Purpose: This study aims to investigate whether the plasma level of glutathione S-transferase P1-1 (GSTP1-1), which is a phase II detoxifying enzyme known to be a resistance factor for anticancer drugs, could be a prognostic factor of de novo non-Hodgkin lymphoma (NHL) in clinical stages (CSs) III and IV.

Experimental Design: Study population consisted of 80 NHL patients with no prior treatment: 12 patients were at CS I, 14 at CS II, 25 at CS III, and 29 at CS IV. All 54 patients at CS III or CS IV were treated with cyclophosphamide, doxorubicin, vincristine, and prednisolone (CHOP). Plasma GSTP1-1 concentration was measured by ELISA. We stained lymph node tissues for GSTP1-1 using anti-GSTP1-1 monoclonal antibody 5F and quantitatively assessed the intensity of immunostaining by using the KS-400 image analyzing system.

Results: There was a significant stepwise increment of plasma GSTP1-1 concentration from CS I to CS IV (P < 0.05). Of the 54 patients with CS III or IV treated with CHOP, 28 (52%) had elevated plasma GSTP1-1 levels. Plasma GSTP1-1 concentration tended to correlate with the intensity of GSTP1-1 expression in lymphoma tissues as assessed by immunostaining (P = 0.07). The CR rates in patients at CS III and CS IV treated by CHOP, 55.2% (14 of 26) and 16.0% (5 of 28) for the low and high plasma GSTP1-1 groups, respectively, were significantly different (P < 0.01). For these two groups, the median survival times were 64 and 25 months, respectively (P < 0.01), and the median times to progression were 58 and 12 months, respectively (P < 0.01). There was no significant correlation between plasma GSTP1-1 concentrations and other NHL prognostic indicators in these patients as determined by univariate and multivariate analyses.

Conclusion: These results showed that plasma GSTP1-1 is a useful prognostic factor for CS III and IV advanced NHL. Thus, it may be a promising strategy to treat NHL concomitantly with anticancer drugs and GSTP1-1-specific inhibitors.

Although therapy for advanced non-Hodgkin’s lymphoma (NHL) has greatly improved over the last two decades, prognosis for approximately half of the patients remains poor. Factors defining prognosis in NHL are those related to tumor burden [clinical stage (CS), lactate dehydrogenase (LDH), extranodal involvement, etc.], status of patients (performance status, age, etc.), and tumor cell properties [histologic aggressiveness, expression of genes relevant to cell survival (p53, bcl-2), drug-resistant factors, etc; refs. 1, 2, 3, 4, 5]. As chemotherapy is inevitable for treatment of NHL at advanced stages such as III and IV, drug resistance factors may be one of the most viable prognosis defining factors. However, the implication of drug resistance factors for prognosis of NHL has not been carefully elucidated.

Drug resistance is generally mediated by metabolic pathways, including phase I and phase II enzymes and the phase III efflux pump system. Phase I enzymes such as cytochrome p450s exert their function through oxidation, reduction, and hydrolysis of xenobiotics. Of the cytochrome p450 isozymes, CYP1B1 is known to be expressed in tumor cells (6), but it is reported to inactivate only one anticancer drug, docetaxel, nonrelevant to NHL treatment (7). The phase III efflux pump, such as P-glycoprotein (multidrug-resistant factor), multidrug resistance-related protein and glutathione S-conjugate pump (GS-X), which actively pump out xenobiotics from the cells, have been shown to be involved in the inactivation of a wide variety of anticancer drugs, including adriamycin, vincristine, and VP-16 (8, 9, 10). However, multidrug-resistant factor is reportedly irrelevant to the prognosis of NHL (11, 12), possibly because it is mainly implicated in acquired drug resistance (13, 14). Multidrug resistance-related protein is also reportedly proven not to be correlative to the prognosis of NHL (15). Phase II enzymes such as aldehyde dehydrogenase-3, UDP-glucuronyltransferase, and glutathione S-transferase, which conjugate xenobiotics to inactivate them, are also known to be resistant factors to certain anticancer drugs. Aldehyde dehydrogenase-3 catalyzes the detoxification of cyclophosphamide, oxazaphosphorine, ifosfamide, etc. (16), and UDP-glucuronyltransferase conjugates mitoxantrone but not adriamycin or vincristine, which are routinely used for treating NHL (17). Glutathione S-transferase, on the other hand, has been shown to function as a multidrug-resistant factor for melphalan and chlorambucil by glutathione conjugation, and for adriamycin, cyclophosphamide and CDDP by a conjugation-unrelated mechanism (18, 19, 20, 21, 22, 23, 24). This enzyme is often increased in malignant tissues, irrespective of prior chemotherapy, and is suggested to serve as a multidrug-resistant factor in both inherent and acquired fashions (19, 25, 26). In fact, there are recent reports that disclosed a correlation between glutathione S-transferase P1-1 (GSTP1-1) expression in tumor tissues and the prognosis in ovarian cancer as well as in head and neck cancer (27, 28).

