RESEARCH ARTICLE

A JH

Gray zone lymphoma with features intermediate between classical Hodgkin lymphoma and diffuse large B-cell lymphoma: Characteristics, outcomes, and prognostication among a large multicenter cohort Andrew M. Evens,1* Jennifer A. Kanakry,2 Laurie H. Sehn,3 Athena Kritharis,1 Tatyana Feldman,4 Aimee Kroll,5 Randy D. Gascoyne,3 Jeremy S. Abramson,6 Adam M. Petrich,7 Francisco J. Hernandez-Ilizaliturri,8 Zeina Al-Mansour,9 Camille Adeimy,10 Jessica Hemminger,11 Nancy L. Bartlett,12 Anthony Mato,4 Paolo F. Caimi,13 Ranjana H. Advani,14 Andreas K. Klein,1 Chadi Nabhan,15 Sonali M. Smith,15 Jesus C. Fabregas,16 Izidore S. Lossos,16 Oliver W. Press,17 Timothy S. Fenske,18 Jonathan W. Friedberg,19 Julie M. Vose,20 and Kristie A. Blum11 Gray zone lymphoma (GZL) with features between classical Hodgkin lymphoma and diffuse large B-cell lymphoma (DLBCL) is a recently recognized entity reported to present primarily with mediastinal disease (MGZL). We examined detailed clinical features, outcomes, and prognostic factors among 112 GZL patients recently treated across 19 North American centers. Forty-three percent of patients presented with MGZL, whereas 57% had nonMGZL (NMGZL). NMGZL patients were older (50 versus 37 years, P 5 0.0001); more often had bone marrow involvement (19% versus 0%, P 5 0.001); >1 extranodal site (27% versus 8%, P 5 0.014); and advanced stage disease (81% versus 13%, P 5 0.0001); but they had less bulk (8% versus 44%, P 5 0.0001), compared with MGZL patients. Common frontline treatments were cyclophosphamide-doxorubicin-vincristine-prednisone 1/2 rituximab (CHOP1/2R) 46%, doxorubicin-bleomycin-vinblastine-dacarbazine 1/2 rituximab (ABVD1/2R) 30%, and doseadjusted etoposide-doxorubicin-cyclophosphamide-vincristine-prednisone-rituximab (DA-EPOCH-R) 10%. Overall and complete response rates for all patients were 71% and 59%, respectively; 33% had primary refractory disease. At 31-month median follow-up, 2-year progression-free survival (PFS) and overall survival rates were 40% and 88%, respectively. Interestingly, outcomes in MGZL patients seemed similar compared with that of NMGZL patients. On multivariable analyses, performance status and stage were highly prognostic for survival for all patients. Additionally, patients treated with ABVD1/2R had markedly inferior 2-year PFS (22% versus 52%, P 5 0.03) compared with DLBCL-directed therapy (CHOP1/2R and DA-EPOCH-R), which persisted on Cox regression (hazard ratio, 1.88; 95% confidence interval, 1.03–3.83; P 5 0.04). Furthermore, rituximab was associated with improved PFS on multivariable analyses (hazard ratio, 0.35; 95% confidence interval, 0.18–0.69; P 5 0.002). Collectively, GZL is a heterogeneous and likely more common entity and often with nonmediastinal presentation, whereas outcomes seem superior when treated with a rituximab-based, DLBCL-specific regimen. C 2015 Wiley Periodicals, Inc. Am. J. Hematol. 90:778–783, 2015. V

Additional Supporting Information may be found in the online version of this article. 1

Tufts Medical Center, Boston, Massachusetts; 2Johns Hopkins University School of Medicine, Baltimore, Maryland; 3Centre for Lymphoid Cancer, British Columbia Cancer Agency, Vancouver, BC, Canada; 4Hackensack University Medical Center, Hackensack, New Jersey; 5Quantitative Health Sciences, The University of Massachusetts (UMASS), Worcester, Massachusetts; 6Massachusetts General Hospital Cancer Center, Boston, Massachusetts; 7Northwestern University Feinberg School of Medicine, Chicago, Illinois; 8Roswell Park Cancer Institute, Buffalo, New York; 9Hematology/Oncology, UMASS, Massachusetts; 10Loyola University Medical Center, Maywood, Illinois; 11Ohio State University, Columbus, Ohio; 12Washington University School of Medicine, St Louis, Missouri; 13Case Western Reserve University, Cleveland, Ohio; 14Stanford University, Palo Alto, California; 15The University of Chicago, Chicago, Illinois; 16University of Miami School of Medicine, Miami, Florida; 17 Hutchinson Cancer Research Center, Seattle, Washington; 18Medical College of Wisconsin, Milwaukee, Wisconsin; 19University of Rochester, Rochester, New York; 20 University of Nebraska Medical Center, Omaha, Nebraska

