The Prevalence of TET2 Gene Mutations in Patients with BCR-ABL-Negative Myeloproliferative Neoplasms (MPN): A Systematic Review and Meta-Analysis

Simple Summary Many molecular biology techniques have been widely used to study the pathogenesis of different diseases, particularly haematologic malignancies which are generally caused by abnormalities in the genome. TET2 gene is one of the commonly found mutated genes in BCR-ABL-negative myeloproliferative neoplasms. However, the prevalence of TET2 gene mutations in the disease remains unclear. Therefore, this study aims to estimate the prevalence of TET2 gene mutations in myeloproliferative neoplasms. The findings may be helpful for future research, diagnoses and the identification of better therapeutic strategies to manage the diseases. Abstract Multiple recurrent somatic mutations have recently been identified in association with myeloproliferative neoplasms (MPN). This meta-analysis aims to assess the pooled prevalence of TET2 gene mutations among patients with MPN. Six databases (PubMed, Scopus, ScienceDirect, Google Scholar, Web of Science and Embase) were searched for relevant studies from inception till September 2020, without language restrictions. The eligibility criteria included BCR-ABL-negative MPN adults with TET2 gene mutations. A random-effects model was used to estimate the pooled prevalence with 95% confidence intervals (CIs). Subgroup analyses explored results among different continents and countries, WHO diagnostic criteria, screening methods and types of MF. Quality assessment was undertaken using the Joanna Briggs Institute critical appraisal tool. The study was registered with PROSPERO (CRD42020212223). Thirty-five studies were included (n = 5121, 47.1% female). Overall, the pooled prevalence of TET2 gene mutations in MPN patients was 15.5% (95% CI: 12.1–19.0%, I2 = 94%). Regional differences explained a substantial amount of heterogeneity. The prevalence of TET2 gene mutations among the three subtypes PV, ET and MF were 16.8%, 9.8% and 15.7%, respectively. The quality of the included studies was determined to be moderate–high among 83% of the included studies. Among patients with BCR-ABL-negative MPN, the overall prevalence of TET2 gene mutations was 15.5%.


Introduction
Myeloproliferative neoplasms (MPN) are a group of rare blood cancers characterised by the clonal expansion of a large number of abnormal haematopoietic stem cells. Classic Philadelphia-negative (BCR-ABL-negative) MPN can be divided into three categories: (i) Cancers 2021, 13, 3078 2 of 20 polycythaemia vera (PV), (ii) essential thrombocythaemia (ET) and (iii) primary myelofibrosis (PMF). MPN can transform into acute myeloid leukaemia (AML) and may be associated with an elevated risk of thrombotic and haemorrhagic events [1,2]. Thrombosis and haemorrhage are the major causes of mortality and morbidity amongst patients with MPN and occur in about 34-39% of cases with PV, 10-29% with ET and 7.2-13.2% of patients with PMF [3].
Three main driver gene mutations, Janus kinase 2 (JAK2), Thrombopoietin receptor (MPL) and Calreticulin (CALR), have been identified in association with MPN and may have an important role in assisting the diagnosis of MPN [4]. In addition, epigenetic modification genes such as TET2, ASXL1, DNMT3A and EZH2 are also commonly mutated in cases of MPN with a frequency of 1-30% [5][6][7][8].
TET2 participates in one of the crucial steps in gene regulation, and mutations in this gene have been identified in 5-20% of people diagnosed with MPN [9]. Somatic missense mutations, somatic nonsense mutations and insertion-deletion mutations are detected in the TET2 gene among MPN patients. All of these mutations are loss-of-function mutations. Malfunction of TET2 protein may lead to the development of MPN and contributes to the disease progression [10,11]. However, some disagreement still exists about the relative significance of these TET2 gene mutations to MPN. Some researchers suggest that TET2 gene mutations are not important for MPN [12,13], whereas others have concluded that these mutations significantly contribute to their phenotype [14,15].
The prevalence of TET2 gene mutations among MPN has not yet been established. This meta-analysis aims to estimate the prevalence of TET2 gene mutations among all BCR-ABL-negative MPN and its three main subtypes.

Materials and Methods
PRISMA guidelines [16] were followed, and a study protocol was registered at the International Prospective Register of Systematic Reviews (PROSPERO), registration number CRD42020212223.

Data Sources and Searches
PubMed, Scopus, ScienceDirect, Google Scholar, Web of Science and Embase databases were searched from their inception till September 2020, without any language restrictions. Detailed search strategies are presented in Table S1. Any published studies or preprints with relevant data were included. Review articles, case reports and opinion articles were excluded. Data presented on websites or reported by press releases and news reports were not considered. Snowball searching was employed to review the references of included studies. Endnote X8 software was used to remove duplicate studies.

