- Email: [email protected]
Immunomodulation with low dose levamisole in patients with colonic polyps
Cancer Detection and Prevention 30 (2006) 94–98 www.elsevier.com/locate/cdp
Immunomodulation with low dose levamisole in patients with colonic polyps Randall F. Holcombe MDa,*, Christine E. McLaren PhDb, Tatjana Milovanovic MSa a
Division of Hematology/Oncology, The Chao Family Comprehensive Cancer Center, University of California, Irvine Medical Center, 101 The City Drive, Bld. 56, Orange, CA 92868, USA b Division of Epidemiology, University of California, Irvine, USA Accepted 22 June 2005
Abstract Background: Levamisole (LMS) has immunomodulatory activity, stimulates the immune system of healthy, normal volunteers and has been proposed previously as a colon cancer preventive agent. Methods: Patients with a history of colonic polyps who are at increased risk of colon cancer received LMS in a placebo-controlled, double-blinded clinical trail with crossover design. Primary endpoints were immunologic and included flow cytometry of peripheral blood mononuclear cells (PBMCs), measurement of interferon-g copy number (IGCN) in PBMCs, and an ex vivo serum immune assay. Results: No differences were seen in the expression of multiple antigens by flow cytometry pre- and postLMS. The IGCN partitioned subjects into two distinct groups defined by a g-distribution which had a differential response to LMS. Those with low basal IGCN had a lower percentage of CD25 expressing PBMCs and responded to low dose LMS by producing more PBMC-derived interferon-g and increasing the expression of CD25 and two NK cell markers, CD16 and CD56. In contrast, subjects with a high basal IGCN responded to low dose LMS with a reduction in PBMC-derived interferon-g and a decrease in the expression of CD25. Conclusions: In aggregate, these responses suggest that LMS may act as an immunostimulatory agent for one group, those with low basal IGCN, and as an immunosuppressive agent for the other. LMS may not be an optimal agent for most patients with colonic polyps and should be avoided in patients with normal immune function. IGCN may be useful as an immunologic surrogate endpoint biomarker in future cancer prevention trials with immunomodulatory agents. # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Immunologic endpoints; Colonic polyps; Levamisole; Interferon-g; Flow cytometry; Cancer prevention; Peripheral blood mononuclear cells; Immunostimulatory activity; Clinical design; Sample acquisition
1. Introduction Levamisole (LMS) is an imidazothiazole anti-helminthic which is effective in reducing the risk of relapse when given adjuvantly to patients with surgically-resected stage III colon cancer [1,2], though it is no longer utilized clinically in this setting. It is thought to act, in part, though stimulation of the cellular immune system. In patients, LMS has been shown to increase the number of circulating natural killer (NK) cells [3], the expression of membrane CD25 (IL2 receptor) [4] and serum levels of soluble IL2 receptor
* Corresponding author. Tel.: +1 714 456 5153; fax: +1 714 456 2242. E-mail address: [email protected] (R.F. Holcombe).
[5]. In vitro, LMS modulates NK cell-mediated tumor lysis [6], tumor cell MHC class I expression [7,8] and IL12dependent Th1 immune responses [9,10]. The role of LMS-mediated immunomodulation in vivo was examined in a dose escalation trial in normal volunteers [10]. Significant increases in the proportion of peripheral blood mononuclear cells (PBMCs) expressing the NK antigen CD16, especially the CD16/CD56(+) subpopulation, was noted. Only small effects on Th1 cellular immune function and serum cytokine levels were seen. Low dosages of LMS, 50 mg/day – 3 times/week, were equally efficacious in stimulating the immune system as clinically utilized dosages which are up to three-fold higher. In contrast to the standard dosages, minimal toxicity was seen at the low, 50 mg/day, dose.
0361-090X/$30.00 # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2005.08.002
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
The appeal among the lay public of utilizing an immunostimulatory agent for cancer prevention is well documented [11]. There is, however, a paucity of scientifically-controlled data to validate this approach. In order to address this, we enrolled patients with a history of colonic polyps onto a placebo-controlled clinical trial with low dose LMS. Patients with colonic polyps are at significantly increased risk for the development of colon cancer, the third leading cause of cancer death for men and women in the United States [12]. While the primary mechanisms involved in colonic polyp formation are not immunologic, there is some evidence that immune function in patients with polyps is not entirely normal [13,14]. Endpoints in this study were immunologic including both general measures of immune activity as well as specific assays of cellular immune function. We report for the first time that low dose LMS is an effective immunostimulant only for patients with evidence of impaired basal cellular immune function and describe a novel assay which partitions patients into two groups with differential responses to LMS.
