Clinical Chemistry 54: 673-681, 2008.
First published February 21, 2008; 10.1373/clinchem.2007.091975
(Clinical Chemistry. 2008;54:673-681.)
© 2008 American Association for Clinical Chemistry, Inc.
|
Molecular Diagnostics and Genetics |
Ex Vivo Simulation of the Action of Antileukemia Drugs by Measuring Apoptosis-Related mRNA in Blood
Masato Mitsuhashi1,2,a,
Katsuya Endo2,
Kazuhiko Obara2,
Hiroshi Izutsu2,
Taishi Ishida3,
Norio Chikatsu3 and
Atsushi Shinagawa3
1 Hitachi Chemical Research Center, Inc., Irvine, CA; 2
Hitachi Chemical Co., Ltd., Hitachi, Japan; 3
Hitachi Ltd., Hitachi General Hospital, Hitachi, Japan.
aAddress correspondence to this author at: Hitachi Chemical Research Center, Inc., 1003 Health Sciences Rd., Irvine, CA. Fax 949-725-2727; e-mail mmitsuhashi{at}HCRcenter.com.
 |
Abstract
|
|---|
Background: In conventional bioassays, isolated cells are suspended in culture media, incubated in vitro for several days, and then characterized with respect to any cellular changes. In developing new molecular tests under physiological ex vivo conditions, we quantified the production of mRNAs for p21 and PUMA (p53 up-regulated modulator of apoptosis), which are involved in cell cycle arrest and apoptosis, respectively.
Methods: We stimulated human whole blood with a chemotherapeutic drug (cytarabine, daunorubicin, mitoxantrone, aclarubicin, etoposide, or idarubicin) for 4 h and then quantified mRNA by assessing mRNA recovery and cDNA-synthesis efficiency in each sample. We also used immunoassay and flow cytometry to investigate nucleosome and annexin V, respectively, as apoptosis markers.
Results: Ex vivo mRNA analysis yielded more positive results than nucleosome and annexin V analyses. The concentrations of cytarabine- and daunorubicin-induced p21 and PUMA mRNAs were significantly lower in acute myelogenous leukemia (AML) patients than in healthy controls (P <0.0001), whereas idarubicin induced significantly greater responses in AML patients than in controls (P = 0.01). The patients had different mRNA-response patterns, which were largely classifiable into 4 groups. Prednisone enhanced cytarabine or mitoxantrone induction of p21 and PUMA mRNAs in 3 (2.6%) of 114 reactions. All 15 patients who achieved complete remission had received at least one drug that produced positive mRNA responses, whereas we observed a lack of mRNA response to the clinically used drugs in all 3 cases in which the therapy failed to induce any hematologic improvement.
Conclusion: This study introduced ex vivo mRNA analysis as a candidate platform for drug-sensitivity tests in leukemia.
|
Introduction
|
|---|
Personalized, or individualized, medicine is a new concept for identifying the suitable treatment for each patient(1). This approach is particularly important for cancer patients, because chemotherapeutic agents are known to induce severe side effects. When a diagnosis is made, drug regimens are generally selected on the basis of the type and characteristics of the cancer cells without testing their sensitivity to the drugs in vitro, and the patients undergo a standard treatment that has been validated statistically. Although the initial induction therapy works in many cases, the identification of suitable drugs becomes critically important when disease relapse occurs.
The sequence variation (single-nucleotide polymorphism) in the genomic DNA responsible for drug metabolism has been extensively studied recently, and these data have been used to predict which patients will not respond to particular drugs(2). Such information is not available for all drugs, however, and although it may predict nonresponders with altered drug metabolism, this approach is not immediately applicable for identifying the drugs that would be effective in particular patients. Furthermore, whether a second or third yet-to-be-identified single-nucleotide polymorphism would compromise or accelerate the drug response is not known.
Microarray technologies have also been used to predict drug responses by enabling the characterization of the expression of thousands of genes in cancer cells to find patterns that distinguish drug responders from nonresponders(3). Creating such a profile is a difficult task, however, and the available data are limited. Furthermore, once specific patterns are identified, it is still difficult to understand how the expression of each gene is involved in a particular drug response. The quality control of thousand of spots in microarray chips is yet another challenge.