This work has been presented at the American Society of Hematology (ASH) 55th Annual Meeting, New Orleans, CA, December 2013 (oral presentation); ASH 56th Annual Meeting, San Francisco, CA, December 2014; and the 13th International Conference on Malignant Lymphoma (ICML), June 2015, Lugano, Switzerland (oral presentation). Conflict of interest: AME—Honoraria: Seattle Genetics, Celgene; Advisory role: Celgene; Speakers’ Bureau: Celgene; Research funding: Millennium. LHS—Honoraria and consulting: Genentech. RDD—Honoraria and consulting: Genentech; Speakers’ Bureau: Seattle Genetics. JSA—Honoraria and consulting: Gilead; Research funding: Genentech; Honoraria, consulting, and Speakers’ Bureau: Seattle Genetics; Consulting: Millennium. NLB—Consulting: Seattle Genetics and Pfizer; Research funding: Seattle Genetics, Bristol-Myers Squibb, Pfizer, ImaginAb, Genentech, Janssen Research Foundation, and AstraZeneca; Honoraria: Millennium Takeda, Eisai Co. Ltd., and General Electric/Clarient; Research funding: Millennium Takeda, Seattle Genetics, Genentech Roche, Allos Therapeutics, Pharmacyclics, Janssen Pharmaceuticals, Idera Pharmaceuticals, Celgene, and Agensys; SMS—Celgene and Janssen Oncology; Consulting or advisory role: Genentech/Roche, Celgene, Onyx, Seattle Genetics, Immunogenix, TG Therapeutics, Gilead Pharmacyclics, and Pharmacyclics; Speakers’ Bureau: Spectrum and Janssen Oncology. OWP—Stock or other ownership: PhaseRx and Emergent Biosolutions; Consulting or advisory role: BIND Biosciences, Algeta, Adaptive Biotechnologies, and Roche; Research funding: PreSage and Genentech. JWF— Consulting or advisory role: Trubion, Kite Pharmaceuticals, and Bayer; Research funding: Seattle Genetics, Millennium, Genentech, and Janssen Pharmaceuticals. JMV— Honoraria: Sanofi Adventis; Consulting or advisory role: Kite Pharma, Foundation Medicine, Seattle Genetics, Bank of America, KEW Group, and Bioconnections; Research funding: Spectrum Pharmaceuticals, Bristol-Myers Squibb, Janssen Biotech, Inc., and Pharmacyclics. KAB—Research funding: Seattle Genetics, Millennium, Pharmacyclics, Janssen, Novartis, Gilead, Celgene, and Cephalon. *Correspondence to: Dr. Andrew M. Evens, Division of Hematology/Oncology, 800 Washington Street, Boston, MA 02111. E-mail: [email protected] Received for publication: 27 May 2015; Revised: 29 May 2015; Accepted: 1 June 2015 Am. J. Hematol. 90:778–783, 2015. Published online: 4 June 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/ajh.24082 C 2015 Wiley Periodicals, Inc. V

778

American Journal of Hematology, Vol. 90, No. 9, September 2015

doi:10.1002/ajh.24082

RESEARCH ARTICLE

Gray Zone Lymphoma

䊏 Introduction

䊏 Results

The World Health Organization (WHO) recognizes a category of B-cell lymphoma unclassifiable with features intermediate between diffuse large B-cell lymphoma (DLBCL) and classical Hodgkin lymphoma (HL), also known as gray zone lymphoma (GZL) [1]. GZL is a relative new entity that is depicted in the WHO as mediastinal GZL (MGZL) without description of nonmediastinal presentation [1–4]. To date, there has been a relative paucity of clinical data describing nonmediastinal GZL (NMGZL). The combined rarity, disease heterogeneity, and lack of clinicopathologic prognostication makes disease management challenging. The majority of clinical data consists of small case series [5]. A recent report of 24 MGZL patients treated with infusional doseadjusted etoposide, doxorubicin and cyclophosphamide with vincristine, prednisone, and rituximab (DA-EPOCH-R) noted 5-year event-free survival of 67% [6]. There remain a lack of data regarding outcomes for GZL patients using HL-specific chemotherapy regimens, and despite the aforementioned data, there are no standard management guidelines for GZL patients in the up-front or relapsed/refractory setting. In addition, further prognostication of GZL is desired. Herein, we report a multicenter collaboration that investigated a large retrospective cohort of patients with GZL, representing all known patients presenting to 19 North American institutions over the last decade. The aims of this study was to examine detailed clinical features, investigate prognostic parameters, and to evaluate outcomes within the context of varying treatments across multiple centers to identify the most optimal therapeutic strategy and to enhance prognostication of this disease entity.

Baseline characteristics

䊏 Methods We performed a multicenter retrospective analysis of GZL patients 18 years of age, who were consecutively diagnosed and treated between 2001 and 2012 across 19 North American academic centers. The Institutional Review Boards of all participating institutions approved the study. All pathology cases/slides were examined by institutional hematopathology experts for confirmation of diagnosis of GZL. Diagnosis of GZL was established by these lymphoma pathology experts as according to WHO definition [1]. In addition, all pathologic cases/descriptions were reviewed by an expert lymphoma pathologist (R.D.G.) for agreement of diagnosis. There were 121 eligible patients initially identified. Of these, 112 had complete data and were entered into a centralized database. Nine cases were excluded: six for inadequate baseline characteristics and/or follow-up data and three for inability to re-examine pathologic slides. Information on diagnostic features, staging investigations, and treatment administered was provided by the patients institutional providers. Receipt of radiation therapy was not considered an event or progression. Bulky disease was defined as any mass lesion 10 cm, and primary refractory disease was defined as lack of remission to frontline therapy or progressive disease documented 45 Gender Male Female Performance status 0–1 2–4 Not available B-symptoms Yes No Not available Bone marrow involvement Absent Present Not available Extranodal sites present 1 >1 Not available Stage I/II III/IV Bulky disease Yes No Albumin Normal Low Not available Hemoglobin (g/dL) Decreased Normal Not available LDH Normal Elevated Not available IPI 0–2 3–5 Not available IPS 0–1 2–3 4–6 Not available

Mediastinal Nonmediastinal P(n 5 48) (n 5 64) valuea

72 64 40 36

42 6

30 34

0.0001

73 65 39 35

28 20

45 19

0.23

97 87 11 10 4 4

53 9 2

44 2 2

0.11

50 45 50 45 12 11

18 22 8

32 28 4

0.54

95 85 12 11 5 4

47 0 1

48 12 4

0.001

89 79 21 19 2 2

44 4 0

45 17 2

0.014

54 48 58 52

42 6

12 52

0.0001

26 23 86 77

21 27

5 59

0.0001

64 57 27 24 21 19

26 10 12

38 17 9

0.82

62 55 43 38 7 6

23 23 2

39 20 5

0.11

58 52 30 27 8 7

29 15 4

29 15 4

0.99

74 66 22 20 10 9

40 3 5

34 19 5

0.001

25 41 15 31

18 13 2 15

7 28 13 16

0.002

22 37 13 28

a Comparing mediastinal versus nonmediastinal. Abbreviations: No, number; LDH, lactate dehydrogenase; IPI, International Prognostic Index; IPS, International Prognostic Score.

American Journal of Hematology, Vol. 90, No. 9, September 2015

779

RESEARCH ARTICLE

Evens et al.

Figure 1. Treatment and patient outcomes. Outcomes following initial chemotherapy and salvage therapy administered to gray zone lymphoma (GZL) patients. Data are based on mediastinal or nonmediastinal presentation. The first row of data summarizes therapeutic regimens given as frontline therapy, with the associated responses immediately below this. This is followed vertically by the number of patients who proceeded to stem cell transplant for relapsed/refractory disease. Finally, outcomes are designated by presence of disease or death at most recent follow-up. †Mediastinal: BEACOPP and HyperCVAD/MA; Non-Mediastinal: BEACOPP (2), HyperCVAD/MA (2), MOPP, MOPP/ABVD, ChIVPP, and ABVD/CHOP.

bone marrow involvement (19% versus 0%, P 5 0.001), more often had >1 extranodal site of disease (27% versus 8%, P 5 0.014), and presented much more commonly with advanced stage disease (81% versus 13%, P 5 0.0001). Because of the increased presence of these adverse factors, patients with NMGZL had worse/higher prognostic scores at initial diagnosis compared with MGZL (IPI 3–5: 30% versus 6%, P 5 0.001; IPS 3–7: 23% versus 4%, P 5 0.0001).

Pathology The most prevalent immunophenotype via immunohistochemistry for all cases were: CD201 (93%), CD301 (89%), CD791 (78%), Pax51 (98%), Oct21 (96%), and MUM11 (100%). For the minority of cases that were CD20-negative, CD79, Pax5, and/or Oct2 staining established B-cell origin. CD15 (46%) and CD45 (67%) staining were more variable, whereas only 11% and 28% of patients had CD10 and EBV positivity (the latter by EBER ISH staining), respectively. Immunophenotype did not appear to differ based on presentation with MGZL versus NMGZL disease (data not shown).

Treatment and response The choice of frontline therapy appeared to impact response. The most common frontline chemotherapy regimens given were cyclophosphamide, vincristine, doxorubicin, and prednisone (CHOP) 1/2 rituximab (n 5 57), doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) 1/2 rituximab (n 5 34), and DA-EPOCH-R (n 5 11). Figure 1 depicts frontline regimens received and includes patient outcome based on presence of mediastinal involvement. The overall response rates and complete remission rates for all patients was 79 of 112 (71%) and 66 of 112 (59%), respectively. Response by the aforementioned chemotherapeutic regimens received were ABVD 61% and 55%, respectively; CHOP 71% and 56%, respectively; and EPOCH 82% and 73%, respectively. Notably, 37 of 112 (33%) of all patients had primary refractory disease (14 of 48 (21%) for MGZL versus 23 of 64 (36%) for NMGZL, P 5 0.54). Seventy-three of 112 (65%) patients received rituximab as part of frontline therapy (33 of 48 (69%) MGZL versus 40 of 64 (63%) NMGZL, P 5 not significant); 91% and 18% of patients who received CHOP and ABVD, respectively, received rituximab. Further, 65 of 100 (65%) CD201 patients received rituximab and five of eight (63%) CD20-negative patients also received rituximab. The overall 780

American Journal of Hematology, Vol. 90, No. 9, September 2015

response rates and complete remission rates for patients who received rituximab as part of front-line therapy were 82% and 73%, respectively, versus 59% and 43%, respectively, without rituximab (P 5 0.02 and P 5 0.008, respectively). Additionally, 37 of 112 (33%) patients had consolidative radiotherapy (MGZL 27 of 48 (56%) versus NMGZL 10 of 64 (16%), P 5 0.001; and stage I/II 30/53 (57%) versus stage III/IV 7/59 (12%), P 5 0.0002). Response rates did not significantly differ by receipt of radiotherapy (data not shown). Overall, 65 (58%) patients relapsed with a median time to relapse of 7 months (range, 1–64 months); the median number of salvage therapies was 3 (range, 1–7). Regimens at first relapse included ifosfamide, carboplatin, and etoposide (n 5 39), EPOCH (n 5 8), etoposide, solumedrol, cytarabine, and cisplatin (n 5 8), brentuximab vedotin (n 5 4), gemcitabine-based (n 5 3), and ABVD (n 5 2); one patient refused salvage therapy. Beyond first relapse, the most common treatments were brentuximab vedotin (n 5 7) and radiation (n 5 5). Sixtyone percent of relapsed patients had a hematopoietic stem cell transplant (HSCT) (38% allogeneic and 62% autologous); 18 of 27 (67%), 20 of 32 (63%), and three of three (100%) of patients who relapsed after frontline ABVD, R-CHOP, and R-EPOCH, respectively, had HSCT at relapse (Fig. 1).

Outcomes At 31-month median follow-up, 2-year PFS and OS for all patients were 40% and 88%, respectively. The 2-year PFS rates for patients with stage I/II versus III/IV disease were 54% versus 30%, respectively, P 5 0.005, and the corresponding OS rates were 94% versus 76%, respectively, P 5 0.007 (Fig. 2). Furthermore, despite a comparative preponderance of early-stage disease and lower IPI, the PFS and OS rates were not significantly different for patients with MGZL versus NMGZL (PFS: 46% versus 36%, respectively, P 5 0.18; OS: 92% versus 85%, respectively, P 5 0.17). In addition, the 2-year OS rate for patients who underwent SCT was 80%.

Prognostication As noted in Table II, baseline characteristics associated with PFS on univariable analysis for all GZL patients were performance status, lactate dehydrogenase, anemia, and stage; stage was prognostic for OS on univariable analysis, whereas performance status was borderline. doi:10.1002/ajh.24082

RESEARCH ARTICLE

Gray Zone Lymphoma

Figure 2. Survival. (A) The 2-year progression-free survival (PFS) and overall survival (OS) for 112 patients with gray zone lymphoma (GZL) were 40% and 88%, respectively. (B) Kaplan–Meier curves comparing GZL patients with mediastinal GZL (MGZL) versus nonmediastinal GZL (NMGZL): 2-year PFS rates were 46% versus 36%, P 5 0.18. The corresponding OS rates were 92% versus 85%, P 5 0.17. Kaplan–Meier curves comparing (C) PFS and (D) OS for GZL patients based on International Prognostic Index (IPI): for IPI 0–2 versus 3–5, 2-year PFS rates were 50% and 24%, respectively (P 5 0.002), and OS were 87% and 76%, respectively (P 5 0.05). Kaplan–Meier curves of (E) PFS and (F) OS based on International Prognostic Score (IPI): 2-year PFS rates were 53% and 25%, respectively (P 5 0.03), and OS were 92% and 77%, respectively (P 5 0.05). (G) The outcome based on frontline therapeutic regimen: 2-year PFS for patients who received a standard frontline DLBCL regimen (i.e., CHOP1/2R and DA-EPOCH-R) versus ABVD1/2R was 52% versus 23% (P 5 0.03). (H) Kaplan–Meier curves for patients who received rituximab as a part of frontline therapy versus not: 2-year PFS were 51% versus 22% (P 5 0.01). In addition, for patients who had hematopoietic stem cell transplant for relapsed disease (I), the 2-year OS rates appeared improved for gray zone lymphoma (GZL) patients with relapsed/refractory disease who underwent hematopoietic stem cell transplantation (HSCT) versus patients who did not (88% versus 67%, P 5 0.01). This finding persisted on multivariable Cox regression analysis controlling for IPI (HR, 0.14; 95% CI, 0.02–0.95, P 5 0.04).

Further, both the IPI and IPS scoring systems were prognostic for PFS and OS as categorical variables (i.e., IPI 3–5 versus 0–2: PFS HR, 2.60 (95% CI, 1.45–4.68; P 5 0.001); OS HR, 4.12 (95% CI, 1.19– 14.26; P 5 0.03); IPS 3–7 versus 0–2: PFS HR, 2.01 (95% CI, 1.12– 3.63; P 5 0.019); OS HR, 4.41 (95% CI, 0.85–22.81; P 5 0.08); and doi:10.1002/ajh.24082

continuous variables (Table II). These survival differences are highlighted in Fig. 2C–F. On multivariable analysis of patient and disease characteristics (Supporting Information Table S1), performance status was the most dominant prognostic factor for PFS (i.e., performance status  2: PFS HR, 2.94; 95% CI, 1.19–7.22; P 5 0.02), whereas American Journal of Hematology, Vol. 90, No. 9, September 2015

781

RESEARCH ARTICLE

Evens et al. TABLE II. Univariable Prognostic Factor Analyses 2-Year PFS

2-Year OS

95% HR confidence limits

Baseline clinical factors Age (continuous) Gender (male vs. female) B symptoms (yes vs. no) ECOG performance status (2–4 vs. 0–1) Hemoglobin < 10.5 g/dL (no vs. yes) LDH (normal vs. increased) ESR (normal vs. increased) Albumin < 4.0 g/dL (no vs. yes) Bone marrow involved (yes vs. no) >1 extranodal site (yes vs. no) Bulky disease (yes vs. no) Stage at diagnosis (3–4 vs. 1–2) Prognostic Score IPI (continuous) Prognostic Score IPS (continuous) Frontline treatment Rituximab ABVD CHOP DA-EPOCH Consolidative radiotherapy

95% HR confidence limits

HR

Low

High

P-value

HR

Low

High

P-value

0.99 0.94 1.51 3.43 1.89 0.52 0.36 0.60 0.95 1.50 0.95 2.05 1.47 1.24

0.97 0.56 0.89 1.72 1.07 0.30 0.10 0.33 0.41 0.81 0.52 1.22 1.19 1.01

1.00 1.58 2.58 6.83 3.34 0.90 1.25 1.09 2.20 2.78 1.73 3.45 1.82 1.52

0.09 0.82 0.13 0.0005 0.03 0.02 0.11 0.09 0.89 0.19 0.88 0.007 0.0004 0.04

1.01 1.90 3.47 1.86 2.72 0.62

0.98 0.52 0.94 0.41 0.75 0.21

1.04 6.89 12.83 8.48 9.89 1.85

a

a

a

a

0.43 0.84 2.10 2.27 1.94 1.72 1.77

0.13 0.11 0.65 0.74 1.19 1.10 1.15

1.42 6.63 6.84 6.94 3.16 2.68 2.73

0.17 0.87 0.22 0.15 0.008 0.01 0.01

0.57 1.69 0.87 0.34 0.65

0.34 1.02 0.53 0.11 0.37

0.94 2.80 1.43 1.09 1.12

0.02 0.04 0.58 0.07 0.12

3.13 0.61 1.18 0.81 0.56

0.69 0.17 0.40 0.11 0.16

14.13 2.22 3.51 6.20 2.05

0.68 0.33 0.06 0.42 0.13 0.39

0.14 0.45 0.76 0.84 0.39

a Too few events to provide an accurate estimate for Cox overall survival models. Abbreviations: PFS, progression-free survival; OS, overall survival; HR, hazard ratio; ECOG, Eastern Cooperative Oncology Group; IPS, International Prognostic Score; LDH, lactate dehydrogenase; ESR, erythrocyte sedimentation rate; IPI, International Prognostic Index; CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone; ABVD, doxorubicin, bleomycin, vinblastine, and dacarbazine; DA-EPOCH, dose-adjusted etoposide, doxorubicin, cyclophosphamide, vincristine, and prednisone.

increased lactate dehydrogenase was borderline. For OS, stage was the only significant prognostic factor (i.e., Stage 3–4: HR, 4.89; 95% CI, 1.00–24.18; P 5 0.05). In terms of the impact of frontline therapy received, 2-year PFS and OS for patients who received R-CHOP were 52% and 84%, respectively; ABVD1/2R 23% and 96%, respectively; and DA-EPOCH-R 68% and 83%, respectively. There was a trend for improved PFS by log rank with DA-EPOCH-R treatment (P 5 0.07); however, this effect was abrogated after controlling for IPI (HR, 0.42; 95% CI, 0.12–1.56; P 5 0.21). Combining all patients who received a standard frontline DLBCL regimen (i.e., CHOP1/2R and DA-EPOCH-R), PFS was improved compared with ABVD1/2R (2-year PFS 52% versus 23%, P 5 0.03) as shown in Fig. 2G–H. Furthermore, this finding persisted on multivariable Cox regression controlling for IPI and receipt of rituximab (HR, 0.41; 95% CI, 0.10–0.75; P 5 0.04). In addition, patients who received rituximab with frontline therapy had improved 2-year PFS (51% versus 22%, respectively, P 5 0.01), which remained significant after controlling for IPI and type of chemotherapy regimen (HR, 0.35; 95% CI, 0.18–0.69; P 5 0.002). Interestingly, CD20 positivity also independently predicted PFS on Cox regression controlling for receipt of rituximab and IPI (rituximab HR, 0.55; 95% CI 0.33–0.93; P 5 0.025; CD20 0.35; 95% CI, 0.16–0.75; P 5 0.007) as seven of the eight (88%) CD20-negative GZL patients experienced progressive disease. Finally, among GZL patients with relapsed/refractory disease, the 2-year OS appeared improved for patients who underwent HSCT (88% versus 67%, P 5 0.01) as depicted in Fig. 2I. The improvement in OS persisted on multivariable Cox regression analysis after controlling for IPI (HR, 0.14; 95% CI, 0.02–0.95; P 5 0.04).

