Abstract
Background/Aim: Genetic variations of the non-coding RNA gene, ANRIL, have been associated with human diseases including cancer, type-2 diabetes, and atherosclerosis. In the present study, we investigated the potential associations of select ANRIL single nucleotide polymorphisms (SNPs) with overall survival and other clinical outcomes in adult patients with hematologic malignancies after allogeneic hematopoietic stem cell transplantation (allo-HSCT). Patients and Methods: Genomic DNA was extracted from whole blood samples from 103 adult patients with hematologic malignancies who had received allo-HSCT followed by oral tacrolimus therapy. The genotypes of four select ANRIL SNPs, rs564398, rs1063192, rs2151280, and rs2157719 were determined using qRT-PCR-based genotyping assays. Results: rs2151280 (C->T) in ANRIL was associated with worse overall survival in these patients (CT/CC vs. TT). Contrarily, rs2151280 and the other select ANRIL SNPs were not associated with death at Day-100 after transplantation, the incidence of graft-versus-host disease (GVHD), acute kidney injury (AKI), and neurotoxicity in the study cohort. Conclusion: rs2151280 represents a potential prognostic biomarker for overall survival in adult patients with hematologic malignancies after allo-HSCT.
- ANRIL
- single nucleotide polymorphism
- ARF-INK4 locus
- hematologic malignancies
- allogeneic hematopoietic stem cell transplantation
Chromosome 9p21 harbors the ARF-INK4 locus, which encodes three tumor suppressive proteins, p15INK4B (P15), p16INK4A (P16) and p14ARF (ARF) (1-5). P15 and P16 specifically inhibit cyclin D-dependent kinase 4 (CDK4)-mediated phosphorylation of retinoblastoma protein (pRB), thus regulating cell cycle progression; ARF inhibits the ability of mouse double minute 2 homolog (MDM2) to suppress p53, and consequently promotes apoptosis or cell cycle arrest (4, 5). Recent studies have demonstrated that the non-coding RNA ANRIL (also designated as CDKN2BAS) is located adjacent to the INK4-ARF locus but is transcribed in the antisense direction with respect to p14ARF (6-9). ANRIL interacts with polycomb repressing complexes 1 and 2 (PRC1 and 2), and negatively regulates the transcription of the entire INK4-ARF locus (10-12). Emerging evidence has revealed that some single nucleotide polymorphisms (SNPs) in ANRIL are associated with susceptibility to the development of human diseases, including different types of cancer, type-2 diabetes and atherosclerosis (9, 13-26). Specifically, rs564398 has frequently been found in patients with leukemia and esophageal squamous cell carcinoma (ESCC). ANRIL rs2151280 has been associated with increased risk of developing lung cancer and basal cell carcinoma (BCC) while rs1063192 and rs2157719 have been correlated with glioma, brain cancer, and ESCC. Additionally, our previous study has demonstrated that rs2151280 was associated with mRNA expression of p15, p16, and ARF in peripheral blood mononuclear cells (PBMCs) from multiple myeloma patients, and with progression-free survival (PFS) following autologous stem cell transplantation (27). Taken together, these findings imply that ANRIL SNPs may impact the development and treatment of human hematologic malignancies.
In this study, we determined the geneotypes of four previously reported cancer-related ANRIL SNPs, rs564398, rs1063192, rs2151280, and rs2157719 using whole blood DNA samples from 103 adult patients with hematologic malignancies and evaluated their associations with patient overall survival (OS) following allogeneic hematopoietic stem cell transplantation (allo-HSCT). Our results showed that ANRIL rs2151280 alone was associated with OS in the study cohort.
Patients and Methods
Study population. The study population was from a single-institution research protocol to study the pharmacokinetics and pharmacogenomics of tacrolimus and their association with clinical outcomes in allo-HSCT patients. One hundred and three adult patients receiving oral tacrolimus for the prevention of GVHD following allo-HSCT at The Ohio State University Wexner Medical Center were prospectively enrolled following approval from the Institutional Review Board (IRB #2012C0021). Inclusion criteria included age greater than 18 years, patients receiving their first allo-HSCT during the study period, and absence of renal or hepatic dysfunction at baseline. Exclusion criteria included HIV positivity, presence of any medical conditions that would interfere with the patient's ability to provide informed consent, hypersensitivity to any components of tacrolimus, clinical history of any solid organ transplantation, previous chronic use of tacrolimus, and pregnancy.
Genomic DNA extraction and genotyping assays. Genomic DNA was extracted from patient's whole blood sample using the DNAeasy Blood mini kit (Qiagen, Valencia, CA, USA) following the manufacturer's instructions. RNase A was included in the purification procedure to remove potential RNA contamination as recommended by the manufacturer. After quantitation using UV spectrometry, DNA samples were stored at -80°C for future use.