In the case of leukemia and lymphoma, the cancer cells exist in the peripheral blood. Because many drugs are administered intravenously and their blood concentrations can be quantified, in vitro simulation is feasible by incubating samples of whole blood with each drug under consideration. In the in vitro tests of drug sensitivity that have previously been used to identify appropriate drugs(4), leukemic cells were isolated, suspended in artificial culture media, and incubated with the drug in vitro for 48 h to characterize any cell damage; however, a drug-sensitivity test is not commonly available for leukemia, because of various technical difficulties and the discrepancies that have been observed between laboratory results and clinical outcomes.
We recently found that mRNAs for p21 and PUMA1
(p53 up-regulated modulator of apoptosis) were the most predictive mRNA markers of leukocyte apoptosis(5) and that the induction of these mRNAs was detectable in human whole blood within 1–4 h of drug incubation(5). p21 is an inhibitor of cyclin-dependent kinase(6), and increased p21 production consequently causes cell cycle arrest(7). p53 induces the production of mRNA for PUMA(8), which liberates BAX to promote apoptosis(9). Thus, analysis of the expression of the genes encoding these 2 proteins may reveal an association with the cytostatic and cytocidal effects of drugs, similar to the bacteriostatic and bacteriocidal effects of antibiotics. In this second in a series of investigations on the clinical applications of the mRNA technology we have developed(10)(11), we reinvestigated the concept of drug-sensitivity testing with new molecular tests carried out under physiological ex vivo conditions.
|
Materials and Methods
|
|---|
blood sampling
Heparinized samples of whole blood were obtained from Hitachi, Ltd., Hitachi General Hospital (Hitachi, Japan), and the Apex Research Institute (Tustin, CA) after approval by the respective institutional review boards. The 40 healthy adult volunteers of Japanese descent used in this study consisted of 10 participants each with ages in the twenties, thirties, forties, and fifties; there were equal numbers of men and women. We also obtained blood samples from 30 healthy adults and 18 patients with various blood malignancies [acute myelogenous leukemia (AML), 8 patients; chronic lymphocytic leukemia, 4 patients; acute lymphocytic leukemia, 3 patients; non-Hodgkin lymphoma, 2 patients; and myelodysplasia, 1 patient] in order to optimize assay conditions during preliminary studies (data not shown). We then studied a new set of 18 AML patients (12 men and 6 women, 30–86 years of age; 17 new cases and 1 relapse case), including 4 cases of acute promyelocytic leukemia. Blood was drawn from these 18 AML patients before chemotherapy, stored at 4 °C, and stimulated with the appropriate drugs on the same day. Blood samples were processed for mRNA analysis after each drug-stimulation treatment.
primers and probes
PCR primers and TaqMan probes were designed with Primer Express (Applied Biosystems) and HYBsimulator (RNAture), as previously described(7). Oligonucleotides were synthesized by Integrated DNA Technologies, Tsukuba Oligo Service, Nippon EGT, and Hokkaido System Science.
blood stimulation
We added 1.4 µL of 50x concentrations (see below) of drugs or control solutions [PBS and dimethyl sulfoxide (DMSO)] to 8-well microtube strips and stored them at –20 °C until use. The drugs and their final concentrations (1:50) used in this study were as follows: cytarabine (AraC), 10 and 100 µmol/L; daunorubicin (DNR), 0.2 and 2 µmol/L; mitoxantrone (MIT), 0.02 and 0.2 µmol/L; aclarubicin (ACR), 0.04 and 2 µmol/L; etoposide (VP-16), 10 and 100 µmol/L; and idarubicin (IDR), 0.02 and 2 µmol/L. We added 70 µL of blood sample to each tube, in triplicate, and incubated the tubes at 37 °C for 4 h with the caps closed. In addition, we added blood samples to tubes containing prednisone (PSL), alone (final concentration, 0.2 µmol/L) or with 0.04 µmol/L ACR, 10 µmol/L AraC, 0.02 µmol/L DNR, 0.02 µmol/L IDR, 0.02 µmol/L MIT, or 10 µmol/L VP-16 (i.e., a total of 24 stimulations, including the PBS, DMSO, and PSL controls).
