Experimental Hematology 2014;-:-–-

1 2 3 4 a,b c a,b , Sebastien Baillyd, Bertrand Toussaintb,c, and 5 Q16 Emmanuel Pourcelot , Candice Trocme , Julie Mondet a,b Pascal Mossuz 6 a 7 Q1 Cellular Hematology Laboratory, DHOGI, IBP CHU Grenoble, France; bTherex-TIMC/Imag, UMR CNRS 5525, Joseph Fourier University, Grenoble, France; cBiochemistry and Enzymology Laboratory, DBTP, IBP CHU Grenoble, France; dParasitology Laboratory, DAI, IBP CHU 8 Grenoble, France 9 10 (Received 10 April 2013; revised 10 January 2014; accepted 17 January 2014) 11 12 13 Studies have shown that the clinical impact of Janus kinase 2 (JAK2) inhibitors in primary 14 myelofibrosis patients is due to the regulation of cytokine levels, suggesting that cytokine profiles 15 might play a critical role in myeloproliferative neoplasms (MPNs) physiopathology. In this study, we compared the plasma cytokine profiles of polycythemia vera (PV) patients and essen16 tial thrombocythemia (ET) patients as a function of their JAK2 V617F status and the presence 17 of thrombohemorrhagic complications. Using a multiplex cytokine assay, cytokine measure18 ments were taken of the plasma of 17 PV patients and 21 ET patients. Twenty-two of these 19 patients (10 PV and 12 ET) experienced at least one thrombohemorrhagic manifestation before 20 diagnosis. We showed that cytokine levels were significantly increased in PV and ET patients 21 compared with normal values and that several positive correlations existed between the cyto22 kine concentrations and the biological parameters in each MPN. The comparison between 23 the cytokine profiles of ET and PV patients showed a statistically significant increase of inter24 Q2 leukin (IL)-4, IL-8, granulocyte macrophage–colony-stimulating factor, interferon -g, mono25 cyte chemotactic protein -1, platelet-derived growth factor-BB, and vascular endothelial 26 Q3 growth factor in the ET group. Only TFN-a and platelet-derived growth factor-BB were specifically impacted by the JAK2 V617F status of the PV and ET patients, respectively, suggesting 27 that there are both JAK2 V617F–driven and JAK2 V617F–independent inflammatory 28 responses in MPN. We also showed that the subgroup of PV patients with vascular complica29 tions displayed significantly different concentrations of IL-12(p70) and granulocyte macro30 phage–colony-stimulating factor compared with patients without vascular complications. 31 Altogether, these data suggest that cytokine measurement might be useful for the clinical and 32 therapeutic stratification of PV and ET patients. Ó 2014 ISEH - Society for Hematology 33 and Stem Cells. Published by Elsevier Inc. 34 35 36 37 increase of the leukocyte count [2]. These disorders may Myeloproliferative neoplasms (MPNs) are hemopathies in 38 evolve into myelofibrosis or, less commonly, into acute leuwhich an acquired hematopoietic stem cell alteration leads 39 kemia [3,4], but the primary causes of morbidity and mortalto a deregulated production of peripheral blood cells by 40 ity in MPN patients are thrombotic complications and, to a the bone marrow [1], which results in abnormal levels of 41 lesser degree, bleeding [3–6]. In fact, PV and ET are associred blood cells, leukocytes, or thrombocytes. Polycythemia 42 ated with a significantly increased risk of arterial or venous vera (PV) and essential thrombocythemia (ET) are the most 43 thrombosis and bleeding complications [7]. common Philadelphia chromosome negative MPNs . PV and 44 The discovery of a gain-of-function mutation of Janus ET are characterized by an increased red cell mass and 45 kinase 2 (JAK2 V617F) and its role in MPNs [8–10] has thrombocyte count, respectively, associated or not with an 46 focused considerable attention on JAK2 inhibitors as poten47 tial therapeutic targets. Clinical studies of primary myelofi48 Offprint requests to: Dr. Emmanuel Pourcelot, Ph.D. student, PharmaD, brosis (PMF) have highlighted an unexpected impact of Laboratoire d’hematologie cellulaire, departement d’Hematologie, 49 these inhibitors. JAK2 inhibitors primarily act on constituOnco-G e n e tique et Immunologie, Institut de Biologie et Pathologie, CHU 50 tional symptoms (asthenia, cachexia, satiety, night sweats), Grenoble, 38043 Grenoble, France; E-mail: [email protected]

Cytokine profiles in polycythemia vera and essential thrombocythemia patients: Clinical implications

