- Email: [email protected]
Stem cell mobilization and harvesting by leukapheresis alters systemic cytokine levels in patients with multiple myeloma
Cytotherapy, 2013; 0: 1e11
Stem cell mobilization and harvesting by leukapheresis alters systemic cytokine levels in patients with multiple myeloma KNUT ANDERS MOSEVOLL1,4, ÇIGDEM AKALIN AKKÖK2, TOR HERVIG2,3, 2 1,2,4 & HÅKON REIKVAM1,2,4 GURO K. MELVE , ØYSTEIN BRUSERUD 1
Department of Medicine, Haukeland University Hospital, Bergen, Norway, 2Department of Immunology and Transfusion Medicine, Haukeland University Hospital, Bergen, Norway, 3The Gades Institute, University of Bergen, Bergen, Norway, and 4Institute of Medicine, Section of Hematology, University of Bergen, Bergen, Norway
Abstract Background aims. Stem cell mobilization and harvesting by peripheral blood leukapheresis in patients with myeloma can alter plasma levels of certain cytokines. In the present study, we investigated the effects of these interventions on a larger group of cytokines. Methods. Plasma cytokine levels were determined in 15 patients with myeloma who were undergoing peripheral blood stem cell harvesting, and we compared the patients with healthy donors who were undergoing platelet apheresis. Results. Several cytokines showed altered levels in patients with myeloma when examined after chemotherapy plus granulocyte colony-stimulating factoreinduced stem cell mobilization. The most striking effect was increased levels of several CCL (CCL2/3/4) and CXCL (CXCL5/8/10/11) chemokines as well as increased thrombopoietin, interleukin 1 receptor antagonist, interleukin-4, granulocyte colony-stimulating factor and hepatocyte growth factor. Stem cell harvesting by apheresis altered the plasma levels of several mediators (CD40 ligand, interleukin 1 receptor antagonist, CCL5 and CXCL5/ 8/10/11). Apheresis in patients with myeloma had divergent effects on these chemokine levels, although they were all still significantly higher than for healthy individuals. Thrombapheresis in healthy individuals had only minor effects on plasma cytokine levels. Stem cell graft supernatants showed high levels of several cytokines, especially CCL and CXCL chemokines. Analyses of chemokine profiles in pre-apheresis plasma and graft supernatants suggested that such profiling can be used to detect prognostically relevant differences between patients. Conclusions. Our results demonstrate that patients with myeloma have an altered cytokine network during stem cell mobilization, and the network is further altered during stem cell harvesting by leukapheresis. These treatment- or procedure-induced alterations involve several mediators known to affect myeloma cell proliferation, migration and survival. Key Words: cytokines, multiple myeloma, stem cell harvesting, thrombapheresis
Introduction High-dose chemotherapy followed by autologous stem cell support is used in the treatment of several malignancies, including multiple myeloma. The stem cells used in this treatment are often harvested from the peripheral blood after mobilization by growth factors, eventually combined with chemotherapy (1,2). This therapeutic strategy is commonly used in younger patients with myeloma. However, it should be emphasized that (i) autotransplantation is used as a part of a treatment strategy consisting of initial induction chemotherapy; (ii) patients who respond or achieve disease stabilization after this treatment then continue to stem cell mobilization with initial chemotherapy followed by administration of granulocyte colony-stimulating factor (G-CSF); (iii) this is followed by stem cell harvesting by
an apheresis procedure, and finally (iv) patients receive high-dose chemotherapy followed by autotransplantation (3e5). The stem cell harvesting is the only step in this regimen that does not include myeloma-directed therapy. The cytokine network has been studied both in the bone marrow and peripheral blood of patients with myeloma (6,7), and certain single cytokines (ie, hepatocyte growth factor [HGF], vascular endothelial growth factor [VEGF], basic fibroblast growth factor [bFGF], CCL3 and interleukin [IL]6) have then shown prognostic correlation (8e12). In a recent study we described that stem cell harvesting altered plasma levels of angioregulatory cytokines, which also are important in regulation of myeloma cell proliferation (3). The aim of the
Correspondence: Dr Øystein Bruserud, Section for Hematology, Institute of Medicine, University of Bergen, Department of Medicine, Haukeland University Hospital N-5021 Bergen, Norway. E-mail: [email protected] (Received 4 June 2012; accepted 9 February 2013) ISSN 1465-3249 Copyright Ó 2013, International Society for Cellular Therapy. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcyt.2013.02.008
2
K. A. Mosevoll et al.
present study was to investigate whether peripheral blood stem cell harvesting by apheresis causes a more extensive alteration of the systemic cytokine network. We investigated the plasma cytokine profile (30 cytokines examined) for patients with myeloma undergoing peripheral blood stem cell harvesting, and the effect of apheresis in the patients was compared with the effect of thrombapheresis in healthy individuals. Methods Study population The study was approved by the local ethics committee (University of Bergen, Norway). Patients and donors were included after written informed consent was given. Fifteen consecutive patients with myeloma (seven men and eight women; median age, 57 years; range, 44e67 years) with newly diagnosed symptomatic disease and 15 healthy volunteer blood donors (seven women and eight men; median age, 46 years; range, 26e62 years) were investigated (peripheral blood plasma samples from 10 and platelet concentrate samples from 15 healthy donors). Apheresis procedures Hematopoietic stem cell mobilization and harvesting. The study included one leukapheresis for each patient, and this was always the first one if the patients had to be harvested several times. All patients received disease-stabilizing induction chemotherapy with either (i) two cycles of intravenous cyclophosphamide 1 g/m2 on day 1 plus oral dexamethasone 40 mg on days 1e4 and 9e12, with 4-week intervals (14 patients), or (ii) two cycles with bortezomib 1.3 mg/m2 on days 1, 4, 8 and 11 and dexamethasone 40 mg orally on days 1e4 and 9e12, with 3-week intervals (one patient). After this treatment, three patients showed complete response, four patients showed partial response, and eight patients had stable disease. All patients received stem cell mobilization with cyclophosphamide 2 g/m2 followed by G-CSF 5 mg/ kg per day. The first apheresis was performed 10 days (12 patients) or 11 days (3 patients) after start of cyclophosphamide. The median peripheral blood concentrations of CD34þ cells at the time of apheresis was 39.9 103/mL (range, 9.7e175), and the median yield of CD34þ cells was 5.3 106/kg (range, 1.1e27.9). Harvesting, cryopreservation and thawing of peripheral blood stem cell grafts were performed as described in detail previously (4,5). Version 7.0 of the Cobe Spectra cell separator (Cobe Laboratories,
Gloucester, United Kingdom) was used, and largevolume leukapheresis with processing of 4 times the patients’ total blood volume was performed. A combination of acid citrate dextrose (ACD-A) (Baxter FA, Lessin, Belgium) at a blood to citrate ratio of 18:1 and 2500 IU heparin/500 mL of ACD-A solution (Leo Pharma AS, Oslo, Norway) was used for anticoagulation. Moreover, ACD-A equivalent to 10% of the estimated harvest volume was added initially to the collection bag. The total leukocyte counts and platelet counts did not differ between the healthy control subjects and patients with myeloma examined immediately before stem cell mobilization, whereas the hemoglobin levels were still decreased in the patients (median, 10.1 g/100 mL; range, 6.8e12.0, P < 0.0005) compared with the control subjects (median, 14.4; range, 12.4e15.8). The median peripheral blood platelet counts were within the normal range both at the start of mobilization and at the time of harvesting, whereas there was an expected increase in total peripheral blood leukocyte counts from the start of mobilization chemotherapy (median, 5.4 109/L; range, 2.5e9.9, P ¼ 0.003) until stem cell harvesting (median, 11.0; range, 2.7e44.6). Preparation of platelet concentrates. Single-donor platelet concentrates were obtained from the 10 healthy blood donors with median total blood volume of 5.6 L (range, 4.3e7.0 L). The donors had not used anti-platelet drugs the last 5 days before donation. The collection was performed with the Fenwal Amicus Cell Separator (Baxter Healthcare Corp, Deerfield, IL, USA), which uses the elutriation principle to provide leukocyte-reduced platelet concentrates. ACD-A was used as anticoagulant at a ratio of 1:9.
Analysis of cytokine plasma levels Preparation of plasma samples. Blood samples were collected immediately before and after apheresis and 24 h after initiation of apheresis. Samples were drawn from the central venous catheter in patients and from the arm not used during apheresis in the platelet donors. Samples were also taken from stem cell autografts and platelet concentrates. Blood samples were collected into Vacutainer 9NC tubes (BectonDickinson, San Jose, CA, USA) with citric acid as anticoagulant. Samples from platelet concentrates and stem cell autografts were transferred to plastic tubes without additive and centrifuged twice at 200g for 15 min at room temperature. Centrifugation was started within 30 min from sampling. The supernatants were thereafter transferred to cryotubes, frozen immediately and stored at 70 C until analyzed.
Cytokine modulation during apheresis Cytokine analyses. Cytokine levels were determined by Luminex analyses (R&D Systems; Abingdon, United Kingdom) and included (i) the interleukins IL1a, IL1b, interleukin 1 receptor antagonist (IL1RA), IL2, IL4, IL5, IL6, IL7, IL8/CXCL8, IL10, IL12, IL13, IL17 and IL22; (ii) the chemokines CCL3/4/5/ 11 and CXCL5/10/11; (iii) bFGF, G-CSF, granulocyte macrophage colony-stimulating factor (GMCSF), VEGF, thrombopoietin (TPO), epithelial growth factor, HGF and leptin; and (iv) the immunomodulatory cytokines interferon-g (IFN-g), CD40 ligand (CD40L), tumor necrosis factor-a (TNFa) (13). In addition, CXCL12 and soluble CXCR4 were determined by means of enzyme-linked immunosorbent assay (ELISA) analyses (CXCL12; Quantikine ELISA kit; R&D Systems. CXCR4; soluble Human CXC-chemokine Receptor 4 ELISA Kit; Cusabio Biotech Co Ltd, Wuhan, China). All analyses were performed in duplicate, strictly according to the manufacturer’s instructions. Bioinformatic and statistical analyses Bioinformatic analyses were performed with the use of the J-Express (MolMine AS, Bergen, Norway) (14). For hierarchical clustering, all values were median variance standardized and log2-transformed. The complete linkage was used as linkage method, and for distance measured the Euclidian correlation was used. Additional statistical analyses were performed with the use of the GraphPad Prism 5 (Graph Pad Software, Inc, San Diego, CA, USA). Pearson correlation for bivariate samples was used for correlation analyses, the Wilcoxon signed rank test was used to compare different groups, and values of P < 0.05 were regarded as statistically significant (13,14). Results Patient symptoms, findings and disease development At the time of diagnosis, the patients had a median hemoglobin level of 11.8 g/100 mL (range, 9.2e14.9 g/100 mL), median serum calcium of 2.39 mmol/L (1.75e3.02 mmol/L) and median serum kreatinin of 76 mmol/L (29e1024 mmol/L). The monoclonal components were Immunoglobulin G for eight patients (two of them also having detectable levels of monoclonal light chains); immunoglobulin A for two patients, and five patients had light chain disease. Osteolytic lesions were detected in nine patients at the time of diagnosis. The following symptoms/signs were regarded as indications for treatment: anemia, nine patients (one having general bone marrow failure); renal failure, five patients; skeletal pain, five patients; hypercalcemia, three patients; spinal cord
3
compression, three patients; and polyneuropathy, two patients. All patients received high-dose melphalan followed by autologous stem cell rescue. After this treatment, three patients achieved a complete response, seven patients achieved a partial response and five patients had disease stabilization. The median follow-up of the patients is now 62 months (range, 50e64 months), and three patients have not yet had progression of their disease after 53, 55 and 64 months, respectively. Eleven patients received anti-myeloma treatment for disease progression; one patient died of other diseases before myeloma progression, and this patient was not included in the statistical analyses of time until progression. Pre-apheresis plasma cytokine levels in patients with myeloma and in healthy individuals Cytokine levels were determined in plasma samples derived immediately before leukapheresis for all patients with myeloma (Table I), and these results were compared with the levels in 10 healthy individuals. The last results are summarized in Table II and are presented in detail in Figure 1. CXCR4, the CCL and the CXCL chemokines (CCL2-5 and CXCL5/8/10/11/12) showed all detectable levels in the patients. The majority of these chemokines were increased in patients, the only exceptions being CCL5 and CCL11, which did not differ. The most significant increases were seen for CCL4 (P < 0.0001), CXCL10 (P ¼ 0.0043) and CXCL11 (P ¼ 0.0025). CXCR4 and CXCL12 were only determined in the patients. In contrast to the chemokines, all the interleukins (IL1a, IL1b, IL2, IL4, IL6, IL10, IL12, IL13 and IL17) showed relatively low or undetectable levels that did not differ between patients and healthy control subjects except for IL4, which showed an increase (Table II, right panel). Finally, the antagonistic IL1RA showed significantly increased levels in patients compared with healthy control subjects (Figure 1). The growth factors GM-CSF and epithelial growth factor showed undetectable levels, whereas leptin levels were relatively high but with no differences between the groups (Table II). In contrast, TPO levels were significantly increased in the patients (P ¼ 0.0091) but with no correlation to the peripheral blood platelet count. Finally, the immunomodulatory cytokines differed; CD40L levels were significantly increased in the patients (P ¼ 0.0088), whereas IFNg showed undetectable levels. The patients received G-CSF for stem cell mobilization, and, as expected, the G-CSF plasma levels were considerably higher for patients than for control subjects. Even though there was a considerable variation in the preapheresis G-CSF levels among