There have been a few reports about GSP1-1 expression with special reference to their chemotherapy responsiveness in NHL. Berhame et al.(29), applying immunohistochemistry, found no significant correlation between chemosensitivity and GSTP1-1 expression, but they dealt with only nine cases. Rodrigeuz et al.(30) obtained similar results using dot blot analysis, but 16 of the 41 cases they studied had anamnesis of prior chemotherapy and therefore are not suitable for analysis of prognostic factors. Ribrag et al.(31) studied GSTP1-1 expression by immunohistochemistry in 69 diffuse large B-cell lymphoma patients and found a significant correlation between the staining intensity and complete remission rate but not between the intensity and overall survival. They included 41 cases at stages I and II, which may not be suitable for studying prognostic factors because curability and prognosis of patients at these early stages may not merely reflect sensitivity of tumor cells to chemotherapy.

We have previously established an ELISA for plasma GSTP1-1 (32, 33). In the present study, therefore, we investigated whether or not the plasma GSTP1-1, which reflects the sum of GSTP1-1 expression in total multiple lesions, could be a predictor of the response to cyclophosphamide, adriamycin, vincristine, and prednisolone (CHOP) treatment and/or the prognosis of untreated NHL (stage III or IV).

Subjects.

This study was approved by the ethics committee of Sapporo Medical University. Eighty patients who were histopathologically diagnosed as having NHL at the 4th Department of Internal Medicine of Sapporo Medical University Hospital and affiliated hospitals during the period of January 1994 to March 1999 were enrolled after strict informed consent was provided by each patient. The study group consisted of 43 males and 37 females, with a mean age of 60 years (27–79). All lymphoma tissues were histologically evaluated by two independent reviewers (M. S. and T. I.) in the Department of Clinical Pathology of Sapporo Medical University. Forty-two patients were diagnosed as having diffuse large B-cell lymphoma, 30 with low-grade B-cell lymphoma, 3 with marginal-zone B-cell lymphoma, and 2 with lymphoblastic lymphoma in accordance with the World Health Organization classification (34). Lymphomas were also classified as indolent type (30 low-grade B-cell lymphomas and 3 marginal-zone B-cell lymphomas), aggressive type (42 diffuse large B-cell lymphomas) or very aggressive type (2 lymphoblastic lymphomas) according to the criteria of the International Lymphoma Study Group (35). The distribution of clinical stages was 12 patients in CS I, 14 in CS II, 25 in CS III, and 29 in CS IV.

Blood samples were kindly provided by forty healthy volunteers with strict informed consent.

Chemotherapy Protocol.

All of the patients at CS III or CS IV were treated with CHOP (750 mg of cyclophosphamide per square meter of body-surface area, 50 mg adriamycin per square meter, and 1.4 mg of vincristine per square meter (up to a maximal dose of 2 mg) on day 1 as well as 40 mg of prednisolone per square meter per day for 5 days; ref. 36) and were treated every 3 weeks for 8 cycles. Treatment was stopped if lymphoma progressed, if the patients declined to continue or at the discretion of the investigator in cases of either intercurrent illness, or where adverse effects were displayed.

Tumor responses were assessed after 8 cycles of chemotherapy or at the end of treatment and were judged as complete response, unconfirmed complete response, partial response, stable disease, or progressive disease according to the response criteria of the International Workshop (37).

Plasma GSTP1-1 Measurement by ELISA.

Before treatment, blood samples were gently drawn from the cubital vein of the patients by syringes that had been wetted with EDTA-2Na and transferred to plastic tubes containing EDTA2Na. They were then immediately centrifuged at 1,200 × g at 4°C for 15 minutes. Plasma samples were separated with a Pasteur pipette, taking care to avoid collecting platelets or blood cells, and filtered through millipore membrane (0.45 μm) to avoid contamination of blasts. Samples were stored at −20°C, and after thawing to room temperature, we subjected them to a sandwich ELISA using monoclonal and polyclonal antibodies established in our laboratory (32, 33).

Immunohistochemical Analysis of GSTP1-1.

Lymph node tissues biopsied from patients before treatment were fixed in formalin and embedded in paraffin. Then we cut the tissues into sections 3-μm thick and stained for GSTP1-1 using the mouse monoclonal antibody 5F (32) or for controls using normal mouse IgG (DAKO, Glostrup, Denmark) for the first antibodies and biotinylated horse antimouse immunoglobulin G (Vector Laboratories, Burlingame, CA) for the second. GSTP1-1 protein was visualized with a Vectastain ABC kit (Vector Laboratories) using 3–3′-diaminobenzidine (Sigma Chemical Co. St. Louis, MO) as a chromogen. Hematoxylin was not used because it would have disturbed the assessment of immunostaining intensity by the imaging analyzer. We also did the immunostaining for CD20 in the serial sections to examine the percentage of lymphoblasts. We carried our image analysis of staining for GSTP1-1 by using an AxioCam photomicroscope and the KS-400 image analyzing system (Carl Zeiss Vision GmbH, Hallbermoss, Germany) in accordance with the method of Kato et al.(38) with minor modifications. In brief, five randomly selected microscopic images 50 × 50 μm2 each, in which the percentages of lymphoblasts were more than 90%, were selected. They were imported into the KS400; brown-stained areas, which represented positively stained GSTP1-1 in each lymphoma cell, were converted into a 255-grade gray scale. The sum of gray scale intensities in each 50 × 50 μm2 square was calculated by using the KS-400 image analyzing program, and the relative intensity was represented by the ratio of the intensity to that of the serial section immunostained by normal mouse IgG. The mean of the five relative intensities in each section was calculated.