0301-472X/$ - see front matter. Copyright Ó 2014 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.exphem.2014.01.006 EXPHEM3095_proof ■ 2-5-2014 16-50-16

51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

2

101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

thereby improving a patient’s quality of life. The inhibitors also mitigate the symptoms of splenomegaly, including reductions in spleen size and the attenuation of symptoms such as abdominal discomfort or early satiety [11–14], without evidence of the regulation of the JAK2 V617F allele burden. Studies have shown that this clinical impact is partially due to the down regulation of cytokine levels. In fact, PMF patients harbor increased levels of the following cytokines, which are all significantly reduced by treatment with JAK1/JAK2 inhibitors: the antagonist receptor of interleukin-1 (IL-1RA), macrophage inflammatory protein (MIP)-1b, tumor necrosis factor (TNF)-a, vascular endothelial growth factor (VEGF), interleukin (IL)-6 and IL-8 [12]. These data have highlighted new issues regarding the potential physiopathologic role of plasma cytokines in MPNs. The hypersensitivity of the erythroid progenitors to erythropoietin (Epo) is a well-known hallmark of PV, and erythroid colonies can be generated in vitro in the absence of exogenous cytokines. The variations of plasma Epo levels can enable the differentiation between primary and secondary erythrocytosis; however, these variations are only the consequence of autonomous erythroid growth and are not related to the clinical phenotype. Boissinot et al. provided evidence that inflammation-linked cytokines were required for the growth of JAK2 V617F–mutated erythroid progenitors [15]. On the other hand, antiinflammatory hepatocyte growth factor (HGF) and IL-11 were also upregulated in PV, and both cytokines contributed to the proliferation of PV erythroblasts [15]. Moreover, Tefferi et al. showed that IL-8, IL-2R, IL-12, IL-15 and interferon-inducible protein (IP)-10 were independent predictive factors of inferior quality of life in PMF patients and that some phenotype cytokine associations could be highlighted [16]. It is also noteworthy that the capacity of some cytokines to modulate hemostasis plays a role in the manifestation of thrombotic events, and one could thereby hypothesize that they might also contribute to the vascular complications of MPN. Altogether, these data encourage further clinical evaluation of plasma cytokines in PV and ET. For this reason, we conducted a comparative study of the plasma cytokine profiles of MPN patients using a multiplex immunoassay as a function of PV, ET, JAK2 V617F status, and the presence (or not) of a thrombohemorrhagic complication before the diagnosis.

Methods Patients and sample preparation Thirty-eight patients diagnosed with a myeloproliferative neoplasm between 2004 and 2007 entered the study at the University Hospital of Grenoble (France). Seventeen patients had polycythemia vera, and 21 patients had essential thrombocythemia. Approvals and informed consent were obtained according to the Declaration of Helsinki. Diagnosis was established based on the pre-2008 group

criteria for polycythemia vera and was verified by the 2008 WHO criteria. All PV patients without the JAK2 V617F mutation were Q4 positive for endogenous erythroid colonies, displayed low Epo serum values, and exhibited no primary cause for polycythemia. To eliminate a diagnosis of prefibrotic PMF, the following biological parameters were verified in the ET patients: absence of fibrous tissue from osteomedullar biopsy and absence of erythroblast myelemia, cytopenia, or cell morphology disorders such as dacryocytosis. To study the impact of previous vascular events on the cytokine profile, we selected PV and ET patients that fit into one of the following two subgroups: patients with previous vascular events (n 5 22) or patients without vascular complications (n 5 16). Similarly, we selected a significant number of JAK2 V617 negative patients (n 5 14) for comparison with JAK2 V617F positive patients (n 5 24). These criteria explain the relatively low percentage of JAK2 V617F positive PV patients (76%) in our cohort compared with the well-known general frequency of the mutation in PV (95%). Peripheral blood (plasma from a citrate tube) was collected upon diagnosis of MPN, so none of the patients were on therapy when the samples were stored. Samples were spun at 2,000 rpm for 10 min (Multifuge Kandro 3S/3S-R, Heraeus, Germany), and plasma was collected and stored at
80 C until testing. Evaluation of plasmatic cytokine levels Cytokine levels were measured using the xMAP immunoassay technology with formatted magnetic beads. Thirteen cytokines (IL-1b, IL-4, IL-6, IL-8, IL-10, IL-12(p70), granulocyte macrophage–colony-stimulating factor [GM-CSF], interferon [IFN]-g, monocyte chemotactic protein [MCP]-1, platelet-derived growth factor [PDGF]-BB, TNF-a, VEGF, and HGF) were evaluated by Bio-Plex Pro Assays (Bio-Rad, Marnes La Coquette, France). All experiments were performed according to the manufacturer’s protocol. All experiments were run in duplicate. Fifty micrograms of each type of 13 coupled beads, which corresponded to each cytokine being studied, were added to each well of a 96-well plate. Beads were washed twice, 50 mL of the eight standard set or samples were diluted 1:4, added to the wells, Q5 and incubated for 30 min with shaking at RT. After washing, Q6 25 mL of diluted detection antibodies were added and incubated for another 30 min. Finally, 50 mL of diluted streptavidinphycoerythrin was added, and the wells were washed three times before adding 125 mL of assay buffer to each well. The plate was read with the Bio-Plex200 System (Bio-Rad), and data acquisition and analysis were accomplished with the Bio-Plex Manager 5.0 (Bio-Rad) software at a low PMT. The fluorescence signal of a Q7 minimum of 50 beads per cytokine was evaluated and recorded. The concentration (pg/mL) of each cytokine was calculated from its standard curve. Statistical analysis All quantitative variables were expressed as medians. The comparison of the data was carried out using the Mann Whitney t test. Correlations between the cytokine levels and the main biological parameters were investigated by the Spearman test. Statistical analyses were performed using StatView 5.0 (SAS Institute, Cary, North Carolina), except for matrix correlations, which were realized with SAS 9.3 (SAS Institute), and graph correlations, which were realized with R 14.0 (R Development Core Team, Vienna, Austria).