Mrna quantification
Absolute mRNA concentrations (copies/microliter blood) were measured by assessing mRNA recovery and cDNA-synthesis efficiency in each sample, as previously described(10)(11). Each gene was amplified individually. The cycle threshold (Ct), the PCR cycle that generated certain amounts of PCR products (as measured by fluorescence), was evaluated with SDS analytical software (Applied Biosystems). We used 10–106 copies/well of calibrator templates in each PCR for calibration. The Ct values for the external control, RNA34, were converted to copy numbers, and percent recovery was calculated for each well(10)(11). For p21 and PUMA, we used the respective calibration curves to convert Ct values to copy numbers and then converted copy numbers to copies/microliter of blood by dividing by percent RNA34 recovery in each sample, according to the method we reported previously(10). We divided each copy number obtained for triplicate drug-treated samples by the mean copy number of the control samples to calculate the mean (SD) values for the fold increase. The Student t-test was used for statistical analyses.
nucleosome analysis
We incubated blood samples from 10 AML patients overnight with the drugs and then used a sandwich immunoassay (Cell Death Detection ELISAPLUS reagent set; Roche Diagnostics) to quantify the amounts of nucleosome, one of the end products of apoptosis. In brief, we mixed 5-µL samples of whole blood with 200 µL of lysis buffer supplied in the reagent set and incubated them at room temperature for 30 min. After centrifugation at 10 000g for 10 min, we mixed 20 µL of supernatant with 80 µL of immunoreagent (an 18:1:1 mixture of incubation buffer, biotin-labeled mouse monoclonal antibody against human histone, and peroxidase-labeled mouse monoclonal antibody against human DNA; all were supplied in the reagent set manufacturer), and transferred this mixture to streptavidin-coated 96-well microplates. After a 2-h incubation at 37 °C, the microplates were washed 3 times with the Incubation Buffer supplied with the reagent set. We then added 100 µL of substrate solution [2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonate), ABTS] to each well and incubated the microplates at room temperature for approximately 20 min. We then measured the absorbance at 405 nm and 490 nm with an ARVO microplate reader (PerkinElmer). The enrichment factor was calculated by dividing the values for the drug-treated samples with those for the vehicle-treated (PBS or DMSO) control samples.
annexin v analysis
We incubated 70-µL samples of blood from 5 AML patients overnight with the drug and then added 630 µL VersaLyse (Beckman Coulter) to lyse erythrocytes. After centrifugation at 400g for 5 min, we suspended the leukocytes in 490 µL binding buffer (Annexin V-FITC Apoptosis Detection Kit, Biovision) and incubated the leukocyte suspension with 5 µL each of reagent set–supplied, undiluted fluorescein isothiocyanate–labeled annexin V (BioVision) and reagent set–supplied, undiluted propidium iodide (BioVision) at room temperature for 5 min in the dark. We then incubated the suspension on ice for an additional 10 min with 20 µL of phycoerythrin-labeled antihuman CD45 monoclonal antibody (Beckman Coulter). We then analyzed annexin V binding via flow cytometry (EPICS XL-MCL; Beckman Coulter) with CD45 gating [excitation, 385 nm (argon laser); emission, 525 nm (for phycoerythrin)] and side scatter and monitored the emission signals at 488 nm (fluorescein isothiocyanate) and 675 nm (propidium iodide).