4
K. A. Mosevoll et al.
Table I. Plasma cytokine profiles in patients with myeloma: A comparison between levels in samples derived immediately before and immediately after stem cell harvesting by leukapheresis. Mediator Chemokines CCL2 CCL3a CCL4a CCL5 CCL11a CXCL5a CXCL8a CXCL10a CXCL11a CXCL12 CXCR4 Interleukins IL1-b IL1RAa IL2 IL4 IL10 IL17 Growth factors G-CSFa TPO Leptin HGF Immunoregulatory cytokines CD40La TNFa
Before apheresis (pg/mL)
After apheresis (pg/mL)
P value
203 194 82 3313 33 150 3.2 406 80 1730 65
(123e566) (20e347) (37e160) (660e17 598) (16e278) (33e288) (nde17.4) (75e2456) (29e340) (916e2507) (nde120)
250 202 81 1768 42 73 5.4 1565 60 1576 45
(101e434) (nde285) (37e186) (780e14 914) (16e318) (33e176) (nde22.5) (25e2874) (13e300) (735e2373) (nde177)
Ns Ns Ns 0.0002 Ns 0.0011 0.034 0.0004 0.0076 Ns Ns
1.0 3723 5.8 20.8 2.0 12.4
(nde16.6) (765e8184) (nde36.9) (8e24.3) (nde6.8) (2e16.7)
3.0 4510 6.7 16,7 2.0 11
(0.5e12.2) (906e11 896) (nde32.1) (8e32.2) (nde7.3) (2e18)
Ns 0.0003 Ns Ns Ns Ns
(4781e32 072) (32e687) (3090e83 293) (1186e44 422)
0.0181 Ns Ns 0.0008
18 366 477 19 956 327
(9861e46 314) (106e1147) (3695e71 109) (40e21 685)
255 (57e1857) 2.0 (nde30.4)
13 472 489 20 817 15 855
191 (39e944) nd (nde31.5)
0.009 Ns
The table represents the results for 15 patients. All results are presented in pg/mL and as the median levels and with the corresponding variation ranges in parentheses. The table presents P values for the comparison between patient levels before and immediately after apheresis. The results for cytokines with undetectable median levels in both samples (Interferon-g, epithelial growth factor and granulocyte macrophage colony-stimulating factor) or undetectable levels in all samples (IL1a, IL6, IL12 and IL13) are not included. nd, not detectable; Ns, not significant; IL, interleukin; IL1RA, interleukin 1 receptor antagonist; G-CSF, granulocyte colony-stimulating factor; TPO, thrombopoietin; HGF, hepatocyte growth factor; CD40L, CD40 ligand; TNFa, tumor necrosis factor-a. a Cytokines with significantly increased levels in peripheral blood stem cell grafts compared with platelet concentrates.
patients, there was no correlation between G-CSF levels and the amount of harvested CD34þ cells. The preharvesting cytokine profile identifies two major patient subsets The overall preharvesting cytokine data were analyzed by hierarchical clustering; two major patient clusters/subsets were then identified including the lower six and the upper nine patients, respectively (Supplementary Figure 1). These two patient clusters differed especially with regard to the cytokine cluster consisting of leptin, CXCL10, IL2 and IL1b, and, at both time points, HGF, IL10, IL17 and CCL3 also differed between the two subsets. Thus, this classification of patients into two major clusters is mainly determined by a limited number of cytokines showing significant differences between these two subsets. The patient clustering was identical when tested immediately after and 24 h after leukapheresis, and the same clusters were identified when distance metrics were based on the Euclidian
correlation tests (data not shown). The median time until disease progression for the two clusters was 21 (upper cluster) and 43 months (lower cluster), respectively, but this difference was not significant. The upper cluster could be further divided into two patient subsets, but also when comparing the three different clusters (upper 5, middle 4 and lower 6 patients), no significant difference in survival was detected between among groups. Effects of stem cell harvesting on plasma cytokine levels: Analyses of single cytokines We determined plasma cytokine levels immediately after the stem cell harvesting, and several cytokines showed altered levels compared with pre-apheresis levels (Table I and Figure 2). Chemokines. The chemokines CCL5, CXCL5 and CXCL11 showed decreased levels immediately after apheresis, whereas CXCL8 and CXCL10 levels were increased. The differences were still significant after
Cytokine modulation during apheresis
5
Table II. Systemic cytokine profiles in plasma samples: A comparison between patients with myeloma immediately before apheresis and healthy control subjects. Mediator Chemokines CCL2 CCL3a CCL4a CCL5 CCL11a CXCL5a CXCL8a CXCL10a CXCL11a Interleukins IL1-b IL1RAa IL2 IL4 IL10 IL17 Growth factors G-CSFa TPO Leptin HGF Immunoregulatory cytokines CD40La TNFa
Myeloma (pg/mL)
Healthy (pg/mL)
203 194 82 3313 33 150 3.2 406 80
(123e566) (20e347) (37e160) (660-17 598) (16e278) (33e288) (nde17.4) (75e2456) (29e340)
149 126.5 14 2318 37 62 nd 71 30.5
1.0 3723 5.8 20.8 2.0 12.4
(nde16.6) (765e8184) (nde36.9) (8e24.3) (nde6.8) (2e16.7)
1.15 207 2.75 12 0.2 7.6
18 366 477 19 956 327
(9861e46 314) (106e1147) (3695e71 109) (40e21 685)
255 (57e1857) 2.0 (nde30.4)
20 238 16 362 44
P value
(37e321) (nde188) (nde34) (1191e6189) (16e176) (30e182) (nde4.8) (17e366) (3.6e120)
0.0285 0.0134 <0.0001 Ns Ns 0.0264 0.0418 0.0043 0.0025
(nde10.8) (156e353) (nde23.9) (nde21.4) (nde2.3) (3e14.5)
Ns <0.0001 Ns 0.0143 Ns Ns
(12e34) (142e437) (6045e64 679) (0e211)
<0.0001 0.0091 Ns 0.001
163 (45e230) nd (nde4.5)
Ns Ns
The table represents the results for 15 patients and 10 healthy control individuals. All results are presented in pg/mL and as the median levels and with the corresponding variation ranges in parentheses. The table presents P values for the comparison between patient levels before leukapheresis and levels in healthy individuals. The results for cytokines with undetectable median levels in all three samples (interferon-g, epithelial growth factor and granulocyte macrophage colony-stimulating factor) or undetectable levels in all samples (IL1a, IL6, IL12 and IL13) are not included. nd, not detectable; Ns, not significant; IL, interleukin; IL1RA, interleukin 1 receptor antagonist; G-CSF, granulocyte colony-stimulating factor; TPO, thrombopoietin; HGF, hepatocyte growth factor; CD40L, CD40 ligand; TNFa, tumor necrosis factor-a. a Cytokines with significantly increased levels in peripheral blood stem cell grafts compared with platelet concentrates.
24 hours only for CCL5 (median, 1613 pg/mL; range, 817e17 206 pg/mL, P ¼ 0.0026) and CXCL5 (78 pg/ mL, 34e225 pg/mL, P ¼ 0.017), and an additional decrease in CCL4 levels was detected after 24 h (67 pg/ mL, 5e115 pg/mL, P ¼ 0.029) compared with the preapheresis levels (82 pg/mL, 37e160 pg/mL). Interleukins. The interleukin levels were not significantly altered either immediately or 24 h after stem cell harvesting. In contrast, IL1RA levels were increased immediately after harvesting. Growth factors. HGF showed an increase immediately after apheresis, but was normalized after 24 hours. Leptin showed increased plasma levels after 24 h (median, 31 773 pg/mL; range, 5520e86 892 pg/ mL; P ¼ 0.015) compared with pre-apheresis levels (19 956 pg/m L, 3695e71 109 pg/mL). The G-CSF levels showed an expected decrease both immediately (13 472 pg/mL, 4781e32 072 pg/mL, P ¼ 0.018) and 24 h (5670 pg/mL, 347e26 009 pg/mL, P < 0.0001) after stem cell harvesting compared
with the pre-apheresis levels (18 366 pg/mL, 9861e46 314 pg/mL). Immunomodulatory cytokines. CD40L levels were decreased immediately after apheresis but normalized within 24 h, whereas IFNg and TNFa levels remained low or undetectable. Thus, harvesting of peripheral blood stem cells alters systemic cytokine levels, especially chemokine levels, and for some cytokines these alterations persist for >24 h. Plasma chemokine profiles in patients with myeloma before and after stem cell harvesting In contrast to most other cytokines, the chemokines (i) showed detectable pre-apheresis plasma levels for most patients; (ii) altered preharvesting levels were observed in patients with myeloma compared with healthy individuals; and (iii) mainly chemokine levels were altered by leukapheresis. For these reasons, we focused on the chemokine levels in the further
6
K. A. Mosevoll et al.
Figure 1. Pre-apheresis peripheral blood plasma cytokine levels in patients with myeloma and healthy individuals. We investigated the levels for 30 soluble mediators; the figure presents the levels for the cytokines (CCL2-4, CXCL5/8/10/11, IL1RA4, thrombopoietin [TPO], granulocyte colony-stimulating factor [G-CSF] and hepatocyte growth factor [HGF]) that showed statistically significant differences between the two groups. The results are presented as the cytokine level for each individual patient (P) and control subject (C), the mean level is indicated in the figure, and the cytokine and the corresponding P-value are indicated at the top of the figure. Undetectable levels are presented as open symbols.