Statistical Analysis.

Correlations between the plasma GSTP1-1 concentration and the clinical stage or the histologic classification were assessed by the Kruskal-Wallis test. The correlation between plasma GSTP1-1 concentration and intensity of GSTP1-1 expression was assessed by the Pearson’s test. The correlation between the CR rates and relative intensity of GSTP1-1 expression was assessed by the χ2 test. The correlations between the CR rates and plasma GSTP1-1 concentration or factors from the International Prognostic Index (1) were assessed by Fisher’s exact probability test. Correlations between the plasma GSTP1-1 concentration and progression-free or overall survival curves were constructed with the Kaplan-Meier method. Multivariate analysis was done with the COX proportional hazards model.

Plasma GSTP1-1 Concentration in NHL.

The mean plasma GSTP1-1 concentration in 40 healthy subjects was 13.8 ± 4.4 ng/mL. The cutoff value was set at 22.6 ng/mL, which was the mean + 2SD (Fig. 1,A). There was a significant stepwise increment of plasma GSTP1-1 concentration from CS I to CS IV (Kruskal-Wallis test, P < 0.05), although no significant difference was observed between those of CS III and CS IV. Of the 54 patients with CS III or IV, 28 (52%) had elevated plasma GSTP1-1 levels. There was no statistical difference between any indolent, aggressive, or very aggressive lymphoma (P = 0.64, Fig. 1 B). Because CS I or CS II NHL is not usually treated with chemotherapy and therefore not relevant for GSTP1-1 study, further investigation was done only on CS III and CS IV NHL.

Immunohistochemistry of GSTP1-1 and Its Quantitative Analysis.

The expression of GSTP1-1 in lymphoma tissue was investigated by immunohistochemical staining in 26 patients with CS III and CS IV. In Fig. 2, three representative cases (case 1, A-C; case 2, D-F; case 3, G-I) with different staining intensities are shown, although all of the histologic types were the same; diffuse large B-cell lymphoma was revealed in A, D, and G. In Fig. 2,B, the lymphoma cells were strongly stained for GSTP1-1 throughout the lesion. The figure with higher magnification revealed that GSTP1-1 is located mainly in cytoplasm and, although weakly, in nuclei (Fig. 2,C). Figure 2,E shows lymphoma tissue in which lymphoma cells were weakly stained for GSTP1-1 throughout the lesions. In contrast, the lymphoma cells in Fig. 2,H were almost negative for GSTP1-1. Figures 2 C, F, and I show higher magnification (50 × 50 μm2) of 2B, 2E, and 2H, respectively. With five of such higher magnifications, randomly selected, the mean GSTP1-1 intensity of lymphoma cells was calculated as described in “Materials and Methods.”

Plasma GSTP1-1 Concentrations and Immunohistochemical GSTP1-1 Expression Levels in Biopsied Lymphoma Tissue.

The relationship between plasma GSTP1-1 concentrations and tissue expression levels of GSTP1-1 was studied (Fig. 3). Although the plasma GSTP1-1 concentrations tended to correlate with relative intensities of immunohistochemical findings in biopsied lymphoma tissue, the correlation was not statistically significant (r = 0.45, P = 0.07).

Immunohistochemical GSTP1-1 Expression and the Prognosis of NHL.

We first investigated the relationship between the relative intensity of GSTP1-1 tissue expression and the prognosis of NHL. We defined the high intensity group as those with relative intensity higher than the median value (10.7) and the low intensity group as those with relative intensity <10.7. The average CR rates were 46.2% (6 of 13) among patients in the low intensity group and 38.5% (5 of 13) among those in the high intensity group, there being no statistically significant difference between them (P = 0.72, Table 1). The median survival time was 44 months for the low intensity group (range, 12–94 months) and 37 months for the high intensity group (range, 6–96 months), the difference being not significant (P = 0.31, Fig. 4).

Plasma GSTP1-1 Concentration and the Prognosis of NHL.

We next investigated the relationship between the plasma GSTP1-1 level and the responsiveness to chemotherapy. The average CR rates were 55.2% (14 of 26) among patients with low GSTP1-1 and 16.0% (5 of 28) among those with high GSTP1-1, the difference being statistically significant (P = 0.01; Table 2). The median overall survival time was 64 months (range, 9–102 months) and 25 months (range, 6–93 months) among patients with low and high plasma GSTP1-1 concentration, respectively (P < 0.01, Fig. 5). Figure 6 shows the influence of plasma GSTP1-1 concentration on the time to progression, with a median time of 58 months (range, 0–100 months) in patients with low GSTP1-1 levels compared with 12 months (range, 0–90 months) in those with high GSTP1-1 levels (P < 0.01).