EXPHEM3095_proof ■ 2-5-2014 16-50-16

156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

3

and one case of bleeding. Vascular complications were 211 Results predominantly venous thrombotic complications in PV 212 (73%) and ischemia complications in ET (70%), but the 213 Clinical characteristics of patients rates of complication were not significantly different in 214 We obtained cytokine profiles from the plasma of 38 each disease. 215 patients, 17 with PV and 21 with ET. Gender ratios were The 38 patients also displayed several clinical pheno216 0.89 and 1.33 for PV and ET, respectively. There was no types in 2013: eight lost sight, one had leukemia, six died 217 statistically significant difference between the PV and ET (two after acute leukemia), and 23 had hematologic stability. 218 groups for age or blood parameters such as leukocyte, 219 lymphocyte, and number of polynuclear neutrophil cells. 220 However, we did observe a statistically significant differCytokine profiles in PV and ET 221 ence between the two groups for red cell counts, platelet The cytokines assessed in this study included: pro222 counts, hemoglobin (Hb) levels, and hematocrit (Ht) levels; inflammatory cytokines (IL-1b, IL-6, IL-8, IL-12(p70), 223 the p value was less than 0.001 (Table 1). TNF-a, and IFN-g), anti-inflammatory cytokines (IL-4, IL-10), one chemokine (MCP-1), and growth factors (GM224 Q8 Twenty-two patients (10 PV, 12 ET) each had at least 225 one thrombohemorrhagic manifestation before the diagCSF, PDGF-BB, VEGF, and HGF). We were able to mea- Q9 226 nosis of MPN. This included arterial and venous thrombosis sure the concentration of every cytokine except IL-1b, 227 228 Table 1. Characteristics and cytokine concentration of MPN patients Q12 229 Group with Group with essential 230 polycythemiavera (n 517) trombocythemia (n 521) Mann Whitney test 231 Median (range) Median (range) P value Characteristics 232 233 Age (years) 64 (26–80) 67 (44–80) 0.988 234 Leukocytes (x106/L) 10.0 (6.6–27.2) 10.7 (5.1–19.8) 0.975 235 Q13 Redcells (x1012/L) 6.15 (5.09–7.95) 4.46 (3.14–5.78) *!0.001 501 (221–936) 872 (310–1443) *!0.002 Platelets (x106/L) 236 Haemoblobin (g/L) 165 (138–205) 135 (105–166) *!0.003 237 Hematocrit (fraction of 1) 0.51 (0.40–0.62) 0.39 (0.32–0.50) *!0.004 Q14 238 Neutrophils (x106/L) 7.10 (4.40–25.0) 6.30 (2.30–15.2) 0.576 239 Lymphocytes (x106/L) 1.50 (0.60–3.00) 2.30 (0.30–3.00) 0.306 240 Prothrombin time (%) 82 (34–100) 78 (35–92) 0.960 Partial Prothrombin time (second) 1.29 (1.18–3.00) 1.25 (0.97–1.60) 0.291 241 Jak2 mutation n(%)[IC95%] 13 (0.76) [50.1–93.2] 11 (0.65) [38.3–85.8] O0.900 242 Vascular complication n(%)[IC95%] 243 Yes 10 (0.59) [0.33–0.82] 12 (0.57) [0.34–0.78] O0.900 244 No 7 (0.41) [0.18–0.67] 9 (0.43) [0.22–0.66] 245 246 Human cytokine Median (range) Median (range) P value 247 IL-1b (pg/mL) !2.19 !2.19 NA 248 IL-4 (pg/mL) 5.25 (1.59–10.6) 8.36 (3.16–16.2) **0.006 249 IL-6 (pg/mL) 11.8 (5.12–62.1) 10.8 (6.04–39.5) 0.357 250 IL-8 (pg/mL) 16.0 (8.66–31.6) 26.1 (8.66–60.8) **0.003 IL-10 (pg/mL) 12.9 (8.51–63.1) 28.8 (7.96–60.7) 0.220 251 IL-12(p70) (pg/mL) 18.6 (9.45–47.5) 34.7 (10.9–65.9) 0.094 252 HGF (pg/mL) 283.6 (115.9–835.5) 277.9 (97.7–923.3) 0.941 253 GM-CSF (pg/mL) 32.5 (14.1–57.5) 45.3 (20.9–82.9) ***0.026 254 IFN-g (pg/mL) 209.5 (109.8–426.4) 334.7 (89.1–921.0) **0.007 255 MCP-1 (pg/mL) 32.2 (7.53–43.1) 36.2 (24.5–59.0) **0.009 PDGF-BB (pg/mL) 1211 (189.1–3716) 2837 (592.7–6597) ***0.015 256 TNF-a (pg/mL) 37.6 (23.9–76.0) 40.4 (21.5–108.1) 0.671 257 VEGF (pg/mL) 50.8 (14.3–381.8) 202.6 (22.5–474.3) ***0.017 258 259 GM-CSF 5 Granulocyte macrophage–colony stimulating factor; HGF 5 hepatocyte growth factor; IFN 5 interferon; IL 5 interleukin; MCP 5 monocyte 260 chemotactic protein; MPN 5 myeloproliferative neoplasm; PDGF 5 platelet derived growth factor; TNF 5 tumor necrosis factor; VEGF 5 vascular endothelial growth factor. 261 Bold values represent XX. 262 Italic values represent XX. 263 *Values represent XX. 264 **Values represent XX. 265 Q15 ***Values represent XX.

