Japanese Journal of Clinical Oncology Advance Access published April 28, 2015 Japanese Journal of Clinical Oncology, 2015, 1–6 doi: 10.1093/jjco/hyv064 Original Article
Original Article
Correlation of C/EBPα expression with response and resistance to imatinib in chronic myeloid leukaemia Sailaja Kagita1, Srihari Uppalapati1, Sadasivudu Gundeti1, and Raghunadharao Digumarti1,2,* 1
Department of Medical Oncology, Nizams Institute of Medical Sciences, Hyderabad, Andhra Pradesh, and 2Homi Bhabha Cancer Hospital and Research Centre, Visakapatnam, Andhra Pradesh, India
*For reprints and all correspondence: Raghunadharao Digumati, Homi Bhabha Cancer Hospital and Research Centre, Aganampudi, Visakapatnam 530053, Andhra Pradesh, India. E-mail:
[email protected] Received 12 January 2015; Accepted 1 April 2015
Abstract Objective: Altered differentiation is a common feature of haematopoietic malignancies with poor prognosis. CAAT/enhancer binding protein alpha (C/EBPα) is a key transcription factor that regulates myeloid differentiation. This study is aimed to know the prognostic value of CAAT/enhancer binding protein alpha expression and correlate its expression with response to imatinib therapy. Methods: We quantified the expression of C/EBPα gene in 126 chronic myeloid leukaemia samples (82 from newly diagnosed and 44 from imatinib-resistant patients) and 20 control samples. C/EBPα mRNA level was measured by real-time quantitative polymerase chain reaction using the ΔΔCT method. Results: C/EBPα expression level was significantly lower in the imatinib-resistant group than in the pretreatment and control group (P = 0.0398). Low CAAT/enhancer binding protein alpha levels in the imatinib-resistant group were significantly associated with advanced phase (P = 0.04), with more peripheral blasts (P = 0.0001), high BCR-ABL levels (P = 0.018) and T315I and P-loop mutations (P = 0.0002). In the pretreatment group, low expression showed association with high EUTOS risk score (P = 0.03) and possible partial cytogenetic response (P = 0.010). Conclusions: Our results suggest that low expression of CAAT/enhancer binding protein alpha might have a role in the response to imatinib and progression of disease in patients with chronic myeloid leukaemia. Key words: chronic myeloid leukaemia, C/EBPα, expression, imatinib, response, BCR-ABL TK mutations
Introduction Chronic myelogenous leukaemia (CML) is a clonal stem cell disorder of myeloid precursors characterized by the presence of BCR-ABL fusion gene. BCR-ABL is a constitutively active tyrosine kinase that perturbs downstream signalling pathways, thereby leading to increased proliferation and survival of leukaemic cells. CML is a triphasic disease, most of the CML patients being diagnosed in chronic phase (CP). As a result of genomic instability, it progresses
to accelerated phase (AP) and thence to the blast crisis (BC). Imatinib (IM) is one of the frontline therapies used for the management of CML. Suboptimal responses/resistance to IM is more common in patients who start IM in late CP and in advanced phases of CML. The mechanism of resistance is complex and may implicate additional or secondary chromosomal changes, which leads to CML evolving in to BC with a poor prognosis and shorter survival (1).
© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email:
[email protected]
1
2
Correlation of C/EBPα expression
Myeloid differentiation is controlled by a coordinated network of several transcription factors that regulate the expression of important differentiation-related genes, including those encoding growth factors and their receptors (2,3). CAAT/enhancer binding protein alpha (C/EBPα) is a key transcription factor. It regulates granulocytic differentiation. C/EBPα is involved with all three haematopoietic check points: cell proliferation, differentiation and apoptosis; hence, impairment of one or more of these pathways may induce leukaemia (4). C/EBPα is expressed at high levels throughout myeloid differentiation and binds to the promoters of specific genes at different stages of myeloid maturation (5). C/EBPα activity is repressed at different levels by PML-RARα (transcriptional), BCR-ABL (translational) and AML-ETO chromosomal translocations ( post-translational) (6–9). Functional inactivation of C/EBPα gene has been associated with leukaemias (10,11). Down-regulation of C/EBPα gene activity/expression inhibits myeloid differentiation and contributes to myeloid transformation. Hence, this study is aimed to evaluate the prognostic value of C/EBPα mRNA expression levels in patients with CML.