2. Patients and experimental methods 2.1. Clinical trial design and sample acquisition Eighteen patients with a history of colonic polyps were identified through the University of California, Irvine (UCI), Division of Gastroenterology and the Chao Family Comprehensive Cancer Center. Each patient signed written informed consent for this study which was approved by the UCI Institutional Review Board. Levamisole (LMS) and an identically-appearing inert placebo tablet were obtained from Johnson & Johnson. Patients were randomized to receive either LMS at 50 mg/day – 3 times/week or placebo on the same schedule, every other week for 12 weeks. After 12 weeks, patients were crossed over to the other arm of the study and treated for an additional 12 weeks. The overall duration of participation in the trial was 24 weeks. Randomization was performed according to a table of random numbers and the key was retained by the study pharmacist. Patients, the treating physicians, and laboratory personnel performing immunologic assays were blinded as to when the patient was receiving active drug versus placebo. Heparinized blood and serum samples for laboratory analyses were obtained at study initiation, after the first 12 weeks (at crossover), and after 24 weeks (at completion of the study). Any toxicity notations were also made at these timepoints.
95
washes cells were incubated for 30 min at 4 8C with FITCor PE-conjugated monoclonal antibodies directed against the following antigens: CD4, CD8, CD16, CD25, CDw29 (co-stain with CD4), CD45RA (co-stain with CD4 to identify naı¨ve helper T-cells), CD45RO (co-stain with CD4 to identify memory helper T-cells), CD56, CD95 (Fas) and CD178 (Fas ligand). After additional washes and fixation with 1% formaldehyde, cells were analyzed by flow cytometry. Controls included unstained cells to assess viability and autofluorescence and cells incubated with conjugated isotype-specific controls. 2.3. Interferon-g copy number (IGCN) in PBMCs PBMCs were obtained as described above and RNA was isolated with a TRIzol reagent from 10 106 cells. cDNA was synthesized in a 20 ml reaction containing 5 mg of total RNA, 200 UMmuLV reverse transcriptase and 500 ng of oligo(dT)20 primers (Promega, Madison, WI). PCR products were analyzed by CytoXpress Quantitative PCR system (BioSource International, Camarillo, CA) with spectrophotometric readouts with a Biorad 96-well Elisa plate reader at 450 nm. The CytoXpress system is based on biotin-labeled primer incorporation into PCR products, followed by hybridization to sequence-specific capture oligos in microplate wells. The captured sequences are detected and quantified in enzyme-streptavidin/substrate reaction and the optical density measured. Primers were utilized to provide a quantitative assessment of the number of copies of interferon-g mRNA in a defined amount of total PBMC RNA. 2.4. Statistics Flow cytometric data and IGCN values were tested for significance with a two-tailed Wilcoxon signed rank (paired, non-parametric) test with significance defined as p < 0.05 (actual values reported). For comparison across groups, an unpaired t-test with Welch’s correction for unequal variances was utilized. Correlations between different data sets were examined with the Spearman’s ‘‘r’’ correlation test. The entire IGCN data set was examined for ‘‘goodness to fit’’ to a g-distribution model by x2 and plotted as observed quantiles versus fitted g-quantiles. A further description of the statistical analysis describing the populations defined by IGCN is included under results.
3. Results
2.2. Flow cytometry
3.1. Flow cytometry
Heparinized blood was diluted 1:1 with Hanks’ buffered salt solution and peripheral blood mononuclear cells (PBMCs) were obtained by centrifugation on a Ficoll/ hypaque gradient (Organon Technika, NC). Following
Results of flow cytometry for the expression of multiple antigens on PBMCs are shown in Table 1. The expression of CD4, multiple subsets of CD4 cells including CD4/CDw29, CD4/CD45RA (naı¨ve), and CD4/CD45RO (memory) and
96
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
Table 1 Flow cytometry of peripheral blood mononuclear cells before and following administration of levamisole, all patients (n = 18) Antigen/CD designation
CD4 CD4/CD45RA CD4/CD45RO CD4/CDw29 CD8 CD16 CD25 CD56 CD95 (Fas) CD178 (Fas-ligand)
Post-levamisolea
Pre-levamisole Percent (+)
S.E.
Percent (+)
S.E.
51.0 43.5 46.8 46.6 27.6 25.8 33.9 24.9 71.9 21.5
2.2 4.1 2.6 2.2 1.8 2.2 3.5 1.9 3.2 2.8
44.0 39.0 41.3 42.6 29.1 25.3 28.8 27.8 60.1 22.9
3.5 3.1 4.0 3.7 2.2 2.3 3.8 2.2 6.7 2.7
a
There were no statistically significant differences between the prelevamisole and the post-levamisole groups.