|
Results
|
|---|
Mrna assay versus conventional apoptosis assays
We defined a 1.5-fold increase as the cutoff value (Fig. 1
, dotted lines), because the majority of the results stayed below this value in both the nucleosome ELISA and the annexin V fluorescence-activated cell sorting (FACS) analysis. The total numbers of analytical data points (for the various combinations of patients and drugs) for the ELISA and the FACS analysis were 198 and 65, respectively. As shown in Fig. 1
(lower-left quadrant in each graph), the lack of mRNA responses for p21 (Fig. 1
, A and C) or PUMA (Fig. 1
, B and D) (x-axis, <1.5-fold increase) corresponded with the lack of responses in the ELISA (Fig. 1
, A and B) and the FACS analysis (Fig. 1
, C and D) (y-axis, <1.5-fold increase) in 65%–70% of the cases. Seven percent of the p21 data points and 6% of the PUMA data points showed dual positive results in both the mRNA analysis and the ELISA (Fig. 1
, A and B, upper-right quadrant), and the frequencies of such dual positives were 15% for the mRNA and FACS analyses for both p21 and PUMA (Fig. 1
, C and D, upper-right quadrant). Only 2%–3% of the results were positive in the ELISA or the FACS analysis without any mRNA response (Fig. 1
, upper-left quadrant), whereas 15%–26% of the data points consisted of positive mRNA responses (p21 or PUMA) without positive results in the ELISA or the FACS analysis (Fig. 1
, lower-right quadrant).
View larger version (41K):
[in this window]
[in a new window]
|
Figure 1. Comparison of ex vivo mRNA analysis and biological apoptosis assays.
Blood samples from 10 (A, B) or 5 (C, D) AML patients were stimulated with various drugs for 4 h. We then quantified p21 (A, C) and PUMA (B, D) mRNAs and calculated the fold increase in mRNA concentration (x-axis), as described in Materials and Methods. The remainder of the blood samples were further incubated overnight, and ELISA and FACS analysis were used to quantify nucleosome (A, B; y-axis) and annexin V (C, D; y-axis), respectively. Dotted lines represent the cutoff value of a 1.5-fold increase. The numbers of data points (various combinations of patients and drugs) for the ELISA and the FACS analysis were 198 and 65, respectively.
|
|
drug-induced P21 and puma Mrna in aml and controls
The results for 40 healthy adult volunteers and 18 AML patients are summarized in Fig. 2
. AraC and DNR induced significantly lower p21 and PUMA mRNA responses in AML cases (Fig. 2
, D and E) than in control samples (Fig. 2
, A and B) (all P <0.0001), although 17 of the AML patients had never been exposed to chemotherapeutic drugs. The dual-negative population was increased for both p21 and PUMA, from 10% to 56% for AraC (Fig. 2
, A and D, lower-left quadrant) and from 15% to 39% for DNR (Fig. 2
, B and E, lower-left quadrant). Contrasting results were obtained for IDR (Fig. 2
, C and F), and AML samples showed significantly higher responses than controls (P = 0.01 for both p21 and PUMA). The dual-negative population decreased from 13% to 6% (Fig. 2
, C and F, lower-left quadrant). One relapse case (Fig. 2
, D–F, arrow) showed results similar to those of other AML cases.
View larger version (31K):
[in this window]
[in a new window]
|
Figure 2. Induction of p21 and PUMA mRNAs in healthy adult control individuals and AML patients.
Blood samples were stimulated with 10 µmol/L AraC (A, D), 2 µmol/L DNR (B, E), or 2 µmol/L IDR (C, F) for 4 h. We then quantified p21 (x-axis) and PUMA (y-axis) mRNAs and calculated the fold increase in mRNA concentration, as described in Materials and Methods. (A–C), Healthy adult volunteers (n = 40; 10 each with ages in the twenties, thirties, forties, and fifties, with equal numbers of males and females). (D–F), AML patients (n = 18). Dotted lines represent the cutoff of a 2-fold increase.
|
|
As shown in Fig. 3A
(upper-left quadrant), 28% (p21) and 39% (PUMA) of patients had positive responses (>2-fold increase) to DNR and a lack of response to AraC, whereas only 11% (p21) and 0% (PUMA) of patients had positive responses to AraC and a lack of response to DNR (lower-right quadrant). Positive responses to IDR were seen in patients with a lack of response to DNR (Fig. 3B
, lower-right quadrant) and AraC (Fig. 3C
, lower-right quadrant), whereas we found no case of a positive response to DNR or AraC with a lack of response to IDR (Fig. 3
, B and C, upper-left quadrant).
effect of psl
Among the 108 data points (i.e., 6 drugs x 18 patients), PSL did not enhance or inhibit the induction of p21 and PUMA mRNA by any drug, with the exception of the 3 cases shown in Fig. 4
. Although stimulation with PSL alone and with the drug alone produced no change compared with the PBS control values, costimulation with PSL plus MIT (Fig. 4A
) or PSL plus AraC (Fig. 4
, B and C) significantly (all P <0.001 over drug alone) induced p21 mRNA (Fig. 4
, open columns) and PUMA mRNA (Fig. 4B
, closed columns).
