Surveillance Imaging after Autologous Hematopoietic Cell Transplantation Predicts Survival in Patients with Diffuse Large B Cell Lymphoma

ARTICLE IN PRESS Biol Blood Marrow Transplant && (2019) &&& &&&

Biology of Blood and Marrow Transplantation journal homepage: www.bbmt.org 1

59

2

60

3

Surveillance Imaging after Autologous Hematopoietic Cell Transplantation Predicts Survival in Patients with Diffuse Large B Cell Lymphoma

4 5 6 7 8 9 10 11 12

17

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

70

40 41 42 43 44 45 46 47 48

51 52 53 54 55

Keywords: Autologous hematopoietic cell transplantation Deauville score Diffuse large B cell lymphoma PET Surveillance

A B S T R A C T The utility of surveillance imaging after autologous hematopoietic cell transplantation (AHCT) in relapsed/refractory diffuse large B cell lymphoma (DLBCL) remains unclear. The purpose of this study was to determine whether surveillance imaging predicts survival after AHCT. At the University of Minnesota, serial imaging for early relapse detection has been used prospectively for all consecutive AHCT recipients treated since 2010. The present analysis included 91 AHCT recipients with DLBCL who underwent 18F-fluorodeoxyglucose positron emission tomography (18F-FDG-PET) scan at day +100 post-AHCT. 18F-FDG-PET parameters included the Deauville (D) 5-point scale, peak standardized uptake values (SUVmax), total legion glycolysis (TLG), and total metabolic tumor volume (TMTV). Survival of patients with clinically symptomatic versus asymptomatic radiographically detected relapsed DLBCL after AHCT was compared. Sixty patients experienced relapse; 35% was detected on day +100 surveillance PET scan. 5-year overall survival (OS) by 18F-FDG-PET scan at day +100 post-AHCT was significantly lower in D4 and D5 patients (37%; 95% confidence interval [CI], 14% to 100% versus 25%; 95% CI, 43% to 89%) compared with patients with D1 and D2 (62%; 95% CI, 43% to 89% versus 62%; 95% CI, 46% to 84%). TLG and TMTV were not prognostic. SUVmax at day +100 varied from 1.5 (D1) to 17.9 (D5). In multivariate analysis, only SUVmax was predictive of relapse and OS; mortality increased 1.8-fold with each SUVmax doubling (hazard ratio [HR], 1.8; 95% CI, 1.3 to 2.3; P < .01). At a median follow-up of 3.3 years (range, 1 to 12 years), lymphoma-related mortality was 1.8-fold higher among patients whose relapse was detected clinically (symptomatic) versus radiographically on surveillance scan (HR, 1.8; 95% CI, .9 to 3.4; P = .08). In patients with relapsed/refractory DLBCL, a routine PET imaging at day +100 post-AHCT detects asymptomatic relapse and high SUVmax identifies patients with poor expected survival of less than 1 year. Identifying this high-risk cohort can potentially highlight patients who might benefit from preemptive interventions to prevent or delay relapse. © 2019 Published by Elsevier Inc. on behalf of the American Society for Transplantation and Cellular Therapy

58

73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95

INTRODUCTION Diffuse large B cell D23X X lymphoma (DLBCL) is an aggressive lymphoma constituting approximately 30% to 40% D24X X of all nonHodgkin lymphomas [1]. Conventional first-line chemotherapy for DLBCL yields a 5-year overall survival (OS) rate of 60% to 70% D25X X [1]; hDowever, 26X X many patients either relapse (30% to 40%) D27X X or have initially refractory disease (»10%) [2 5]. Those patients with disease sensitive to salvage chemotherapy may benefit from high-dose chemotherapy followed by autologous hematopoietic cell transplantation (AHCT).

Financial disclosure: See Acknowledgments on page 6. *Correspondence and reprint requests: Veronika Bachanova, MD, PhD, Division of Hematology, Oncology, and Transplantation, University of Minnesota Medical Center, 420 Delaware Street SE, MMC 480, Minneapolis, MN 55455, USA. E-mail addresses: [email protected], [email protected] (V. Bachanova).

The pDrognosis 28X X for patients who relapse after AHCT is unfavorable [5,6]. Survival is particularly poor for patients who fail within the 1-D D29X X year 30X X post-AHCT period [6]. Several recent studies aimed to determine high-risk features and preDtransplantation 31X X characteristics that portend a particularly unfavorable prognosis [6-9]. Gisselbrecht et al [9] reported that a duration of initial remission <12 months or not achievement of complete D32X X remission with primary therapy predicted poor response rate and inferior event-free survival. Costa et al [7] showed that patients with DLBCLD3X X with MYC translocation, intermediate high/high, National Comprehensive Cancer Network-International Prognostic IndexD34X X at the time of primary therapy failure and the presence of primary progression experienced a 2-year OS of only 14D% 35X X post-AHCT. On the basis of these findings, there are ongoing efforts to incorporate novel risk factors to better risk-Dstratify 36X X patients who may benefit from experimental strategies.

56 57

69

72 Article history: Received 10 July 2019 Accepted 12 October 2019

49 50

68

71

37 39

65

Division of Hematology, Oncology, and Transplantation, Department of Medicine, University of Minnesota Medical Center, Minneapolis, Minnesota, USA

36 38

64

67

18 19

63

1X XMalvi XD SavaniD2X X, D3X XMehmet GencturkD4X X, D5X XRyan ShanleyD6X X, D7X XZuzan CayciD8X X, D9X XChristopher WilkeD10X X, D1X XErica D. WarlickD12X X, D13X XFiona HeD14X X, D15X XMurali JanakiramD16X X, D17X XDaniel J. WeisdorfD18X X, D19X XClaudio G. BrunsteinD20X X, D21X XVeronika BachanovaD2X X*

14 16

62

66

Q1

13 15

61

96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114

https://doi.org/10.1016/j.bbmt.2019.10.013 1083-8791/© 2019 Published by Elsevier Inc. on behalf of the American Society for Transplantation and Cellular Therapy

115 116

ARTICLE IN PRESS 2

117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140

M. Savani et al. / Biol Blood Marrow Transplant && (2019) &&& &&&

There are D37X X limited data on D38X X the evaluation of post-AHCT surveillance imaging tools for early relapse detection. 18F-fluorodeoxyglucose (D39X18 X F-FDG) positron emission tomography/computed tomography (PET/CT) has become a standard imaging modality for staging, monitoring treatment response, and guiding optimal treatment strategies in FDG-avid lymphomas [10-12]. PET/CT before AHCT has shown utility in identifying patients with DLBCLD40X X at high risk for treatment failure and also serve as D41X X a prognostic tool for predicting D42X X long-term clinical outcomes [10,13]. The 5-point Deauville (D) scale uses the mediastinal and liver blood pool as reference points, and pre-AHCT D-score has served as a reliable parameter for predicting D43X X outcomes in DLBCL AHCT reciDpients 4X X [10,13-15]. Although our D45X X center and others usDe 46X X routine surveillance imaging post-AHCT, there are currently no evidence-based guidelines for an optimal D47X X follow-up strategy for patients with D48X X DLBCL in D49X X remission post-AHCT. Moreover, the prognostic implications of specific day +100 PET parameters remain unclear. Novel interventions are now available for patients who relapsDe 50X X after AHCT, making early disease recognition critical to enable intervention before D51X X the onset of symptomD D52X X 53X X progression or advanced disease. We evaluated the impact of surveillance 18 D54X X F-FDG-PET D5X X semiquantitative metabolic parameters at day +100 post-AHCT on subsequent clinical outcomes.

141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158

D56XMETHODS X Study Design We performed a single-center retrospective analysis to study the impact of day +100 PET/CT on AHCT outcomes. We examined 131 consecutive adult patients (age 18 yearsD)57X X with relapsed/refractory DLBCL and preDtransplantation 58X X PET/CT treated with AHCT at the University of Minnesota Medical Center between 2000 D59X X and 2016. D60X X Ninety-one patients underwent a preDplanned 61X X day +100 post-AHCT (interquartile range, 97 toD62X X 103 days) surveillance PET/CT. Twelve patients (9%) relapsed before day D63X X +100, and for 28 patients, we did not have access to their images for performing D64X X Deauville assessment. The primary objective was to analyze post-AHCT day +100 18F-FDG-PET parameters including D-Dscore, 65X X peak standardized uptake values (SUVmax), total lesion glycolysis (TLG), and total metabolic tumor volume (TMTV) as predictors of relapse and survival after AHCT. Secondary objectives were to determine the associations of patient and transplantation factors with the risk of having abnormal PET/CT at day +100. In addition, we compared outcomes D6X X according to relapse presentation: symptomatic clinical, defined by palpable lymphadenopathy in addition to the presence of B symptoms versus asymptomatic radiographic evidence of relapse. Relapse was confirmed by biopsy in all cases. The study was approved by the University of Minnesota’s Institutional Review BoardD,67X X and all subjects provided signed informed D68X X consent to collect data.

159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181

Transplantation Procedures and Surveillance All patients were treated on prospective protocols using the conditioning regimen BEAM (carmustine, etoposide, cytarabine, and melphalan) or cyclophosphamideD69X 6 X 0 mg/kg i.v. for 2 D70X X days with D71X X total body irradiation of 1350 cGyD.72X X All patients underwent baseline staging PET/CT before stem D73X X cell mobilization and collection D74X X and AHCT. Post-transplantation standard monitoring for all patients included surveillance CT scan of the chest/abdomen/pelvis (CAP) and/or neck with i.v.D75X X and oral contrast at day +28, 18F-FDG PET/CT at day +100,D76X aX nd CT CAP again at 6, 12, and 24 months. Consolidative Radiation TherapyD7X X Consolidative radiation therapy (CRT)D78X X was recommended to those individuals with persistent nodal masses D2 79X X cm on day +28 CT scan, with sites clinically suspiciDous 80X X for residual disease, or considered at high risk D81X X of relapseD.82X X Not all individuals meeting these criteria D83X X underwent CRT D84X X due to previous radiation, comorbidities, or progressive disease on post-AHCT imaging. PET/CT Imaging Technique and Analysis Most scans were performed on a Biograph mCT Flow 64-4R PET/CT system (Siemens Healthineers, Malvern, PAD)85X X at 60 minutes after i.v.D86X iX njection of D87X18 X F-FDG D8X X (370 to 555 D89X X MBq), using continuous table motion acquisition mode. The tDable 90X X speed was set atD91X X 1.4 mm/second from the vertex to pelvis and 2.9 mm/second from pelvis to feet. For some patients, scans were performed with scanners that had similar D92X X specifications at outside institutions and were uploaded to a picture archiving and communication system for interpretation.

All PET/CT images were retrospectively re-reviewed by a nuclear medicine radiologist (M.G), who was blinded to clinical outcomes. Treatment response was assessed using the Lugano criteria, and each hypermetabolic lesion was scored visually using the 5-point Deauville criteria [16]. In addition to this qualitative analysis, quantitative analysis was performed in all patients with a measurable PET-positive lesion of any size using commercially available software (Syngo.Via; Siemens Healthcare, Forchheim, Germany). The vDolume 93X X of interest was selected by the nuclear radiologist for each FDG-avid lesion, whDereas 94X X the SUVmax, TMTV, and TLG were automatically calculated by the software. TMTV was calculated as the sum of all individual lesion volumes for each PET study and reported in cm3. For patients with multiple lesions, the 6 largest lesions were selected for analysis. TLG was calculated as the sum of products of SUVmax (for each region of interest) and metabolic tumor volume in all lesions. For TMTV and TLG calculations, 41% of the SUVmax was defined as the threshold [15].

182

Statistical Analysis ODSD 95X X 96X X from the date of AHCT was estimated by the Kaplan-Meier method. Relapse/progression from the date of AHCT was estimated by the cumulative incidence function, considering death before relapse as a competing risk. The primary analysis was to study single variable associations between postAHCT PET parameters using Cox regression for OS and Fine and Gray D97X X regression relapse [17]. No adjustment for multiple tests were performed. The primary variables of interest were D-score, SUVmax, TMTV, and TLG. All were measured by PET both pre-AHCT and post-AHCT, and single variable models were fit for each pre-AHCT variable to predict OS and relapse. In addition, associations of post-AHCT PET variables with OS were analyzed, conditional on survivDal 98X X until post-AHCT PET. Exploratory analysis demonstrated right-skewed distributions for SUV, TMTV, and TLG, and a relative increase in their value produced a better model fit than an additive difference. Thus, these variables were log-transformed for analysis, and HRD9X X estimates are reported per doubling of the variable’s value in standard units. Secondary variables analyzed for OS and relapse were sexD,10X X age, conditioning regimen, disease status at transplantation, Karnofsky PDerformance 10X X SDtatus 102X X (KPS), number of CD34 cells infused, year of AHCT, and receipt of CRTD D103X X .104X X Logistic regression analysis of the odds of having a post-AHCT D-score of D4 or D5 was performed for the pre-AHCT clinical and PET variables. A Cox multiple regression model was used to estimateD105X Xthe risk of death after relapse, predicted by clinical or radiographic detection of relapse, and the time from AHCT to relapse detection. All statistical analyseDs106X X wDere 107X X performed using RD108X vX ersion 3.4 [18].

195

RESULTS Patient Characteristics Patient, disease, and transplantation characteristics of 91 patients treated with AHCT for DLBCL between 2000 and 2016 D109X X D10Xare X summarized in Table 1. The median age at D1X X transplantation was 60 years (range, 19 to 73 D12X X years), and 59% of the patients were male. Patients in first complete remission (CR1; n D13X X = 14) included 5 with transformed lymphomaD,14X X 6 with aggressive B D15Xcell X lymphoma with C-MYC and BCL-2/BCL-6 reDarrangementD 16X X ,17X X and 3 with other aggressive histologies (intravascular B cell lymphoma, Richter’s transformationD). 18X X All patients underwent pre-AHCT PET/CT imaging at a median of 44 days before AHCT D19X X (range, 32 to 55 D120X X days). The conditioning regimen included mostly BEAM and cyclophosphamideD/total 12X X body irradiationD.12X X Disease status at transplantation included 18% of patients in D123XCR1D X ,124X X and 13% with D125X X chemosensitive primary induction failureD/126X X relapse. KPSD127X Xwas 90% in 74% of the cohort. Ten patients underwent CRTD.128X X All surviving patients underwent surveillance day +100 PET/CT (rangeD,129X X 97 to 103 D130X X days post-DAHCT). 13X X The mDedian 132X X duration of follow-up was 3.3 years (range, 1 to 12 D13X X years).

183 184 185 186 187 188 189 190 191 192 193 194 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236

Outcomes The 5D-year 134X X OS was 51% (95% cDonfidence 135X X iDnterval 136X X [DCI],D 137X X 138X 4 X 1% to 63%), and the cumulative incidence of relapse after AHCT was 50% (95% CI, 39% to 59%) (Figure 1A andD139X X B). OS by day preAHCT is shown in Figure 1C. Twelve patients (9%) relapsed before day D140X X +100 and were not included in our analyses. In univariate analysis, OSD14X X or risk of relapse was not affected by year of AHCT, sex, age, type of conditioning regimen, KPSD,142X X or receipt D143X X CRTD.14X X Patients with primary refractory disease or refractory of relapseD145X X before AHCT had a 2-fold lower OS (HR, 2.26; 95% CI,

237 238 239 240 241 242 243 244 245 246

ARTICLE IN PRESS M. Savani et al. / Biol Blood Marrow Transplant && (2019) &&& &&&

247 248

Table 1 Patient and Disease Characteristics

249

Characteristic

250

Male sex

251

Age, yr (median, 58)

252 253 254 255 256 257 258 259

Table 2 Univariate Analysis of Pretransplantation Variables on OS Patients, n (%)

Variable

75 (57)

316

1.63

.9-2.9

.09

317

80%

.83

.4-1.8

.64

90%

.64

.4-1.2

.14

PIF, relapse-resistant

2.26

1.2-6.4

.03

PR2+, CR2+

1.68

.9-3.2

.12

1.4

.5-3.7

.5

326

.96

.5-1.7

.9

327

.89

.4-1.8

.74

328

Sex, male/female* KPS; 100%*

57 (44)

100

49 (38)

Disease status at transplantation

Disease status at transplantation; CR*

CR1

18 (13)

261

PIF/relapse-resistant

24 (19)

Cy/VP16/carmustine

262

PR2+, CR2+

89 (68)

Fractionated TBI/Cy

Conditioning regimen

315

.07

74 (56)

90

P Value

.98-1.7

57 (44) 23 (18)

314

95% CI

1.29

61 70-80

313

Age (per 10 yr)

<60 KPS

312 OS

HR

260

263

3

Conditioning regimen; BEAM*

CRT, no versus yes*

318 319 320 321 322 323 324 325

264

BEAM

63 (48)

265

Fractionated TBI/Cy

60 (46)

2011-2013

.86

.2-3.2

.82

330

266

Cytoxan/VP16-etoposide/carmustine

8 (6)

2014-2016

1.03

.3-3.6

.96

331

267

Stem cell source

268

PBSCs

269

Marrow with or without PBSCs

270 271 272 273 274 275 276 277

CRT

124 (95)

D2

1.75

.5-5.7

.36

333

7 (5)

D3

2.78

.4-16.8

.27

334

21(16)

D4

2.33

.6-8.6

.21

335

D5

5.23

1.4-19.7

.01

Year of transplantation 2005-2010

34 (26)

2011-2013

44 (34)

2014-2016

53 (40)

CR2+ indicates second or higher complete response; PBSCs, peripheral blood stem cells; PIF, primary induction failure; PR2+, second equal or higher partial response; TBI, total body irradiation; Cy, cyclophosphamide.

* Reference category.

Relapse/progression disease detection

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

18XD18 X F-FDG-PET D182X X before AHCT and Day +100 Post-AHCT 18 Of 131 patients in this cohort, 91 had a preDplanned 183X X D184X X F-FDGD185X X PET scan at day +100 post-AHCT. D-score and quantitative metabolic parameters TMTV, SUV, and TLG are outlined in Table 4. D-scores for the 91 patients with day +100 D186X18 X F-FDG-PET D187X X were D1 in 22%D D18X X ,189X X D2 in D190X X 55%D,19X X D3 in 0%D D192X X ,193X X D4 in 10%D D194X X ,195X X and D5 in 13%D D196X X .197X X

337 339

279

1.2 to 6.4). D146X X A Deauville score of 5 on pre-AHCT PET was associated with a 5-fold greaterD147X X mortality risk compared with a D148X X Dscore of 1 (95% CI, .1D D149X X 4 150X X to 19.75; D15X X PD= 152X X .01) D153X X (Table 2). Older patients had slightly, but not significantly, worse OS in this cohort (DHRD 154X X 15X X per 10-Dyear 156X X age increase,D157X 1 X .3; 95% D158X X CI, .9 D159X X to 1.7%D;160X X PD= 16X X .07). D162X X Overall, there were 60 relapse/progression events,D163X X 58 of which occurred within 5 years post-AHCT. More than one-half of the relapses were radiographically detected (55%), and 35% (n = 21D)164X X had relapse detected on D165X X day +100 imaging. The cDlinical 16X X characteristics of patients experiencing relapse/progression post-AHCT are summarized in Table 3. The mDedian 167X X time to relapse was 195 days for asymptomatic D168X X radiographically detected relapseD169X X and 176 days for clinical D170X X symptomatic relapseD.17X X At the time of post-AHCT relapse, the majority (76%) did not experience B symptoms, D172X X including fevers, chills, and D173X X night sweats. Extranodal disease was common (73%), and lactate dehydrogenase was D174X X elevated above the upper limit of normal in 65% of relapses. The HR for death was 2-fold higher (HR, 1.8; D175X95% X CI, .9 D176X X to 3.4D;17X X PD= 178X X .08) D179X X if relapse was detected clinically versus only radiographically. Day +100 PET was positive in 21 patients;D180X X 16 patients received subsequent chemotherapy, and 4 underwent allo-HCT using umbilical cord blood or related donor. Out of 60 patients with relapse/progression events, 49 (81%) went on to receive further chemotherapy, and 11 underwent subsequent allogeneic HCT in subsequent remission.

336 338

Table 3 Clinical Characteristics of Patients Experiencing Relapse/Progression PostAHCT (N = 60) Relapse Parameter

281

332

D-score; D1*

278 280

329

Years of transplantation; 2005-2010*

Patients, n (%)

340 341 342 343 344

Radiographically detected on surveillance

33 (55)

345

Clinically diagnosed

21 (35)

346

6 (10)

347

Unknown

348

International Prognostic Index at relapse 0-2

23 (37)

3-5

24 (39)

Unknown

13 (24)

Ann Arbor stage at relapse 1-2

20 (36)

3-4

35 (64)

Lactate dehydrogenase upper limit of normal

349 350 351 352 353 354 355 356

Yes

17 (35)

357

No

31 (65)

358

B symptoms at relapse

359

Yes

13 (24)

360

No

41 (76)

361 362

Extranodal disease at relapse Yes

40 (73)

363

No

15 (27)

364 365 366

The mDean 198X X SUV by D-score varied from 1.53 (range, D19X X 1.1 to 1.8) D20X X for D1 to 17.9 (Drange, 201X X 8.5 to 49.1) D20X X for D5. Only 21 patients had measurable disease allowing calculatD D203X X 204X X ion of TMTV and TLG. Patients with D1 and D2 had no measurable disease, making TLG and TMTV 0. Patients with D5 (n = 12) had significantly higher TMTV compared D205X X with D4 D206X X (n = 9) (137 cm3D207X X versusD208X X 12.5 cm3) (Table 4). Out of 9D209X X patients with D4 on day +100 18 D210X X F-FDGD21X X PET, 5 had biopsy-Dproven 21X X relapse, 2 immediately started salvage chemotherapy, and 2D213X were D214X X monitored and demonstrated remission on a subsequent scanD.215X X Post-AHCT SUVmax and D-score were

367 368 369 370 371 372 373 374 375 376

ARTICLE IN PRESS M. Savani et al. / Biol Blood Marrow Transplant && (2019) &&& &&&

4

Table 4 Day +100 Post-AHCT 18F-FDG PET Parameters by D Score

377 378

442 443

379

Deauville Score

Patients, n (%)

380

Dl

20 (22)

381

D2

50 (55)

382

D4

9 (10)

5.8 (3 5-7.6)

D5

12 (13)

17.9 (8.5-49)

383 384

Mean TLG (range)

Mean TMTV, cm3 (range)

444

1.5 (1.1-1.8)

0

0

445

2.5 (1.9-3.5)

0

0

446

12.5

447

Mean SUV (range)

28.8 (3-86) 1322 (8-8441)

137

385

451

predictive of relapse and OS (Figure 1D and F). D216X X The rDisk 217X X of overall mortality was 4-foldD218X X higher in D5 patients relative to D1 patients (reference; HR, D219X X 4.10;D20X X 95% CI =1.56 to 10.77; P D21X X  .01). D2X X The rDisk 23X X of mortality increased with each doubling of SUVmax (HR, 1.7;D24X X 95% CI, 1.3 to 2.3; P D25X X  .01), D26X X but tumor bulk (TLG) or TMTV did not predict survival (data not shown).

387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 411 412 413 414

Q2

Association of Day +100 PET with PDre-AHCT 27X X Clinical Variables on PET/CT Scan Approximately one-D D28X X 29X X quarter of patients with abnormal day +100 18 D230X X F-FDG-PET D231X X had hypermetabolic lymphoma lesions at pre-AHCT D23X18 X F-FDG-PET D23X X (n = 21), whereas 75% D234X X patients were in metabolic complete remission preDtransplantation. 235X X We divided patients into 4 groups based on both pre-AHCT (PET-Dnegative 236X X versusD237X X PET-positiveD)238X X and day +100 PET/CT imaging (negativeD239X X vDersus 240X X positiveD)241X X and compared OS in 4 groups: -/-, D24X X -/+, +/+, and +/-D.243X X Patients (n = 63) with both negative pre-AHCT and negative day +100 post- AHCT PET scan (-/-)D24X X had a 5-year OS of 76%, compared with D245X X 46% for those with both D246X X scans positive (+/+) and 30% for those with pre-D D247X X 248X X AHCT negative and day + 100 post-DAHCT 249X X positive (-/+)D250X (X Figure 1E). We nDextD 251X X 25X Xanalyzed pre-AHCT variables and the odds of having day +100 post-DAHCT 253X X D4D254X Xor D5. The soleD25X Xvariable predictive of worse OS was disease status pre-AHCT: PIR/refra ctory XX relapse had an OR of D256X X 7 (95% CI, 1% to 45%; D257X X PD= 258X X .04) D259X X compared with patients D260X X in CR1. None of other baseline pre-AHCT clinical variables were associated with an abnormal day +100 PET scan D261X X (data not showDn). 26X X

415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441

449 450

386

410

448

DISCUSSION Our single-center study sought to elucidate the prognostic value of surveillance PET/CT assessment at day +100 on survival in patients undergoing AHCT for DLBCL. We found that approximately one-Dthird 263X X of transplant recipients had an abnormal day +100 PET/CT, and that this was associated with poor survival. Patients with D4 and D5 lesions on day +100 PET/CT experienced significantly higher lymphomarelated mortality compared with patients D264X X without these radiographic features. In addition, D5 was associated with a particularly short median survival of only 1.2 years. These results suggest that early post-AHCT PET/CT imaging can help identifyD D265X X 26X X patients with particularly poor prognosis who may be considered for experimental therapeutic interventions such as chimeric antigen receptor (CAR) T cell D267X X therapy before D268Xdeveloping X disease-Dassociated 269X X symptoms. In addition to Deauville, we found that SUV served as an alternative quantitative parameter that can be used for clinical prognostic care. TLG and TMTV parameters could be calculated only in patients with D4 and D5, and their lack of prognostic value could be attributed to our small cohort. Unlike TLG and TMTV, which were unmeasurable in patients without radiographically enlarged tumor bulk, SUVmax is a continuous variable measurDable 270X X even in patients with lower D-score and PET scans with subDcentimeter 271X X lesions. It is notable that all PET scans were re-read by a radiologist for this analysis, and that

D-score and SUVmax values were assigned retrospectively. Of note, out of the 9D27X Xpatients with D4 on day +100 D273X18 X F-FDG-PET, D274X X 2D275X X patients had D276X X a false-Dpositive 27X X PET scan with low-level SUV. This finding again highlights the importance of quantitative PET parameters and caution in their interpretation. Biopsy D278X X is highly recommended to confirm the relapse in asymptomatic patient. Given thDe 279X X use of day D280X X +100 PETD281X X to surveil for relapse, the prognostic value of SUVmax in this setting is novel and should be validated in future studies. Owing to the heterogeneity in visual analysis of PET scans with modest interobserver reproducibility, SUV-based assessment of therapeutic response has been used D28X X in other clinical settings as well [19,20]. In a prospective study of newly diagnosed DLBCL, SUVmax before and after 2 cycles of chemotherapy was used to determine prognosis [20]; an SUVmax change >D67% 283X X predicted a favorable 2-year event-free survival (EFS) (D21% 284X X versus 79%). The prognostic value of pre-AHCT Deauville score on PFS has been well described by several groups [10,13,14]. A retrospective study evaluating 32 patients with DLBCL who underwent AHCT found a 3-year D285X X OS of 84% and 25% for patients with D1 to DD3 286X X and D4, respectively, along with high preDtransplantation 287X X SUVmax (>6) and a D28X X trend towardD289X X inferior PFS [10]. Our study also found that patients who present with clinical symptoms related to relapsed disease had a trend towardD290X X 2-fold lower survivalD291X X compared with D29X X patients with disease radiographically detected on D293X X surveillance imaging. In contrast, a recent multicenter retrospective study of 160 DLBCL AHCT recipients, in which 45 patients relapsed after day +100 postAHCT; the relapse was detected clinically or radiographically in 29% and 71% of patients, respectively [21]. The mDedian 294X X postAHCT OS was not significantly different between the 2D D295X X 296X cX ohorts; however, tDhat 297X X study had variable surveillance practices across institutions, and the reasonD298X X for imaging at specific time D29X X D30X X points was unknown. In contrast, oDur 301X X analysisD302X X studied consecutive AHCT recipients treated on prospective protocols with preDplanned 30X X day +100 PET/CT imaging for all. In addition, we incorporated quantitative parameters of Deauville and SUV, and thDus 304X X examined the biological nature of relapse more granularlyD.305X X The discrepancy between the 2D306X X studies also highlights that surveillance 18 D307X X F-FDG D308X X PET/CT results need be interpreted quantitatively to provide more meaningful clinical value. We acknowledge that our findings are limited by the study’s retrospective design andD309X Xlimited sample size. Nonetheless, oDverallD 310X X 31X X our study highlights the importance of SUV and Deauville post-DAHCT 312X X as significant parameters in predicting clinical outcomes and in identifying high-Drisk 31X X patients who D314X X may be candidates for novel therapies, such as CARD D315X X 316X XT cell D317X X therapy. Relapse after AHCT, whether detected clinically or radiographically, is currently considered a Food and Drug AdministrationD-approved 318X X indication for CAR T D319X X cell therapy. Furthermore, various CAR T cell D320X X products are being evaluated in randomized phase 3 trials in patients failing front-line chemotherapy for DLBCL and are being compared to standard of

452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506

ARTICLE IN PRESS M. Savani et al. / Biol Blood Marrow Transplant && (2019) &&& &&&

5

507

572

508

573

509

574

510

575

511

576

512

577

513

578

514

579

515

580

516

581

517

582

518

583

519

584

520

585

521

586

522

587

523

588

524

589

525

590

526

591

527

592

528

593

529

594

530

595

531

596

532

597

533

598

534

599

535

600

536

601

537

602

538

603

539

604

540

605

541

606

542

607

543

608

544

609

545

610

546

611

547

612

548

613

549

614

550

615

551

616

552

617

553

618

554

619

555

620

556

621

557

622

558

623

559

624

560

625

561

626

562 563 564

Figure 1. Outcomes of patients with DLBCL treated with AHCT. (A) OS. (B) Cumulative incidence of relapse. (C) OS by pre-AHCT PET SUVmax divided into tertials. (D) OS by day +100 post-AHCT PET SUVmax divided into tertials. (E) OS by pre-/day +100 post-AHCT PET scan. (F) OS by D-score on day +100 post-AHCT PET scan. In (C) to (F), patients who relapsed before day +100 are excluded from the analysis.

567

care second-line treatment including AHCT (ClinicalTrials.gov; identifiers NCT03570892 and NCT03391466). D321X X

568 569 570 571

628 629 630

565 566

627

CONCLUSIONS Our findings highlight the utility of performing day +100 PET in AHCT recipients who exhibit no symptoms of disease

progression, with the goal of identifying relapse before the manifestation of clinical D32X X symptoms. We suggest studying this approach in larger prospective trials and incorporating D-score and SUV in pretransplantationD32X X and post-transplantation 18 D324X X FD325X X FDG PET interpretation in an effort to more effectively useD326X X risk-adapted therapies for DLBCL.

631 632 633 634 635 636

ARTICLE IN PRESS 6

637 638 639 640 641 642

M. Savani et al. / Biol Blood Marrow Transplant && (2019) &&& &&&

ACKNOWLEDGMENTS Financial disclosure: This work was supported by National Institutes of Health Grants 5P30CA077598-21 and P01 5P01CA065493-24. Conflict of interest statement: There are no conflicts of interest to report.