CD8 were examined. CD16 and CD56, NK cell markers, CD25, membrane IL2 receptor Fas (CD95) and Fas-ligand (CD178) were also tested. There were no statistically significant differences for any of these antigens between the pre-LMS and post-LMS groups. 3.2. Interferon-g copy number (IGCN) IGCN was examined in PBMCs from patients enrolled on the study both pre- and post-LMS. The results of these analyses are shown in Fig. 1 (two left-sided bars). There was a trend toward decreased IGCN in PBMCs following LMS, but this did not reach statistical significance. IGCN of Jurkat T-cells was also examined following incubation with preand post-LMS patient serum but no statistically significant changes were seen (data not shown).
the remainder (12280 4748). Triplicate assays were repeated to confirm this observation. The population data did not fit a normal distribution; the tails of the observed distribution varied significantly from the expected normal tails. This was confirmed by both the Kolmogorov–Smirnov normality test ( p < 0.01) and the Anderson–Darling normality test (A2: 2.143; p < 0.001). Basal IGCN partitioned subjects into two distinct groups defined by a gdistribution with curve parameters a = 0.5126 and b = 20783.12. Four of 18 and six of 18 subjects fell below the 99% and 95% confidence intervals, respectively, suggesting impaired basal immune function in these subjects. 3.4. LMS induced immune responses in IGCN-defined subgroups Patients with low basal IGCN exhibited a large increase in IGCN following LMS (234 95 to 5140 4748) while subjects with a high basal IGCN demonstrated a decline following LMS (12280 4748 to 3380 963, p = 0.019; Fig. 1). There was a highly significant correlation between basal IGCN levels and the percent-change in IGCN following LMS ( p = 0.001). Consistent with the IGCN data, and changes seen following LMS, the expression of CD25 was significantly lower in the low basal IGCN subgroup than in the high basal IGCN subgroup (23.1 4.1% versus 38.0 4.0%, p = 0.021). The subgroup with low initial CD25 expression exhibited an increase following LMS (23.1 4.1% to 28.2 3.6%) while the subgroup with high initial CD25 expression exhibited a decline (38.4 4.0% to 29.1 5.5%; p = 0.06). The low basal IGCN subgroup also demonstrated an increase in the proportion of PBMCs expressing the NK cell markers CD56
3.3. Subject partitioning based on IGCN It was apparent that the basal levels of PBMC IGCN were very low in a several patients (234 95), and quite high in
Fig. 1. Interferon-g copy number (IGCN) in peripheral blood mononuclear cells (PBMCs) from patients pre- and post-levamisole. Levamisole-related changes in all patients are depicted in the columns on the left. The middle and right columns depict levamisole-realted changes for subgroups divided according to the basal PBMC IGCN values (see text). * p = 0.02 vs. pre-LMS level.
Fig. 2. Levamisole-related changes in the expression of CD25, CD56 and CD16 in patient PBMCs grouped according to basal PBMC IGCN values. ** p = 0.02 vs. CD25, low, pre-LMS; *p = 0.03 vs. CD56, low, pre-LMS.
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
(21.4 2.5% to 29.4 2.3%; p = 0.037) and CD16 (21.7 3.2% to 26.9 5.3%). CD56 and CD16 expression did not change significantly following LMS in the high basal IGCN subgroup. LMS-related changes in the expression of CD25, CD56 and CD16 in patient PBMCs, grouped according to basal PBMC IGCN values, is shown in Fig. 2.
4. Discussion Measurement of interferon-g production by PBMCs, in this study by a competitive PCR Elisa-based assay (IGCN), partitions subjects into two distinct groups which have a differential response to LMS. Interferon-g is produced by cells of the monocyte/macrophage lineage as well as T-helper-1 (Th1) cells which regulate the cellular immune response [15]. Increases in interferon-g production generally reflect a more robust cellular immune response and these, Th1-type immune responses are felt to be important in antitumor immunity [16]. In this study, subjects with a low basal IGCN level had a lower percentage of CD25 expressing PBMCs and responded to low dose LMS by producing more PBMC-derived interferon-g and increasing the expression of CD25 and two NK cell markers, CD16 and CD56. NK cells also play a critical role in immune surveillance against tumors [17,18]. In contrast, subjects with a high basal IGCN responded to low dose LMS with a reduction in PBMCderived interferon-g and a decrease in the expression of CD25. In aggregate, these responses suggest that LMS may act as an immunostimulatory agent for one group, those with low basal IGCN, and as an immunosuppressive agent for the other. The concept that a single immunomodulatory agent, including LMS, can cause both immunostimulation or immunodepression in different settings or under different conditions, has been reported previously [19,20]. In an earlier study [10], LMS increased CD16 and CD16/ CD56 expression in normal volunteers. In this current report, overall CD16 was not increased and the increases in CD16 seen in the low basal IGCN subgroup were modest. However, since basal IGCN was not measured in the earlier study it is difficult to directly compare these results. The sample populations in the two studies are similar except for the history of colonic polyps in this current trial. There was no clinical evidence of basal immune suppression for either group. Overall, LMS had minimal immunostimulatory activity except in patients with low basal IGCN. In colon cancer patients, LMS has significant immunostimulatory activity when utilized in conjunction with five-fluorouracil in the adjuvant treatment of patients with stage III disease [21]. This may be because cancer patients, especially those on chemotherapy, have significant immune system suppression [22]. Individuals with low basal IGCN may represent a subset of the population with some degree of immune suppression whose immune systems are primed to respond positively to LMS. Because of the modest degree of immunostimulation
97
noted with LMS, and concerns that subjects with intact immune function may actually experience suppression of cellular immune activity, LMS may not be an optimal agent for most patients with a history of colonic polyps. It should be avoided in patients with normal immune function but may be a useful immunostimulatory agent for cancer patients, other patients with impaired cellular immune responses, or for patients with familial polyposis syndromes who have documented low basal IGCN values. This study reports for the first time the use of a novel assay, IGCN, which defines two subgroups of individuals and, specifically, may identify subjects with impaired basal cellular immune function. IGCN may be a useful surrogate endpoint biomarker (SEB) for cancer prevention trials with immunomodulatory agents, especially in immunocompromized populations. If LMS is utilized in future trials, IGCN will be an important assay to assess its effects on patient immune function.