EXPHEM3095_proof ■ 2-5-2014 16-50-16

266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320

4

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

significant difference between the PV and ET patients in the 321 which had serum levels below the quantification limit of JAK2 V617F-negative cohort. 322 2.19 pg/mL. The median value of all of the other cytokines 323 was determined and used for comparison between the 324 groups (Table 1). The coefficient of variation of all of the Cytokine profile according to vascular events 325 cytokines was below 10%. Data on vascular events were available for 32 patients. 326 All of our patients displayed cytokine levels that were We compared patients with vascular complications with 327 higher than the normal physiologic levels of human cytopatients without vascular complications. The comparison 328 kines when measured using the Bio-Plex Pro cytokine of the two subgroups did not show any statistically significant 329 assays in the same conditions [18]. Except for IL-1b, all differences for age, JAK2 mutational status, or biological 330 of the cytokines were significantly increased in both MPNs. parameters (leukocytes, red cells, platelets, neutrophils and 331 The analysis of each cytokine level within the MPNs lymphocytes number, hemoglobin, and hematocrit). Except 332 showed different profiles for PV and ET, with a statistically for IL-12(p70), which was increased in MPN patients 333 significant difference for IL-4, IL-8, GM-CSF, IFN-g, without vascular complications (p 5 0.047), there were no 334 MCP-1, PDGF-BB, and VEGF (Table 1). It is noteworthy significant differences in other cytokine levels between 335 that each cytokine concentration was higher in the ET patients with and without vascular complications (Fig. 3A). 336 patients compared with the PV patients. The comparison of cytokine profiles within the PV group 337 and the ET group according to the vascular complications 338 status of each patient revealed significant differences for Impact of JAK2 status 339 IL-12(p70) (p 5 0.025) and GM-CSF (p 5 0.040) in the There was no significant difference between the cytokine 340 PV patients (Fig. 3B). Both parameters were increased in concentrations of the patients with the JAK2 V617F mutation 341 patients without complications. The median and range for (positive) and the cytokine levels of the wild type patients 342 IL-12(p70) and GM-CSF were 14.1 pg/mL (9.45–26.5) and (negative). However, separate analyses of the results of 343 24.0 pg/mL (14.1–46.0), respectively. The median and range each MPN revealed statistically significant differences 344 for PV patients with complications versus PV patients between the positive and negative patients for TNF-a 345 without complications were 29.7 pg/mL (13.2–47.5) and (p 5 0.042) and PDGF-BB (p 5 0.019) in PVand ET, respec346 39.0 pg/mL (27.5–57.5), respectively. No statistically signiftively (Fig. 1A–B). The median and the range for TNF-a be347 icant difference was observed in the ET population. tween the positive and negative PV patients was 13.9 pg/mL 348 Next, we compared the cytokine concentrations of PV Q10 (range, 1.64–44.6) and 37.6 pg/mL (23.9–76.0). The median 349 and ET patients who have had a vascular complication. A and range for PDGF-BB was 2838 pg/mL (997–6538) and 350 significant increase in IL-4, IL-8, GM-CSF, IFN-g, MCP-1 1140 pg/mL (593–1813) for ET-positive and -negative 351 and VEGF was measured in the ET population (Fig. 4). patients, respectively. 