Plasma GSTP1-1 Concentrations and Prognostic Factors.

The components of the international prognostic index for NHL include age, serum LDH, performance status, clinical stage, and extranodal involvement. In this study, the GSTP1-1 levels were compared with NHL prognostic indicators, excluding clinical stage. Univariate analysis showed no significant correlation between plasma GSTP1-1 concentration and any of the NHL prognostic indicators, demonstrating an independency of GSTP1-1 from the known prognostic indicators (Table 3). Multivariate analysis also showed a significant correlation between the plasma GSTP1-1 concentration and the overall survival or progression free survival (Table 4).

In the present study, we successfully showed that plasma GSTP1-1 is a useful prognostic factor for NHL at stages III and IV. Specifically, we showed that overall and disease-free survival rates as well as complete response rates of patients with plasma GSTP1-1 above normal levels were lower than those within normal levels. These findings are compatible with the fact that GSTP1-1 is a resistant factor against two major drugs, cyclophosphamide and adriamycin (21, 22, 23, 24), used in the standard regimen (CHOP) for NHL.

GSTP1-1 is known to conjugate melphalan and chlorambucil with glutathione and export them through the GSX pump. GSTP1-1 has also been shown to endow resistance to 4-HC, the active form of cyclophosphamide, and adriamycin through a conjugation-unrelated mechanism (18, 19, 39). Although the mechanism of the conjugation-unrelated resistance has not yet been elucidated in detail, it has been proposed recently that GSTP1-1 endows resistance to apoptosis through inhibition of JUN kinase pathway (18). It is also another plausible mechanism that GSTP1-1 sequestrates these drugs in cytosol by direct binding (40).

Furthermore, we evidenced an independency of plasma GSTP1-1 from other previously accepted prognostic factors, which include age, serum LDH, performance status, clinical stage, and extranodal involvement; this may be reasonably accepted because none of these factors are related to drug resistance whereasGSTP1-1 is.

An obvious advantage of using GSTP1-1 in circulation rather than in tissue as a prognostic marker is, needless to say, the objectiveness in quantitative analysis. Even more importantly, GSTP1-1 expression in a biopsied lymphoma tissue does not necessarily represent that in multiple lesions because malignant tissues are generally heterogeneous (41, 42), whereas plasma GSTP1-1 level is believed to reflect the sum of GSTP1-1 expression in each lymphoma tissue. Should an advanced patient whose biopsy specimen is negative for GSTP1-1 undergo CHOP therapy, the prognosis may be poor because other lymphoma tissues express high levels of GSTP1-1 and are resistant to CHOP. In fact, in the present study, expression of GSTP1-1 in lymphoma tissue, which did not correlate with plasma GSTP1-1 levels, was not a good indicator of chemotherapy responsiveness or prognosis, in agreement with previous reports (29, 30, 31).

The postulation that plasma GSTP1-1 reflects the sum of GSTP1-1 expression may be supported by the fact that GSTP1-1 concentrations in culture media of various tumor cell lines well correlate with those of their respective cell lysates (data not shown). With regard to the mechanism of GSTP1-1 being released from cells, it may not be simple leakage because GSTP1-1 in the medium exists in monomer form unlike the intracellular form, which is a homodimer (43).

Analyses of the gene expression profile of diffuse large B-cell lymphoma by using DNA microarray have been reported recently by three independent groups (44, 45, 46). Arizadeh et al.(44) and Rosenwald et al.(45) used a common customized cDNA microarray “lymphochip” that did not include the cDNA of the GSTP1-1 gene. Shipp et al.(46) used a customized oligonucleotide microarray “HU6800,” which could detect GSTP1-1 gene expression, although their gene expression profile, which determined the prognosis of NHL, did not include that of the GSTP1-1 gene. However, the stage distribution of their cases was uncertain. More importantly, their investigation was carried out with a biopsied specimen, which does not necessarily represent scattered multiple lesions of NHL at advanced stages. It has been also suggested that the microarray analysis they used was less sensitive than conventional methods such as real-time PCR and Northern blotting. For example, Rajeevan et al.(47) reported that differences <4-fold detected by the microarray method were not always reproducible by real-time PCR.

Incidentally, a report by Schisselbauer et al.(48) indicated that glutathione S-transferase in blasts of chronic lymphatic leukemia was relevant to chlorambucil resistance. This apparently contradicts our present claim that the concentration of glutathione S-transferase in tumor cells may not represent that of total tumor mass in the body because of tumor heterogeneity. However, in chronic lymphatic leukemia, needless to say, blasts are circulating and, unlike lymphoma, heterogeneous sampling could be avoided. Therefore, their results were compatible with ours analyzed by plasma GSTP1-1 measurement in NHL.

Nonetheless, the plasma GSTP1-1 concentration is thus considered to be a useful prognostic factor for NHL at advanced stages.

Fig. 1.

Plasma GSTP1-1 concentrations in patients with NHL of various clinical stages (A) and histologic types (B). The cutoff value was set at 22.6 ng/mL (mean + 2SD). There was a significant stepwise increment from CS I to CS IV (Kruskal-Wallis test, P < 0.05). There was no significant difference among the histologic types.

