Diagnostic yield of next-generation sequencing in suspect primary immunodeficiencies diseases: a systematic review and meta-analysis

To determine the diagnostic yield of Next-generation sequencing (NGS) in suspect Primary Immunodeficiencies Diseases (PIDs). This systematic review was conducted following PRISMA criteria. Searching Pubmed and Web of Science databases, the following keywords were used in the search: (“Next-generation sequencing”) OR “whole exome sequencing” OR “whole genome sequencing”) AND (“primary immunodeficiency disease” OR “PIDs”). We used STARD items to assess the risk of bias in the included studies. The meta-analysis included 29 studies with 5847 patients, revealing a pooled positive detection rate of 42% (95% CI 0.29–0.54, P < 0.001) for NGS in suspected PID cases. Subgroup analyses based on family history demonstrated a higher detection rate of 58% (95% CI 0.43–0.71) in patients with a family history compared to 33% (95% CI 0.21–0.46) in those without (P < 0.001). Stratification by disease types showed varied detection rates, with Severe Combined Immunodeficiency leading at 58% (P < 0.001). Among 253 PID-related genes, RAG1, ATM, BTK, and others constituted major contributors, with 34 genes not included in the 2022 IUIS gene list. The application of NGS in suspected PID patients can provide significant diagnostic results, especially in patients with a family history. Meanwhile, NGS performs excellently in accurately diagnosing disease types, and early identification of disease types can benefit patients in treatment.


Introduction
Primary Immunodeficiency Diseases (PIDs), also recognized as Inborn Errors of Immunity (IEI), constitute a diverse set of monogenic disorders that impact the immune system.The clinical spectrum of PIDs encompasses various phenotypes, including infections, autoinflammation, autoimmunity, allergies, malignancies, and more [9].Presently, over 430 genes associated with PIDs have been identified, predominantly inherited in a monogenic manner.The list of disease-causing genes continues to expand with ongoing discoveries.The International Union of Immunological Societies (IUIS) classifies IEI into 10 categories, addressing diseases with overlapping phenotypes (citation).These categories range from Combined Immunodeficiencies to Congenital Immune Deficiencies (citation).Given the broad clinical spectrum exhibited by PIDs, achieving an accurate diagnosis based solely on clinical manifestations poses a significant challenge.Hence, there is an urgent need in clinical practice for methods that are safe, rapid, and accurate for testing and diagnosing PIDs.
Before the advent of advanced technologies, Sanger sequencing was the primary method employed in clinical practice.However, its limitations became evident in the context of PIDs due to genetic pleiotropy and heterogeneity.Sanger sequencing, being time-consuming, laborintensive, inefficient, and relatively expensive, faced challenges in identifying disease-causing genes [19,23] Genome Sequencing (WGS) [39].
TGP is a method focusing on a limited set of genes relevant to specific phenotypes or disease groups and is used for patients with well-defined clinical phenotypes [16,31].WES, targeting the exome region, can identify genetic variations associated with classical or atypical phenotypes, expanding the understanding of new phenotypes and genes.Diagnostic yields from WES vary widely (15-79%) depending on patient types and clinical phenotypes (citation).WGS, covering the entire genome, provides better resolution for detecting copy number variations (CNVs) and structural variations.However, it introduces the challenge of interpreting a substantial number of variants of uncertain significance, especially in non-coding regions [10,20].
As technology advances and costs decrease, NGS plays an increasingly vital role in primary PID diagnosis.Nevertheless, literature reports indicate significant variation in the diagnostic rates of NGS in PIDs.Hence, there is a crucial need to comprehensively evaluate the technical performance and diagnostic efficacy of NGS in PID patients across diverse populations and diseases.This meta-analysis aims to fulfill this need through a comprehensive evaluation.

Search strategy
A comprehensive search for pertinent studies was systematically conducted in the Pubmed and Web of Science databases up to November 1st, 2023.The search strategy employed the following Boolean operators: ("next-generation sequencing" OR "whole exome sequencing" OR "whole genome sequencing") AND ("primary immunodeficiency disease" OR "PIDs").A total of 958 studies were retrieved from Pubmed, and 836 studies were identified from Web of Science.No additional pertinent studies were discerned in the bibliographies of the encompassed studies.A transparent depiction of the search and literature screening process is presented in Fig. 1.

Literature selection
In adherence to predefined inclusion and exclusion criteria, studies were meticulously selected for the meta-analysis.Inclusion criteria stipulated: (1) a study population exhibiting clinically evident manifestations of PIDs; (2) incorporation of genetic sequencing methodologies and provision of pertinent genetic testing information; (3) explicit reporting of both the total number of patients and the number of positive patients detected; and (4) documentation in the English language.Exclusion criteria comprised: (1) studies exclusively focused on the detection of copy number variations (CNVs); and (2) exclusion of duplicate publications, conference proceedings, case reports, reviews, and unpublished studies.