352 The corresponding medians and ranges for patients with PV Next, we focused on the patients carrying the JAK2 muta353 versus patients with ET were: IL-4 (4.87 pg/mL [1.59–8.50] tion and observed a statistically significant difference between 354 vs. 7.52 pg/mL [6.13–13.3]), IL-8 (15.5 pg/mL [9.45–25.7] PV and ET patients for IL-4 (p 5 0.007), IL-8 (p 5 0.008), 355 vs. 23.8 pg/mL [15.1–42.7]), GM-CSF (24.0 pg/mL [14.1– IFN-g (p 5 0.010), and PDGF-BB (p 5 0.012), with corre356 46.0] vs. 44.7 pg/mL [27.2–72.8]), IFN-g (179.3 pg/mL sponding median and range values of 4.48 pg/mL (1.59– 357 10.0), 14.4 pg/mL (8.66–28.6), 149.1 pg/mL (109.8–426.4), [109.8–426.4] vs. 277.8 pg/mL [209.5–635.7]), MCP-1 358 and 1144 pg/mL (189.1–3717) for JAK2 positive PV patients (31.2 pg/mL [7.53–43.1] vs. 37.7 pg/mL [29.4–56.4]), and 359 and 7.52 pg/mL (5.20–14.5), 23.0 pg/mL (15.1–50.9), 316.0 VEGF (46.1 pg/mL [14.3–128.9] vs. 197.8 pg/mL [25.8– 360 pg/mL (191.3–742.0), and 2838 pg/mL (966.6–6538) for 474.3]). Curiously, no statistically significant difference was 361 JAK2 positive ET patients (Fig. 2). There was no statistically observed for patients without vascular complications. Q11 362 363 364 365 366 367 368 369 370 371 372 373 374 Figure 1. (A, B) A box plot diagram of the distribution of cytokine concentrations (pg/mL) according to the JAK2 status in the plasma of patients with 375 polycythemia vera and essential thrombocythemia. The circles represent outliers above the 90th percentile and below the 10th percentile.

EXPHEM3095_proof ■ 2-5-2014 16-50-16

376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485

5

Figure 2. A box plot diagram of the distribution of the cytokine concentrations (pg/mL) in the plasma of essential thrombocythemia (ET) and polycythemia vera (PV) patients with the JAK2 V617F mutation. The circles represent outliers above the 90th percentile and below the 10th percentile.

Correlation of biological markers Next, we looked for correlations between the cytokine levels and the main biological parameters. Several positive correlations (above 0.5) were characterized between the cytokine concentrations and the biological parameters in each MPN subgroup. In the PV cohort, a link was observed

between each of the following pairs of parameters: IL-4 and Hb (r 5 0.53), IL-6 and lymphocyte count (r 5 0.55), TNF-a and Ht (r 5 0.56), TNF-a and lymphocyte count (r 5 0.60), MCP-1 and Hb (r 50.72), and MCP-1 and Ht (r 5 0.57) (Fig. 5A). In the ET population, polynuclear cell count had a correlation with HGF (r 5 0.55), IL-6

Figure 3. (A, B) A box plot diagram of the distribution of cytokine concentrations (pg/mL) according to the vascular complications (VC) status in the plasma of all of the MPN patients and for the set of patients with polycythemia vera. The circles represent outliers above the 90th percentile and below the 10th percentile.