Fig. 1.

Plasma GSTP1-1 concentrations in patients with NHL of various clinical stages (A) and histologic types (B). The cutoff value was set at 22.6 ng/mL (mean + 2SD). There was a significant stepwise increment from CS I to CS IV (Kruskal-Wallis test, P < 0.05). There was no significant difference among the histologic types.

Close modal
Fig. 2.

Immunohistochemical staining for GSTP1-1 in lymphoma tissues. A, H&E of lymphoma tissue. B, representative lymphoma tissue strongly stained for GSTP1-1. C, 50 × 50 μm2 field from the panel B. D, H&E of lymphoma tissue. E, representative lymphoma tissue weekly stained for GSTP1-1. F, 50 × 50 μm2 field from the panel E. G, H&E of lymphoma tissue. H, representative lymphoma tissue negative for GSTP1-1. I, 50 × 50 μm2 field from panel G.

Fig. 2.

Immunohistochemical staining for GSTP1-1 in lymphoma tissues. A, H&E of lymphoma tissue. B, representative lymphoma tissue strongly stained for GSTP1-1. C, 50 × 50 μm2 field from the panel B. D, H&E of lymphoma tissue. E, representative lymphoma tissue weekly stained for GSTP1-1. F, 50 × 50 μm2 field from the panel E. G, H&E of lymphoma tissue. H, representative lymphoma tissue negative for GSTP1-1. I, 50 × 50 μm2 field from panel G.

Close modal
Fig. 3.

Relationship between plasma GSTP1-1 concentration and immunohistochemical GSTP1-1 expression level in biopsied lymphoma tissues. The plasma GSTP1-1 concentration tended to correlate with relative intensity of immunohistochemical staining; the correlation was not statistically significant (r = 0.45, P = 0.07).

Fig. 3.

Relationship between plasma GSTP1-1 concentration and immunohistochemical GSTP1-1 expression level in biopsied lymphoma tissues. The plasma GSTP1-1 concentration tended to correlate with relative intensity of immunohistochemical staining; the correlation was not statistically significant (r = 0.45, P = 0.07).

Close modal
Fig. 4.

Kaplan-Meier analysis of overall survival in high and low intensity groups of GSTP1-1 expression in NHL. Relative intensity of immunohistochemical staining in biopsied lymphoma tissue of each case was calculated as described in “Materials and Methods.” There was no significant difference in overall survival between high and low intensity groups (P = 0.31).

Fig. 4.

Kaplan-Meier analysis of overall survival in high and low intensity groups of GSTP1-1 expression in NHL. Relative intensity of immunohistochemical staining in biopsied lymphoma tissue of each case was calculated as described in “Materials and Methods.” There was no significant difference in overall survival between high and low intensity groups (P = 0.31).

Close modal
Fig. 5.

Kaplan-Meier analysis of overall survival in high and low plasma GSTP1-1 groups of NHL. There was a significant difference in overall survival between the two groups (P < 0.01).

Fig. 5.

Kaplan-Meier analysis of overall survival in high and low plasma GSTP1-1 groups of NHL. There was a significant difference in overall survival between the two groups (P < 0.01).

Close modal
Fig. 6.

Kaplan-Meier analysis of progression-free survival in high and low plasma GSTP1-1 groups of NHL. There was a significant difference in progression-free survival between the two groups (P < 0.01).

Fig. 6.

Kaplan-Meier analysis of progression-free survival in high and low plasma GSTP1-1 groups of NHL. There was a significant difference in progression-free survival between the two groups (P < 0.01).

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Yoshiro Niitsu, Fourth Department of Internal Medicine, Sapporo Medical University School of Medicine, South-1, West-16, Chuo-ku, Sapporo, 060-8543, Japan. Phone: 81-11-611-2111, ext. 3260; Fax: 81-11-612-7987; E-mail: niitsu@sapmed.ac.jp

Table 1

Relationship between the relative intensity of GSTP1-1 expression at the time of diagnosis and therapeutic outcome in CS III and IV NHL

Relative intensity of GSTP1-1 expressionOverall (%)
≤10.7 (%)>10.7 (%)
Complete response * 6/13 (46.2) 5/13 (38.5) 11/26 (42.3) 
Unconfirmed complete response 3/13 (23.1) 3/13 (23.1) 6/26 (23.1) 
Partial response 1/13 (7.7) 2/13 (15.4) 3/26 (11.5) 
Stable disease 1/13 (7.7) 0/13 (0.0) 1/26 (3.8) 
Progressive disease 2/13 (15.4) 3/13 (23.1) 5/26 (19.2) 
Relative intensity of GSTP1-1 expressionOverall (%)
≤10.7 (%)>10.7 (%)
Complete response * 6/13 (46.2) 5/13 (38.5) 11/26 (42.3) 
Unconfirmed complete response 3/13 (23.1) 3/13 (23.1) 6/26 (23.1) 
Partial response 1/13 (7.7) 2/13 (15.4) 3/26 (11.5) 
Stable disease 1/13 (7.7) 0/13 (0.0) 1/26 (3.8) 
Progressive disease 2/13 (15.4) 3/13 (23.1) 5/26 (19.2) 
*

There was no significant difference of complete response rates between the low and high intensity groups.