Data extraction and quality assessment
A rigorous evaluation of potentially relevant articles commenced with an independent screening of titles and abstracts by three authors.Subsequent data extraction, in concordance with predefined inclusion criteria, encompassed key parameters such as the first author, publication year, detection method, total number of patients, number of positive patients, and genetic locus information.The assessment of the risk of bias in the included studies adhered to the Standards for Reporting of Diagnostic Accuracy Studies (STARD) criteria.Any discrepancies in the extraction and assessment processes were resolved through meticulous discussion.In instances where consensus remained elusive, a third reviewer assumed the role of an arbitrator.

Analytical procedures
All statistical analyses were conducted utilizing the meta package in Stata 14.A meta-analysis of the gene testing detection rate for PIDs was undertaken utilizing a singlegroup rate approach.Effect size (ES) and a 95% confidence interval (CI) were computed employing a random-effects model.Subgroup analyses discerned the detection rate of gene testing for distinct populations and varied diseases.Heterogeneity was quantified employing the chi-square Q test and I 2 test.A fixed-effects model was employed when I 2 was less than 50% or the p value of the Q test exceeded 0.05.Conversely, in instances where I 2 surpassed 50% or the p value of the Q test fell below 0.05, a random-effects model was applied.
Fig. 2 The overall physical examination rate of primary immunodeficiency diseases  1.

Genetic landscape of PID-related genes
A total of 253 PID-related genes were examined, with notable pathogenic mutations mainly involving RAG1, ATM, BTK, LRBA, DOCK8, STAT3, IL2RG, JAK3, RAG2, and WAS, each accounting for more than 3% of the cases, with RAG1 accounting for 6% (Fig. 5).Notably, 34 genes were not included in the updated 2022 IUIS gene list, and additional information on these genes can be found in Table 2.

Quality assessment of included studies
Quality assessment utilizing a modified version of the Standards for Reporting of Diagnostic Accuracy (STARD) specific to this project indicated high-quality studies (Fig. 6).As the analysis is a single-group rate analysis with descriptive results, publication bias assessment was deemed unnecessary.