detailed analyses of our results. We then compared chemokine plasma profiles for all 15 patients before and after apheresis (Figure 3, upper left); the plasma profile was altered by the apheresis, and this was reflected in the chemokine clustering. However, the three chemokines CXCL5/CCL5/CXCL11, which represent the only chemokines with decreased plasma levels immediately after leukapheresis, clustered together both before and after the procedure. The clustering after 24 h was similar to the preharvesting clustering (data not shown). Taken together, these observations suggest that even though stem cell harvesting alters the serum levels of individual cytokines, the effects on the chemokine profile and clustering are limited. Finally, both the pre-apheresis and postapheresis analyses identified two major patient
clusters, but the time until disease progression did not differ between these patient subsets. Cytokine levels in supernatants from peripheral blood stem cell grafts and platelet concentrates: The chemokine profile identifies a patient subset with prolonged time until progression We compared the supernatant levels of individual cytokines for stem cell grafts from patients with myeloma and platelet concentrates from healthy individuals (see Table III, supplementary information and Supplementary Figures 3 and 4). The most important difference between stem cell grafts and platelet concentrates was altered levels for several chemokines, and, for this reason, we investigated the
Cytokine modulation during apheresis
7
Figure 2. Peripheral blood plasma cytokine levels before and immediately after stem cell harvesting from multiple patients with myeloma. We investigated the levels for 30 soluble mediators, and the figure presents the levels for the cytokines (CCL5, CXCL5/8/10/11, interleukin 1 receptor antagonist [IL1RA], granulocyte colony-stimulating factor [G-CSF], hepatocyte growth factor [HGF] and CD40L) that showed statistically significant differences before and after apheresis. The results are presented as the cytokine level for each individual patient and control subject, the mean levels are indicated, and the cytokine and the corresponding P-value are indicated at the top of the figure. Mean values are marked in bold.
chemokine profile of the peripheral blood stem cell grafts more in detail (Figure 3, lower left). The graft chemokine profile and clustering differed from the peripheral blood plasma clustering (Figure 3, upper versus lower left), suggesting that the graft chemokine profile is affected by the harvesting procedure and does not simply reflect the plasma chemokine profile. Furthermore, two major patient clusters were then identified; a lower cluster with five patients (including the three patients without disease progression after 53, 55 and 64 months) and an upper cluster with 10 patients including the one patient who died early after transplantation without disease progression. The five patients in this lower cluster had significantly longer time until disease progression (Figure 3, lower right; Gehan-BreslowWilcoxon test, P ¼ 0.0108) than the other patients, and these five patients clustered either in the same main cluster or close to each other (Figure 3, upper two figures) when analyzing preharvesting and postharvesting chemokine plasma levels. However, the observation that these five patients did not cluster in the same main cluster when analyzing the plasma chemokine cluster (Figure 3, upper part), further
supports our hypothesis that the graft chemokine profile reflects additional biological characteristics of prognostic relevance and does not simply mirror the plasma chemokine profiles. Discussion The aim of our study was to investigate whether apheresis procedures induce alterations in the systemic cytokine levels. We investigated a wide range of immunoregulatory and hematopoiesisregulating cytokines, and we compared cytokine profiles for stem cell harvesting in patients with myeloma and platelet harvesting in healthy blood donors. Our studies demonstrate that apheresis procedures can alter systemic cytokine profiles, but the magnitude and duration of these alterations differ between our two groups. We examined a group of relatively young, consecutive patients with myeloma. All patients had symptoms at the start of induction therapy, but all of them responded to the treatment or had stabilization of the disease. The patients were mobilized with chemotherapy followed by G-CSF treatment; for
8
K. A. Mosevoll et al.
Figure 3. Chemokine clusters in multiple patients with myeloma. (upper part of figure) The peripheral blood plasma chemokine profile for patients with myeloma before and immediately after apheresis is shown. We investigated the levels of 10 chemokines and CXCR4, and we used unsupervised hierarchical cluster analysis to characterize the chemokine profile. The figure presents the results for patients with myeloma before (upper left) and immediately after (upper right) stem cell harvesting. The five patients that formed a separate cluster with prolonged time until disease progression in the analysis of graft chemokine profiles (see lower part of the figure) are marked with blue in the right part of the figures. (lower part of figure) The chemokine profile for peripheral blood stem cell autografts derived from patients with myeloma is shown. The levels of the same 10 chemokines and CXCR4 were determined, and unsupervised hierarchical cluster analysis was used to characterize the chemokine profile. The figure presents the results for the autografts derived from the 15 patients. Two different clusters were identified (left), and patients in cluster 1 (blue) showed a significantly longer time until myeloma progression than the other patients (lower right).
this reason, it was not possible to have a control group of healthy individuals receiving an identical mobilization treatment, and our patients were therefore compared with healthy donors undergoing thrombapheresis. An extensive cytokine network modulation was only seen for the patients with myeloma but not for the platelet donors; we therefore conclude that this apheresis-induced modulation is not a characteristic for all types of apheresis but depends on the clinical context. This difference between patients with myeloma and healthy individuals can be caused by (i) initial chemotherapy before mobilization or residual myeloma disease, (ii) cyclophosphamide followed by hematopoietic regeneration or (iii) the G-CSF treatment. In our opinion, it is not possible to use control groups to clarify the relative contribution of each of these three factors; even G-CSF mobilization in healthy individuals will differ because the kinetics of stem cell mobilization will differ and the period of G-CSF treatment will usually be shorter. However, previous studies have demonstrated that G-CSF therapy alone will increase HGF and VEGF levels (15,16), and this treatment may be at least partly responsible for the increased preapheresis levels of these two cytokines.
In this unique clinical context, especially the HGF levels but not the VEGF levels are further increased by the leukapheresis in patients with myeloma (3). High systemic levels of HGF, VEGF, bFGF, CCL3 and IL6 as well as the soluble IL6 receptor have all been linked to an adverse prognosis in patients with multiple myeloma (7e12). However, to the best of our knowledge, the possible prognostic impact of more extensive cytokine profiles has not been investigated in previous studies. Furthermore, several chemokines showed increased plasma levels in patients with myeloma before stem cell harvesting. These high levels are probably caused by a combined effect of the disease and the recent cyclophosphamide treatment followed by G-CSFesupported hematopoietic regeneration. Many chemokines showed increased levels before apheresis and/or were altered by the leukapheresis, and these chemokines can bind to the myeloma-expressed receptors CCR1 (CCL3, CCL5), CCR2 (CCL2), CCR5 (CCL4), CXCR1 (CXCL8) and CXCR2 (CXCL5, CXCL8) (17e26). All these receptors appear to be involved in the pathogenesis of multiple myeloma (17e23), and many of them can also bind several ligands, emphasizing the importance of investigating chemokine
Cytokine modulation during apheresis
9
Table III. Supernatant cytokine levels for peripheral blood stem cell grafts prepared from multiple patients with myeloma and for platelet concentrates derived from healthy blood donors. Mediator
Patients with myeloma PBSC autograft
Chemokines (pg/mL) CCL2 58 CCL3 193 CCL4 88 CCL5 22 519 CCL11 50 CXCL5 625 CXCL8 2.5 CXCL10 703 CXCL11 156 Interleukins (pg/mL) 4.1 IL1-b IL1RA 10 493 IL2 9.2 IL4 17.2 IL10 1.2 IL17 9.6 Growth factors (pg/mL) G-CSF 12 906 TPO 483 Leptin 16 943 EGF nd HGF 24 427 Immunoregulatory cytokines (pg/mL) CD40L 1185
Healthy individuals platelet concentrate
(41e308) (20e314) (25e298) (13 588ehigh) (16e284) (121e2598) (nde23.9) (107e2922) (23e432)
60 110 28 26 659 17 3977 16 8 38.5
(1e15.7) (2731e26 000) (nde36.8) (8e32.1) (nde5.3) (2e20)
P value
(29e80) (nde181) (nde88) (18 701ehigh) (nde60) (1886e5383) (nde30.7) (5e60) (13e196)
Ns 0.024 0.0008 Ns 0.0035 <0.0001 0.019 <0.0001 0.0025
2.05 120.5 5.45 20.8 0.1 8.95
(0e9.3) (65e188) (nde18.3) (12.2e24.4) (nde2.1) (2e15.6)
Ns <0.0001 Ns Ns Ns Ns
(8863e41 139) (233e828) (3207e70 527) (nde36.9) (2062e50 000)
20 393 5010 49.1 72
(10e29) (170e530) (2028e65 872) (8.4e103) (18e211)
<0.0001 Ns Ns Ns <0.0001
(192e2877)
7025 (3615e15 246)
<0.0001
The table represents the results for 15 patients and 10 healthy control individuals. All results are presented in pg/mL and as the median levels and with the corresponding variation ranges in parentheses. The table presents the cytokine levels for peripheral blood stem cell grafts and platelet concentrates. The results for cytokines with undetectable median levels in both samples (interferon-g, epithelial growth factor and granulocyte macrophage colony-stimulating factor) or undetectable (nd, not detectable) levels in all samples (IL1a, IL6, IL12 and IL13) are not included. CXCL12 and CXCR4 were only determined in the patients and are not included in the table (nd, not detectable; high, above detection range, two and three samples, respectively). PBSC, peripheral blood stem cell Ns, not significant; IL, interleukin; IL1RA, interleukin 1 receptor antagonist; G-CSF, granulocyte colonystimulating factor; EGF, epithelial growth factor; TPO, thrombopoietin; HGF, hepatocyte growth factor; CD40L, CD40 ligand.
profiles and not only single chemokine levels. However, exceptional chemokines showed apheresisinduced decreases in plasma levels (CCL5, CXCL5, CXCL11), whereas others showed increased postapheresis levels (CXCL8, CXCL10). Previous studies have also shown a strong increase in the proangiogenic myeloma growth factor HGF after leukapheresis (3). Finally, cyclophosphamide may be given to our patients as a part of the mobilization regimen, but it is also a well-known anti-myeloma drug (27). It is not known whether its anti-myeloma effect during mobilization is clinically relevant. Thus, it is difficult to predict the final effect of cyclophosphamide-based mobilization and modulation of the cytokine/chemokine network on myeloma cell growth and survival during the period of stem cell mobilization and harvesting. This issue is further complicated by variation of myeloma cell expression of chemokine receptors between patients and even within the same patient (26). Several chemokines that showed altered plasma levels either before or after apheresis may be important for the pathogenesis of multiple myeloma. First,
myeloma cells express both CXCR3A and CXCR3B receptors, and CXCR3A ligation by CXCL10 (and possibly also CXCL11) inhibits CD95-induced apoptosis and IL6-dependent myeloma cell proliferation (17). Second, another study confirmed that plasma cells express CXCR3 and in addition the CXCR1, CXCR2 (both being receptors for CXCL8, CXCR1 also binding CXCL5) and CCR2 (receptor for CCL2). Another study found that CXCL8 stimulated myeloma cell proliferation, whereas CXCL8 as well as CXCL11 (CXCR3 ligation) and CCL2 (CCR2 ligation) increased myeloma cell migration (18). Third, CXCL10 appears to increase the adherence of myeloma cells to bone marrow stromal cells (19). Finally, CXCR3 ligation by CXCL9, CXCL10 and CXCL11 induces downstream intracellular signalling leading to altered myeloma cell migration and increased release of matrix metalloproteinase (MMP)2 and MMP9 (20). Thus, most of these studies suggest that the apheresis-induced reduction in plasma chemokine levels would be expected to mediate anti-myeloma effects and thereby counteract the potential myeloma-