Table 2

Relationship between the plasma GSTP1-1 levels at the time of diagnosis and therapeutic outcome in CS III and IV NHL

Plasma GSTP1-1 level (ng/ml)Overall (%)
≤22.6 (%)>22.6 (%)
Complete response * 14/26 (55.2) 5/28 (16.0) 20/54 (37.0) 
Unconfirmed complete response 6/26 (24.1) 7/28 (28.0) 14/54 (25.9) 
Partial response 2/26 (6.9) 8/28 (28.0) 9/54 (16.7) 
Stable disease 1/26 (23.3) 2/28 (4.0) 2/54 (3.7) 
Progressive disease 3/26 (10.3) 6/28 (24.0) 9/54 (16.7) 
Plasma GSTP1-1 level (ng/ml)Overall (%)
≤22.6 (%)>22.6 (%)
Complete response * 14/26 (55.2) 5/28 (16.0) 20/54 (37.0) 
Unconfirmed complete response 6/26 (24.1) 7/28 (28.0) 14/54 (25.9) 
Partial response 2/26 (6.9) 8/28 (28.0) 9/54 (16.7) 
Stable disease 1/26 (23.3) 2/28 (4.0) 2/54 (3.7) 
Progressive disease 3/26 (10.3) 6/28 (24.0) 9/54 (16.7) 
*

There was a significant difference of complete response rates between the low and high plasma GSTP1-1 groups (P < 0.01 by χ2 test).

Table 3

Univariate association of international prognostic index with GSTP1-1 high and low groups in CS III, IV NHL

CharacteristicGSTP1-1 (ng/ml)P
≤22.6>22.6
Age   0.44 
 <61 yr 15 20  
 ≥61 yr 11  
Extranodal involvement   0.29 
 <2 14 12  
 ≥2 12 16  
Performance status   0.62 
 ≤1 15 19  
 >1 11  
Serum LDH level   0.29 
 ≤1 × normal 12  
 >1 × normal 14 20  
CharacteristicGSTP1-1 (ng/ml)P
≤22.6>22.6
Age   0.44 
 <61 yr 15 20  
 ≥61 yr 11  
Extranodal involvement   0.29 
 <2 14 12  
 ≥2 12 16  
Performance status   0.62 
 ≤1 15 19  
 >1 11  
Serum LDH level   0.29 
 ≤1 × normal 12  
 >1 × normal 14 20  
Table 4

Multivariate analysis of prognostic variables in patients with CS III and IV NHL

VariableOverall survivalProgression free survival
Hazard95% CIPHazard95% CIP
GSTP1-1 (>22.6) 3.472 1.307–9.224 0.0202 2.611 1.091–6.249 0.0311 
Extranodal involvement (≥2) 2.817 1.079–7.234 0.0314 2.141 0.915–5.008 0.0791 
PS (>1) 2.423 0.915–6.414 0.0658 1.468 0.605–3.563 0.4159 
LDH (>1 × normal) 2.527 0.854–7.897 0.0813 1.851 0.578–3.643 0.1989 
Age (≥61) 1.851 0.704–4.866 0.1934 0.956 0.422–2.166 0.6142 
VariableOverall survivalProgression free survival
Hazard95% CIPHazard95% CIP
GSTP1-1 (>22.6) 3.472 1.307–9.224 0.0202 2.611 1.091–6.249 0.0311 
Extranodal involvement (≥2) 2.817 1.079–7.234 0.0314 2.141 0.915–5.008 0.0791 
PS (>1) 2.423 0.915–6.414 0.0658 1.468 0.605–3.563 0.4159 
LDH (>1 × normal) 2.527 0.854–7.897 0.0813 1.851 0.578–3.643 0.1989 
Age (≥61) 1.851 0.704–4.866 0.1934 0.956 0.422–2.166 0.6142 

Abbreviations: PS, performance status; CI, confidence interval.

We thank Dr. Ohtani for valuable advice on immunohistochemical staining. Thanks to K. Litton, bachelor of English literature, for correction of the English in this manuscript.