EXPHEM3095_proof ■ 2-5-2014 16-50-16

486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540

6

Figure 4. A box plot diagram of the distribution of the cytokine concentrations (pg/mL) in the plasma of essential thrombocythemia (ET) and polycythemia vera (PV) patients with vascular complications. The circles represent outliers above the 90th percentile and below the 10th percentile.

(r 5 0.58), IL-12 (r 5 0.61), GM-CSF (r 5 0.50), and VEGF (r 5 0.60). Moreover, PDGF-BB levels had a correlation with red cell counts (r 5 0.57) (Fig. 5B).

Discussion The study of inhibitors of JAK2 has shed light on the functional role of inflammatory cytokines in PMF. The roles of some cytokines in PV and ET have been previously

described, but few studies have focused on cytokine profiles and their clinical and biological correlations. In this study, we performed multiple cytokine immunoassays using multiparametric bead-based xMAP technology. This technology allows the simultaneous measurement of up to 100 parameters in several hours with high sensitivity and specificity in a unique small volume sample of plasma. Although quantitative differences in the concentrations of some cytokines were reported according to the measurement method, the

print & web 4C=FPO

541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

Figure 5. (A, B) The correlation between cytokines and biological parameters in polycythemia vera and essential thrombocythemia. The color intensity and size of each circle are related to the value of the spearman coefficient close to 1 (dark blue) or -1 (dark red). The red rectangles correspond to a coefficient above 0.5.

EXPHEM3095_proof ■ 2-5-2014 16-50-17

596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705

relative differences were comparable [17]. Moreover, the performance of the Bio-Plex kit used in this study was validated for precision and accuracy [18]. To our knowledge, this study was the first time that a global cytokine analysis was conducted on samples from patients with PV or ET. Our first observation was the confirmation of an abnormal expression of several cytokines in patients with PV and ET compared with healthy patients. Except for IL-1b, all of the cytokines evaluated in both MPNs displayed higher levels compared with normal physiologic values. Overexpression of some of the cytokines in patients with PV, ET, or both MPNs has been previously reported. For example, it has already been shown that PV patients have increased levels of IL-6, IL-8, IL-11, and HGF [15,19,20] and that both PV and ET patients can have higher levels of VEGF [21]. Our data confirmed the increase of the levels of these cytokines and identified the increase of others, suggesting that some cytokines implicated in the inflammatory response or in the control of hematopoiesis might participate in PV and ET physiopathology. Our data, in addition to published results on PMF, demonstrated that the presence of an inflammatory response is a common feature of Philadelphia negative myeloproliferative neoplasms. Prospective additional studies are needed to further characterize the dynamics of this response in MPN patients. Our second observation was the identification of a cytokine pattern that enabled the differentiation between PV and ET patients. This pattern consisted of the following seven cytokines, which were overexpressed in ET patients compared with PV patients: IL-4, IL-8, GM-CSF, IFN-g, MCP-1, PDGF-BB, and VEGF. These dissimilar expressions of inflammatory cytokines as well as certain growth factors in PV and ET patients suggest that the impact of JAK2 inhibitors may be different for PV and ET patients. Next, we evaluated the impact of the JAK2 V617F status on cytokine profiles to further explain these differences. No difference in cytokine expression was shown when all of the V617F positive MPNs patients were compared with the V617F negative patients. However, the individual analysis of the results of each MNP revealed that TNF-a in PV patients and PDGF-BB in ET patients were differentially expressed in positive and negative patients. TNF-a levels were previously reported to have a correlation with the JAK2 V617F allele burden [22], which may explain why we only observed differences in PV patients that had the highest allele burden rates. Specific increases in PDGFBB levels in the positive ET patients most likely reflect the role of JAK2 in thrombopoietin-regulated megakaryopoiesis and thereby in PDGF-BB synthesis by megakaryocytes. This result suggests that the increase in PDGF-BB levels in ET patients could potentially be a biomarker of JAK2 activation. Therefore, a PDGF-BB assay may help to identify the ET patients who would benefit from treatment with JAK2 inhibitors. Interestingly, the comparison