Genotypes of four select ANRIL SNPs, rs564398, rs2151280, rs1063192, and rs2157719 were determined as previously described (27).
Statistical analyses. Demographic and clinical characteristics of patients were analyzed using descriptive statistics. Characteristics were compared using Student's t-tests, χ2 tests or Fisher's exact tests where appropriate.
The consistency between the distribution of each ANRIL SNP in the study cohort and the Hardy-Weinberg principle and the potential nonrandom allele associations between selected SNPs were analyzed using χ2 tests.
Patients without known death/alive information were censored at the date of last follow up. Time to death, the time from transplant (Time 0) until death, was analyzed using Kaplan-Meier curves followed by the log-rank test. Multivariate analysis was conducted using the proportional hazard regression model of Cox (Cox PH regression) and a stepwise backward approach was applied for model reduction. The final proportional hazard regression model included factors with p-values less than 0.05 and was justified as previously described (28).
Potential associations between ANRIL SNPs and the incidence of death at Day-100 post transplant, GVHD, AKI, and neurotoxicity were evaluated using competing risks regression in which death was a competitive risk factor for GVHD, AKI, and neurotoxicity (29). All statistical analyses were conducted using R3.5.0 (R Foundation for Statistical Computing). p-Values were two-sided, and unless specified, p-values values of less than 0.05 were regarded as statistically significant.
Results
Patients. Table I summarizes the demographic characteristics of 103 patients included in this study. The mean age at transplant of this patient cohort was 55.6 years (±11.8 years), and 46 patients (44.7%) were older than 65 years of age. Additionally, 64 patients (62.1%) were male while 88 patients were Caucasian (85.4%). Sixty-seven patients (65%) were newly diagnosed at time of study entry. Disease diagnosis varied and 41 (49.8%), 8 (7.7%), 8 (7.7%), and 7 (6.8%) patients diagnosed with AML, ALL, DLBCL, and myelofibrosis, respectively. Eighty-nine patients (86.3%) received matched transplants, including matched sibling and matched unrelated donors, while 1 patient received a mismatched unrelated transplant. The remaining patients received cells from umbilical cord donors (13 of 103, 12.8%).
No significant associations were found between age, gender, race, risk score, the ECOG performance score, comorbidity score, baseline SCr, baseline CrCL, and weight-adjusted starting tacrolimus dose. The one exception was a correlation between ECOG performance score and comorbidity score. Patients with higher comorbidity scores (≥3) were more likely to have higher ECOG scores (1 and 2) than patients with low comorbidity scores (0-2) (p=0.007). This is consistent with previous studies showing that the ECOG score significantly increased in cancer patients with comorbidities in comparison with patients without comorbidities (30).
ANRIL SNPs in patients. A number of ANRIL SNPs have recently been associated with the development and treatment of human cancers. Four of these SNPs, ss564398, rs2151280, rs1063192, and rs2157719, were investigated in this study due to their reported correlations with human cancers (13-16, 23-27).
Table II summarizes the genotypes of the 4 ANRIL SNPs in this study cohort. The minor allele frequencies (MAFs) for these four SNPs were approximately 40%, and the genotype distributions of these SNPs followed the Hardy-Weinberg principle (all p-values >0.9) (27). For rs2151280, 31.1% (n=32), 49.5% (n=51), and 19.4% (n=20) of patients were homozygous for the major allele (genotype: CC; ancestral allele), heterozygous (genotype: TC), and homozygous for the minor allele (genotype: TT; variant), respectively. Further analyses showed that all four ANRIL SNPs were in significant LD (linkage disequilibrium) with each other in our study cohort (all p values <0.01), primarily ascribed to their close chromosomal proximity (27).
rs2151280 is associated with overall survival after allo-HSCT. We then evaluated the potential associations between patient characteristics and OS in the study cohort using univariate and multivariate survival analyses (Table III). Of the 103 patients, 50 (48.5%) died during the study period. The median time to follow-up was 1084 days (date of transplant time was defined as Time 0). Survival analyses using univariate models showed statistically significant associations with OS between age (>65 vs. ≤65; HR=2.21, 95%CI=1.26-3.88, p=0.006) and comorbidity score (≥3 vs. 0-2; HR=2.09, 95%CI=1.13-3.88, p=0.02). Moreover, ANRIL rs2151280 (CT/CC vs. TT) and disease type (AML vs. Other) were associated with OS with borderline significance (for rs2151280, HR=0.55, 95%CI=0.29-1.04, p=0.07; for disease type, HR=1.66, 95%CI=0.95-2.89, p=0.075), whereas rs564398 (CC vs. CT/TT) and risk score (1-3 vs. ≥ 4) tended to be associated with OS (for rs564398, HR=0.42, 95%CI=0.13-1.35, p=0.15; for risk score, HR=0.63, 95%CI=0.35-1.14, p=0.14). The median time to death for rs2151280 TT patients was 379 days. In contrast, the median time to death for rs2151280 TC/CC patients was >1300 days; thus, median survival was not achieved during the study period (Figure 1). Furthermore, no association was observed between OS and race, disease stage (relapsed vs newly diagnosed), baseline renal function (baseline CrCL), and starting tacrolimus dose (all p-values >0.20). In addition, there are no statistically significant differences in the characteristics of the patients harboring rs2151280 TC/CC and TT genotypes (data not shown).