1
Shipp MA. A Predictive model for aggressive non-Hodgkin’s lymphoma. The International non-Hodgkin’s lymphoma prognostic factors project.
N Engl J Med
1993
;
329
:
987
-94.
2
Hermans J, Krol AD, van Groningen K, et al International prognostic index for aggressive non-Hodgkin’s lymphoma is valid for all malignancy grades.
Blood
1995
;
86
:
1460
-3.
3
Hermine O, Haioun C, Lepage E, et al Prognostic significance of bcl-2 protein expression in aggressive non-Hodgkin’s lymphoma. Groupe d’Etude des Lymphomes de l’Adulte (GELA).
Blood
1996
;
87
:
265
-72.
4
Ichikawa A, Kinoshita T, Watanabe T, et al Mutations of the p53 gene as a prognostic factor in aggressive B-cell lymphoma.
N Engl J Med
1997
;
337
:
529
-34.
5
Wilson WH, Teruya-Feldstein J, Fest T, et al Relationship of p53, bcl-2, and tumor proliferation to clinical drug resistance in non-Hodgkin’s lymphomas.
Blood
1997
;
89
:
601
-9.
6
Murray GI, Taylor MC, McFadyen MC, et al Tumor-specific expression of cytochrome P450 CYP1B1.
Cancer Res
1997
;
57
:
3026
-31.
7
McFadyen MC, McLeod HL, Jackson FC, Melvin WT, Doehmer J, Murray GI. Cytochrome P450 CYP1B1 protein expression: a novel mechanism of anticancer drug resistance.
Biochem Pharmacol
2001
;
62
:
207
-12.
8
Gottesman MM, Pastan I. Biochemistry of multidrug resistance mediated by the multidrug transporter.
Annu Rev Biochem
1993
;
62
:
385
-427.
9
Tan B, Piwnica-Worms D, Ratner L. Multidrug resistance transporters and modulation.
Curr Opin Oncol
2000
;
12
:
450
-8.
10
Roninson IB, Chin JE, Choi KG, et al Isolation of human mdr DNA sequences amplified in multidrug-resistant KB carcinoma cells.
Proc Natl Acad Sci USA
1986
;
83
:
4538
-42.
11
Finnegan MC, Royds J, Goepel JR, Lorigan P, Hancock BW, Goyns MH. MDR-1 expression in non-Hodgkin’s lymphomas is unrelated to treatment intensity or response to therapy.
Leuk Lymphoma
1995
;
18
:
297
-302.
12
Kaczorowski S, Ochocka M, Kaczorowska M, Aleksandrowicz R, Matysiakl M, Karwacki M. Is P-glycoprotein a sufficient marker for multidrug resistance in vivo? Immunohistochemical staining for P-glycoprotein in children and adult leukemia: correlation with clinical outcome.
Leuk Lymphoma
1995
;
20
:
143
-52.
13
Goldstein LJ, Galski H, Fojo A, et al Expression of a multidrug resistance gene in human cancers.
J Natl Cancer Inst (Bethesda)
1989
;
81
:
116
-24.
14
Holzmayer TA, Hilsenbeck S, Von Hoff DD, Roninson IB. Clinical correlates of MDR1 (P-glycoprotein) gene expression in ovarian and small-cell lung carcinomas.
J Natl Cancer Inst (Bethesda)
1992
;
84
:
1486
-91.
15
Zhan Z, Sandor VA, Gamelin E, et al Expression of the multidrug resistance-associated protein gene in refractory lymphoma: quantitation by a validated polymerase chain reaction assay.
Blood
1997
;
89
:
3795
-800.
16
Sladek NE, Sreerama L, Rekha GK. Constitutive and overexpressed human cytosolic class-3 aldehyde dehydrogenases in normal and neoplastic cells/secretions.
Adv Exp Med Biol
1995
;
372
:
103
-14.
17
Wolf CR, Macpherson JS, Smyth JF. Evidence for the metabolism of mitozantrone by microsomal glutathione transferases and 3-methylcholanthrene-inducible glucuronosyl transferases.
Biochem Pharmacol
1986
;
35
:
1577
-81.
18
Townsend DM, Tew KD. The role of glutathione-S-transferase in anti-cancer drug resistance.
Oncogene
2003
;
22
:
7369
-75.
19
Hayes JD, Pulford DJ. The glutathione S-transferase supergene family: regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance.
Crit Rev Biochem Mol Biol
1995
;
30
:
445
-600.
20
Cullen KJ, Newkirk KA, Schumaker LM, Aldosari N, Rone JD, Haddad BR. Glutathione S-transferase pi amplification is associated with cisplatin resistance in head and neck squamous cell carcinoma cell lines and primary tumors.
Cancer Res
2003
;
63
:
8097
-102.
21
Chen G, Waxman DJ. Identification of glutathione S-transferase as a determinant of 4-hydroperoxycyclophosphamide resistance in human breast cancer cells.
Biochem Pharmacol
1995
;
49
:
1691
-701.
22
Berhane K, Mannervik B. Inactivation of the genotoxic aldehyde acrolein by human glutathione transferases of classes alpha, mu, and pi.
Mol Pharmacol
1990
;
37
:
251
-4.
23
Lee WP, Lee CL, Lin HC. Glutathione S-transferase and glutathione peroxidase are essential in the early stage of adriamycin resistance before P-glycoprotein overexpression in HOB1 lymphoma cells.
Cancer Chemother Pharmacol
1996
;
38
:
45
-51.
24
Black SM, Beggs JD, Hayes JD, et al Expression of human glutathione S-transferases in Saccharomyces cerevisiae confers resistance to the anticancer drugs adriamycin and chlorambucil.