Patients and methods This study includes 126 CML patients: 82 are newly diagnosed and 44 are IM-resistant. Blood was collected from patients of CML attending the Department of Medical Oncology, NIMS, between 2011 and 2013. The study was approved by the institutional ethics committee and an informed consent was obtained from every participant. The median age at presentation was 35 years (range 9–69 years) for new and 39 years (range, 18–78 years) for IM-resistant cases. The male-to-female ratio varied in the two groups: it was 1.15 in new and 2.38 in resistant cases.
RNA extraction, cDNA synthesis and quantitative mRNA analysis using the ΔΔCT method Total RNA extraction was done using the TRIzol method, after RBC lysis. RNA (1 µg) was reverse-transcribed into complementary DNA (cDNA), using high capacity reverse transcription kit (Applied Biosystems, Foster city, CA). Commercially available primers and Taqman probe from Applied Biosystems (Taqman Pre-Developed Assay Reagents for gene expression) were used for quantification of C/EBPα mRNA expression (Hs00269972_S1). Actin ß gene (Hs99999903_M1) served as endogenous control. Control cDNA was included in the RT-PCR. The reaction mixture of 20 µl contained 0.25 µl of primer (250 nmol/l) and probe (150 nmol/l), 10.0 µl of expression assay mix, 7.0 µl nuclease free water and 2 µl of cDNA. The thermal cycling conditions were as follows: 2 min at 50°C, 5 min at 95°C, followed by 45 cycles 95°C for 15 s and 60°C for 1 min. All the samples were run in duplicate to minimize handling errors. Assays included negative controls in all stages of reactions to assess specificity and to rule out contamination. Relative quantification was performed by the ΔΔCT method and fold change was measured relative to the reference gene in terms of cycle threshold (CT).
features in both new and resistant groups. The mean haemoglobin (Hb) and mean platelets were found to be higher in the pretreatment group compared with the IM-resistant group (P = 0.048; P = 0.029; Figure 1). There was slight variation observed with total leucocyte count (TLC), granulocytes, lymphocytes and monocytes between the two groups (Table 2). The proportion of granulocytes at each differential stage from myeloblasts to mature granulocytes was described in the two groups; the mean blasts were observed to be more in the pretreatment compared with the IM-resistant group (P = 0.032) (Table 3). A total of 126 CML samples (82 newly diagnosed and 44 IM resistant) and 20 control samples were analysed for C/EBPα mRNA expression, using quantitative real-time PCR. The mean expression observed in the control group was 145.72 ± 40.29, in the pretreatment group 219.9 ± 41.99 and in the resistant group 80.90 ± 24.29. The expression level was significantly lower in the IM-resistant group than that in the pretreatment group and control group (P = 0.0398) (Table 4).
C/EBPα expression versus clinical variables C/EBPα expression levels were correlated with different clinical variables of pretreatment CML and IM-resistant CML groups (Table 4). With respect to the phase of the disease, low expression levels were
Table 1. Clinical characteristics
Clinical phase Chronic Accelerated + BC EUTOS risk High Low Sokal risk High Intermediate Low
New cases (n = 82)
IM-resistant cases (n = 44)
P value
No.
%
No.
%
67 15
81.70 18.29
35 9
79.54 20.45
0.768
25 57
30.48 69.51
13 31
29.54 57.40
0.913
32 30 20
39.02 36.58 24.39
12 15 17
27.27 34.09 38.63
0.206
IM, imatinib; BC, blast crisis.
Statistical analysis Two-sided unpaired Student’s t-test and ANOVA tests were performed to determine the statistical significance of experimental results. Probability (P) values 3% peripheral blasts (PB) had higher levels of C/EBPα expression (P = 0.04). The BCR-ABL expression levels were correlated with C/EBPα expression and are presented in Table 4. Based on the median BCR-ABL levels, patients were grouped into two categories: less than median and more than median. In both pretreatment and IM-resistant groups, patients with more than median BCR-ABL levels had lower C/EBPα expression (Figure 4). In the IM-resistant group, C/EBPα expression was significantly lower in patients having more than median BCRABL levels (P = 0.018).
Pretreatment group (new cases)—C/EBPα expression versus cytogenetic response at 6 months after IM initiation Out of 82 pretreatment cases, 67 patients were diagnosed in CP and 15 in advanced phases of CML. Among these 15 advanced cases, 10 cases
died within 6 months of IM therapy. All patients were started on IM treatment, after 6 months of IM initiation 57.31% (47/82) patients achieved complete cytogenetic response (CCyR), 20.73% (17/82) had partial cytogenetic response (PCyR), 12.19% (10/82) died and 9.75% (8/82) lost follow-up (LFU). The mean C/EBPα expression in complete cytogenetic responders and partial responders was 342.0 ± 85 and 213.1 ± 76.77, respectively. The expression was found to be significantly lower in partial cytogenetic responders compared with those with CCyR (P = 0.010) (Table 4).
IM-resistant group—C/EBPα expression versus BCR-ABL TK (tyrosine kinase) domain mutations Forty-four cases were refractory to IM therapy at the standard dose of 400 mg daily. Out of 44 resistant cases, 26 had primary resistance and 18 had secondary resistance (14 had loss of CCyR, 2 each had loss of complete haematological response and complete molecular response). These patients were screened for BCR-ABL tyrosine kinase domain mutations, using the direct sequencing method. Of 44 resistant cases, 45.45% (20/44) patients carried TK domain mutations. When C/EBPα expression level was compared in patients with the presence and absence of BCR-ABL kinase domain mutations, expression was statistically lower in patients with P-loop and T315I mutation and other mutations versus patients with no mutations (P = 0.0002; Table 5).
Discussion C/EBPα regulates haematopoietic stem cell development and loss of its function might lead to differentiation block in myeloid cells (10). C/EBPα alterations comprise genomic mutations, promoter methylation, reduced mRNA expression and aberrant post-translational modifications. They are often observed in human myeloid leukaemias (10,12). Suppression of C/EBPα expression or function is associated with myeloid leukaemia in humans (9,10). C/EBPα mutations have been reported in 4–11% of adult AML patients (13) and promoter methylation is reported in 12–51% of AML patients (12). Earlier studies demonstrated that CEBPα expression was significantly lower in myeloid leukaemias (8,14), breast cancer (15), lung and skin adenocarcinomas (16), epithelial tumours (17,18), pancreatic cancer (19), gastric carcinoma (20) and cervical cancers (21). C/EBPα is expressed only in granulocytic cells and its expression levels are quite different between myeloblasts and mature granulocytes. We observed that the mean TLC, platelet counts, neutrophils and blasts were found higher in new cases compared with resistant cases. In this study, C/EBPα mRNA expression was analysed in pretreatment and IM-resistant CML patients. Expression was statistically
4
Correlation of C/EBPα expression
Table 4. Association of C/EBPα expression between new and IM-resistant CML cases Parameter
New cases (n = 82) Mean ± SD
Cases 136.3 ± 27.90 Clinical phase Chronic 309.8 ± 69.25 Accelerated+ BC 300.2 ± 98.08 EUTOS risk High 225.0 ± 79.53 Low 354.5 ± 81.19 Sokal risk High 278.9 ± 78.01 Intermediate 287.5 ± 106.8 Low 350.3 ± 123.5 Peripheral blasts 3% 330.2 ± 81.55 BCR-ABL expression < Median 137.1 ± 47.19 > Median 132.4 ± 35.76 Cytogenetic response at 6 months CCyR 342.0 ± 85 PCyR 213.1 ± 76.77
IM-resistant cases (n = 44) Total
P value
Mean ± SD
Total
P value
82
–
80.90 ± 24.29
45
0.01
67 15
0.12
72.44 ± 23.49 38.40 ± 22.72
35 9
0.04
25 57
0.03
106.9 ± 49.33 48.35 ± 21.88
13 31
0.065
19.43 ± 12.05 70.72 ± 30.11 68.80 ± 31.32
12 15 17
32 30 20
0.89
0.002
28 54
0.04
93.15 ± 33.43 23.87 ± 10.66
24 19
0.0001
33 42
0.21
70.95 ± 30.83 41.64 ± 16.10
18 22
0.018
47 17
0.010
C/EBPα, cCAAT/enhancer binding protein alpha; CML, chronic myelogenous leukaemia; CCyR, complete cytogenetic response; PCyR, partial cytogenetic response.
Figure 2. C/EBPα expression versus clinical phase in the IM-resistant group.
Figure 3. C/EBPα expression versus Sokal risk in the IM-resistant group.
lower in the IM-resistant group compared with the pretreatment group (P = 0.018). In the pretreatment group, lower C/EBPα levels showed significant association with patients having EUTOS high-risk and PCyR. EUTOS score predicts CCyR at 18 months in patients being treated with tyrosine kinase inhibitor (TKI) therapy (22). We observed significant association of reduced C/EBPα expression levels and IM-resistant group with all the clinical parameters. Previous studies reported that expression of genes is frequently altered in resistant and advanced phase patients, which might be due to the presence of additional epigenetic abnormalities and increased BCR-ABL levels (1,23). A recent study from our site observed significant association
between the CML phases and C/EBPα gene promoter hypermethylation (P = 0.017) (24). In the IM-resistant group, low C/EBPα levels were significantly associated with patients in advanced phase, and patients having >3% peripheral blasts (Figure 5). In the pretreatment group, patients with high EUTOS risk scores had significant lower levels. Earlier studies reported that in initial CP, BCR-ABL provides a survival advantage in myeloid progenitors without affecting differentiation (Figure 6). During blastic transformation, the BCR-ABL expression specifically increases in haematopoietic stem cells and committed myeloid progenitor clones with increasingly malignant characteristics (14) (Figure 7). In addition, loss of C/EBPα in CML BC depends on
Jpn J Clin Oncol, 2015
5
Figure 4. C/EBPα expression versus BCR-ABL levels in the IM-resistant group. Figure 6. C/EBPα expression versus BCR-ABL kinase domain mutations in the IM-resistant group.
Table 5. C/EBPα expression versus BCR-ABL kinase domain mutations in IM-resistant cases Mean ± SD BCR-ABL kinase domain mutations P-loop and T315I mutations 7.605 ± 2.338 Other mutations 26.30 ± 15.35 Absence of mutation 40.16 ± 10.56
Total
P value
11 9 24
0.0002
Figure 7. C/EBPα expression versus EUTOS risk in the pretreatment group.
Figure 5. C/EBPα expression versus peripheral blasts (PB) in the IM-resistant group.
the BCR-ABL activity (10). BCR-ABL kinase inhibitor (IM) restores C/EBPα expression and granulocytic differentiation in the earlier phases of CML (8) (Figure 8). Our results are similar to those reported by Tavor et al., (2003) who reported that the KCL cell line, derived from BCR-ABL positive CML BC patients, expressed non-functional C/EBPα protein (25). Halmos et al. (2002) observed reduced C/EBPα expression in advanced stages of lung adenocarcinoma and poorly differentiated cancers (16,26). The expression levels were higher in the pretreatment group of patients having >3% blasts versus