References [1] Moertel CG, Fleming TR, McDonald JS, Haller DG, Laurie JA, Goodman PJ, et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. New Engl J Med 1990;322:352–8. [2] Laurie JA, Moertel CG, Fleming TR, Wieand HS, Leigh JE, Rubin J, et al. Surgical adjuvant therapy of large-bowel carcinoma: an evaluation of levamisole and the combination of levamisole and fluorouracil. J Clin Oncol 1989;7:1447–56. [3] Holcombe RF, Stewart RM, Betzing KW, Kannan K. Alteration in lymphocyte phenotype associated with administration of adjuvant levamisole and 5-fluorouracil. Cancer Immunol Immunother 1994;34:394–8. [4] DeBrander M, Vandebroek J, Wassenaar H, De Cree J, Braisier A, Demoen B, et al. Immunological alterations induced by adjuvant treatment of postoperative colon carcinoma Duke’s B or C with levamisole in combination with 5-FU. Anticancer Res 1995;15:2271–7. [5] Janik J, Kopp WC, Smith JW, Longo DL, Alvord G, Sharfman WH, et al. Dose-related immunologic effects of levamisole in patients with cancer. J Clin Oncol 1993;11:125–35. [6] Clarke GR, Burton RC, Smart YC. The antitumor effect of levamisole in mice are mediated by NC-1.1+ cells. Cancer Immunol Immunother 1997;45:115–8. [7] Goodrich KH, Alvarez X, Holcombe RF. Effect of levamisole on major histocompatibility complex (MHC) class I expression in colorectal and breast carcinoma cell lines. Cancer 1993;72:225–30. [8] Kimball ES, Fisher MC. Levamisole effects on major histocompatibility complex and adhesion molecule expression and on myeloid cell adhesion to human colon tumor cell lines. J Natl Cancer Inst 1996;88:109–16. [9] Holcombe RF, Stewart RM. Augmentation of IL12-dependent Th1 cytokine production by levamisole. Allergy Clin Immunol Int 1997;9:119–20. [10] Holcombe RF, Milovanovic T, Stewart RM, Brodhag TM. Investigating immunomodulation for colon cancer prevention: results of an in vivo dose escalation trial of levamisole with immunologic endpoints. Cancer Detect Prev 2001;25:183–91. [11] Biological therapies for cancer: questions and answers in NCI on-line, August 16, 2004. http://cis.nci.nih.gov/fact/7_2.htm. [12] Cancer facts and figures, 2004 at American Cancer Society Inc., 2004. http://www.cancer.org/. [13] Jacobson-Brown P, Neuman MG. Colon polyps and cytokines: emerging immunological mechanisms. Rom J Gastroenterol 2003;12:207–14.
98
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
[14] Robbins TG, Weinstein RJ, Demers RY. Lymphocytopenia, T-lyphocyte subsets, and colorectal polyps in automotive pattern and model markers. J Occup Med 1991;33:510–5. [15] Murphy KM, Reiner SL. The lineage decisions of helper T-cells. Nat Rev Immunol 2002;2(12):933–44. [16] Dredge K, Marriott JB, Todryk SM, Dalgleish AG. Adjuvants and the promotion of Th1-type cytokines in tumour immunotherapy. Cancer Immunol Immunother 2002;51(10):521–31. Epub September 20, 2002. [17] Smyth MJ, Thia KY, Street SE, Cretney E, Trapani JA, Taniguchi M, et al. Differential tumor surveillance by natural killer (NK) and NKTcells. J Exp Med 2000;191:661–8. [18] Martiniello R, Burton RC, Smart YC. Ly-6C+ natural T (NT) cells mediate immune surveillance against NK-sensitive and NK-resistant
[19]
[20] [21]
[22]
transplantable tumours in certain strains of mice. Int J Cancer 1996;66:532–7. Gozzo JJ, Hanson RH. Varied in vivo properties of levamisole (LMS) on the immune response in mice. Transpl Proc 1979;11: 1420–1. Bliznakov EG. Immunostimulation or immunodepression? Biomedicine 1977;26:73–6. Holcombe RF, Jacobson J, Dakhil SR, Stewart RM, Betzing KS, Kannan K, et al. Association of immune parameters with clinical outcome in stage III colon cancer: results of Southwest oncology group protocol 9009. Cancer Immunol Immunother 1999;48:533–9. Block KI, Boyd B, Gonzalez N, Vojdani A. The immune system in cancer. Integr Cancer Ther 2002;1:294–316.