10
K. A. Mosevoll et al.
stimulatory effect of increased preapheresis chemokine levels and the previously described apheresis-induced HGF increase (21). CD40L levels were altered by stem cell harvesting. CD40 is involved in immunoregulation, and in vitro studies suggest that it can also induce apoptosis in myeloma cells (28e30). In addition, the modulation of TPO levels suggests that the regulation of normal hematopoiesis may also be affected by the apheresis (31). However, it should be emphasized that additional studies are required to clarify whether the variations have any clinical impact. We compared the chemokine profiles for patients with myeloma and for the healthy blood donors and found that the pre-apheresis chemokine profile differed between the two groups. There was also a difference in the effect of apheresis between patients and control subjects with regard to the overall chemokine profile, although the apheresis procedures had relatively small effects on the chemokine profiles both for patients and blood donors. Thus, the effect of apheresis on the chemokine network in these two groups is reflected mainly in the absolute chemokine levels. The overall chemokine profile was altered by apheresis; these effects were most clearly seen in patients with myeloma, whereas thrombapheresis in healthy individuals had only minor effects. Thus, the balance between cytokine effects in the systemic (plasma) cytokine network is altered by stem cell harvesting. Furthermore, the graft cytokine profile does not simply reflect the plasma profile, and very high levels of several cytokines are observed in the stem cell grafts together with an altered chemokine profile and clustering (see above). Reinfusion of the graft may therefore alter the posttransplant cytokine network, especially the chemokine balance, through infusion of cytokines together with the stem cells. We also investigated whether there was any association between chemokine levels and time until myeloma relapse. We strongly emphasize that these results should be interpreted with great precaution because our study included few patients. We observed that the analysis of chemokine profiles in the graft supernatants identified a subset of patients with significantly longer time to relapse compared with the other patients. These patients were all included in the same main patient subset when the preharvesting chemokine profile was analyzed. Our observations must be confirmed in larger prospective studies, but they suggest that analysis of cytokine profiles deserves further investigation in future clinical studies. However, one should not conclude that reinfusion of graft cytokines directly affects prognosis in autotransplanted patients with myeloma. In our opinion, it seems more likely that apheresis-induced modulation
of chemokine levels and prolonged time until relapse are two different parameters reflecting the same unique clinical/biological context in these patients. The cytokine network modulations described in this study are relatively-short-lasting, but, despite this, they may be of clinical relevance. A short-lasting initial event may be sufficient to initiate a biological process, and the hypothesis of angiogenic switch to initiate growth of cancer metastasis from dormant malignant cells is based on this (32). We have documented that the cytokine-modulating effects of apheresis depend on the clinical context; if the apheresis-induced alterations and/or the later reinfusion of cytokines together with the stem cells occur at critical time points, they may synergize with the specific clinical context and thereby become clinically relevant. Our observations illustrate that stem cell mobilization, harvesting and reinfusion have several additional and largely unknown effects that may predict or possibly affect the later course of the myeloma disease. To conclude, our study demonstrates that the stem cell harvesting by leukapheresis alters the systemic cytokine network in patients with myeloma, and our results also suggest that analysis of cytokine profiles (and especially chemokine profiles) may become a useful tool in the prognostic evaluation of these patients. However, the biological impact of this cytokine modulation and the possible clinical use of cytokine profiling need further evaluation in larger clinical studies. Previous studies have suggested that anti-myeloma treatment induces specific alterations that can be detected in bone marrow plasma (6), and it would be a clinical advantage if plasma profiling instead of bone marrow samples could be used in the follow-up of these patients. Acknowledgments We thank Kristin Rye Paulsen for technical assistance. This work was supported by the Norwegian Cancer Society and the Solveig and Ove Lunde Foundation. Disclosure of interests: The authors have no commercial, proprietary, or financial interest in the products or companies described in this article. References 1. Ljungman P, Urbano-Ispizua A, Cavazzana-Calvo M, Demirer T, Dini G, Einsele H, et al. Allogeneic and autologous transplantation for haematological diseases, solid tumours and immune disorders: definitions and current practice in Europe. Bone Marrow Transplant. 2006;37:439e49. 2. Gratwohl A, Baldomero H, Schwendener A, Rocha V, Apperley J, Frauendorfer K, et al. The EBMT activity survey
Cytokine modulation during apheresis
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
16.
17.
18.
2007 with focus on allogeneic HSCT for AML and novel cellular therapies. Bone Marrow Transplant. 2009;43:275e91. Akkok CA, Hervig T, Stamnesfet S, Nesthus I, Melve GK, Lassalle P, et al. Effects of peripheral blood stem cell apheresis on systemic cytokine levels in patients with multiple myeloma. Cytotherapy. 2011;13:1259e68. Akkok CA, Liseth K, Nesthus I, Lokeland T, Tefre K, Bruserud O, et al. Autologous peripheral blood progenitor cells cryopreserved with 5 and 10 percent dimethyl sulfoxide alone give comparable hematopoietic reconstitution after transplantation. Transfusion. 2008;48:877e83. Liseth K, Ersvaer E, Abrahamsen JF, Nesthus I, Ryningen A, Bruserud O. Long-term cryopreservation of autologous stem cell grafts: a clinical and experimental study of hematopoietic and immunocompetent cells. Transfusion. 2009;49:1709e19. Cao Y, Luetkens T, Kobold S, Hildebrandt Y, Gordic M, Lajmi N, et al. The cytokine/chemokine pattern in the bone marrow environment of multiple myeloma patients. Exp Hematol. 2010;38:860e7. Lauta VM. A review of the cytokine network in multiple myeloma: diagnostic, prognostic, and therapeutic implications. Cancer. 2003;97:2440e52. Jakob C, Sterz J, Zavrski I, Heider U, Kleeberg L, Fleissner C, et al. Angiogenesis in multiple myeloma. Eur J Cancer. 2006; 42:1581e90. Iwasaki T, Sano H. Predicting treatment responses and disease progression in myeloma using serum vascular endothelial growth factor and hepatocyte growth factor levels. Leukemia Lymphoma. 2003;44:1347e51. Terpos E, Politou M, Viniou N, Rahemtulla A. Significance of macrophage inflammatory protein-1 alpha (MIP-1alpha) in multiple myeloma. Leukemia Lymphoma. 2005;46: 1699e707. Kyle RA. Prognostic factors in multiple myeloma. Stem Cells. 1995;13(Suppl 2):56e63. Joshua DE, Brown RD, Gibson J. Prognostic factors in myeloma: what they tell us about the pathophysiology of the disease. Leukemia Lymphoma. 1994;15:375e81. Reikvam H, Mosevoll KA, Melve GK, Gunther CC, Sjo M, Bentsen PT, et al. The pretransplantation serum cytokine profile in allogeneic stem cell recipients differs from healthy individuals, and various profiles are associated with different risks of posttransplantation complications. Biol Blood Marrow Transplant. 2012;18:190e9. Stavrum AK, Petersen K, Jonassen I, Dysvik B. Analysis of gene-expression data using J-Express. Curr Protoc Bioinformatics. 2008 Mar;Chapter 7:Unit 7.3. Fujii K, Ishimaru F, Kozuka T, Matsuo K, Nakase K, Kataoka I, et al. Elevation of serum hepatocyte growth factor during granulocyte colony-stimulating factor-induced peripheral blood stem cell mobilization. Br J Haematol. 2004;124:190e4. Ohki Y, Heissig B, Sato Y, Akiyama H, Zhu Z, Hicklin DJ, et al. Granulocyte colony-stimulating factor promotes neovascularization by releasing vascular endothelial growth factor from neutrophils. FASEB J. 2005;19:2005e7. Giuliani N, Bonomini S, Romagnani P, Lazzaretti M, Morandi F, Colla S, et al. CXCR3 and its binding chemokines in myeloma cells: expression of isoforms and potential relationships with myeloma cell proliferation and survival. Haematologica. 2006;91:1489e97. Pellegrino A, Ria R, Di Pietro G, Cirulli T, Surico G, Pennisi A, et al. Bone marrow endothelial cells in multiple myeloma secrete CXC-chemokines that mediate
19.
20.
21.
22.
23. 24.
25.
26.
27.
28.
29.
30.
31. 32.
11
interactions with plasma cells. Br J Haematol. 2005;129: 248e56. Nguyen AN, Stebbins EG, Henson M, O’Young G, Choi SJ, Quon D, et al. Normalizing the bone marrow microenvironment with p38 inhibitor reduces multiple myeloma cell proliferation and adhesion and suppresses osteoclast formation. Exp Cell Res. 2006;312:1909e23. Pellegrino A, Antonaci F, Russo F, Merchionne F, Ribatti D, Vacca A, et al. CXCR3-binding chemokines in multiple myeloma. Cancer Lett. 2004;207:221e7. Mahtouk K, Tjin EP, Spaargaren M, Pals ST. The HGF/ MET pathway as target for the treatment of multiple myeloma and B-cell lymphomas. Biochim Biophys Acta. 2010;1806: 208e19. Olsnes AM, Hatfield KJ, Bruserud O. The chemokine system and its contribution to leukemogenesis and treatment responsiveness in patients with acute myeloid leukemia. J BUON. 2009;14(Suppl. 1):S131e40. Vallet S, Anderson KC. CCR1 as a target for multiple myeloma. Exp Opin Therapeut Targets. 2011;15:1037e47. Abe M, Hiura K, Wilde J, Moriyama K, Hashimoto T, Ozaki S, et al. Role for macrophage inflammatory protein (MIP)-1alpha and MIP-1beta in the development of osteolytic lesions in multiple myeloma. Blood. 2002;100:2195e202. Menu E, De Leenheer E, De Raeve H, Coulton L, Imanishi T, Miyashita K, et al. Role of CCR1 and CCR5 in homing and growth of multiple myeloma and in the development of osteolytic lesions: a study in the 5TMM model. Clin Exp Metast. 2006;23:291e300. Trentin L, Miorin M, Facco M, Baesso I, Carraro S, Cabrelle A, et al. Multiple myeloma plasma cells show different chemokine receptor profiles at sites of disease activity. Br J Haematol. 2007;138:594e602. Schey SA, Morgan GJ, Ramasamy K, Hazel B, Ladon D, Corderoy S, et al. The addition of cyclophosphamide to lenalidomide and dexamethasone in multiply relapsed/refractory myeloma patients: a phase I/II study. Br J Haematol. 2010;150:326e33. Antoniades C, Bakogiannis C, Tousoulis D, Antonopoulos AS, Stefanadis C. The CD40/CD40 ligand system: linking inflammation with atherothrombosis. J Am Coll Cardiol. 2009;54: 669e77. Tong AW, Seamour B, Chen J, Su D, Ordonez G, Frase L, et al. CD40 ligand-induced apoptosis is Fas-independent in human multiple myeloma cells. Leukemia Lymphoma. 2000; 36:543e58. Pellat-Deceunynck C, Amiot M, Robillard N, Wijdenes J, Bataille R. CD11a-CD18 and CD102 interactions mediate human myeloma cell growth arrest induced by CD40 stimulation. Cancer Res. 1996;56:1909e16. de Graaf CA, Metcalf D. Thrombopoietin and hematopoietic stem cells. Cell Cycle. 2011;10:1582e9. Indraccolo S, Stievano L, Minuzzo S, Tosello V, Esposito G, Piovan E, et al. Interruption of tumor dormancy by a transient angiogenic burst within the tumor microenvironment. Proc Natl Acad Sci U S A. 2006;103:4216e21.