View larger version (21K):
[in this window]
[in a new window]
|
Figure 4. Effects of PSL.
Blood samples from 18 AML patients were stimulated for 4 h with 0.04 µmol/L ACR, 10 µmol/L AraC, 0.2 µmol/L DNR, 0.02 µmol/L IDR, 0.02 µmol/L MIT, or 10 µmol/L VP-16 in combination with 0.2 µmol/L PSL or a PBS control (24 stimulations, including the PBS, DMSO, and PSL controls). In drug tests with PSL, PSL was added together with the drug. After each treatment, we quantified p21 ( ) and PUMA ( ) mRNAs and calculated the fold increase in mRNA concentration, as described in Materials and Methods. Data are presented as the mean (SD) of measurements of triplicate aliquots of whole blood.
|
|
individual patterns of drug responses
Individuals differed with respect to their responses to the various drugs, but these responses could be categorized into 4 patterns. As shown in Fig. 5A
, 22% of AML cases showed positive responses to both AraC and DNR and not to ACR. Another 22% showed positive responses to AraC, DNR, and ACR (Fig. 5B
). Two cases (11%) showed positive responses to DNR and ACR and no response to AraC (Fig. 5C
). Three cases (17%) showed no responses to AraC, DNR, and ACR (Fig. 5D
).
View larger version (32K):
[in this window]
[in a new window]
|
Figure 5. Individual patterns of drug responses.
Blood samples from 18 AML patients were stimulated for 4 h with AraC, DNR, MIT, ACR, VP-16, or IDR. We then quantified p21 () and PUMA () mRNAs and calculated the fold increase in mRNA concentration, as described in Materials and Methods. Data are presented as the mean (SD) of measurements of triplicate aliquots of whole blood. Numbers in parentheses indicate final drug concentrations (in µmol/L).
|
|
case studies
Of the 18 AML cases, 13 cases (15 treatments) were eligible for further analysis. The drugs used clinically were all tested by ex vivo mRNA analysis before chemotherapy, and the clinical follow-up was completed. We also added 4 other cases from our preliminary studies, although some of the information (doses and PUMA response) was missing. All 15 patients who achieved complete remission (CR) had positive mRNA responses to at least one of the drugs (positive match) (Fig. 6
, circles). For 3 cases (2A, 10, and 16; Fig. 6
) in which the therapy failed to induce a CR, the drugs used clinically all induced no mRNA response (negative match). In case 10, AraC was excluded from the analysis because it had been used clinically at low doses (Fig. 6
, hatched box). Patient 11A used drugs (AraC, DNR, VP-16) that induced a positive mRNA response, but they failed to induce CR, although hematologic improvements were observed (mismatch). A 
2 test of the 15 positive matches, 3 negative matches, and 1 mismatch was statistically significant (P <0.0003). Even when the results of the 4 preliminary cases (14–17; Fig. 6
) were excluded, the data still showed statistical significance (P = 0.002).
View larger version (42K):
[in this window]
[in a new window]
|
Figure 6. Case studies of ex vivo mRNA analysis.