Immunomodulation with low dose levamisole in patients with colonic polyps Randall F. Holcombe MDa,*, Christine E. McLaren PhDb, Tatjana Milovanovic MSa a
Division of Hematology/Oncology, The Chao Family Comprehensive Cancer Center, University of California, Irvine Medical Center, 101 The City Drive, Bld. 56, Orange, CA 92868, USA b Division of Epidemiology, University of California, Irvine, USA Accepted 22 June 2005
Abstract Background: Levamisole (LMS) has immunomodulatory activity, stimulates the immune system of healthy, normal volunteers and has been proposed previously as a colon cancer preventive agent. Methods: Patients with a history of colonic polyps who are at increased risk of colon cancer received LMS in a placebo-controlled, double-blinded clinical trail with crossover design. Primary endpoints were immunologic and included flow cytometry of peripheral blood mononuclear cells (PBMCs), measurement of interferon-g copy number (IGCN) in PBMCs, and an ex vivo serum immune assay. Results: No differences were seen in the expression of multiple antigens by flow cytometry pre- and postLMS. The IGCN partitioned subjects into two distinct groups defined by a g-distribution which had a differential response to LMS. Those with low basal IGCN had a lower percentage of CD25 expressing PBMCs and responded to low dose LMS by producing more PBMC-derived interferon-g and increasing the expression of CD25 and two NK cell markers, CD16 and CD56. In contrast, subjects with a high basal IGCN responded to low dose LMS with a reduction in PBMC-derived interferon-g and a decrease in the expression of CD25. Conclusions: In aggregate, these responses suggest that LMS may act as an immunostimulatory agent for one group, those with low basal IGCN, and as an immunosuppressive agent for the other. LMS may not be an optimal agent for most patients with colonic polyps and should be avoided in patients with normal immune function. IGCN may be useful as an immunologic surrogate endpoint biomarker in future cancer prevention trials with immunomodulatory agents. # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Immunologic endpoints; Colonic polyps; Levamisole; Interferon-g; Flow cytometry; Cancer prevention; Peripheral blood mononuclear cells; Immunostimulatory activity; Clinical design; Sample acquisition
1. Introduction Levamisole (LMS) is an imidazothiazole anti-helminthic which is effective in reducing the risk of relapse when given adjuvantly to patients with surgically-resected stage III colon cancer [1,2], though it is no longer utilized clinically in this setting. It is thought to act, in part, though stimulation of the cellular immune system. In patients, LMS has been shown to increase the number of circulating natural killer (NK) cells [3], the expression of membrane CD25 (IL2 receptor) [4] and serum levels of soluble IL2 receptor
* Corresponding author. Tel.: +1 714 456 5153; fax: +1 714 456 2242. E-mail address: [email protected] (R.F. Holcombe).
[5]. In vitro, LMS modulates NK cell-mediated tumor lysis [6], tumor cell MHC class I expression [7,8] and IL12dependent Th1 immune responses [9,10]. The role of LMS-mediated immunomodulation in vivo was examined in a dose escalation trial in normal volunteers [10]. Significant increases in the proportion of peripheral blood mononuclear cells (PBMCs) expressing the NK antigen CD16, especially the CD16/CD56(+) subpopulation, was noted. Only small effects on Th1 cellular immune function and serum cytokine levels were seen. Low dosages of LMS, 50 mg/day – 3 times/week, were equally efficacious in stimulating the immune system as clinically utilized dosages which are up to three-fold higher. In contrast to the standard dosages, minimal toxicity was seen at the low, 50 mg/day, dose.
0361-090X/$30.00 # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2005.08.002
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
The appeal among the lay public of utilizing an immunostimulatory agent for cancer prevention is well documented [11]. There is, however, a paucity of scientifically-controlled data to validate this approach. In order to address this, we enrolled patients with a history of colonic polyps onto a placebo-controlled clinical trial with low dose LMS. Patients with colonic polyps are at significantly increased risk for the development of colon cancer, the third leading cause of cancer death for men and women in the United States [12]. While the primary mechanisms involved in colonic polyp formation are not immunologic, there is some evidence that immune function in patients with polyps is not entirely normal [13,14]. Endpoints in this study were immunologic including both general measures of immune activity as well as specific assays of cellular immune function. We report for the first time that low dose LMS is an effective immunostimulant only for patients with evidence of impaired basal cellular immune function and describe a novel assay which partitions patients into two groups with differential responses to LMS.