Supplementary data Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.jcyt.2013. 02.008
Stem cell mobilization and harvesting by leukapheresis alters systemic cytokine levels in patients with multiple myeloma KNUT ANDERS MOSEVOLL1,4, ÇIGDEM AKALIN AKKÖK2, TOR HERVIG2,3, 2 1,2,4 & HÅKON REIKVAM1,2,4 GURO K. MELVE , ØYSTEIN BRUSERUD 1
Department of Medicine, Haukeland University Hospital, Bergen, Norway, 2Department of Immunology and Transfusion Medicine, Haukeland University Hospital, Bergen, Norway, 3The Gades Institute, University of Bergen, Bergen, Norway, and 4Institute of Medicine, Section of Hematology, University of Bergen, Bergen, Norway
Abstract Background aims. Stem cell mobilization and harvesting by peripheral blood leukapheresis in patients with myeloma can alter plasma levels of certain cytokines. In the present study, we investigated the effects of these interventions on a larger group of cytokines. Methods. Plasma cytokine levels were determined in 15 patients with myeloma who were undergoing peripheral blood stem cell harvesting, and we compared the patients with healthy donors who were undergoing platelet apheresis. Results. Several cytokines showed altered levels in patients with myeloma when examined after chemotherapy plus granulocyte colony-stimulating factoreinduced stem cell mobilization. The most striking effect was increased levels of several CCL (CCL2/3/4) and CXCL (CXCL5/8/10/11) chemokines as well as increased thrombopoietin, interleukin 1 receptor antagonist, interleukin-4, granulocyte colony-stimulating factor and hepatocyte growth factor. Stem cell harvesting by apheresis altered the plasma levels of several mediators (CD40 ligand, interleukin 1 receptor antagonist, CCL5 and CXCL5/ 8/10/11). Apheresis in patients with myeloma had divergent effects on these chemokine levels, although they were all still significantly higher than for healthy individuals. Thrombapheresis in healthy individuals had only minor effects on plasma cytokine levels. Stem cell graft supernatants showed high levels of several cytokines, especially CCL and CXCL chemokines. Analyses of chemokine profiles in pre-apheresis plasma and graft supernatants suggested that such profiling can be used to detect prognostically relevant differences between patients. Conclusions. Our results demonstrate that patients with myeloma have an altered cytokine network during stem cell mobilization, and the network is further altered during stem cell harvesting by leukapheresis. These treatment- or procedure-induced alterations involve several mediators known to affect myeloma cell proliferation, migration and survival. Key Words: cytokines, multiple myeloma, stem cell harvesting, thrombapheresis
Introduction High-dose chemotherapy followed by autologous stem cell support is used in the treatment of several malignancies, including multiple myeloma. The stem cells used in this treatment are often harvested from the peripheral blood after mobilization by growth factors, eventually combined with chemotherapy (1,2). This therapeutic strategy is commonly used in younger patients with myeloma. However, it should be emphasized that (i) autotransplantation is used as a part of a treatment strategy consisting of initial induction chemotherapy; (ii) patients who respond or achieve disease stabilization after this treatment then continue to stem cell mobilization with initial chemotherapy followed by administration of granulocyte colony-stimulating factor (G-CSF); (iii) this is followed by stem cell harvesting by
an apheresis procedure, and finally (iv) patients receive high-dose chemotherapy followed by autotransplantation (3e5). The stem cell harvesting is the only step in this regimen that does not include myeloma-directed therapy. The cytokine network has been studied both in the bone marrow and peripheral blood of patients with myeloma (6,7), and certain single cytokines (ie, hepatocyte growth factor [HGF], vascular endothelial growth factor [VEGF], basic fibroblast growth factor [bFGF], CCL3 and interleukin [IL]6) have then shown prognostic correlation (8e12). In a recent study we described that stem cell harvesting altered plasma levels of angioregulatory cytokines, which also are important in regulation of myeloma cell proliferation (3). The aim of the
Correspondence: Dr Øystein Bruserud, Section for Hematology, Institute of Medicine, University of Bergen, Department of Medicine, Haukeland University Hospital N-5021 Bergen, Norway. E-mail: [email protected] (Received 4 June 2012; accepted 9 February 2013) ISSN 1465-3249 Copyright Ó 2013, International Society for Cellular Therapy. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcyt.2013.02.008
2
K. A. Mosevoll et al.
present study was to investigate whether peripheral blood stem cell harvesting by apheresis causes a more extensive alteration of the systemic cytokine network. We investigated the plasma cytokine profile (30 cytokines examined) for patients with myeloma undergoing peripheral blood stem cell harvesting, and the effect of apheresis in the patients was compared with the effect of thrombapheresis in healthy individuals. Methods Study population The study was approved by the local ethics committee (University of Bergen, Norway). Patients and donors were included after written informed consent was given. Fifteen consecutive patients with myeloma (seven men and eight women; median age, 57 years; range, 44e67 years) with newly diagnosed symptomatic disease and 15 healthy volunteer blood donors (seven women and eight men; median age, 46 years; range, 26e62 years) were investigated (peripheral blood plasma samples from 10 and platelet concentrate samples from 15 healthy donors). Apheresis procedures Hematopoietic stem cell mobilization and harvesting. The study included one leukapheresis for each patient, and this was always the first one if the patients had to be harvested several times. All patients received disease-stabilizing induction chemotherapy with either (i) two cycles of intravenous cyclophosphamide 1 g/m2 on day 1 plus oral dexamethasone 40 mg on days 1e4 and 9e12, with 4-week intervals (14 patients), or (ii) two cycles with bortezomib 1.3 mg/m2 on days 1, 4, 8 and 11 and dexamethasone 40 mg orally on days 1e4 and 9e12, with 3-week intervals (one patient). After this treatment, three patients showed complete response, four patients showed partial response, and eight patients had stable disease. All patients received stem cell mobilization with cyclophosphamide 2 g/m2 followed by G-CSF 5 mg/ kg per day. The first apheresis was performed 10 days (12 patients) or 11 days (3 patients) after start of cyclophosphamide. The median peripheral blood concentrations of CD34þ cells at the time of apheresis was 39.9 103/mL (range, 9.7e175), and the median yield of CD34þ cells was 5.3 106/kg (range, 1.1e27.9). Harvesting, cryopreservation and thawing of peripheral blood stem cell grafts were performed as described in detail previously (4,5). Version 7.0 of the Cobe Spectra cell separator (Cobe Laboratories,
Gloucester, United Kingdom) was used, and largevolume leukapheresis with processing of 4 times the patients’ total blood volume was performed. A combination of acid citrate dextrose (ACD-A) (Baxter FA, Lessin, Belgium) at a blood to citrate ratio of 18:1 and 2500 IU heparin/500 mL of ACD-A solution (Leo Pharma AS, Oslo, Norway) was used for anticoagulation. Moreover, ACD-A equivalent to 10% of the estimated harvest volume was added initially to the collection bag. The total leukocyte counts and platelet counts did not differ between the healthy control subjects and patients with myeloma examined immediately before stem cell mobilization, whereas the hemoglobin levels were still decreased in the patients (median, 10.1 g/100 mL; range, 6.8e12.0, P < 0.0005) compared with the control subjects (median, 14.4; range, 12.4e15.8). The median peripheral blood platelet counts were within the normal range both at the start of mobilization and at the time of harvesting, whereas there was an expected increase in total peripheral blood leukocyte counts from the start of mobilization chemotherapy (median, 5.4 109/L; range, 2.5e9.9, P ¼ 0.003) until stem cell harvesting (median, 11.0; range, 2.7e44.6). Preparation of platelet concentrates. Single-donor platelet concentrates were obtained from the 10 healthy blood donors with median total blood volume of 5.6 L (range, 4.3e7.0 L). The donors had not used anti-platelet drugs the last 5 days before donation. The collection was performed with the Fenwal Amicus Cell Separator (Baxter Healthcare Corp, Deerfield, IL, USA), which uses the elutriation principle to provide leukocyte-reduced platelet concentrates. ACD-A was used as anticoagulant at a ratio of 1:9.
Analysis of cytokine plasma levels Preparation of plasma samples. Blood samples were collected immediately before and after apheresis and 24 h after initiation of apheresis. Samples were drawn from the central venous catheter in patients and from the arm not used during apheresis in the platelet donors. Samples were also taken from stem cell autografts and platelet concentrates. Blood samples were collected into Vacutainer 9NC tubes (BectonDickinson, San Jose, CA, USA) with citric acid as anticoagulant. Samples from platelet concentrates and stem cell autografts were transferred to plastic tubes without additive and centrifuged twice at 200g for 15 min at room temperature. Centrifugation was started within 30 min from sampling. The supernatants were thereafter transferred to cryotubes, frozen immediately and stored at 70 C until analyzed.
Cytokine modulation during apheresis Cytokine analyses. Cytokine levels were determined by Luminex analyses (R&D Systems; Abingdon, United Kingdom) and included (i) the interleukins IL1a, IL1b, interleukin 1 receptor antagonist (IL1RA), IL2, IL4, IL5, IL6, IL7, IL8/CXCL8, IL10, IL12, IL13, IL17 and IL22; (ii) the chemokines CCL3/4/5/ 11 and CXCL5/10/11; (iii) bFGF, G-CSF, granulocyte macrophage colony-stimulating factor (GMCSF), VEGF, thrombopoietin (TPO), epithelial growth factor, HGF and leptin; and (iv) the immunomodulatory cytokines interferon-g (IFN-g), CD40 ligand (CD40L), tumor necrosis factor-a (TNFa) (13). In addition, CXCL12 and soluble CXCR4 were determined by means of enzyme-linked immunosorbent assay (ELISA) analyses (CXCL12; Quantikine ELISA kit; R&D Systems. CXCR4; soluble Human CXC-chemokine Receptor 4 ELISA Kit; Cusabio Biotech Co Ltd, Wuhan, China). All analyses were performed in duplicate, strictly according to the manufacturer’s instructions. Bioinformatic and statistical analyses Bioinformatic analyses were performed with the use of the J-Express (MolMine AS, Bergen, Norway) (14). For hierarchical clustering, all values were median variance standardized and log2-transformed. The complete linkage was used as linkage method, and for distance measured the Euclidian correlation was used. Additional statistical analyses were performed with the use of the GraphPad Prism 5 (Graph Pad Software, Inc, San Diego, CA, USA). Pearson correlation for bivariate samples was used for correlation analyses, the Wilcoxon signed rank test was used to compare different groups, and values of P < 0.05 were regarded as statistically significant (13,14). Results Patient symptoms, findings and disease development At the time of diagnosis, the patients had a median hemoglobin level of 11.8 g/100 mL (range, 9.2e14.9 g/100 mL), median serum calcium of 2.39 mmol/L (1.75e3.02 mmol/L) and median serum kreatinin of 76 mmol/L (29e1024 mmol/L). The monoclonal components were Immunoglobulin G for eight patients (two of them also having detectable levels of monoclonal light chains); immunoglobulin A for two patients, and five patients had light chain disease. Osteolytic lesions were detected in nine patients at the time of diagnosis. The following symptoms/signs were regarded as indications for treatment: anemia, nine patients (one having general bone marrow failure); renal failure, five patients; skeletal pain, five patients; hypercalcemia, three patients; spinal cord
3
compression, three patients; and polyneuropathy, two patients. All patients received high-dose melphalan followed by autologous stem cell rescue. After this treatment, three patients achieved a complete response, seven patients achieved a partial response and five patients had disease stabilization. The median follow-up of the patients is now 62 months (range, 50e64 months), and three patients have not yet had progression of their disease after 53, 55 and 64 months, respectively. Eleven patients received anti-myeloma treatment for disease progression; one patient died of other diseases before myeloma progression, and this patient was not included in the statistical analyses of time until progression. Pre-apheresis plasma cytokine levels in patients with myeloma and in healthy individuals Cytokine levels were determined in plasma samples derived immediately before leukapheresis for all patients with myeloma (Table I), and these results were compared with the levels in 10 healthy individuals. The last results are summarized in Table II and are presented in detail in Figure 1. CXCR4, the CCL and the CXCL chemokines (CCL2-5 and CXCL5/8/10/11/12) showed all detectable levels in the patients. The majority of these chemokines were increased in patients, the only exceptions being CCL5 and CCL11, which did not differ. The most significant increases were seen for CCL4 (P < 0.0001), CXCL10 (P ¼ 0.0043) and CXCL11 (P ¼ 0.0025). CXCR4 and CXCL12 were only determined in the patients. In contrast to the chemokines, all the interleukins (IL1a, IL1b, IL2, IL4, IL6, IL10, IL12, IL13 and IL17) showed relatively low or undetectable levels that did not differ between patients and healthy control subjects except for IL4, which showed an increase (Table II, right panel). Finally, the antagonistic IL1RA showed significantly increased levels in patients compared with healthy control subjects (Figure 1). The growth factors GM-CSF and epithelial growth factor showed undetectable levels, whereas leptin levels were relatively high but with no differences between the groups (Table II). In contrast, TPO levels were significantly increased in the patients (P ¼ 0.0091) but with no correlation to the peripheral blood platelet count. Finally, the immunomodulatory cytokines differed; CD40L levels were significantly increased in the patients (P ¼ 0.0088), whereas IFNg showed undetectable levels. The patients received G-CSF for stem cell mobilization, and, as expected, the G-CSF plasma levels were considerably higher for patients than for control subjects. Even though there was a considerable variation in the preapheresis G-CSF levels among