Biochem J
1990
;
268
:
309
-15.
25
Ban N, Takahashi Y, Takayama T, et al Transfection of glutathione S-transferase (GST)-pi antisense complementary DNA increases the sensitivity of a colon cancer cell line to adriamycin, cisplatin, melphalan, and etoposide.
Cancer Res
1996
;
56
:
3577
-82.
26
Mulder TP, Manni JJ, Roelofs HM, Peters WH, Wiersma A. Glutathione S-transferases and glutathione in human head and neck cancer.
Carcinogenesis (Lond)
1995
;
16
:
619
-24.
27
Shiga H, Heath EI, Rasmussen AA, et al Prognostic value of p53, glutathione S-transferase pi, and thymidylate synthase for neoadjuvant cisplatin-based chemotherapy in head and neck cancer.
Clin Cancer Res
1999
;
5
:
4097
-104.
28
Hirazono K, Shinozuka T, Kuroshima Y, Itoh H, Kawai K. Immunohistochemical expression of glutathione S-transferase pi (GST-pi) and chemotherapy response in malignant ovarian tumors.
J of Obstet Gynaecol
1995
;
21
:
305
-12.
29
Berhane K, Hao XY, Christensson B, Hansson J, Ringborg U, Mannervik B. The expression of glutathione transferase isoenzymes in human malignant lymphoma biopsies.
Acta Oncol
1995
;
34
:
35
-41.
30
Rodriguez C, Commes T, Robert J, Rossi JF. Expression of P-glycoprotein and anionic glutathione S-transferase genes in non-Hodgkin’s lymphoma.
Leuk Res
1993
;
17
:
149
-56.
31
Ribrag V, Koscielny S, Carpiuc I, et al Prognostic value of GST-pi expression in diffuse large B-cell lymphomas.
Leukemia
2003
;
17
:
972
-7.
32
Takahashi Y, Hirata Y, Saito T, et al Enzyme immunoassay for human serum glutathione S-transferase π using monoclonal and polyclonal antibodies.
Cancer J
1989
;
2
:
225
-9.
33
Niitsu Y, Takahashi Y, Saito T, et al Serum glutathione-S-transferase-pi as a tumor marker for gastrointestinal malignancies.
Cancer (Phila)
1989
;
63
:
317
-23.
34
Jaffe ES, Harris NL, Stein H, Vardiman JW. .
World Health Organization classification of tumours: pathology and genetics of tumours of haematopoietic and lymphoid tissues
2001
International Agency for Research on Cancer Lyon, France
35
Harris NL, Jaffe ES, Stein H, et al A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group.
Blood
1994
;
84
:
1361
-92.
36
Fisher RI, Gaynor ER, Dahlberg S, et al Comparison of a standard regimen (CHOP) with three intensive chemotherapy regimens for advanced non-Hodgkin’s lymphoma.
N Engl J Med
1993
;
328
:
1002
-6.
37
Cheson BD, Horning SJ, Coiffier B, et al Report of an international workshop to standardize response criteria for non-Hodgkin’s lymphomas. NCI Sponsored International Working Group.
J Clin Oncol
1999
;
17
:
1244
Review Erratum in: J Clin Oncol 2000;18:2351.
38
Kato J, Kobune M, Nakamura T, et al Normalization of elevated hepatic 8-hydroxy-2′-deoxyguanosine levels in chronic hepatitis C patients by phlebotomy and low iron diet.
Cancer Res
2001
;
61
:
8697
-8702.
39
Gaudiano G, Koch TH, Lo Bello M, et al Lack of glutathione conjugation to adriamycin in human breast cancer MCF-7/DOX cells. Inhibition of glutathione S-transferase p1-1 by glutathione conjugates from anthracyclines.
Biochem Pharmacol
2000
;
60
:
1915
-23.
40
Beckett GJ, Hayes JD. Glutathione S-transferases: biomedical applications.
Adv Clin Chem
1993
;
30
:
281
-380.
41
Fidler IJ, Hart IR. Biological diversity in metastatic neoplasms: origins and implications.
Science (Wash D C)
1982
;
217
:
998
-1003.
42
Kerbel RS. Implications of immunological heterogeneity of tumours.
Nature (Lond)
1979
;
280
:
358
-60.
43
Kura T, Takahashi Y, Takayama T, et al Glutathione S-transferase-pi is secreted as a monomer into human plasma by platelets and tumor cells.
Biochim Biophys Acta
1996
;
1292
:
317
-23.
44
Alizadeh AA, Eisen MB, Davis RE, et al Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling.
Nature (Lond)
2000
;
403
:
503
-11.
45
Rosenwald A, Wright G, Chan WC, et al Lymphoma/Leukemia Molecular Profiling Project. The use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma.
N Engl J Med
2002
;
346
:
1937
-47.
46
Shipp MA, Ross KN, Tamayo P, et al Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning.
Nat Med
2002
;
8
:
68
-74.
47
Rajeevan MS, Vernon SD, Taysavang N, Unger ER. Validation of array-based gene expression profiles by real-time (kinetic) RT-PCR.
J Mol Diagn
2001
;
3
:
26
-31.
48
Schisselbauer JC, Silber R, Papadopoulos E, Abrams K, LaCreta FP, Tew KD. Characterization of glutathione S-transferase expression in lymphocytes from chronic lymphocytic leukemia patients.
Cancer Res
1990
;
50
:
3562
-8.