2. Patients and experimental methods 2.1. Clinical trial design and sample acquisition Eighteen patients with a history of colonic polyps were identified through the University of California, Irvine (UCI), Division of Gastroenterology and the Chao Family Comprehensive Cancer Center. Each patient signed written informed consent for this study which was approved by the UCI Institutional Review Board. Levamisole (LMS) and an identically-appearing inert placebo tablet were obtained from Johnson & Johnson. Patients were randomized to receive either LMS at 50 mg/day – 3 times/week or placebo on the same schedule, every other week for 12 weeks. After 12 weeks, patients were crossed over to the other arm of the study and treated for an additional 12 weeks. The overall duration of participation in the trial was 24 weeks. Randomization was performed according to a table of random numbers and the key was retained by the study pharmacist. Patients, the treating physicians, and laboratory personnel performing immunologic assays were blinded as to when the patient was receiving active drug versus placebo. Heparinized blood and serum samples for laboratory analyses were obtained at study initiation, after the first 12 weeks (at crossover), and after 24 weeks (at completion of the study). Any toxicity notations were also made at these timepoints.
95
washes cells were incubated for 30 min at 4 8C with FITCor PE-conjugated monoclonal antibodies directed against the following antigens: CD4, CD8, CD16, CD25, CDw29 (co-stain with CD4), CD45RA (co-stain with CD4 to identify naı¨ve helper T-cells), CD45RO (co-stain with CD4 to identify memory helper T-cells), CD56, CD95 (Fas) and CD178 (Fas ligand). After additional washes and fixation with 1% formaldehyde, cells were analyzed by flow cytometry. Controls included unstained cells to assess viability and autofluorescence and cells incubated with conjugated isotype-specific controls. 2.3. Interferon-g copy number (IGCN) in PBMCs PBMCs were obtained as described above and RNA was isolated with a TRIzol reagent from 10 106 cells. cDNA was synthesized in a 20 ml reaction containing 5 mg of total RNA, 200 UMmuLV reverse transcriptase and 500 ng of oligo(dT)20 primers (Promega, Madison, WI). PCR products were analyzed by CytoXpress Quantitative PCR system (BioSource International, Camarillo, CA) with spectrophotometric readouts with a Biorad 96-well Elisa plate reader at 450 nm. The CytoXpress system is based on biotin-labeled primer incorporation into PCR products, followed by hybridization to sequence-specific capture oligos in microplate wells. The captured sequences are detected and quantified in enzyme-streptavidin/substrate reaction and the optical density measured. Primers were utilized to provide a quantitative assessment of the number of copies of interferon-g mRNA in a defined amount of total PBMC RNA. 2.4. Statistics Flow cytometric data and IGCN values were tested for significance with a two-tailed Wilcoxon signed rank (paired, non-parametric) test with significance defined as p < 0.05 (actual values reported). For comparison across groups, an unpaired t-test with Welch’s correction for unequal variances was utilized. Correlations between different data sets were examined with the Spearman’s ‘‘r’’ correlation test. The entire IGCN data set was examined for ‘‘goodness to fit’’ to a g-distribution model by x2 and plotted as observed quantiles versus fitted g-quantiles. A further description of the statistical analysis describing the populations defined by IGCN is included under results.
3. Results
2.2. Flow cytometry
3.1. Flow cytometry
Heparinized blood was diluted 1:1 with Hanks’ buffered salt solution and peripheral blood mononuclear cells (PBMCs) were obtained by centrifugation on a Ficoll/ hypaque gradient (Organon Technika, NC). Following
Results of flow cytometry for the expression of multiple antigens on PBMCs are shown in Table 1. The expression of CD4, multiple subsets of CD4 cells including CD4/CDw29, CD4/CD45RA (naı¨ve), and CD4/CD45RO (memory) and
96
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
Table 1 Flow cytometry of peripheral blood mononuclear cells before and following administration of levamisole, all patients (n = 18) Antigen/CD designation
CD4 CD4/CD45RA CD4/CD45RO CD4/CDw29 CD8 CD16 CD25 CD56 CD95 (Fas) CD178 (Fas-ligand)
Post-levamisolea
Pre-levamisole Percent (+)
S.E.
Percent (+)
S.E.
51.0 43.5 46.8 46.6 27.6 25.8 33.9 24.9 71.9 21.5
2.2 4.1 2.6 2.2 1.8 2.2 3.5 1.9 3.2 2.8
44.0 39.0 41.3 42.6 29.1 25.3 28.8 27.8 60.1 22.9
3.5 3.1 4.0 3.7 2.2 2.3 3.8 2.2 6.7 2.7
a
There were no statistically significant differences between the prelevamisole and the post-levamisole groups.
CD8 were examined. CD16 and CD56, NK cell markers, CD25, membrane IL2 receptor Fas (CD95) and Fas-ligand (CD178) were also tested. There were no statistically significant differences for any of these antigens between the pre-LMS and post-LMS groups. 3.2. Interferon-g copy number (IGCN) IGCN was examined in PBMCs from patients enrolled on the study both pre- and post-LMS. The results of these analyses are shown in Fig. 1 (two left-sided bars). There was a trend toward decreased IGCN in PBMCs following LMS, but this did not reach statistical significance. IGCN of Jurkat T-cells was also examined following incubation with preand post-LMS patient serum but no statistically significant changes were seen (data not shown).
the remainder (12280 4748). Triplicate assays were repeated to confirm this observation. The population data did not fit a normal distribution; the tails of the observed distribution varied significantly from the expected normal tails. This was confirmed by both the Kolmogorov–Smirnov normality test ( p < 0.01) and the Anderson–Darling normality test (A2: 2.143; p < 0.001). Basal IGCN partitioned subjects into two distinct groups defined by a gdistribution with curve parameters a = 0.5126 and b = 20783.12. Four of 18 and six of 18 subjects fell below the 99% and 95% confidence intervals, respectively, suggesting impaired basal immune function in these subjects. 3.4. LMS induced immune responses in IGCN-defined subgroups Patients with low basal IGCN exhibited a large increase in IGCN following LMS (234 95 to 5140 4748) while subjects with a high basal IGCN demonstrated a decline following LMS (12280 4748 to 3380 963, p = 0.019; Fig. 1). There was a highly significant correlation between basal IGCN levels and the percent-change in IGCN following LMS ( p = 0.001). Consistent with the IGCN data, and changes seen following LMS, the expression of CD25 was significantly lower in the low basal IGCN subgroup than in the high basal IGCN subgroup (23.1 4.1% versus 38.0 4.0%, p = 0.021). The subgroup with low initial CD25 expression exhibited an increase following LMS (23.1 4.1% to 28.2 3.6%) while the subgroup with high initial CD25 expression exhibited a decline (38.4 4.0% to 29.1 5.5%; p = 0.06). The low basal IGCN subgroup also demonstrated an increase in the proportion of PBMCs expressing the NK cell markers CD56
3.3. Subject partitioning based on IGCN It was apparent that the basal levels of PBMC IGCN were very low in a several patients (234 95), and quite high in
Fig. 1. Interferon-g copy number (IGCN) in peripheral blood mononuclear cells (PBMCs) from patients pre- and post-levamisole. Levamisole-related changes in all patients are depicted in the columns on the left. The middle and right columns depict levamisole-realted changes for subgroups divided according to the basal PBMC IGCN values (see text). * p = 0.02 vs. pre-LMS level.
Fig. 2. Levamisole-related changes in the expression of CD25, CD56 and CD16 in patient PBMCs grouped according to basal PBMC IGCN values. ** p = 0.02 vs. CD25, low, pre-LMS; *p = 0.03 vs. CD56, low, pre-LMS.
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
(21.4 2.5% to 29.4 2.3%; p = 0.037) and CD16 (21.7 3.2% to 26.9 5.3%). CD56 and CD16 expression did not change significantly following LMS in the high basal IGCN subgroup. LMS-related changes in the expression of CD25, CD56 and CD16 in patient PBMCs, grouped according to basal PBMC IGCN values, is shown in Fig. 2.
4. Discussion Measurement of interferon-g production by PBMCs, in this study by a competitive PCR Elisa-based assay (IGCN), partitions subjects into two distinct groups which have a differential response to LMS. Interferon-g is produced by cells of the monocyte/macrophage lineage as well as T-helper-1 (Th1) cells which regulate the cellular immune response [15]. Increases in interferon-g production generally reflect a more robust cellular immune response and these, Th1-type immune responses are felt to be important in antitumor immunity [16]. In this study, subjects with a low basal IGCN level had a lower percentage of CD25 expressing PBMCs and responded to low dose LMS by producing more PBMC-derived interferon-g and increasing the expression of CD25 and two NK cell markers, CD16 and CD56. NK cells also play a critical role in immune surveillance against tumors [17,18]. In contrast, subjects with a high basal IGCN responded to low dose LMS with a reduction in PBMCderived interferon-g and a decrease in the expression of CD25. In aggregate, these responses suggest that LMS may act as an immunostimulatory agent for one group, those with low basal IGCN, and as an immunosuppressive agent for the other. The concept that a single immunomodulatory agent, including LMS, can cause both immunostimulation or immunodepression in different settings or under different conditions, has been reported previously [19,20]. In an earlier study [10], LMS increased CD16 and CD16/ CD56 expression in normal volunteers. In this current report, overall CD16 was not increased and the increases in CD16 seen in the low basal IGCN subgroup were modest. However, since basal IGCN was not measured in the earlier study it is difficult to directly compare these results. The sample populations in the two studies are similar except for the history of colonic polyps in this current trial. There was no clinical evidence of basal immune suppression for either group. Overall, LMS had minimal immunostimulatory activity except in patients with low basal IGCN. In colon cancer patients, LMS has significant immunostimulatory activity when utilized in conjunction with five-fluorouracil in the adjuvant treatment of patients with stage III disease [21]. This may be because cancer patients, especially those on chemotherapy, have significant immune system suppression [22]. Individuals with low basal IGCN may represent a subset of the population with some degree of immune suppression whose immune systems are primed to respond positively to LMS. Because of the modest degree of immunostimulation
97
noted with LMS, and concerns that subjects with intact immune function may actually experience suppression of cellular immune activity, LMS may not be an optimal agent for most patients with a history of colonic polyps. It should be avoided in patients with normal immune function but may be a useful immunostimulatory agent for cancer patients, other patients with impaired cellular immune responses, or for patients with familial polyposis syndromes who have documented low basal IGCN values. This study reports for the first time the use of a novel assay, IGCN, which defines two subgroups of individuals and, specifically, may identify subjects with impaired basal cellular immune function. IGCN may be a useful surrogate endpoint biomarker (SEB) for cancer prevention trials with immunomodulatory agents, especially in immunocompromized populations. If LMS is utilized in future trials, IGCN will be an important assay to assess its effects on patient immune function.