4
K. A. Mosevoll et al.
Table I. Plasma cytokine profiles in patients with myeloma: A comparison between levels in samples derived immediately before and immediately after stem cell harvesting by leukapheresis. Mediator Chemokines CCL2 CCL3a CCL4a CCL5 CCL11a CXCL5a CXCL8a CXCL10a CXCL11a CXCL12 CXCR4 Interleukins IL1-b IL1RAa IL2 IL4 IL10 IL17 Growth factors G-CSFa TPO Leptin HGF Immunoregulatory cytokines CD40La TNFa
Before apheresis (pg/mL)
After apheresis (pg/mL)
P value
203 194 82 3313 33 150 3.2 406 80 1730 65
(123e566) (20e347) (37e160) (660e17 598) (16e278) (33e288) (nde17.4) (75e2456) (29e340) (916e2507) (nde120)
250 202 81 1768 42 73 5.4 1565 60 1576 45
(101e434) (nde285) (37e186) (780e14 914) (16e318) (33e176) (nde22.5) (25e2874) (13e300) (735e2373) (nde177)
Ns Ns Ns 0.0002 Ns 0.0011 0.034 0.0004 0.0076 Ns Ns
1.0 3723 5.8 20.8 2.0 12.4
(nde16.6) (765e8184) (nde36.9) (8e24.3) (nde6.8) (2e16.7)
3.0 4510 6.7 16,7 2.0 11
(0.5e12.2) (906e11 896) (nde32.1) (8e32.2) (nde7.3) (2e18)
Ns 0.0003 Ns Ns Ns Ns
(4781e32 072) (32e687) (3090e83 293) (1186e44 422)
0.0181 Ns Ns 0.0008
18 366 477 19 956 327
(9861e46 314) (106e1147) (3695e71 109) (40e21 685)
255 (57e1857) 2.0 (nde30.4)
13 472 489 20 817 15 855
191 (39e944) nd (nde31.5)
0.009 Ns
The table represents the results for 15 patients. All results are presented in pg/mL and as the median levels and with the corresponding variation ranges in parentheses. The table presents P values for the comparison between patient levels before and immediately after apheresis. The results for cytokines with undetectable median levels in both samples (Interferon-g, epithelial growth factor and granulocyte macrophage colony-stimulating factor) or undetectable levels in all samples (IL1a, IL6, IL12 and IL13) are not included. nd, not detectable; Ns, not significant; IL, interleukin; IL1RA, interleukin 1 receptor antagonist; G-CSF, granulocyte colony-stimulating factor; TPO, thrombopoietin; HGF, hepatocyte growth factor; CD40L, CD40 ligand; TNFa, tumor necrosis factor-a. a Cytokines with significantly increased levels in peripheral blood stem cell grafts compared with platelet concentrates.
patients, there was no correlation between G-CSF levels and the amount of harvested CD34þ cells. The preharvesting cytokine profile identifies two major patient subsets The overall preharvesting cytokine data were analyzed by hierarchical clustering; two major patient clusters/subsets were then identified including the lower six and the upper nine patients, respectively (Supplementary Figure 1). These two patient clusters differed especially with regard to the cytokine cluster consisting of leptin, CXCL10, IL2 and IL1b, and, at both time points, HGF, IL10, IL17 and CCL3 also differed between the two subsets. Thus, this classification of patients into two major clusters is mainly determined by a limited number of cytokines showing significant differences between these two subsets. The patient clustering was identical when tested immediately after and 24 h after leukapheresis, and the same clusters were identified when distance metrics were based on the Euclidian
correlation tests (data not shown). The median time until disease progression for the two clusters was 21 (upper cluster) and 43 months (lower cluster), respectively, but this difference was not significant. The upper cluster could be further divided into two patient subsets, but also when comparing the three different clusters (upper 5, middle 4 and lower 6 patients), no significant difference in survival was detected between among groups. Effects of stem cell harvesting on plasma cytokine levels: Analyses of single cytokines We determined plasma cytokine levels immediately after the stem cell harvesting, and several cytokines showed altered levels compared with pre-apheresis levels (Table I and Figure 2). Chemokines. The chemokines CCL5, CXCL5 and CXCL11 showed decreased levels immediately after apheresis, whereas CXCL8 and CXCL10 levels were increased. The differences were still significant after
Cytokine modulation during apheresis
5
Table II. Systemic cytokine profiles in plasma samples: A comparison between patients with myeloma immediately before apheresis and healthy control subjects. Mediator Chemokines CCL2 CCL3a CCL4a CCL5 CCL11a CXCL5a CXCL8a CXCL10a CXCL11a Interleukins IL1-b IL1RAa IL2 IL4 IL10 IL17 Growth factors G-CSFa TPO Leptin HGF Immunoregulatory cytokines CD40La TNFa
Myeloma (pg/mL)
Healthy (pg/mL)
203 194 82 3313 33 150 3.2 406 80
(123e566) (20e347) (37e160) (660-17 598) (16e278) (33e288) (nde17.4) (75e2456) (29e340)
149 126.5 14 2318 37 62 nd 71 30.5
1.0 3723 5.8 20.8 2.0 12.4
(nde16.6) (765e8184) (nde36.9) (8e24.3) (nde6.8) (2e16.7)
1.15 207 2.75 12 0.2 7.6
18 366 477 19 956 327
(9861e46 314) (106e1147) (3695e71 109) (40e21 685)
255 (57e1857) 2.0 (nde30.4)
20 238 16 362 44
P value
(37e321) (nde188) (nde34) (1191e6189) (16e176) (30e182) (nde4.8) (17e366) (3.6e120)
0.0285 0.0134 <0.0001 Ns Ns 0.0264 0.0418 0.0043 0.0025
(nde10.8) (156e353) (nde23.9) (nde21.4) (nde2.3) (3e14.5)
Ns <0.0001 Ns 0.0143 Ns Ns
(12e34) (142e437) (6045e64 679) (0e211)
<0.0001 0.0091 Ns 0.001
163 (45e230) nd (nde4.5)
Ns Ns
The table represents the results for 15 patients and 10 healthy control individuals. All results are presented in pg/mL and as the median levels and with the corresponding variation ranges in parentheses. The table presents P values for the comparison between patient levels before leukapheresis and levels in healthy individuals. The results for cytokines with undetectable median levels in all three samples (interferon-g, epithelial growth factor and granulocyte macrophage colony-stimulating factor) or undetectable levels in all samples (IL1a, IL6, IL12 and IL13) are not included. nd, not detectable; Ns, not significant; IL, interleukin; IL1RA, interleukin 1 receptor antagonist; G-CSF, granulocyte colony-stimulating factor; TPO, thrombopoietin; HGF, hepatocyte growth factor; CD40L, CD40 ligand; TNFa, tumor necrosis factor-a. a Cytokines with significantly increased levels in peripheral blood stem cell grafts compared with platelet concentrates.
24 hours only for CCL5 (median, 1613 pg/mL; range, 817e17 206 pg/mL, P ¼ 0.0026) and CXCL5 (78 pg/ mL, 34e225 pg/mL, P ¼ 0.017), and an additional decrease in CCL4 levels was detected after 24 h (67 pg/ mL, 5e115 pg/mL, P ¼ 0.029) compared with the preapheresis levels (82 pg/mL, 37e160 pg/mL). Interleukins. The interleukin levels were not significantly altered either immediately or 24 h after stem cell harvesting. In contrast, IL1RA levels were increased immediately after harvesting. Growth factors. HGF showed an increase immediately after apheresis, but was normalized after 24 hours. Leptin showed increased plasma levels after 24 h (median, 31 773 pg/mL; range, 5520e86 892 pg/ mL; P ¼ 0.015) compared with pre-apheresis levels (19 956 pg/m L, 3695e71 109 pg/mL). The G-CSF levels showed an expected decrease both immediately (13 472 pg/mL, 4781e32 072 pg/mL, P ¼ 0.018) and 24 h (5670 pg/mL, 347e26 009 pg/mL, P < 0.0001) after stem cell harvesting compared
with the pre-apheresis levels (18 366 pg/mL, 9861e46 314 pg/mL). Immunomodulatory cytokines. CD40L levels were decreased immediately after apheresis but normalized within 24 h, whereas IFNg and TNFa levels remained low or undetectable. Thus, harvesting of peripheral blood stem cells alters systemic cytokine levels, especially chemokine levels, and for some cytokines these alterations persist for >24 h. Plasma chemokine profiles in patients with myeloma before and after stem cell harvesting In contrast to most other cytokines, the chemokines (i) showed detectable pre-apheresis plasma levels for most patients; (ii) altered preharvesting levels were observed in patients with myeloma compared with healthy individuals; and (iii) mainly chemokine levels were altered by leukapheresis. For these reasons, we focused on the chemokine levels in the further
6
K. A. Mosevoll et al.
Figure 1. Pre-apheresis peripheral blood plasma cytokine levels in patients with myeloma and healthy individuals. We investigated the levels for 30 soluble mediators; the figure presents the levels for the cytokines (CCL2-4, CXCL5/8/10/11, IL1RA4, thrombopoietin [TPO], granulocyte colony-stimulating factor [G-CSF] and hepatocyte growth factor [HGF]) that showed statistically significant differences between the two groups. The results are presented as the cytokine level for each individual patient (P) and control subject (C), the mean level is indicated in the figure, and the cytokine and the corresponding P-value are indicated at the top of the figure. Undetectable levels are presented as open symbols.