Discussion
Primary Immunodeficiency Diseases (PIDs) represent a class of inherent immunodeficiency disorders, attributed to genetic mutations impacting distinct facets of the immune system.This diverse group manifests through recurrent infections, autoimmunity, autoinflammation, Fig. 4 The detection rate of different types of primary immunodeficiency diseases hypersensitivity reactions, and malignancies [9].PIDs exhibit genetic heterogeneity, with different mutations in the same gene yielding varied clinical and immune phenotypes, while distinct gene mutations may produce similar clinical outcomes [42].The broad clinical spectrum, coupled with genetic heterogeneity and pleiotropy, renders the diagnosis and treatment of PIDs challenging.A precise molecular diagnosis is crucial for accurate recognition and tailored intervention in clinical practice.
Our study reveals a positive detection rate of 42% (95% CI 0.29-0.54,P < 0.001) for Next-Generation Sequencing (NGS) in patients displaying clinical symptoms associated with PIDs.Previous literature reports a variable detection rate for PIDs, ranging from 15 to 80%.In our analysis of 5847 cases, the observed 42% detection rate falls within an intermediate range.Three key factors contribute to this result: Firstly, the expansive array of clinical manifestations related to PIDs has led to a surge in suspected cases.Secondly, the limited nucleotide coverage of NGS diminishes the potential for detecting pathogenic mutations.Additionally, our study excludes cases diagnosed through Copy Number Variation (CNV) detection [17,35,44], a method reported in the literature to offer a substantial number of additional genetic diagnoses [35].Due to the complexity of these factors, determining an average diagnostic rate poses challenges.Enhancing the accuracy of this rate involves the meticulous collection of comprehensive clinical data.
Our subgroup analysis, differentiating patients based on the presence or absence of a family history and various disease types, yielded intriguing insights.Notably, the detection rate among patients with a family history stands at an impressive 58% (95% CI 0.43-0.71,P < 0.001), while their counterparts without a family history exhibit a lower detection rate of 33% (95% CI 0.21-0.46,P < 0.001).This stark contrast underscores the substantial impact of familial factors on the detection rate of Primary Immunodeficiency Diseases (PIDs).Supporting this finding, existing literature posits that a majority of PIDs follow an autosomal recessive (AR) inheritance pattern.Identifying individuals with a singular clinical and immune phenotype within consanguineous families emerges as a pivotal strategy for uncovering novel pathogenic genes [2].Our study, incorporating data from 10 relevant articles [2,4,5,13,15,18,32,38,25,26] involving patients with a family history, aligns with this genetic landscape.Among these, 2 cases inherited PIDs as autosomal recessive (AR) traits, 1 as a combination of AR and X-linked (XL) traits, 6 as a blend of AR, autosomal dominant (AD), and XL traits, and 1 did not specify the genetic features.This consistency with prior research underscores the crucial role of families with consanguinity in unraveling intricate phenotypes associated with PIDs.Exploring consanguineous  Our investigation underscores distinctive detection rates across various primary immunodeficiency diseases (PIDs).Severe Combined Immunodeficiency (SCID) emerges with the highest detection rate, followed by Common Variable Immunodeficiency (CVID), Hyper-IgE Syndrome (HIES), and Combined Immunodeficiency (CID).Clinical manifestations of Severe Combined Immunodeficiency (SCID) denote a complex spectrum of disorders characterized by impaired T lymphocyte development, impacting the quantity and functionality of B cells and NK cells [34].SCID stands out as a profoundly severe subset within the broader PID landscape, marked by early-onset dermatitis, dermal complications, persistent enteritis, pneumonia, oral candidiasis, and other distinctive manifestations [43].Notably, more than 50% of SCID patients have been reported to harbor mutations in the RAG1 or RAG2 genes [33,41].In the absence of immune reconstitution, the survival prognosis for SCID patients beyond 6-12 months is exceedingly poor.Nonetheless, this patient cohort commonly demonstrates a favorable response to allogeneic hematopoietic stem cell transplantation (HSCT) [41].Hence, early disease recognition and expeditious intervention hold the potential to substantially augment patient survival rates.
Similarly, diseases such as Common Variable Immunodeficiency (CVID), Hyper-IgE Syndrome (HIES), and Combined Immunodeficiency (CID) constitute prevalent categories within the PID spectrum.The extensive range of diseases within PIDs manifests shared clinical manifestations, necessitating timely and precise identification of disease types.Next-generation sequencing (NGS) emerges as a critical tool for optimizing treatment effectiveness and providing early benefits to patients.However, it is crucial to acknowledge that the literature data incorporated into our study, while informative, is not derived from largescale studies.The potential introduction of slight errors in the results emphasizes the need for more extensive clinical data to validate our conclusions thoroughly.Large-scale studies will enhance the robustness of our findings, contributing to a more comprehensive understanding of the disease-specific landscape within PIDs.
A critical facet of our study involves scrutinizing primary immunodeficiency diseases (PIDs)-related genes updated by the International Union of Immunological Societies (IUIS) in 2022 [40].Our analysis identified 34 genes absent from the IUIS list, prompting a deeper investigation.Consulting the Online Mendelian Inheritance in Man (OMIM) database revealed six genes with significant relevance to PIDs' clinical manifestations.CD40L's potential involvement in Immunodeficiency and hyper-IgM, RECQL4's associations with Baller-Gerold syndrome, RAPADILINO syndrome, and Rothmund-Thomson syndrome, and IL7RA's connection to severe combined immunodeficiency exemplify the intricate genetic landscape.Additionally, SP1NK5, ITGA2B, and CR2 exhibit diverse implications, linking to Netherton syndrome, bleeding disorders, and common variable immunodeficiency, respectively.Despite the potential of these genes as targets for clinical testing, their validation is hindered by limited clinical data, underscoring the need for comprehensive validation efforts.
This article still has certain limitations.Firstly, the metaanalysis included a total of 29 articles, and some of the sequencing results in these articles were not validated using Sanger sequencing.Secondly, we have excluded positive cases determined through CNVs detection when entering the data, which may provide accurate information.Thirdly, some of the included literature had a small number of patients tested (< 20), which could affect the pooled results and lead to minor errors.More clinical sequencing data is required to validate the relevant conclusions.

Conclusion
The application of NGS in suspected PID patients can provide significant diagnostic results, especially in patients with a family history.Meanwhile, NGS performs excellently in accurately diagnosing disease types, and early identification of disease types can benefit patients in treatment.
Author contributions YYC, DRL and JWY contributed to the study design, while JLX, MX and QQ contributed to the data collection.Statistical analyses and interpretation of results were performed by YYC and DRL, whereas WLY drafted the manuscript and edited the language.All authors contributed to the article and approved the submitted version.
. The emergence of next-generation high-throughput sequencing technologies marks a transformative era in PID diagnosis.Next-generation sequencing (NGS) has not only expanded the known gene list associated with PIDs but has also introduced a faster and more cost-effective means of evaluating Yingying Chen and Dongrui Li have contributed equally to this work.the genome.Particularly in cases lacking clear candidate genes, NGS becomes indispensable.NGS encompasses various techniques, including Targeted Gene Panel (TGP) sequencing, Whole Exome Sequencing (WES), and Whole

Fig. 6
Fig. 6 Quality assessment of included studies

Table 1
Characteristics of included studies

Table 2
The specific information of genes that were not included in the updated 2022 IUIS gene list