References [1] Moertel CG, Fleming TR, McDonald JS, Haller DG, Laurie JA, Goodman PJ, et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. New Engl J Med 1990;322:352–8. [2] Laurie JA, Moertel CG, Fleming TR, Wieand HS, Leigh JE, Rubin J, et al. Surgical adjuvant therapy of large-bowel carcinoma: an evaluation of levamisole and the combination of levamisole and fluorouracil. J Clin Oncol 1989;7:1447–56. [3] Holcombe RF, Stewart RM, Betzing KW, Kannan K. Alteration in lymphocyte phenotype associated with administration of adjuvant levamisole and 5-fluorouracil. Cancer Immunol Immunother 1994;34:394–8. [4] DeBrander M, Vandebroek J, Wassenaar H, De Cree J, Braisier A, Demoen B, et al. Immunological alterations induced by adjuvant treatment of postoperative colon carcinoma Duke’s B or C with levamisole in combination with 5-FU. Anticancer Res 1995;15:2271–7. [5] Janik J, Kopp WC, Smith JW, Longo DL, Alvord G, Sharfman WH, et al. Dose-related immunologic effects of levamisole in patients with cancer. J Clin Oncol 1993;11:125–35. [6] Clarke GR, Burton RC, Smart YC. The antitumor effect of levamisole in mice are mediated by NC-1.1+ cells. Cancer Immunol Immunother 1997;45:115–8. [7] Goodrich KH, Alvarez X, Holcombe RF. Effect of levamisole on major histocompatibility complex (MHC) class I expression in colorectal and breast carcinoma cell lines. Cancer 1993;72:225–30. [8] Kimball ES, Fisher MC. Levamisole effects on major histocompatibility complex and adhesion molecule expression and on myeloid cell adhesion to human colon tumor cell lines. J Natl Cancer Inst 1996;88:109–16. [9] Holcombe RF, Stewart RM. Augmentation of IL12-dependent Th1 cytokine production by levamisole. Allergy Clin Immunol Int 1997;9:119–20. [10] Holcombe RF, Milovanovic T, Stewart RM, Brodhag TM. Investigating immunomodulation for colon cancer prevention: results of an in vivo dose escalation trial of levamisole with immunologic endpoints. Cancer Detect Prev 2001;25:183–91. [11] Biological therapies for cancer: questions and answers in NCI on-line, August 16, 2004. http://cis.nci.nih.gov/fact/7_2.htm. [12] Cancer facts and figures, 2004 at American Cancer Society Inc., 2004. http://www.cancer.org/. [13] Jacobson-Brown P, Neuman MG. Colon polyps and cytokines: emerging immunological mechanisms. Rom J Gastroenterol 2003;12:207–14.
98
R.F. Holcombe et al. / Cancer Detection and Prevention 30 (2006) 94–98
[14] Robbins TG, Weinstein RJ, Demers RY. Lymphocytopenia, T-lyphocyte subsets, and colorectal polyps in automotive pattern and model markers. J Occup Med 1991;33:510–5. [15] Murphy KM, Reiner SL. The lineage decisions of helper T-cells. Nat Rev Immunol 2002;2(12):933–44. [16] Dredge K, Marriott JB, Todryk SM, Dalgleish AG. Adjuvants and the promotion of Th1-type cytokines in tumour immunotherapy. Cancer Immunol Immunother 2002;51(10):521–31. Epub September 20, 2002. [17] Smyth MJ, Thia KY, Street SE, Cretney E, Trapani JA, Taniguchi M, et al. Differential tumor surveillance by natural killer (NK) and NKTcells. J Exp Med 2000;191:661–8. [18] Martiniello R, Burton RC, Smart YC. Ly-6C+ natural T (NT) cells mediate immune surveillance against NK-sensitive and NK-resistant
[19]
[20] [21]
[22]
transplantable tumours in certain strains of mice. Int J Cancer 1996;66:532–7. Gozzo JJ, Hanson RH. Varied in vivo properties of levamisole (LMS) on the immune response in mice. Transpl Proc 1979;11: 1420–1. Bliznakov EG. Immunostimulation or immunodepression? Biomedicine 1977;26:73–6. Holcombe RF, Jacobson J, Dakhil SR, Stewart RM, Betzing KS, Kannan K, et al. Association of immune parameters with clinical outcome in stage III colon cancer: results of Southwest oncology group protocol 9009. Cancer Immunol Immunother 1999;48:533–9. Block KI, Boyd B, Gonzalez N, Vojdani A. The immune system in cancer. Integr Cancer Ther 2002;1:294–316.