detailed analyses of our results. We then compared chemokine plasma profiles for all 15 patients before and after apheresis (Figure 3, upper left); the plasma profile was altered by the apheresis, and this was reflected in the chemokine clustering. However, the three chemokines CXCL5/CCL5/CXCL11, which represent the only chemokines with decreased plasma levels immediately after leukapheresis, clustered together both before and after the procedure. The clustering after 24 h was similar to the preharvesting clustering (data not shown). Taken together, these observations suggest that even though stem cell harvesting alters the serum levels of individual cytokines, the effects on the chemokine profile and clustering are limited. Finally, both the pre-apheresis and postapheresis analyses identified two major patient
clusters, but the time until disease progression did not differ between these patient subsets. Cytokine levels in supernatants from peripheral blood stem cell grafts and platelet concentrates: The chemokine profile identifies a patient subset with prolonged time until progression We compared the supernatant levels of individual cytokines for stem cell grafts from patients with myeloma and platelet concentrates from healthy individuals (see Table III, supplementary information and Supplementary Figures 3 and 4). The most important difference between stem cell grafts and platelet concentrates was altered levels for several chemokines, and, for this reason, we investigated the
Cytokine modulation during apheresis
7
Figure 2. Peripheral blood plasma cytokine levels before and immediately after stem cell harvesting from multiple patients with myeloma. We investigated the levels for 30 soluble mediators, and the figure presents the levels for the cytokines (CCL5, CXCL5/8/10/11, interleukin 1 receptor antagonist [IL1RA], granulocyte colony-stimulating factor [G-CSF], hepatocyte growth factor [HGF] and CD40L) that showed statistically significant differences before and after apheresis. The results are presented as the cytokine level for each individual patient and control subject, the mean levels are indicated, and the cytokine and the corresponding P-value are indicated at the top of the figure. Mean values are marked in bold.
chemokine profile of the peripheral blood stem cell grafts more in detail (Figure 3, lower left). The graft chemokine profile and clustering differed from the peripheral blood plasma clustering (Figure 3, upper versus lower left), suggesting that the graft chemokine profile is affected by the harvesting procedure and does not simply reflect the plasma chemokine profile. Furthermore, two major patient clusters were then identified; a lower cluster with five patients (including the three patients without disease progression after 53, 55 and 64 months) and an upper cluster with 10 patients including the one patient who died early after transplantation without disease progression. The five patients in this lower cluster had significantly longer time until disease progression (Figure 3, lower right; Gehan-BreslowWilcoxon test, P ¼ 0.0108) than the other patients, and these five patients clustered either in the same main cluster or close to each other (Figure 3, upper two figures) when analyzing preharvesting and postharvesting chemokine plasma levels. However, the observation that these five patients did not cluster in the same main cluster when analyzing the plasma chemokine cluster (Figure 3, upper part), further
supports our hypothesis that the graft chemokine profile reflects additional biological characteristics of prognostic relevance and does not simply mirror the plasma chemokine profiles. Discussion The aim of our study was to investigate whether apheresis procedures induce alterations in the systemic cytokine levels. We investigated a wide range of immunoregulatory and hematopoiesisregulating cytokines, and we compared cytokine profiles for stem cell harvesting in patients with myeloma and platelet harvesting in healthy blood donors. Our studies demonstrate that apheresis procedures can alter systemic cytokine profiles, but the magnitude and duration of these alterations differ between our two groups. We examined a group of relatively young, consecutive patients with myeloma. All patients had symptoms at the start of induction therapy, but all of them responded to the treatment or had stabilization of the disease. The patients were mobilized with chemotherapy followed by G-CSF treatment; for
8
K. A. Mosevoll et al.
Figure 3. Chemokine clusters in multiple patients with myeloma. (upper part of figure) The peripheral blood plasma chemokine profile for patients with myeloma before and immediately after apheresis is shown. We investigated the levels of 10 chemokines and CXCR4, and we used unsupervised hierarchical cluster analysis to characterize the chemokine profile. The figure presents the results for patients with myeloma before (upper left) and immediately after (upper right) stem cell harvesting. The five patients that formed a separate cluster with prolonged time until disease progression in the analysis of graft chemokine profiles (see lower part of the figure) are marked with blue in the right part of the figures. (lower part of figure) The chemokine profile for peripheral blood stem cell autografts derived from patients with myeloma is shown. The levels of the same 10 chemokines and CXCR4 were determined, and unsupervised hierarchical cluster analysis was used to characterize the chemokine profile. The figure presents the results for the autografts derived from the 15 patients. Two different clusters were identified (left), and patients in cluster 1 (blue) showed a significantly longer time until myeloma progression than the other patients (lower right).
this reason, it was not possible to have a control group of healthy individuals receiving an identical mobilization treatment, and our patients were therefore compared with healthy donors undergoing thrombapheresis. An extensive cytokine network modulation was only seen for the patients with myeloma but not for the platelet donors; we therefore conclude that this apheresis-induced modulation is not a characteristic for all types of apheresis but depends on the clinical context. This difference between patients with myeloma and healthy individuals can be caused by (i) initial chemotherapy before mobilization or residual myeloma disease, (ii) cyclophosphamide followed by hematopoietic regeneration or (iii) the G-CSF treatment. In our opinion, it is not possible to use control groups to clarify the relative contribution of each of these three factors; even G-CSF mobilization in healthy individuals will differ because the kinetics of stem cell mobilization will differ and the period of G-CSF treatment will usually be shorter. However, previous studies have demonstrated that G-CSF therapy alone will increase HGF and VEGF levels (15,16), and this treatment may be at least partly responsible for the increased preapheresis levels of these two cytokines.
In this unique clinical context, especially the HGF levels but not the VEGF levels are further increased by the leukapheresis in patients with myeloma (3). High systemic levels of HGF, VEGF, bFGF, CCL3 and IL6 as well as the soluble IL6 receptor have all been linked to an adverse prognosis in patients with multiple myeloma (7e12). However, to the best of our knowledge, the possible prognostic impact of more extensive cytokine profiles has not been investigated in previous studies. Furthermore, several chemokines showed increased plasma levels in patients with myeloma before stem cell harvesting. These high levels are probably caused by a combined effect of the disease and the recent cyclophosphamide treatment followed by G-CSFesupported hematopoietic regeneration. Many chemokines showed increased levels before apheresis and/or were altered by the leukapheresis, and these chemokines can bind to the myeloma-expressed receptors CCR1 (CCL3, CCL5), CCR2 (CCL2), CCR5 (CCL4), CXCR1 (CXCL8) and CXCR2 (CXCL5, CXCL8) (17e26). All these receptors appear to be involved in the pathogenesis of multiple myeloma (17e23), and many of them can also bind several ligands, emphasizing the importance of investigating chemokine
Cytokine modulation during apheresis
9
Table III. Supernatant cytokine levels for peripheral blood stem cell grafts prepared from multiple patients with myeloma and for platelet concentrates derived from healthy blood donors. Mediator
Patients with myeloma PBSC autograft
Chemokines (pg/mL) CCL2 58 CCL3 193 CCL4 88 CCL5 22 519 CCL11 50 CXCL5 625 CXCL8 2.5 CXCL10 703 CXCL11 156 Interleukins (pg/mL) 4.1 IL1-b IL1RA 10 493 IL2 9.2 IL4 17.2 IL10 1.2 IL17 9.6 Growth factors (pg/mL) G-CSF 12 906 TPO 483 Leptin 16 943 EGF nd HGF 24 427 Immunoregulatory cytokines (pg/mL) CD40L 1185
Healthy individuals platelet concentrate
(41e308) (20e314) (25e298) (13 588ehigh) (16e284) (121e2598) (nde23.9) (107e2922) (23e432)
60 110 28 26 659 17 3977 16 8 38.5
(1e15.7) (2731e26 000) (nde36.8) (8e32.1) (nde5.3) (2e20)
P value
(29e80) (nde181) (nde88) (18 701ehigh) (nde60) (1886e5383) (nde30.7) (5e60) (13e196)
Ns 0.024 0.0008 Ns 0.0035 <0.0001 0.019 <0.0001 0.0025
2.05 120.5 5.45 20.8 0.1 8.95
(0e9.3) (65e188) (nde18.3) (12.2e24.4) (nde2.1) (2e15.6)
Ns <0.0001 Ns Ns Ns Ns
(8863e41 139) (233e828) (3207e70 527) (nde36.9) (2062e50 000)
20 393 5010 49.1 72
(10e29) (170e530) (2028e65 872) (8.4e103) (18e211)
<0.0001 Ns Ns Ns <0.0001
(192e2877)
7025 (3615e15 246)
<0.0001
The table represents the results for 15 patients and 10 healthy control individuals. All results are presented in pg/mL and as the median levels and with the corresponding variation ranges in parentheses. The table presents the cytokine levels for peripheral blood stem cell grafts and platelet concentrates. The results for cytokines with undetectable median levels in both samples (interferon-g, epithelial growth factor and granulocyte macrophage colony-stimulating factor) or undetectable (nd, not detectable) levels in all samples (IL1a, IL6, IL12 and IL13) are not included. CXCL12 and CXCR4 were only determined in the patients and are not included in the table (nd, not detectable; high, above detection range, two and three samples, respectively). PBSC, peripheral blood stem cell Ns, not significant; IL, interleukin; IL1RA, interleukin 1 receptor antagonist; G-CSF, granulocyte colonystimulating factor; EGF, epithelial growth factor; TPO, thrombopoietin; HGF, hepatocyte growth factor; CD40L, CD40 ligand.
profiles and not only single chemokine levels. However, exceptional chemokines showed apheresisinduced decreases in plasma levels (CCL5, CXCL5, CXCL11), whereas others showed increased postapheresis levels (CXCL8, CXCL10). Previous studies have also shown a strong increase in the proangiogenic myeloma growth factor HGF after leukapheresis (3). Finally, cyclophosphamide may be given to our patients as a part of the mobilization regimen, but it is also a well-known anti-myeloma drug (27). It is not known whether its anti-myeloma effect during mobilization is clinically relevant. Thus, it is difficult to predict the final effect of cyclophosphamide-based mobilization and modulation of the cytokine/chemokine network on myeloma cell growth and survival during the period of stem cell mobilization and harvesting. This issue is further complicated by variation of myeloma cell expression of chemokine receptors between patients and even within the same patient (26). Several chemokines that showed altered plasma levels either before or after apheresis may be important for the pathogenesis of multiple myeloma. First,
myeloma cells express both CXCR3A and CXCR3B receptors, and CXCR3A ligation by CXCL10 (and possibly also CXCL11) inhibits CD95-induced apoptosis and IL6-dependent myeloma cell proliferation (17). Second, another study confirmed that plasma cells express CXCR3 and in addition the CXCR1, CXCR2 (both being receptors for CXCL8, CXCR1 also binding CXCL5) and CCR2 (receptor for CCL2). Another study found that CXCL8 stimulated myeloma cell proliferation, whereas CXCL8 as well as CXCL11 (CXCR3 ligation) and CCL2 (CCR2 ligation) increased myeloma cell migration (18). Third, CXCL10 appears to increase the adherence of myeloma cells to bone marrow stromal cells (19). Finally, CXCR3 ligation by CXCL9, CXCL10 and CXCL11 induces downstream intracellular signalling leading to altered myeloma cell migration and increased release of matrix metalloproteinase (MMP)2 and MMP9 (20). Thus, most of these studies suggest that the apheresis-induced reduction in plasma chemokine levels would be expected to mediate anti-myeloma effects and thereby counteract the potential myeloma-