Seventeen cases (19 treatments) were eligible for the analysis. The drugs used clinically are boxed, with dashed-line boxes indicating low-dose treatments. Doses (second column) are in micromoles per liter. +++, ++, +, and – indicate percent change >300%, 200%–300%, 150%–200% (all with P <0.05), and <150%, respectively. In the clinical responses row, CR indicates complete remission; +, hematologic improvement; and –, no improvement. In the patient’s profile section, WBC indicates leukocyte count (x103/µL blood); APL, acute promyelocytic leukemia; t-AML, therapy-related AML; 1st, initial case; and R, relapse.
|
|
In case 2, AraC and DNR failed to induce CR, and these 2 drugs did not induce an mRNA response (Fig. 6
, patient 2A). Subsequently, the second treatment regimen consisting of AraC and VP-16 induced CR, and VP-16 induced positive responses for both p21 and PUMA mRNAs (Fig. 6
, patient 2B). In case 11B, therapy with ACR and low-dose AraC (dashed-line box) successfully induced CR. Patients 7 and 8 (Fig. 6
) also showed weak p21 mRNA responses to ACR, but these ACR responses corresponded to CRs. Because high doses of AraC and DNR failed to induce p21 and PUMA mRNA responses in case 2, we also analyzed other BAX family mRNAs, as is described in our previous report(5). This study demonstrated that high doses of AraC, DNR, and MIT did not induce apoptosis-related mRNAs (GADD, SUMO, Apaf1, Bfl1, BclII, Bim, Bik, Bid, Bad, Bcl-xs, Bak, and Bax) (data not shown).
|
Discussion
|
|---|
p21 and PUMA mRNAs are markers of leukocytes’ cytostatic and cytocidal functions, respectively. Although the genes encoding these mRNAs are well characterized at the molecular level, this report is the first to describe the use of both p21 and PUMA mRNAs in drug-sensitivity tests for leukemia.
The ex vivo mRNA analyses yielded more positive results than the biological assays (nucleosome ELISA and annexin V FACS) (Fig. 1
, lower-right quadrants). This finding indicates that the mRNA assay may have better analytical sensitivity than these 2 conventional assays, or that it may produce more false-positive reactions. Although biological assays detect actual apoptosis, a finding of apoptosis is not always the gold standard, given that apoptosis may happen artificially during a lengthy in vitro incubation. If antiapoptotic cascades [such as Bcl-2, and so forth(12)] are more dominant than proapoptotic ones or if drug resistance–related proteins such as P-glycoprotein(13) are produced in substantial quantities, PUMA induction may not be sufficient to induce apoptosis. Alternatively, post-mRNA cascades may be interrupted artificially or physiologically in whole blood. In contrast, the observation of a lack of mRNA responses along with positive results in biological assays (Fig. 1
, upper-left quadrants) may be due to direct activation of mitochondria or caspases, or to secondary apoptosis occurring during an overnight incubation. Although we found PUMA mRNA to be the most sensitive apoptosis marker in leukocytes(5) and leukemic cells, other Bax family genes or yet-to-be-identified proapoptotic mRNAs may be induced(14). Thus, the results of ex vivo mRNA analyses cannot be ignored because of discrepancy with the results of biological assays.
AraC and DNR induced significantly lower p21 and PUMA mRNA responses in leukemia patients than in healthy controls (P <0.0001; Fig. 2
), although patient blood samples were drawn before chemotherapy. Furthermore, these patients (except patient 1 in Fig. 6
) were new cases and had not received any chemotherapy. The results are not technical artifacts, because the responses to IDR were in the opposite direction (Fig. 2
). The interpretation of the results was straightforward in cases 2 and 6 (Fig. 6
), because the proportion of leukemic blasts in whole blood was >90% (Fig. 6
). Although these 2 patients had no response to AraC, the different responses to DNR and IDR suggest variation among the leukemic cells. The patient in case 5 (Fig. 6
) had no response to DNR and a very weak response to AraC, whereas the proportion of leukemic cells was only 5% (Fig. 6
). This finding is not surprising, because such weak responders or nonresponders have been found in healthy adult populations(7). Moreover, nonresponders to one drug were not always resistant to other drugs (Figs. 3
, 5
, and 6
). This result is reasonable because in vitro resistance to AraC is generally explained by the inactivation of deoxycytidine kinase, whereas resistance to DNR has been explained as an overproduction of multidrug efflux pumps, such as P-glycoprotein or multidrug-resistance protein(13). Although both DNR and IDR are anthracycline antibiotics, these 2 drugs are not identical in terms of their drug resistance, and verapamil reverses cytotoxicity due to DNR, but not that due to IDR(15). Identification of such discrepancies on an individual basis will be a powerful clinical research tool for the analysis of drug sensitivity and resistance in humans.