7

of the subgroups of positive ET and positive PV patients showed significant differences for four (IL-4, IL-8, IFN-g, and PDGF-BB) of the seven cytokines initially described as differentiating PV and ET. These data indicate that the JAK2 V617F mutation may be partially responsible for the differences in the cytokine profiles observed in PV and ET patients. Similar to PDGF-BB, these four cytokines may lead to the identification of a JAK2 V617F–dependent profile. Taken together, our results suggest that there is most likely a JAK2 V617F–driven and a JAK2–independent inflammatory response in MPN patients. The correlation between cytokine levels and blood cell counts shows some interesting trends. In PV patients, the red mass parameters such as Hb and Ht had correlations with IL-4, MCP-1, and TNF-a, suggesting a correlation between red cell mass and the regulation of some cytokines. Cytokine levels in ET patients primarily had correlations with the polynuclear neutrophil count, not with platelets as we had expected. Several cytokines, including IL-6, which is implicated in platelet regulation and participates with IL-12 in the inflammatory response, were significantly associated with the polynuclear cell count. Taking into account that the polynuclear neutrophil count is a surrogate marker in ET, these data suggest that, unlike PV, cytokine alterations in ET are less driven by myeloproliferation. Leukocytosis could possibly be a prognostic marker in ET. Therefore, cytokine levels could also be considered as prognostic biomarkers for ET. On the other hand, the Spearman correlation coefficients in PV and ET were weak. One explanation for this may be that one cytokine level does not necessarily correlate with only one biological parameter. Altogether, these data show that the mechanisms of inflammatory responses and the origins of cytokine alterations are not the same for every patient and could be different in each MPN. The current estimate of the individual vascular risk related to a particular therapeutic choice (cytoreductive drugs associated or not with antiplatelet) [4] is still based on a combination of risk factors such as age and previous thrombotic history. That is why the identification of biological markers that will predict vascular complications and risk would be a great asset. In our study, a lower IL-12(p70) level was associated with vascular complications in every patient in the entire MPN cohort. This may suggest that IL-12 has a protector effect on complications; however, this hypothesis requires additional study. In fact, similar results were described in MPN-free patients with venous thrombosis, although no association with the risk determinants for venous thrombosis was apparent [23]. Similarly the concentrations of IL-12 (p70) and GM-CSF were lower in the PV group with complications (mainly venous thrombosis). One explanation may be that these reductions reflect a dysfunction in the cells involved in the synthesis of these cytokines. Previous studies have already reported that a decrease in IL-10 levels relates to

EXPHEM3095_proof ■ 2-5-2014 16-50-20

706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760

8

761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815

E. Pourcelot et al./ Experimental Hematology 2014;-:-–-

a dysfunction of the endothelial cells in idiopathic venous thrombosis [24]. Therefore, the reduction of GM-CSF levels could reflect a dysfunction of the cellular microenvironment that could impede hemostatic balance. We did not observe any increased cytokine levels in the groups of patients with vascular complications, although an increase of IL-6, IL-8, MCP-1, and TNF-a in MPN-free patients with vascular complications has been reported [23,25–27]. However, Christiansen et al. suggested that an alteration of the inflammatory profile is more a result, rather than a cause, of thrombosis [26]. This hypothesis is in accord with our data, as our samples were not collected in the acute phase of vascular manifestation (samples of blood were obtained a minimum of two months after the complication). However, the absence of cytokine differences between PV and ET patients without vascular manifestations indirectly reinforces the hypothesis of cytokine involvement in vascular complications. Finally, we showed that MPN patients harbored significantly increased cytokine levels (including IL-6, IL-8, and TNF-a) due to their hematologic disease. In this context, a modification of the inflammatory cytokines related to venous thrombosis may be masked by the specific pattern of MPN. In conclusion, this original preliminary study showed that PV and ET patients display specific alterations of their plasmatic cytokine pattern that could be due in part to their JAK2 V617F status. Moreover, we highlighted the potential predictive power of IL-12(p70) levels, which could help to stratify vascular risk in PV and ET patients. Altogether, our data emphasize the existence of a global inflammatory reaction with cytokine overexpression in these two types of MPN, suggesting that cytokine profiles as observed in PMF might be useful for a better stratification of PV and ET patients, especially for therapeutic decisions involving JAK2 inhibitors.

Acknowledgments This work was supported by the clinical research department of Grenoble Hospital, Grenoble, France. We thank Dr. Laurent Pelletier (Plateforme Transcriptome et Proteome Clinique) for the use of the Bio-Plex products.

Conflict of interest disclosure No financial interest/relationships with financial interest relating to the topic of this article have been declared.

References 1. Adamson JW, Fialkow PJ, Murphy S, Prchal JF, Steinmann L. Polycythemia vera: stem-cell and probable clonal origin of the disease. N Engl J Med. 1976;295:913–916. 2. Campbell PJ, Green AR. The myeloproliferative disorders. N Engl J Med. 2006;355:2452–2466.

