants of the NLRP7 gene are associated with reproductive failures. The reported index patient had a frameshift mutation in NLRP7 (NM_001127255.1: c.2010_2011del, p.(Phe671Glnfs*18)), resulting in a stop codon. In silico prediction tools suggested nonsense-mediated mRNA decay as the mechanism for the translated mRNA product. Qian et al.⁽¹⁹⁾ also reported a female patient with a heterozygous NLRP7 variant (c.295G>T, p.Glu99*) who had one stillbirth and three normal pregnancies. In the current study, we found heterozygous NLPR7 variants in 2 cases. One of them had no reproductive failures, but she had an affected daughter with neurological problems (epilepsy, hypotonia, cerebellar atrophy), and her sister had a history of recurrent molar pregnancies. Another heterozygous carrier in our study had a missense mutation in the NLRP7 gene and had a history of recurrent pregnancy losses and one molar pregnancy. Our case is also compatible with previously reported cases of heterozygous NLRP7 mutations associated with reproductive failure.

The homozygous NLRP7 variant (c.2471+1G>A) was first reported in a Pakistani patient with spontaneous abortion and complete hydatidiform mole⁽²⁰⁾. Kocabey et al.⁽²¹⁾ reported homozygous NLRP7 c.2471+1G>A splice site variant in two Turkish patients with recurrent molar pregnancies. In the current study, we reported two different patients with homozygous NLRP7 variants and one patient with a heterozygous NLRP7 variant, all with the c.2471+1G>A mutation. One of the patients with a homozygous variant has a history of recurrent hydatidiform mole and one ectopic pregnancy. Another case with a homozygous variant has a history of recurrent hydatidiform mole (more than two hydatidiform moles). The patient with a heterozygous NLRP7 variant (c.2471+1G>A) also had a LP heterozygous NLRP7 variant (c.368G>A). The compound heterozygous patient had recurrent hydatidiform moles, and her three sisters also had recurrent molar pregnancies.

Wang et al.⁽⁵⁾ reported a case with a homozygous NLRP7 c.2147delC variant. The patient had 4 CHM and 2 PTD. Our study reported a patient with a heterozygous NLRP7 variant (c.2147delC) who had no recurrent molar pregnancy, but she had a sister with a history of recurrent hydatidiform mole. Human epidermal growth factor receptor 2-year-old boy had a heterozygous NLRP7 variant (c.2147delC) and neurological abnormalities.

We reported three patients with a homozygous KHDC3L variant (c.322_325delGACT) and one patient with a heterozygous KHDC3L variant (c.572C>A). Biallelic KHDC3L variants are associated with early embryonic arrest, recurrent hydatidiform moles, and recurrent pregnancy loss^{(11, 19)}. The KHDC3L variant (c.322_325delGACT) was reported in several studies^{(10, 13, 22, 23)}. Fatemi et al.⁽⁸⁾ reported a large pedigree with a homozygous KHDC3L (c.322_325delGACT) variant and a history of recurrent molar pregnancies. Wang et al.⁽¹³⁾ reported another patient with four recurrent molar pregnancies (complete hydatidiform mole). Wang et al.⁽¹³⁾ reported another patient with four recurrent molar pregnancies who was compound heterozygous for KHDC3L variants (c.1A>G and c.322_325delGACT). In the current study, we found a homozygous KHDC3L (c.322_325delGACT) variant in two patients with recurrent molar pregnancies. Our patient with a heterozygous KHDC3L variant (c.572C>A) had recurrent pregnancy losses but no hydatidiform mole.

Previous studies suggest that patients with biallelic KHDC3L mutations may have a more severe phenotype compared to patients with biallelic NLRP7 mutations. Fatemi et al.⁽⁸⁾ reported that patients with a biallelic truncating KHDC3L variant, could not have a successful pregnancy and suffer from PL and HM; however, heterozygosity of the same variant does not result in HM and causes recurrent pregnancy losses⁽⁹⁾. Our patient with a heterozygous missense KHDC3L mutation had recurrent pregnancy losses but no HM, and the phenotype is consistent with the previously reported heterozygous cases.

Limited studies were reported in Turkish patients with KHDC3L and NLRP7 gene variants. To date, no KHDC3L gene variant has been reported in Turkish patients with recurrent hydatidiform mole. Some of the variants detected in the NLRP7 gene in our study were reported for the first time in Turkish patients. Kocabey et al.⁽²¹⁾ reported a homozygous NLRP7 splice site variant (c.2471+1G>A) and homozygous frameshift variant (c.2571dupC) in three Turkish patients with recurrent molar pregnancies. Balci et al.⁽²⁴⁾ found a homozygous NLRP7 variant (c.3024_3025insC) in a Turkish patient with six hydatidiform moles and five missed abortions⁽²⁴⁾.

The likelihood of a normal pregnancy in women with a history of recurrent molar pregnancies has been reported to be very low^{(23, 25)}. Several studies have demonstrated that oocyte donation is the best treatment option for women carrying biallelic NLRP7 variants; however, some women with these variants have been reported to have live births from their own oocytes, albeit rarely^{(25, 26)}. Although oocyte donation is considered the most effective treatment for NLRP7 homozygous individuals, live birth cannot be achieved in all these patients. The reported live birth from women with biallelic NLRP7 variants was attributed to different types of mutations. Missense, splice variants, or protein-truncating mutations of the NLRP7 gene are expected to cause a milder functional effect^{(5, 24, 25)}.

Study Limitations

The current study has some limitations. Mutations in the NLRP7 and KHDC3L genes are responsible for most RHM cases; however, other genes have been reported in rare cases. We sequenced only the NLRP7 and KHDC3L genes, but sequencing of other responsible genes is necessary to exclude the presence of mutations in NLRP7-KHDC3L mutation-negative cases. Another limitation of the current study is the lack of population-based data on the prevalence of NLRP7 and KHDC3L mutations in the Turkish population. As no prior large-scale epidemiological studies exist, a power analysis could not be performed. Additionally, only a few published reports have described these mutations in Turkish patients. Nevertheless, this study includes the largest Turkish cohort to date screened for NLRP7 and KHDC3L mutations, and offers preliminary but valuable data that may guide future studies aiming to establish national prevalence rates and explore genotype–phenotype correlations. Furthermore, there was no information available on treatment options or live birth rates in cases with mutations. Further studies with large cohorts are needed to understand the etiopathogenesis of RHM and develop effective treatment options.

Conclusion

Genetic counseling for RHM associated with pathogenic variants in the NLRP7 and KHDC3L genes is a highly complex process that plays a crucial role in the clinical management of patients with RHM. Given the markedly low likelihood of successful pregnancy outcomes in patients with biallelic variants of these genes, it is essential to provide comprehensive counseling regarding reproductive options and the limited probability of achieving a live birth. Although there is no treatment addressing the genetic causes in RHMs with a genetic diagnosis, oocyte donation is the best reproductive method for pregnancy success. Most biallelic variants in NLRP7 and KHDC3L are incompatible with live birth; therefore, studies aimed at identifying clinically relevant genetic variants will aid in the genetic counseling of these patients. Reporting additional cases with confirmed pathogenic variants will not only expand the understanding of genotype-phenotype correlations but also guide patients toward appropriate reproductive strategies, such as oocyte donation. Moreover, such data will help to prevent the patients from unnecessary IVF cycles, medications, and pregnancy expectations.