In clinical use, many drugs are used in combination (Fig. 6
). It will be interesting to use such combinations with our method, because the effects of different drugs on these markers may be additive or synergistic. Because our assay system consumes only 50–70 µL of blood per reaction, such analyses are technically feasible. In fact, we analyzed the effects of PSL as the first step (Fig. 4
). Although the majority of the data showed no effect of PSL on drug-induced p21 and PUMA mRNA responses, 3 (2.6%) of 114 data points showed significant enhancement (Fig. 4
). This result indicates that drug responses are not uniform and that interindividual variation exists.
High and low doses were used to identify nonresponders and responders, respectively (Fig. 6
). We considered any drug that produced a positive response at low doses and no response at high doses to be possibly toxic; however, we found no such cases. When both high and low doses produced no response, we interpreted the patient to be resistant to the drug, whereas we interpreted double positives as indicating sensitivity to the drug. Interpretation was not straightforward for the cases that showed a positive response at high doses and no response at low doses, however, because we did not know whether the low doses we used (Fig. 6
) represented therapeutic doses for each of the patients. Additionally, we did not know the blood concentration of each drug for each patient. In the future, we recommend that appropriate doses be determined empirically through large clinical studies.
Although the number of patients in this study was small, we have obtained valuable information. All of the patients in CR cases received at least one drug that produced a positive mRNA response, and we observed no mRNA responses for clinically used drugs in all cases where such therapy failed to induce any hematologic improvement. This finding indicates that the ex vivo assay may have potential as a drug-sensitivity test. For cases 2, 3, 6, and 9 (Fig. 6
), ACR may be a candidate drug because the positive p21 responses were similar to those of cases 7, 8, and 11B, although other drugs successfully induced CR. Although the patient in case 5 achieved CR, DNR may be excluded from the therapy because of the weak mRNA responses. Removal of an ineffective drug may reduce the side effects that occur during the chemotherapy and may also decrease costs.
Clinical demand exists for drug-sensitivity tests that can identify the drugs most suitable for each patient. The development of such tests is a large challenge, however. Although the numerous previous attempts have used a wide variety of technologies, none has been accepted as a routine clinical test. This study was a preliminary one, and the number of patients was very small. Moreover, institutional review board restrictions required the study to be a retrospective analysis, not a prospective one; however, this study has successfully introduced the concept of ex vivo mRNA analysis as a candidate platform for drug-sensitivity testing, and this approach has shown sufficient technical and clinical promise for subsequent larger clinical studies.
|
Acknowledgments
|
|---|
Grant/Funding Support: This study was financially supported by Hitachi Chemical Research Center, Hitachi Chemical Co., Ltd., and Hitachi Ltd., Hitachi General Hospital.
Financial Disclosures: Authors are employees of Hitachi Chemical Research Center (M.M.), Hitachi Chemical Co., Ltd. (K.E., K.O., H.I.), and Hitachi Ltd., Hitachi General Hospital (T.I., N.C., A.S.).
Acknowledgments: We thank Y. Oka, S. Nemoto, N. Sasaki, and the staffs of the clinical laboratories of Hitachi General Hospital; M.J. Tonkon, C. Fox, C. McGinty, and J. Haire (Apex Research Institute); T. Otake, H. Ito, and B. Maekawa (Hitachi Chemical); and Y. Hasegawa and H. Kojima (Division of Hematology, Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Japan) for their support in the blood drawing and clinical section of this study. Results for part of the study were presented at the annual meeting of the Japanese Society of Hematology in 2005 in Yokohama, Japan.
|
Footnotes
|
|---|
1 Nonstandard abbreviations: PUMA, p53 up-regulated modulator of apoptosis; AML, acute myelogenous leukemia; PBS, phosphate-buffered saline; DMSO, dimethyl sulfoxide; AraC, cytarabine; DNR, daunorubicin; MIT, mitoxantrone; ACR, aclarubicin; VP-16, etoposide; IDR, idarubicin; PSL, prednisone; Ct, cycle threshold; FACS, fluorescence-activated cell sorting; CR, complete remission. 
|
References
|
|---|
- Jain KK. Personalized medicine. Curr Opin Mol Ther 2002;4:548-558.[Web of Science][Medline]
[Order article via Infotrieve]
- Di Paolo A, Danesi R, Del Tacca M. Pharmacogenetics of neoplastic diseases: new trends. Pharmacol Res 2004;49:331-342.[CrossRef][Web of Science][Medline]
[Order article via Infotrieve]
- Chiaretti S, Li X, Gentleman R, Vitale A, Vignetti M, Mandelli F, et al. Gene expression profile of adult T-cell acute lymphocytic leukemia identifies distinct subsets of patients with different response to therapy and survival. Blood 2004;103:2771-2778.[Abstract/Free Full Text]
- Zittoun R, Bouchard M, Facquet-Danis J, Percie-du-Sert M, Bousser J. Prediction of the response to chemotherapy in acute leukemia. Cancer 1975;35:507-513.[Medline]
[Order article via Infotrieve]
- Mitsuhashi M, Endo K, Obara K, Izutsu H, Ishida T, Chikatsu N, Shinagawa A. Quantification of drug-induced mRNA in human whole blood ex vivo. Clin Med: Blood Disorders 2008;1:1-11.
- Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ. The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases. Cell 1993;75:805-816.[CrossRef][Web of Science][Medline]
[Order article via Infotrieve]
- Halevy O, Novitch BG, Spicer DB, Skapek SX, Rhee J, Hannon GJ, et al. Correlation of terminal cell cycle arrest of skeletal muscle with induction of p21 by MyoD. Science 1995;267:1018-1021.[Abstract/Free Full Text]
- Villunger A, Michalak EM, Coultas L, Müllauer F, Böck G, Ausserlechner MJ, et al. p53- and drug-induced apoptotic responses mediated by BH3-only proteins puma and noxa. Science 2003;302:1036-1038.[Abstract/Free Full Text]
- Vousden KH. Apoptosis. p53 and PUMA: a deadly duo. Science 2005;309:1685-1686.[Abstract/Free Full Text]
- Mitsuhashi M, Tomozawa S, Endo K, Shinagawa A. Quantification of mRNA in whole blood by assessing recovery of RNA and efficiency of cDNA synthesis. Clin Chem 2006;52:634-642.[Abstract/Free Full Text]
- Mitsuhashi M. Ex vivo simulation of drug action: quantification of drug-induced mRNA as a bridge between preclinical and clinical trials. Clin Chem 2007;53:148-149.[Free Full Text]
- Tallman MS, Gilliland DG, Rowe JM. Drug therapy for acute myeloid leukemia. Blood 2005;106:1154-1163.[Abstract/Free Full Text]
- Veuger MJ, Honders MW, Spoelder HE, Willemze R, Barge RM. Inactivation of deoxycytidine kinase and overexpression of P-glycoprotein in AraC and daunorubicin double resistant leukemic cell lines. Leuk Res 2003;27:445-453.[CrossRef][Web of Science][Medline]
[Order article via Infotrieve]
- Del Gaizo Moore V, Brown JR, Certo M, Love TM, Novina CD, Letai A. Chronic lymphocytic leukemia requires BCL2 to sequester prodeath BIM, explaining sensitivity to BCL2 antagonist ABT-737. J Clin Invest 2007;117:112-121.[CrossRef][Medline]
[Order article via Infotrieve]
- Lotfi K, Zackrisson AL, Peterson C. Comparison of idarubicin and daunorubicin regarding intracellular uptake, induction of apoptosis, and resistance. Cancer Lett 2002;178:141-149.[CrossRef][Web of Science][Medline]
[Order article via Infotrieve]
Top
Abstract
Introduction
, p21 mRNA data; 
, PUMA mRNA data.