3. Marchioli R, Finazzi G, Landolfi R, et al. Vascular and neoplastic risk in a large cohort of patients with polycythemia vera. J Clin Oncol. 2005;23:2224–2232. 4. Spivak JL. Polycythemia vera: myths, mechanisms, and management. Blood. 2002;100:4272–4290. 5. Passamonti F, Rumi E, Pietra D, et al. Relation between JAK2 (V617F) mutation status, granulocyte activation, and constitutive mobilization of CD34þ cells into peripheral blood in myeloproliferative disorders. Blood. 2006;107:3676–3682. 6. Landolfi R. Bleeding and thrombosis in myeloproliferative disorders. Curr Opin Hematol. 1998;5:327–331. 7. Elliott MA, Tefferi A. Thrombosis and haemorrhage in polycythaemia vera and essential thrombocythaemia. Br J Haematol. 2005; 128:275–290. 8. Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med. 2005;352: 1779–1790. 9. Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet. 2005;365:1054–1061. 10. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature. 2005;434:1144–1148. 11. Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis. N Engl J Med. 2012;366:787–798. 12. Verstovsek S, Kantarjian H, Mesa RA, et al. Safety and efficacy of INCB018424, a JAK1 and JAK2 inhibitor, in myelofibrosis. 2010. N Engl J Med. 2010;363:1117–1127. 13. Pardanani A, Gotlib JR, Jamieson C, et al. Safety and efficacy of TG101348, a selective JAK2 inhibitor, in myelofibrosis. J ClinOncol. 2011;29:789–796. 14. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebocontrolled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012;366:799–807. 15. Boissinot M, Cleyrat C, Vilaine M, et al. Anti-inflammatory cytokines hepatocyte growth factor and interleukin-11 are over-expressed in Polycythemia vera and contribute to the growth of clonal erythroblasts independently of JAK2V617F. Oncogene. 2010;30:990–1001. 16. Tefferi A, Vaidya R, Caramazza D, et al. Circulating interleukin (IL)8, IL-2R, IL-12, and IL-15 levels are independently prognostic in primary myelofibrosis: a comprehensive cytokine profiling study. J Clin Oncol. 2011;29:1356–1363. 17. Chowdhury F, Williams A, Johnson P. Validation and comparison of two multiplex technologies, Luminex and Mesoscale Discovery, for human cytokine profiling. J Immunol Methods. 2009;340:55–64. 18. Kokkonen H, Soderstrom I, Rocklov J, et al. Up-regulation of cytokines and chemokines predates the onset of rheumatoid arthritis. Arthritis Rheum. 2010;62:383–391. 19. Wickenhauser C, Thiele J, Lorenzen J, et al. Polycythemia vera megakaryocytes but not megakaryocytes from normal controls and patients with smokers polyglobuly spontaneously express IL-6 and IL-6R and secrete IL-6. Leukemia. 1999;13:327–334. 20. Hermouet S, Godard A, Pineau D, et al. Abnormal production of interleukin (IL)-11 and IL-8 in polycythaemia vera. Cytokine. 2002;20: 178–183. 21. Musolino C, Calabro L, Bellomo G, et al. Soluble angiogenic factors: implications for chronic myeloproliferative disorders. Am J Hematol. 2002;69:159–163. 22. Fleischman AG, Aichberger KJ, Luty SB, et al. TNFalpha facilitates clonal expansion of JAK2V617F positive cells in myeloproliferative neoplasms. Blood. 2011;118:6392–6398. 23. Reitsma PH, Rosendaal FR. Activation of innate immunity in patients with venous thrombosis: the Leiden Thrombophilia Study. J Thromb Haemost. 2004;2:619–622.

EXPHEM3095_proof ■ 2-5-2014 16-50-20

816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870

E. Pourcelot et al./ Experimental Hematology 2014;-:-–871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935

24. Poredos P, Jezovnik MK. In patients with idiopathic venous thrombosis, interleukin-10 is decreased and related to endothelial dysfunction. Heart Vessels. 2011;26:596–602. 25. vanAken BE, Reitsma PH, Rosendaal FR. Interleukin 8 and venous thrombosis: evidence for a role of inflammation in thrombosis. Br J Haematol. 2002;116:173–177.

9

26. Christiansen SC, Naess IA, Cannegieter SC, et al. Inflammatory cytokines as risk factors for a first venous thrombosis: a prospective population-based study. PLoS Med. 2006;3:e334. 27. vanAken BE, den Heijer M, Bos GM, van Deventer SJ, Reitsma PH. Recurrent venous thrombosis and markers of inflammation. Thromb Haemost. 2002;83:536–539.

EXPHEM3095_proof ■ 2-5-2014 16-50-20

936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000