643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666

REFERENCES 1. Li S, Young KH, Medeiros LJ. Diffuse large B-cell lymphoma. Pathology. 2018;50:74–87. 2. Coiffer B, Lepage E, Briere J, et al. CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med. 2002;346:235–242. € mper L, et al. CHOP-like chemotherapy with 3. Pfreundschuh M, Kuhnt E, Tru or without rituximab in young patients with good-prognosis diffuse large-B-cell lymphoma: 6-year results of an open-label randomised study of the MabThera International Trial (MInT) Group. Lancet Oncol. 2011;12:1013–1022. 4. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol. 2007;25:579–586. 5. Gisselbrecht C, Van Den Neste E. How I manage patients with relapsed/refractory diffuse large B cell lymphoma. Br J Haematol. 2018;182:633–643. 6. Nagle SJ, Woo K, Schuster SJ, et al. Outcomes of patients with relapsed/ refractory diffuse large B-cell lymphoma with progression of lymphoma after autologous stem cell transplantation in the rituximab era. Am J Hematol. 2013;88:890–894. 7. Costa LJ, Maddocks K, Epperla N, et al. Diffuse large B-cell lymphoma with primary treatment failure: ultra-high risk features and benchmarking for experimental therapies. Am J Hematol. 2017;92:161–170. 8. Vardhana SA, Sauter CS, Matasar MJ, et al. Outcomes of primary refractory diffuse large B-cell lymphoma (DLBCL) treated with salvage chemotherapy and intention to transplant in the rituximab era. Br J Haematol. 2017;176:591–599. 9. Gisselbrecht C, Glass B, Mounier N, et al. Salvage regimens with autologous transplantation for relapsed large B-cell lymphoma in the rituximab era. J Clin Oncol. 2010;28:4184–4190.

10. Winter A, Rybicki L, Shah SN, et al. Prognostic value of pre-transplant PET/ CT in patients with diffuse large B-cell lymphoma undergoing autologous stem cell transplantation. Leuk Lymphoma. 2018;59:1195–1201. 11. Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for initial evaluation, staging and response assessment of Hodgkin and nonHodgkin lymphoma: the Lugano classification. J Clin Oncol. 2014;32: 3059–3068. 12. Allen-Auerbach M, de Vos S, Czernin J. The impact of fluorodeoxyglucosepositron emission tomography in primary staging and patient management in lymphoma patients. Radiol Clin North Am. 2008;46:199–211. vii. 13. Qiao W, Zhao J, Xing Y, Wang C, Wang T. Predictive value of [18F]fluoro2-deoxy-D-glucose positron emission tomography for clinical outcome in patients with relapsed/refractory diffuse large B-cell lymphoma prior to and after autologous stem cell transplant. Leuk Lymphoma. 2014; 55:276–282. 14. Sauter CS, Matasar MJ, Meikle J, et al. Prognostic value of FDG-PET prior to autologous stem cell transplantation for relapsed and refractory diffuse large B-cell lymphoma. Blood. 2015;125:2579–2581. zka V, Gawande RS, Cayci Z, et al. Positron emission tomography15. Procha based assessment of metabolic tumor volume predicts survival after autologous hematopoietic cell transplantation for Hodgkin lymphoma. Biol Blood Marrow Transplant. 2018;24:64–70. 16. Meignan M, Gallamini A, Meignan M, Gallamini A, Haioun C. Report on the first international workshop on interim-PET-scan in lymphoma. Leuk Lymphoma. 2009;50:1257–1260. 17. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509. 18. R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2017. 19. Casasnovas RO, Meignan M, Berriolo-Riedinger A, et al. SUVmax reduction improves early prognosis value of interim positron emission tomography scans in diffuse large B-cell lymphoma. Blood. 2011;118:37–43. 20. Lin C, Itti E, Haioun C, et al. Early 18F-FDG PET for prediction of prognosis in patients with diffuse large B-cell lymphoma: SUV-based assessment versus visual analysis. J Nucl Med. 2007;48:1626–1632. 21. Epperla N, Shah N, Hamadani M, et al. Impact of routine surveillance imaging on outcomes of patients with diffuse large B-cell lymphoma after autologous hematopoietic cell transplantation. Clin Lymphoma Myeloma Leuk. 2016;16:672–678.

702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731

667

732

668

733

669

734

670

735

671

736

672

737

673

738

674

739

675

740

676

741

677

742

678

743

679

744

680

745

681

746

682

747

683

748

684

749

685

750

686

751

687

752

688

753

689

754

690

755

691

756

692

757

693

758

694

759

695

760

696

761

697

762

698

763

699

764

700

765

701

766