Study Selection
Study eligibility was determined by screening the title and abstract of the articles of interest. Two authors (Y.C.C. and M.A.I.) independently examined full-text reports of potentially relevant studies for inclusion. Any disagreements were resolved by consensus.

Extraction of Data
Data were independently extracted by two authors (Y.C.C. and M.A.I). The following data were obtained from each eligible study and inserted into a customised Excel spreadsheet: author surname, publication year, study design, study location, type of MPN, number of patients with MPN, demographic characteristics of patients including age and sex, clinical characteristics of the MPN patients including haemoglobin level, leucocyte and platelet counts, the total number of mutated ASXL1 and the screening method used to identify TET2 gene mutations and diagnostic criteria employed for MPN diagnoses.

Quality Assessment
A random-effects model was used to estimate the pooled prevalence of the TET2 gene mutations amongst patients with MPN, including 95% confidence intervals (Cis). Two authors (Y.C.C. and M.A.I.) independently assessed the quality of included studies using the Joanna Briggs Institute critical appraisal tools [13]. Study quality was categorised into three groups: low-quality or high risk of bias, moderate quality or moderate risk of bias, and high-quality or low risk of bias with overall scores of <50%, 50-69% and ≥70%, respectively [17].

Publication Bias
Funnel plots presenting estimates of prevalence plotted against standard error measures were used to assess the likelihood of publication bias. When a minimum of 10 studies were available, an Egger's test was conducted to assess publication bias based on funnel plot asymmetry.

Data Synthesis and Sensitivity Analysis
The I 2 statistic was used to gauge the heterogeneity between studies, with I 2 > 75% indicating substantial heterogeneity. The statistical significance of study heterogeneity was also assessed using Cochran's Q test; p < 0.05 was considered statistically heterogeneous. To help identify the outlier studies and the sources of heterogeneity, a Galbraith plot was constructed. Prevalence estimates were explored with sensitivity analyses. Three strategies were followed for these analyses: (i) studies with small sample sizes (<100) were excluded, (ii) low-quality studies were excluded and (iii) outlier studies were excluded. In each case, the results were then compared to the overall prevalence estimate. Metaprop codes in meta (version 4.15-1) and metaphor (version 2.4-0) packages of R (version 3.6.3) and RStudio (version 1.3.1093) were used for the analyses and graphs [18].

Study Selection
The search generated 758 potentially relevant studies. After excluding 558 studies (duplicates n = 450; review articles n = 67; non-human studies n = 31; and case reports, n = 10), 200 full-text studies were examined and 35 studies met the inclusion criteria and were included in the review ( Figure 1). to determine MPN diagnoses. Many studies confirmed TET2 gene mutations with either next-generation sequencing (NGS) or Sanger sequencing, which have higher sensitivity in detecting mutations compared with other methods, such as high-resolution melting (HRM) analysis [19]. One study was published in Chinese Mandarin and was translated into English (Y.C.C.).

Quality Assessment
Detailed quality assessments of the included studies are presented in Tables S2 and  S3. Most studies were judged to be of high quality (68.6%), while the remainder were considered to be of either moderate (14.3%) or low quality (17.7%).

Publication Bias
The results from the funnel plots and Egger's tests suggest that there is only a small likelihood of publication bias (Figures 3 and S6).

Sensitivity Analyses
In the sensitivity analyses, only minor differences (ranging from 4.0% lower to 1.8% higher) were observed in the pooled prevalence estimates of TET2 gene mutations among cases of MPN compared to the main findings (Table 3 and Figure S7). A Galbraith plot was performed, and four outlier studies [29,30,37,39] were identified (Figure 4). Cancers 2021, 13, x 12 of 21   Table S3. Most studies were judged to be of high quality (68.6%), while the remainder were considered to be of either moderate (14.3%) or low quality (17.7%).

Publication Bias
The results from the funnel plots and Egger's tests suggest that there is only a small likelihood of publication bias (Figure 3 and Figure S6).