10
K. A. Mosevoll et al.
stimulatory effect of increased preapheresis chemokine levels and the previously described apheresis-induced HGF increase (21). CD40L levels were altered by stem cell harvesting. CD40 is involved in immunoregulation, and in vitro studies suggest that it can also induce apoptosis in myeloma cells (28e30). In addition, the modulation of TPO levels suggests that the regulation of normal hematopoiesis may also be affected by the apheresis (31). However, it should be emphasized that additional studies are required to clarify whether the variations have any clinical impact. We compared the chemokine profiles for patients with myeloma and for the healthy blood donors and found that the pre-apheresis chemokine profile differed between the two groups. There was also a difference in the effect of apheresis between patients and control subjects with regard to the overall chemokine profile, although the apheresis procedures had relatively small effects on the chemokine profiles both for patients and blood donors. Thus, the effect of apheresis on the chemokine network in these two groups is reflected mainly in the absolute chemokine levels. The overall chemokine profile was altered by apheresis; these effects were most clearly seen in patients with myeloma, whereas thrombapheresis in healthy individuals had only minor effects. Thus, the balance between cytokine effects in the systemic (plasma) cytokine network is altered by stem cell harvesting. Furthermore, the graft cytokine profile does not simply reflect the plasma profile, and very high levels of several cytokines are observed in the stem cell grafts together with an altered chemokine profile and clustering (see above). Reinfusion of the graft may therefore alter the posttransplant cytokine network, especially the chemokine balance, through infusion of cytokines together with the stem cells. We also investigated whether there was any association between chemokine levels and time until myeloma relapse. We strongly emphasize that these results should be interpreted with great precaution because our study included few patients. We observed that the analysis of chemokine profiles in the graft supernatants identified a subset of patients with significantly longer time to relapse compared with the other patients. These patients were all included in the same main patient subset when the preharvesting chemokine profile was analyzed. Our observations must be confirmed in larger prospective studies, but they suggest that analysis of cytokine profiles deserves further investigation in future clinical studies. However, one should not conclude that reinfusion of graft cytokines directly affects prognosis in autotransplanted patients with myeloma. In our opinion, it seems more likely that apheresis-induced modulation
of chemokine levels and prolonged time until relapse are two different parameters reflecting the same unique clinical/biological context in these patients. The cytokine network modulations described in this study are relatively-short-lasting, but, despite this, they may be of clinical relevance. A short-lasting initial event may be sufficient to initiate a biological process, and the hypothesis of angiogenic switch to initiate growth of cancer metastasis from dormant malignant cells is based on this (32). We have documented that the cytokine-modulating effects of apheresis depend on the clinical context; if the apheresis-induced alterations and/or the later reinfusion of cytokines together with the stem cells occur at critical time points, they may synergize with the specific clinical context and thereby become clinically relevant. Our observations illustrate that stem cell mobilization, harvesting and reinfusion have several additional and largely unknown effects that may predict or possibly affect the later course of the myeloma disease. To conclude, our study demonstrates that the stem cell harvesting by leukapheresis alters the systemic cytokine network in patients with myeloma, and our results also suggest that analysis of cytokine profiles (and especially chemokine profiles) may become a useful tool in the prognostic evaluation of these patients. However, the biological impact of this cytokine modulation and the possible clinical use of cytokine profiling need further evaluation in larger clinical studies. Previous studies have suggested that anti-myeloma treatment induces specific alterations that can be detected in bone marrow plasma (6), and it would be a clinical advantage if plasma profiling instead of bone marrow samples could be used in the follow-up of these patients. Acknowledgments We thank Kristin Rye Paulsen for technical assistance. This work was supported by the Norwegian Cancer Society and the Solveig and Ove Lunde Foundation. Disclosure of interests: The authors have no commercial, proprietary, or financial interest in the products or companies described in this article. References 1. Ljungman P, Urbano-Ispizua A, Cavazzana-Calvo M, Demirer T, Dini G, Einsele H, et al. Allogeneic and autologous transplantation for haematological diseases, solid tumours and immune disorders: definitions and current practice in Europe. Bone Marrow Transplant. 2006;37:439e49. 2. Gratwohl A, Baldomero H, Schwendener A, Rocha V, Apperley J, Frauendorfer K, et al. The EBMT activity survey
Cytokine modulation during apheresis
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
16.
17.
18.
2007 with focus on allogeneic HSCT for AML and novel cellular therapies. Bone Marrow Transplant. 2009;43:275e91. Akkok CA, Hervig T, Stamnesfet S, Nesthus I, Melve GK, Lassalle P, et al. Effects of peripheral blood stem cell apheresis on systemic cytokine levels in patients with multiple myeloma. Cytotherapy. 2011;13:1259e68. Akkok CA, Liseth K, Nesthus I, Lokeland T, Tefre K, Bruserud O, et al. Autologous peripheral blood progenitor cells cryopreserved with 5 and 10 percent dimethyl sulfoxide alone give comparable hematopoietic reconstitution after transplantation. Transfusion. 2008;48:877e83. Liseth K, Ersvaer E, Abrahamsen JF, Nesthus I, Ryningen A, Bruserud O. Long-term cryopreservation of autologous stem cell grafts: a clinical and experimental study of hematopoietic and immunocompetent cells. Transfusion. 2009;49:1709e19. Cao Y, Luetkens T, Kobold S, Hildebrandt Y, Gordic M, Lajmi N, et al. The cytokine/chemokine pattern in the bone marrow environment of multiple myeloma patients. Exp Hematol. 2010;38:860e7. Lauta VM. A review of the cytokine network in multiple myeloma: diagnostic, prognostic, and therapeutic implications. Cancer. 2003;97:2440e52. Jakob C, Sterz J, Zavrski I, Heider U, Kleeberg L, Fleissner C, et al. Angiogenesis in multiple myeloma. Eur J Cancer. 2006; 42:1581e90. Iwasaki T, Sano H. Predicting treatment responses and disease progression in myeloma using serum vascular endothelial growth factor and hepatocyte growth factor levels. Leukemia Lymphoma. 2003;44:1347e51. Terpos E, Politou M, Viniou N, Rahemtulla A. Significance of macrophage inflammatory protein-1 alpha (MIP-1alpha) in multiple myeloma. Leukemia Lymphoma. 2005;46: 1699e707. Kyle RA. Prognostic factors in multiple myeloma. Stem Cells. 1995;13(Suppl 2):56e63. Joshua DE, Brown RD, Gibson J. Prognostic factors in myeloma: what they tell us about the pathophysiology of the disease. Leukemia Lymphoma. 1994;15:375e81. Reikvam H, Mosevoll KA, Melve GK, Gunther CC, Sjo M, Bentsen PT, et al. The pretransplantation serum cytokine profile in allogeneic stem cell recipients differs from healthy individuals, and various profiles are associated with different risks of posttransplantation complications. Biol Blood Marrow Transplant. 2012;18:190e9. Stavrum AK, Petersen K, Jonassen I, Dysvik B. Analysis of gene-expression data using J-Express. Curr Protoc Bioinformatics. 2008 Mar;Chapter 7:Unit 7.3. Fujii K, Ishimaru F, Kozuka T, Matsuo K, Nakase K, Kataoka I, et al. Elevation of serum hepatocyte growth factor during granulocyte colony-stimulating factor-induced peripheral blood stem cell mobilization. Br J Haematol. 2004;124:190e4. Ohki Y, Heissig B, Sato Y, Akiyama H, Zhu Z, Hicklin DJ, et al. Granulocyte colony-stimulating factor promotes neovascularization by releasing vascular endothelial growth factor from neutrophils. FASEB J. 2005;19:2005e7. Giuliani N, Bonomini S, Romagnani P, Lazzaretti M, Morandi F, Colla S, et al. CXCR3 and its binding chemokines in myeloma cells: expression of isoforms and potential relationships with myeloma cell proliferation and survival. Haematologica. 2006;91:1489e97. Pellegrino A, Ria R, Di Pietro G, Cirulli T, Surico G, Pennisi A, et al. Bone marrow endothelial cells in multiple myeloma secrete CXC-chemokines that mediate
19.
20.
21.
22.
23. 24.
25.
26.
27.
28.
29.
30.
31. 32.
11
interactions with plasma cells. Br J Haematol. 2005;129: 248e56. Nguyen AN, Stebbins EG, Henson M, O’Young G, Choi SJ, Quon D, et al. Normalizing the bone marrow microenvironment with p38 inhibitor reduces multiple myeloma cell proliferation and adhesion and suppresses osteoclast formation. Exp Cell Res. 2006;312:1909e23. Pellegrino A, Antonaci F, Russo F, Merchionne F, Ribatti D, Vacca A, et al. CXCR3-binding chemokines in multiple myeloma. Cancer Lett. 2004;207:221e7. Mahtouk K, Tjin EP, Spaargaren M, Pals ST. The HGF/ MET pathway as target for the treatment of multiple myeloma and B-cell lymphomas. Biochim Biophys Acta. 2010;1806: 208e19. Olsnes AM, Hatfield KJ, Bruserud O. The chemokine system and its contribution to leukemogenesis and treatment responsiveness in patients with acute myeloid leukemia. J BUON. 2009;14(Suppl. 1):S131e40. Vallet S, Anderson KC. CCR1 as a target for multiple myeloma. Exp Opin Therapeut Targets. 2011;15:1037e47. Abe M, Hiura K, Wilde J, Moriyama K, Hashimoto T, Ozaki S, et al. Role for macrophage inflammatory protein (MIP)-1alpha and MIP-1beta in the development of osteolytic lesions in multiple myeloma. Blood. 2002;100:2195e202. Menu E, De Leenheer E, De Raeve H, Coulton L, Imanishi T, Miyashita K, et al. Role of CCR1 and CCR5 in homing and growth of multiple myeloma and in the development of osteolytic lesions: a study in the 5TMM model. Clin Exp Metast. 2006;23:291e300. Trentin L, Miorin M, Facco M, Baesso I, Carraro S, Cabrelle A, et al. Multiple myeloma plasma cells show different chemokine receptor profiles at sites of disease activity. Br J Haematol. 2007;138:594e602. Schey SA, Morgan GJ, Ramasamy K, Hazel B, Ladon D, Corderoy S, et al. The addition of cyclophosphamide to lenalidomide and dexamethasone in multiply relapsed/refractory myeloma patients: a phase I/II study. Br J Haematol. 2010;150:326e33. Antoniades C, Bakogiannis C, Tousoulis D, Antonopoulos AS, Stefanadis C. The CD40/CD40 ligand system: linking inflammation with atherothrombosis. J Am Coll Cardiol. 2009;54: 669e77. Tong AW, Seamour B, Chen J, Su D, Ordonez G, Frase L, et al. CD40 ligand-induced apoptosis is Fas-independent in human multiple myeloma cells. Leukemia Lymphoma. 2000; 36:543e58. Pellat-Deceunynck C, Amiot M, Robillard N, Wijdenes J, Bataille R. CD11a-CD18 and CD102 interactions mediate human myeloma cell growth arrest induced by CD40 stimulation. Cancer Res. 1996;56:1909e16. de Graaf CA, Metcalf D. Thrombopoietin and hematopoietic stem cells. Cell Cycle. 2011;10:1582e9. Indraccolo S, Stievano L, Minuzzo S, Tosello V, Esposito G, Piovan E, et al. Interruption of tumor dormancy by a transient angiogenic burst within the tumor microenvironment. Proc Natl Acad Sci U S A. 2006;103:4216e21.
Supplementary data Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.jcyt.2013. 02.008