Further multivariate analyses revealed that rs2151280, age, and comorbidity remained significantly associated with OS in this patient cohort (with p-values <0.05). After adjustment for age and comorbidity, the HR between the rs2151280 CT/CC group and the rs2151280 TT group was 0.44 (95%CI=0.22-0.85, p=0.015), suggesting that patients with rs2151280 TT genotype have higher risk for death after allo-HSCT. The adjusted HRs for age (Age >65 vs. Age ≤65) and comorbidity (≥3 vs. 0-2) were 2.24 (95%CI=1.27-3.94, p=0.005) and 2.57 (95%CI=1.35-4.89, p=0.004), respectively. These findings are consistent with previous studies associating advanced age and presence of comorbidities with poor prognosis (30). Taken together, rs2151280 TT appears to be a potential prognostic factor (in addition to age and comorbidity score) for OS of leukemia patients after allo-HSCT.
Impact of ANRIL SNPs on other clinical outcomes in patients with allo-HSCT. Lastly, we explored the potential associations between the aforementioned ANRIL SNPs and the incidence of death at Day-100, GVHD, AKI, and neurotoxicity in the study cohort. Our results showed that none of the select ANRIL SNPs was associated with death at Day-100. Similarly, no statistically significant correlation was observed between these genetic alterations and GVHD or AKI (all p-values >0.20). Instead, gender, comorbidity and baseline CrCL were statistically associated with AKI (p=0.015, 0.012, and 0.0006, respectively).
Discussion
In the current study, we revealed that ANRIL rs2151280 was significantly associated with OS in adult patients with hematologic malignancies after allo-HSCT. Patients harboring the rs2151280 TT genotype tend to have poorer OS. The underlying mechanism could be ascribed to ANRIL-mediated transcription suppression on the entire ARF-INK4 gene (7, 11, 12). Our previous study had shown that rs2151280 SNP leads to elevated expression of ANRIL in PBMCs from multiple myeloma patients, and such increase in ANRIL expression simultaneously downregulated the mRNA expression of p15, p16, ARF in PBMCs (27). Presumably, patients homozygous for rs2151280 TT would have decreased expression of ARF, which further negatively impacts p53-mediated apoptosis in patients with hematologic malignancies. Similarly, reduced expression of p16 and p15 in rs2151280 TT patients could down-regulate pRB-mediated cell-cycle control, promoting tumor progression or relapse. Hence, both p53 and pRB pathways in patients harboring rs2151280 TT could be impaired (27, 31), thus influencing the clinical outcomes in adult patients with hematologic malignancies.
There are several limitations in this study. First, the patient cohort was relatively small, and statistical comparisons were performed between two “unbalanced” groups (20 rs2161280 TT patients vs. 83 rs2151280 TC+CC patients). Moreover, the study cohort included diverse hematologic malignancies, such as AML, ALL, DLBCL, and myelofibrosis. The molecular mechanisms underlying these hematologic malignancies as well as their clinical outcomes vary considerably (32, 33). Accordingly, a large cohort of patients would be required to stratify patients based on specific hematologic malignancy. Second, this was a retrospective study and only whole blood DNA specimens were available for genotyping and genetic alteration analyses, precluding analysis of rs2151280 TT association with mRNA and protein expression of p15, p16, and ARF in blood. Lastly, the impact of ANRIL SNPs other than the select four were not determined. Regardless of these limitations, our results provide a foundation for better understanding the role of genomic variants in ANRIL in patients with hematologic malignancies after allogeneic hematopoietic stem cell transplantation.
Acknowledgements
This study was partly supported by a research grant from NIH, NIGMS/U01 GM092655 (WS) and a startup fund from The Ohio State University College of Pharmacy (MJP).
Footnotes
Authors' Contributions
MJP and WS contributed to the study design; JL, NS, XZ, JJ, ZV performed experiments and collected data; JL and MP analyzed the data and prepared the manuscript; MJP, WS, MP, and CH contributed to the clinical trial; all co-authors reviewed and approved the final version of the manuscript for submission.
This article is freely accessible online.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
- Received July 26, 2020.
- Revision received August 6, 2020.
- Accepted August 7, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved