Protein signatures as potential surrogate biomarkers for stratification and prediction of treatment response in chronic myeloid leukemia patients
- Authors:
- Published online on: July 7, 2016 https://doi.org/10.3892/ijo.2016.3618
- Pages: 913-933
-
Copyright: © Alaiya et al. This is an open access article distributed under the terms of Creative Commons Attribution License.
Abstract
Introduction
Chronic myeloid leukemia (CML) is unequivocally distinguishable from other myeloproliferative disorders by the presence of a reciprocal translocation of chromosomes 9 and 22 (1–3). Although the Philadelphia chromosome is detected in 90–95% of CML patients, evidence of the BCR-ABL rearrangement is also usually detected in the subgroup of Philadelphia chromosome-negative CML patients (4–6).
The presence of BCR-ABL in CML patients and the requirement of kinase activity for BCR-ABL function make this an attractive target for selective kinase inhibitors.
The old traditional therapy of newly diagnosed chronic phase-CML patients includes busulfan and hydroxyurea and most of the patients will stay in a chronic phase for approximately 3–5 years (7,8). Treatment of CML later evolved to where the goal was prolongation of the chronic phase through induction of karyotypic remission and possibly molecular remission using Alfa-interferon therapy with or without cytosine arabinoside. Thereafter, imatinib mesylate (IM) a tyrosine kinase inhibitor (TKI) was introduced as potential molecular therapy for CML (7,9). IM is capable of inhibiting BCR-ABL kinase activity by blocking ABL tyrosine kinase action through the binding and subsequent inactivation of the ATP-binding sites of ABL tyrosine kinase in leukemic cells (9,10). Since its introduction, several clinical trials have demonstrated the efficacy of IM and new generation TKIs in the treatment of CML, including patients with interferon-refractory CP-CML, as well as patients with CML in blast crisis (11).
Approximately more than 50% of CML patients treated with imatinib achieve a complete cytogenetic response (11,12). CML progression while on imatinib is usually due to the emergence of imatinib-resistant BCR-ABL mutant cells.
The relatively unpredictable biological behavior is a major challenge in its management as the chronic phase of CML is less aggressive and has very favorable prognosis with an excellent 5-year survival rate. By contrast, the biologically aggressive blast phase of CML is often rapidly fatal (2). Currently, there is no recognized prognostic value for the baseline BCR-ABL level, furthermore, there are variations in sensitivity or dependability of RQ-PCR assays across different laboratories (13). There is therefore a need to develop molecular markers for selection of choice of therapy at the time of diagnosis and to identify patients that are more likely to achieve a sustained remission, and patients who are more likely to develop resistance to imatinib therapy.
New analytical tools in proteomics are emerging that give new insights into biological processes that may speed up the discovery of potential biomarkers. Quantitative molecular variations may be used for the development of methods for tumor classification based on large amounts of gene expression data generated by 2-DE analysis of proteins (14,15).
The main aim of the present study is towards discovery of objective markers that predict patients’ response status and selection of appropriate choice of therapy at the onset of disease diagnosis. It focuses on the analysis of global peripheral blood plasma and bone marrow plasma protein expression profiles among CP-CML patients who achieved LT-MMR on imatinib compared with patients without MMR as well as whether or not they remain on TKI or switch to second generation TKI or requiring alternative therapy.
The endpoint is to identify disease-specific/disease-associated protein biomarkers seen in bone marrow tissue as well as in peripheral blood plasma. This would subsequently allow monitoring of such biomarker proteins in peripheral blood, rather than bone marrow, demanding less invasive procedures for objective prediction of individual’s best treatment options and prognostic monitoring of CML patients.
Materials and methods
All bone marrow samples were obtained by aspiration procedure via posterior iliac crest under local anesthesia. Because of limited amount of materials for analysis, the cells were not flow cytometry sorted, rather unsorted bone marrows as well as unsorted peripheral blood plasma were collected and prepared for analysis.
Bone marrow and plasma, samples obtained at diagnosis and prior to initiation of treatment from 37 patients with newly diagnosed CP-CML were subjected to expression proteome analysis using combined gel-based 2-DE and label-free in-solution quantitative liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). Patients selections into those that achieved or did not achieve MMR was based on patients with serial positive or negative responses to treatment at different time-points (3, 6, 12 and 24 months, respectively). Patients that responded at a time-point but failed to respond at the next time-point were not included in the analysis. However, patients that did not achieve MMR at 3 months, but subsequently achieved MMR at 6, 12 and 24 months were included. Because there was fewer number of patients with MMR at 3 months, the focus of our analyzed time-points were at 6, 12 and 24 months. Twenty-five patients consisting 13 with major molecular response and 12 without major molecular response were analyzed. In addition, patients that failed tyrosine kinase inhibitor (TKI) were analyzed. Four additional patients samples not included in the proteomics analysis were used in the western blot analysis. The overview of experimental design is shown in Fig. 1 and the clinical characteristics of all patients were as indicated in Table I.
Sample preparation protocols for proteomic analysis
All the patients with primary diagnosis of CML were recruited in Oncology Center at KFSH&RC. From each of the patients, 10 ml of peripheral EDTA-anti-coagulated blood (plasma) was taken. Where possible, bone marrow aspirations were obtained from the same patients in addition to peripheral blood samples.
All samples were subjected to extensive pre-analysis cleanup using human albumin removal protocols (Agilent Technologies). Written and signed informed consents were obtained from all patients and the Institution’s Research Advisory Council, under the Office of Research Affairs, approved the study (RAC# 2050-040).
Protein separation by high resolution two dimensional gel electrophoresis, (2-DE) scanning and image analysis
Equivalent amount of 50 mg total proteins for each analyzed sample was dissolved in 350 ml volume of rehydration buffer [2% (v/v) IPG-buffer 4–7 linear] and loaded onto an 11-cm IPG-strip 4–7 linear (Bio-Rad Laboratories). This gave better overview of gel separated protein spots across the entire chosen pH window and gel images were visualized by SYPRO Ruby fluorescent staining. Stained gels were scanned using a Typhoon Trio Imager (GE) and data were analyzed using the Progenesis SameSpots software (version 7.1.0; Nonlinear Dynamics, Ltd., Newcastle, UK). Gel images were compared for qualitative and quantitative differences. In addition, the protein expression profiles were used to assess the level of individual variability and only samples with similar phenotypic changes were used for sample pools for LC/MS/MS (due to low through-put analysis) as detailed below. Polypeptide quantities were calculated based on the normalized total integrated density volume.
Protein in solution-digestion
The plasma samples were diluted and protein concentrations of all samples were normalized as previously described (16). Briefly, for analytical runs, equal amount of protein was taken from each sample to generate a pool of patient as one group. The samples within same sample cohort were pooled due to low through-put of LC/MS/MS analysis platform. However, samples were initially screened using 2-DE for homogeneity within the same analysis group. For each analysis sample group, 200 μg complex protein mixture was taken and exchanged twice with 500 μl of 0.1% RapiGest (Waters Corp., Manchester, UK). Protein concentrations of between 0.50 and 1 μg/μl was achieved at the end of digestion. Details of digestion protocols are as previously described (16,17). Briefly, proteins were denatured in 0.1% RapiGest SF at 80°C for 15 min, reduced in 10 mM DTT at 60°C for 30 min, and alkylated in 10 mM Iodoacetamide (IAA) for 40 min at room temperature in the dark. Samples were trypsin digested at 37°C overnight. Samples were diluted with aqueous 0.1% formic acid prior to LC/MS analysis in order to achieve a load of ~2 μg on analytical column. All samples were spiked with yeast alcohol dehydrogenase (ADH; P00330) as internal standard to the digests in order for absolute quantitation.
Protein identification by mass spectrometry: LC-MSE analysis
The digested peptides were subjected to 1-Dimensional Nano Acquity liquid chromatography coupled with tandem mass spectrometry on Synapt G2 (Waters Corp.). Expression proteomics data were generated between sample groups using both qualitative and quantitative protein changes. The ESI-MS analysis and instrument settings were optimized on the tune page as previously reported (16).
A total of 2 μl sample injection representing ~1 μg protein digests was loaded on-column and samples were infused using the Acquity sample manager with mobile phase consisting of A1 99% water +1% acetonitrile + 0.1% formic acid and B1 acetonitrile + 0.1% formic acid with sample flow rate of 0.450 μl/min. Data acquisition using iron mobility separation experiments (HDMSE) were performed and data were acquired over a range of m/z 50–2000 Da with a total acquisition time of 115 min. All samples were analyzed in triplicate runs (triplicate runs were repeated on two different occasions as a measure of reproducibility) and data were acquired using the MassLynx programs (version. 4.1, SCN833; Waters) operated in resolution and positive polarity modes. ProteinLynx Global Server (PLGS) 2.2 and Progenesis QI for proteomics (Progenesis QIfp version 2.0.5387) (Nonlinear Dynamics/Waters) were used for all automated data processing and database searching. The generated peptide masses were searched against two-unified non-redundant databases (Uniprot/Swiss-Prot Human protein sequence database) using the PLGS 2.5 and Progenesis QIfp for protein identification (Waters).
Data analysis and informatics
Progenesis QI v.2.0.5387 for proteomics was used to process and search the data to accurately quantify and identify proteins that are significantly changing between sample groups. The human database containing thousands of reviewed non-redundant entries were downloaded from UniProt/Swiss-Prot and search algorithm was applied as previoudly described (18). The criteria used for the database search were as previously described (16). Normalized label-free quantification was achieved using Progenesis QI software. The generated differentially expressed data was filtered to show only statistically (ANOVA), significantly regulated proteins (P≤0.05) and a fold change >1.5. In addition, ‘Hi3’ absolute quantification was performed using ADH as an internal standard to give an absolute amount of each identified protein. These options are available as incorporated into the Progenesis QIfp (Nonlinear Dynamics/Waters).
Results
Changes in protein expression between patients with/without major molecular response at 6 months
A total of 73 protein spots on 2-DE gels differed significantly between patients that achieved MMR from those who did not achieve MMR (P<0.05 and at least 1.5-fold difference). The locations of these protein spots are shown as marked on a representative 2-DE map derived from a sample with MMR in Fig. 2A. Even though the identifications of these protein spots were not done, their quantitative expression fingerprints from 2-DE analysis pattern accurately predicts 13 individuals that achieved MMR at 6 months from 12 subjects without MMR (No-MMR) using principal component analysis (PCA) (Fig. 2B).
These findings are similar to what was observed with PCA plot generated from non-gel LC/MS/MS analysis platform, as some of the results were independently validated using the label free quantitative liquid chromatography tandem mass spectrometry as detailed below.
LC/MS/MS analysis of peripheral blood for prognostic monitoring of early CML treatment response
Peripheral blood samples were evaluated for early treatment response at 6 month and prediction of treatment options towards personalized medicine. Approximately 115 protein species were identified, of which only 64 were significantly differentially expressed between MMR and No-MMR sample groups. (> 1.5- to ∞-fold change, p<0.05). These proteins predict accurately patients with MMR vs. No-MMR patients using unsupervised Hierarchical Cluster Analysis (Fig. 3).
Evaluation of bone marrow and peripheral blood protein profiles for prognostic monitoring of prolonged and sustained treatment response vs. persistent no-major molecular response
Some of the patients have been followed for more than 24 months. Patients who have been consistent over a long-term in achieving and maintaining MMR from 6 months until 24 months were labeled as LT-MMR, while patients that have been persistent with No-MMR from 6 months until 24 months were called P-No-MMR. We believe that the ability to select early responders from 6 months all through 24 months would be very helpful to identify markers that would accurately predict patients with risk of delayed or suboptimal response further than 6 months. These cohorts of patients were considered as important in an effort to provide the possibility to identify surrogate biomarkers to evaluate long-term treatment response and discovery of disease-specific/disease-associated proteins for objective prognostic monitoring of CML patients.
Equal amounts of total peripheral blood plasma proteins from 10 LT-MMR patients were pooled and compared for their protein expressions among 10 other samples from P-No-MMR patients using quantitative label-free LC/MS/MS expression proteome analysis.
Approximately 700 proteins representing 280 unique protein species were identified (due to different protein isoforms). Only 164 of the 280 proteins were significantly differentially expressed between LT-MMR and P-No-MMR sample groups (>1.5- to ∞-fold change; P<0.05) and accurately predict patients with major molecular response (LT-MMR) vs. No-major molecular response (P-No-MMR) using unsupervised principal component analysis (Fig. 4A). The list of identified differentially expressed proteins in PBP is described in Table IIA.
 | Table IIThe identified differentially expressed proteins in peripheral blood plasma (PBP) and bone marrow plasma (BMP) from CML patients with major molecular response (MMR), No-MMR, On-tyrosine kinase inhibitor (On-TKI) and NOT-on-TKI. |
Similar to peripheral blood samples, >700 proteins representing 250 unique protein species were identified when similar analysis was done on bone marrow pooled samples from 8 LT-MMR patients and 8 P-No-MMR patients. One hundred and thirty-eight of the total identified proteins were significantly differentially expressed between LT-MMR and P-No-MMR bone marrow sample groups (>1.5- to ∞-fold change, P<0.05; Table IIB). These proteins predict accurately LT-MMR patients vs. P-No-MMR patients using unsupervised principal component analysis (Fig. 4B). These results were subsequently evaluated for comparisons with the patterns obtained in early treatment response at 6 months. Notably, the pattern and accuracy of clustering of samples is very similar to that observed with the hierarchical cluster analysis plots at 6 months (Fig. 3).
Protein fingerprinting for prediction of treatment options for individualized therapy
Towards achieving the goal of personalized medicine, the above observed differentially expressed proteins between samples derived from LT-MMR patients vs. P-No-MMR patients were evaluated for their potential for objective prediction of treatment options for some of these cohorts of CML patients. Interestingly, the panel of 164 and 138 differentially expressed protein datasets derived from peripheral blood plasma (PBP) and bone marrow (BM) respectively, also discriminates patients that stay on IM after 1 year of treatment from patients that ultimately required alternative treatment options (second generation TKI/others) (Fig. 5). Following >2 years of follow-up of these patients the same dataset of potential protein biomarkers could still accurately separate all analyzed sample groups into their respective molecular response and treatment sub groups, indicating their usefulness for treatment monitoring as well as prediction of best choice of therapy for individual patient. Some of the identified proteins were implicated in hematological diseases as potential biomarkers using ingenuity pathway analysis (IPA) (Fig. 6). Functional annotations/disease affiliations of some of these proteins implicated in CML are further described under discussion below.
Identification of protein changes in BM as a reflection of detectable changes in peripheral blood
One of the main goals of this study was to identify/develop disease-specific/disease-associated protein biomarkers seen in bone marrow tissue as well as in peripheral blood plasma. This would subsequently allow monitoring of such biomarker proteins in peripheral blood, rather than bone marrow, demanding less invasive procedures for objective prediction of individual’s best treatment options and prognostic monitoring of CML patients. We therefore explored the possibility whether the proteins that are significantly differentially expressed in bone marrow do also show similar expression pattern in peripheral blood. With this in mind, we calculated how many of the 164 differentially expressed proteins in peripheral blood and the 138 protein dataset in bone marrow are common to both body compartments. We found that only 54 proteins (~35%) were in common between the two 164 and 138 datasets as described above. This set of 54 proteins was then subjected to unsupervised hierarchical clustering and correspondence analyses. As shown in Fig. 7, all sample groups were distinctively separated into four response subtypes using unsupervised hierarchical cluster analysis. The common proteins between the two body fluid compartments were highlighted in bold in Table II.
Validation by western blot analysis of some of the identified proteins
In an attempt to validate some of the differentially expressed proteins, we have used immunoblotting analysis. Nine individual samples consisting of 4 samples not included in the proteomics analysis and 5 other samples from the proteomics analyzed sample groups were tested for their expression of haptoglobin and hemoglobin using specific antibodies against these proteins. The expression levels of these proteins across all sample groups were consistent with the average protein normalized levels seen with label-free quantitative LC/MS/MS analysis (Fig. 8). Large scale validation of the majority of these proteins was beyond the scope of this study in order to develop limited panel of markers for clinical trial in a later study.
Discussion
Clinical and molecular diagnosis of most hematological malignancies including CML can be accurately made; however, prediction of treatment response elude the currently available tools for patient care.
A subset of significantly differentially expressed proteins from both peripheral blood and bone marrow were selected for their ability to discriminate samples derived from CML patients that responded differently to initial first line treatment with imatinib. Our strategy of proteomics mining of BM and PBP from the same individual patient would provide unique possibility to identify biomarkers from both sources thus, entailing less invasive procedures.
Report of microarray analysis of peripheral blood and bone marrow of CML samples in blast crisis cells, has been shown with demonstrable biological changes between two bodily fluids (19). Our analysis of peripheral blood samples of 164 differentially expressed proteins show that all samples were correctly classified and similar result was observed with 138 protein changes in bone marrow samples as shown in Fig. 4. Only 54 proteins were shown to be commonly differentially expressed between blood dataset and bone marrow protein dataset in the present study, supporting our notion that it might be possible to identify significant changes in the bone marrow of CML patients that are measurable at peripheral blood compartment for routine diagnostics.
We have attempted to use both the BMP and PBP data-sets that accurately predict patients MMR status for possible prediction of patients that continue to stay on IM after 1 year of treatment vs. those that ultimately required alternative treatment options (second generation TKI/others). Thus, the expression of the 158 protein changes in BM between MMR and No-MMR were further evaluated in 16 unrelated patients that stay on TKI after 1 year of imatinib treatment from patients that ultimately required alternative treatment options (second generation TKI/others). We found four distinct clusters with samples with MMR and No-MMR being very closely separated (not as distinct as in Fig. 4), while patients that stay on TKI (i.e. after 1 year of imatinib) treatment were distantly separated from patients that ultimately required alternative treatment options (second generation TKI/others) as shown in Fig. 5, meaning that it will be challenging to construct a universal model for management of CML patients and that prognostic datasets need to be created for each specific response type.
We have used two independent proteomics analysis platforms in the present study. The expression profiles of 2-DE protein spots successfully discriminated two sample groups of CML patients with MMR and No-MMR. We recognized the inherent limitation of 2-DE based studies (20–22) hence, we have in addition used label-free quantitative protein expression using high definition liquid chromatography tandem mass spectrometry (LC/MS/MS) to extensively map the proteome of bone marrow as well as peripheral blood samples.
Previous studies have used multivariate statistical algorithms and artificial learning models to predict cancer prognosis and for grading different solid tumors (15,23–28). The majority of these studies reported varying degrees of sensitivity and specificity based on evaluation of different clinical parameters (20,24).
Gene expression studies on hematological disease have been largely carried out by analysis of DNA or RNA microarrays. These genomics studies have indicated the potentials of large scale analysis of gene expression towards better understanding the molecular basis of leukemogenesis and that this information could potentially be useful in the classification of subtypes of hematological malignancies (19,29,30). In a recent study of CLL samples, Alsagaby and colleagues used combined transcriptomics and proteomics analyses to unravel the heterogeneity of gene expression patterns as well attempting to identify proteins that are implicated in prognosis of chronic lymphocytic leukemia (31). Recent studies have attempted to evaluate protein changes between imatinib sensitive and resistance samples (32) as well as to better understand the molecular mechanism in therapy resistance at the level of bone marrow extracellular fluid in CML (33).
Our initial analysis of 64 differentially expressed proteins of peripheral blood for prognostic monitoring of early CML treatment response at 6 months was encouraging and led us into extensive analysis of samples with sustained long-term MMR against patients that persistently could not achieve MMR.
Some of the identified proteins in the bone marrow of the 138 dataset for the prolonged and sustained MMR vs. persistent No-MMR were further evaluated for their functional characteristics and their hematological disease relevance using ingenuity pathway analysis (IPA). In the canonical pathway analysis of network signaling of identified proteins, only 48 of the 138 identified differentially expressed proteins were represented in the IPA database. The analysis of the identified proteins is composed of multiple networks of which, one is implicated in hematological disorders. The cellular localization, interconnections and functional annotation as well as the expression profile of some of these 48 identified molecules are as detailed in Fig. 6A. A review of some of these molecules showed that they mostly regulate among others: cell-to-cell signaling and interaction, hematological system development and function, aggregation of blood cells, coagulation, as well as quantity of granulocytes as indicated in Fig. 6. Among the identified proteins in this study is TYRO3 protein tyrosine kinase, a member of TAM family of receptor tyrosine kinases (RTKs) and known for their role as regulator of cellular proliferation, migration and survival processes, as well as maintenance of blood coagulation equilibrium (34). We observed connection of TYRO 3 in AKT/P13K pathway; similar to that previously described (34–36).
The S100A8 is a calcium-binding protein of the S100 family and have been described to be associated with myeloid differentiation (37). We observed a more than 9-fold differential expression of S100A8 and in the network connecting with RAS, TGFb, MAPK and MMP. The S-100 protein has been previously reported as a useful marker in juvenile chronic myeloid leukemia (JCML) as well as myeloid leukemia cutis (LC) (38,39).
Overexpression of MYC has been associated with CML with poor response to imatinib (40,41). We observed a more than 25-fold differential expression of MYC associated factor x in this study.
Altogether our findings indicate that rather than the use of a single marker, analyses of a panel of protein markers have the potential to provide better insight into complex biologic processes towards better prognostication of CML patients.
We recognize the limitation of this study as samples were prospectively collected and patients observed over the years for their treatment responses. One other issue with this study is the low number of patients enrolled in different clinical and molecular response groups; hence we have limited the analysis to evaluation of patients based on MMR and whether or not they are on IM or alternative treatment option (second generation TKI/others).
In conclusion, we have identified protein signatures capable of prediction of molecular response and choice of therapy for CML patients at 6 months and beyond using expression proteomics as objective stratification of CML patients for treatment options.
Although these results are very promising, we recognized that analysis of much larger materials of patients with similar treatments and responses will be necessary to validate if clustering analysis can be used as a routine prognostic tool for CML patients.
These proteins might be valuable once validated, to complement the currently existing parameters for reliable and objective prediction of disease progression, monitoring treatment response and clinical outcome of CML patients as a model of personalized medicine.
Acknowledgements
We thank Dr Abdelilah Aboussekhra for critical review, as well as Mr. Melvin Velasco, Mr. Parvez Siddiqui, Mr. Romeo Caracas and Ms. Tusneem Elhassan for technical assistant. We acknowledge the assistant and support of Mr. Faisal Al Otaibi and the logistics and purchasing department, RC, KFSH&RC. The present study was supported by the Research Center Administration, KFSH&RC, Riyadh, Saudi Arabia (RAC# 2050 040).
Abbreviations:
|
CML |
chronic myeloid leukemia |
|
CMR |
complete molecular response |
|
CP |
chronic phase |
|
DAS |
dasatinib |
|
IM |
imatinib mesylate |
|
MCyR |
major cytogenetic response |
|
MMR |
major molecular response |
|
No-MMR |
no-major molecular response |
|
LT-MMR |
long-term-MMR |
|
P-No-MMR; persistent-no-MMR; TKI |
tyrosine kinase inhibitors |
|
2-DE |
two-dimensional gel electrophoresis |
|
LC-MS/MS |
liquid chromatography coupled with tandem mass spectrometry |
References
|
Bartram CR, de Klein A, Hagemeijer A, van Agthoven T, Geurts van Kessel A, Bootsma D, Grosveld G, Ferguson-Smith MA, Davies T, Stone M, et al: Translocation of c-ab1 oncogene correlates with the presence of a Philadelphia chromosome in chronic myelocytic leukaemia. Nature. 306:277–280. 1983. View Article : Google Scholar : PubMed/NCBI | |
|
Faderl S, Talpaz M, Estrov Z, O’Brien S, Kurzrock R and Kantarjian HM: The biology of chronic myeloid leukemia. N Engl J Med. 341:164–172. 1999. View Article : Google Scholar : PubMed/NCBI | |
|
Goldman JM and Melo JV: Chronic myeloid leukemia--advances in biology and new approaches to treatment. N Engl J Med. 349:1451–1464. 2003. View Article : Google Scholar : PubMed/NCBI | |
|
Kantarjian HM, Kurzrock R and Talpaz M: Philadelphia chromosome-negative chronic myelogenous leukemia and chronic myelomonocytic leukemia. Hematol Oncol Clin North Am. 4:389–404. 1990.PubMed/NCBI | |
|
Martiat P, Michaux JL and Rodhain J: Philadelphia-negative (Ph-) chronic myeloid leukemia (CML): Comparison with Ph+ CML and chronic myelomonocytic leukemia. The Groupe Français de Cytogénétique Hématologique. Blood. 78:205–211. 1991.PubMed/NCBI | |
|
Santucci MA, Saglio G and Tura S: Pathogenesis and progression of chronic myeloid leukemia. Haematologica. 81:63–76. 1996.PubMed/NCBI | |
|
Goldman JM: Chronic myeloid leukemia: A historical perspective. Semin Hematol. 47:302–311. 2010. View Article : Google Scholar : PubMed/NCBI | |
|
Kantarjian HM, O’Brien S, Smith TL, Rios MB, Cortes J, Beran M, Koller C, Giles FJ, Andreeff M, Kornblau S, et al: Treatment of Philadelphia chromosome-positive early chronic phase chronic myelogenous leukemia with daily doses of interferon alpha and low-dose cytarabine. J Clin Oncol. 17:284–292. 1999.PubMed/NCBI | |
|
Druker BJ, Sawyers CL, Kantarjian H, Resta DJ, Reese SF, Ford JM, Capdeville R and Talpaz M: Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med. 344:1038–1042. 2001. View Article : Google Scholar : PubMed/NCBI | |
|
Gambacorti-Passerini C, le Coutre P, Mologni L, Fanelli M, Bertazzoli C, Marchesi E, Di Nicola M, Biondi A, Corneo GM, Belotti D, et al: Inhibition of the ABL kinase activity blocks the proliferation of BCR/ABL+ leukemic cells and induces apoptosis. Blood Cells Mol Dis. 23:380–394. 1997. View Article : Google Scholar | |
|
Kantarjian HM, Cortes JE, O’Brien S, Giles F, Garcia-Manero G, Faderl S, Thomas D, Jeha S, Rios MB, Letvak L, et al: Imatinib mesylate therapy in newly diagnosed patients with Philadelphia chromosome-positive chronic myelogenous leukemia: High incidence of early complete and major cytogenetic responses. Blood. 101:97–100. 2003. View Article : Google Scholar | |
|
Kantarjian H, Talpaz M, O’Brien S, Garcia-Manero G, Verstovsek S, Giles F, Rios MB, Shan J, Letvak L, Thomas D, et al: High-dose imatinib mesylate therapy in newly diagnosed Philadelphia chromosome-positive chronic phase chronic myeloid leukemia. Blood. 103:2873–2878. 2004. View Article : Google Scholar : PubMed/NCBI | |
|
Hughes TP and Branford S: Monitoring disease response to tyrosine kinase inhibitor therapy in CML. Hematology Am Soc Hematol Educ Program. 2009.477–487. 2009.PubMed/NCBI | |
|
Alaiya AA1, Franzén B, Hagman A, Dysvik B, Roblick UJ, Becker S, Moberger B, Auer G and Linder S: Molecular classification of borderline ovarian tumors using hierarchical cluster analysis of protein expression profiles. Int J Cancer. 98:895–899. 2002. View Article : Google Scholar : PubMed/NCBI | |
|
Alaiya AA, Franzén B, Hagman A, Silfverswärd C, Moberger B, Linder S and Auer G: Classification of human ovarian tumors using multivariate data analysis of polypeptide expression patterns. Int J Cancer. 86:731–736. 2000. View Article : Google Scholar : PubMed/NCBI | |
|
Alaiya A, Fox J, Bobis S, Matic G, Shinwari Z, Barhoush E, Márquez M, Nilsson S and Holmberg AR: Proteomic analysis of soft tissue tumor implants treated with a novel polybisphosphonate. Cancer Genomics Proteomics. 11:39–49. 2014.PubMed/NCBI | |
|
Al-Moghrabi N, Nofel A, Al-Yousef N, Madkhali S, Bin Amer SM, Alaiya A, Shinwari Z, Al-Tweigeri T, Karakas B, Tulbah A, et al: The molecular significance of methylated BRCA1 promoter in white blood cells of cancer-free females. BMC Cancer. 14:8302014. View Article : Google Scholar : PubMed/NCBI | |
|
Li GZ, Vissers JP, Silva JC, Golick D, Gorenstein MV and Geromanos SJ: Database searching and accounting of multiplexed precursor and product ion spectra from the data independent analysis of simple and complex peptide mixtures. Proteomics. 9:1696–1719. 2009. View Article : Google Scholar : PubMed/NCBI | |
|
Nowicki MO, Pawlowski P, Fischer T, Hess G, Pawlowski T and Skorski T: Chronic myelogenous leukemia molecular signature. Oncogene. 22:3952–3963. 2003. View Article : Google Scholar : PubMed/NCBI | |
|
Alaiya A, Al-Mohanna M and Linder S: Clinical cancer proteomics: Promises and pitfalls. J Proteome Res. 4:1213–1222. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Celis JE: Gel-based proteomics: What does MCP expect? Mol Cell Proteomics. 3:9492004.PubMed/NCBI | |
|
Celis JE and Gromov P: High-resolution two-dimensional gel electrophoresis and protein identification using western blotting and ECL detection. EXS. 88:55–67. 2000.PubMed/NCBI | |
|
Alaiya AA, Franzén B, Fujioka K, Moberger B, Schedvins K, Silfversvärd C, Linder S and Auer G: Phenotypic analysis of ovarian carcinoma: polypeptide expression in benign, borderline and malignant tumors. Int J Cancer. 73:678–683. 1997. View Article : Google Scholar : PubMed/NCBI | |
|
Alaiya AA, Franzén B, Moberger B, Silfverswärd C, Linder S and Auer G: Two-dimensional gel analysis of protein expression in ovarian tumors shows a low degree of intratumoral heterogeneity. Electrophoresis. 20:1039–1046. 1999. View Article : Google Scholar : PubMed/NCBI | |
|
Erler BS, Hsu L, Truong HM, Petrovic LM, Kim SS, Huh MH, Ferrell LD, Thung SN, Geller SA and Marchevsky AM: Image analysis and diagnostic classification of hepatocellular carcinoma using neural networks and multivariate discriminant functions. Lab Invest. 71:446–451. 1994.PubMed/NCBI | |
|
Goldschmidt D, Decaestecker C, Berthe JV, Gordower L, Remmelink M, Danguy A, Pasteels JL, Salmon I and Kiss R: The contribution of image cytometry and artificial intelligence-related methods of numerical data analysis for adipose tumor histopathologic classification. Lab Invest. 75:295–306. 1996.PubMed/NCBI | |
|
Ravdin PM, Clark GM, Hilsenbeck SG, Owens MA, Vendely P, Pandian MR and McGuire WL: A demonstration that breast cancer recurrence can be predicted by neural network analysis. Breast Cancer Res Treat. 21:47–53. 1992. View Article : Google Scholar : PubMed/NCBI | |
|
Tewari A and Narayan P: Novel staging tool for localized prostate cancer: A pilot study using genetic adaptive neural networks. J Urol. 160:430–436. 1998. View Article : Google Scholar : PubMed/NCBI | |
|
Albano F, Zagaria A, Anelli L, Coccaro N, Impera L, Minervini CF, Minervini A, Rossi AR, Tota G, Casieri P, et al: Gene expression profiling of chronic myeloid leukemia with variant t(9;22) reveals a different signature from cases with classic translocation. Mol Cancer. 12:362013. View Article : Google Scholar : PubMed/NCBI | |
|
Verhaak RG, Wouters BJ, Erpelinck CA, Abbas S, Beverloo HB, Lugthart S, Löwenberg B, Delwel R and Valk PJ: Prediction of molecular subtypes in acute myeloid leukemia based on gene expression profiling. Haematologica. 94:131–134. 2009. View Article : Google Scholar : | |
|
Alsagaby SA, Khanna S, Hart KW, Pratt G, Fegan C, Pepper C, Brewis IA and Brennan P: Proteomics-based strategies to identify proteins relevant to chronic lymphocytic leukemia. J Proteome Res. 13:5051–5062. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Pizzatti L, Panis C, Lemos G, Rocha M, Cecchini R, Souza GH and Abdelhay E: Label-free MSE proteomic analysis of chronic myeloid leukemia bone marrow plasma: Disclosing new insights from therapy resistance. Proteomics. 12:2618–2631. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Gjertsen BT and Wiig H: Investigation of therapy resistance mechanisms in myeloid leukemia by protein profiling of bone marrow extracellular fluid. Expert Rev Proteomics. 9:595–598. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Linger RM, Keating AK, Earp HS and Graham DK: TAM receptor tyrosine kinases: Biologic functions, signaling, and potential therapeutic targeting in human cancer. Adv Cancer Res. 100:35–83. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Sun WS, Fujimoto J and Tamaya T: Clinical implications of coexpression of growth arrest-specific gene 6 and receptor tyrosine kinases Axl and Sky in human uterine leiomyoma. Mol Hum Reprod. 9:701–707. 2003. View Article : Google Scholar : PubMed/NCBI | |
|
Zhu S, Wurdak H, Wang Y, Galkin A, Tao H, Li J, Lyssiotis CA, Yan F, Tu BP, Miraglia L, et al: A genomic screen identifies TYRO3 as a MITF regulator in melanoma. Proc Natl Acad Sci USA. 106:17025–17030. 2009. View Article : Google Scholar : PubMed/NCBI | |
|
Kerkhoff C, Klempt M and Sorg C: Novel insights into structure and function of MRP8 (S100A8) and MRP14 (S100A9). Biochim Biophys Acta. 1448:200–211. 1998. View Article : Google Scholar | |
|
Thomas CG, Patel RM and Bergfeld WF: Cytophagic and S-100 protein immunoreactive myeloid leukemia cutis. J Cutan Pathol. 37:390–395. 2010. View Article : Google Scholar | |
|
Ng CS, Lam TK, Chan JK, Hui PK, Ng HK, Szeto SC and Feng CS: Juvenile chronic myeloid leukemia. A malignancy of S-100 protein-positive histiocytes. Am J Clin Pathol. 90:575–582. 1988. View Article : Google Scholar : PubMed/NCBI | |
|
Gómez-Casares MT, García-Alegria E, López-Jorge CE, Ferrándiz N, Blanco R, Alvarez S, Vaqué JP, Bretones G, Caraballo JM, Sánchez-Bailón P, et al: MYC antagonizes the differentiation induced by imatinib in chronic myeloid leukemia cells through downregulation of p27KIP1. Oncogene. 32:2239–2246. 2013. View Article : Google Scholar | |
|
Vaqué JP, Navascues J, Shiio Y, Laiho M, Ajenjo N, Mauleon I, Matallanas D, Crespo P and León J: Myc antagonizes Ras-mediated growth arrest in leukemia cells through the inhibition of the Ras-ERK-p21Cip1 pathway. J Biol Chem. 280:1112–1122. 2005. View Article : Google Scholar |