Sensitivity Analyses
In the sensitivity analyses, only minor differences (ranging from 4.0% lower to 1.8% higher) were observed in the pooled prevalence estimates of TET2 gene mutations among cases of MPN compared to the main findings (Table 3 and Figure S7). A Galbraith plot was performed, and four outlier studies [29,30,37,39] were identified (Figure 4).
The results appear to confirm the observation that epigenetic regulators like the TET2 gene have mutated more frequently among those patients with PV (p = 0.05), compared with those with either MF (p = 0.02) or ET (p = 0.023) [27]. According to a meta-analysis that analysed the frequency of three main genes (JAK2, MPL and CALR) in MPN from 2000 to 2018 [66], for the most common gene mutation JAK2 V617F, TET2 showed a lower prevalence as compared to JAK2 V617F in PV (46.7-100.0%), ET (31.3-72.1%) and MF (25.0-85.7%). For the MPL gene, our results displayed a higher proportion in PV (16.8% vs. 0%) but similar percentages in ET (9.8% vs. 0.9-12.5%) and MF (15.7% vs. 0-17.7%). Finally, in relation to the last common driver gene, CALR, TET2 manifested a higher prevalence in PV (16.8% vs. 0%), a lower prevalence in ET (9.8% vs. 12.6-50%) and a similar proportion in MF (15.7% vs. 10-100%). From these results, it appears that the TET2 gene mutations have distributed more evenly across MPN subcategories in contrast to the three main driver mutation genes [47].
This study has several notable strengths. To our knowledge, no meta-analysis has previously investigated the prevalence of TET2 gene mutations in patients with MPN. This meta-analysis included studies from six databases using robust search strategies without any language restrictions. All the sensitivity analyses produced similar results to the overall findings, suggesting that the main result is likely to be robust and credible.
Nevertheless, there are a few limitations. Several meta-analyses exhibited high heterogeneity, indicating considerable variability among the results from the included studies. After excluding the four outlier studies identified by the Galbraith plot, heterogeneity was reduced from 94% to 63% across all MPN studies, 60% to 54% for PV, 62% to 48% for ET and 89% to 18% for MF, suggesting that these four studies were an important source of heterogeneity. Several factors may further explain this heterogeneity. One of the outlier studies [29] recorded a very low prevalence of TET2 gene mutations (0.4%). This may be due to the use of a different method (HRM analysis) that may be less sensitive to identifying the mutations compared with most other studies. Notably, a similar result was also observed in one of the two other studies [35,44] that also employed the same analytical technique. Different etiological exposures might occur in different regions, resulting in differences in the prevalence estimates across the different studies [67]. In support of this hypothesis, a lower prevalence was recorded in Australia and Italy, whereas a higher result was identified in China and the United States. Variations in the use of different diagnostic guidelines may have also affected the estimates of prevalence and further contributed to the heterogeneity of results between studies. The discovery of the JAK2 gene mutations in 2005 and their subsequent inclusion in the diagnostic criteria [68] for MPN, PV and ET but not MF [2,69]  Another limitation of this meta-analysis is that the prevalence of MPN may be underestimated in some studies. Patients with MPN can be relatively symptom-free for many years so people, with little contact with health services, can remain undiagnosed for long periods [70]. Estimates of the prevalence of TET2 mutations in MPN may be underestimated, particularly in less-developed countries or among disadvantaged groups in well-developed countries.
Finally, the included studies largely focused on the allele frequencies of the main driver mutations (JAK2, MPL and CALR) and did not permit any analysis of the allelic frequencies of the TET2 mutant allele in MPN.

Conclusions
This meta-analysis provides the most comprehensive currently available estimate of the overall prevalence of TET2 gene mutations (15.5%) among patients with MPN. However, substantial heterogeneity was evident among the results included in this meta-analysis, likely related to factors such as regional differences in patients included in studies and variations in the diagnostic criteria employed by the studies. This heterogeneity suggests that caution should be employed with using the estimates of prevalence.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/cancers13123078/s1, Figure S1: prevalence of TET2 gene mutations based on continents, Figure S2: Prevalence of TET2 gene mutations based on countries, Figure S3: Prevalence of TET2 gene mutations based on the WHO classification and diagnostic criteria of MPN, Figure S4: Prevalence of TET2 gene mutations based on different methods used to detect TET2 gene mutations in MPN, Figure S5: Subgroup analysis. Prevalence of TET2 gene mutations in patients with (A) PMF and (B) SMF, Figure S6: Funnel plots estimating the prevalence of TET2 gene mutations in patients with (A) polycythaemia vera, (B) essential thrombocythaemia and (C) myelofibrosis, Figure S7: Sensitivity analyses, Table S1: Search strategies, Table S2: Quality assessment of the included cross-sectional studies, Table S3: Quality assessment of the included case-control studies.

Data Availability Statement:
The data presented in this study are available within the article and supplementary materials.

Conflicts of Interest:
The authors declare no conflict of interest.