䊏 Discussion Most lymphomas can be accurately classified based on morphology, immunophenotype, and molecular characteristics. However, some 782

American Journal of Hematology, Vol. 90, No. 9, September 2015

patients present with features intermediate between different lymphoma subtypes. The term GZL was first used in 1998 at the “Workshop on Hodgkin’s disease and related diseases” to designate lymphomas at the border of classical HL and other entities [12]. Diagnosis and treatment of these cases has been very challenging in part given their mixed features and relative rarity of these neoplasms. We report here a multicenter cohort of 112 GZL patients treated over a recent 11-year period. To the best of our knowledge, this is the largest clinical series of GZL reported to date and the most detailed description of the entity of NMGZL. In 2005, the National Institutes of Health formally described the pathologic findings of 20 GZL cases diagnosed and treated over a 20year period [13]. All cases in this series had mediastinal involvement. Published in 2008, the 4th edition of the WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues first included GZL with mediastinal involvement as a formal diagnostic entity [1]. This has been followed by several pathologic analyses of GZL examining immunophenotypic, chromosomal, and molecular changes [3], and also methylation profiling GZL signatures that appear distinct from both classical HL and primary mediastinal DLBCL [14]. Nine of the 27 cases in the former pathologic study had no evidence of mediastinal disease [3]; however, nonmediastinal presentation of GZL has not been more fully characterized. In our series, the majority of GZL patients had no evidence of mediastinal involvement. Furthermore, there were several important clinical differences among patients with mediastinal and nonmediastinal presentation of disease. This included NMGZL patients having less frequent occurrence of bulk disease, but being significantly older and presenting more often with extranodal involvement and advancedstage disease. Due in part to these characteristics, NMGZL patients had significantly worse/higher prognostic scores (IPS and IPI) compared with MGZL patients. Interestingly, despite presenting with lower risk features/prognostic scores, outcomes for MGZL appeared similar to the doi:10.1002/ajh.24082

RESEARCH ARTICLE

Gray Zone Lymphoma

NMGZL patient cohort. This may have been in part due to the much higher incidence of bulky disease in MGZL patients; however, a potential biologic difference cannot be fully excluded. There are scant recommendations toward a consensus therapeutic approach for GZL. Wilson et al. [6] recently reported an important single institution study of 24 patients with MGZL who were treated with DAEPOCH-R over a 13-year time period. On initial analysis (i.e., at 4 years), the authors noted a PFS rate of 30% and OS rate of 83% [15]; an updated analysis reported 5-year event-free survival and OS rates of 62% and 74%, respectively. The two most common therapeutic approaches in this study were CHOP1/R and ABVD1/2R given to 51% and 30% of patients, respectively, with 10% of patients treated with DA-EPOCH-R. Furthermore, approximately two-thirds of all patients received rituximab as part of frontline therapy. Relapse rates were relatively high, which was especially apparent when patients were treated with HL-specific therapy (i.e., ABVD1/R) despite prior data adding rituximab in classical HL [16]. Treatment with a DLBCL-like regimen (i.e., DA-EPOCH-R or R-CHOP) appeared to be associated with improved PFS. In addition, use of rituximab was an independent factor associated with enhanced outcome on Cox regression analyses as was the presence of CD20, which also predicted a favorable outcome (independent of rituximab). Interestingly, GZL patients with relapsed/refractory disease were salvaged with fair success, and OS rates were overall favorable. Among all patients who relapsed, 61% proceeded to HSCT, of which the majority were autologous. HSCT appeared to be associated with improved outcomes; however, this may reflect that mostly fit patients with chemotherapy-sensitive disease likely proceeded to SCT; therefore, this is a selected population. Continued and longer follow-up of these patients are warranted in part to rule out late relapse. Nevertheless, the fairly wide discrepancy between PFS and OS indicates that many patients required second-line therapy (often HSCT) to achieve longterm remission and that intensification of first-line regimens may be worth pursuing in future studies. The lack of formalized central pathologic review is a limitation of this analysis. GZL is a relatively recently described pathologic entity and requires expert pathologic review. GZL is a diagnostic entity included in the 2008 WHO book with detailed diagnostic criteria. Further, it should be highlighted that all slides/cases included in this study

䊏 References 1. Jaffe ES, Stein H, Swerdlow SH, et al. B-cell lymphoma, unclassifiable, with features intermediated between diffuse large B-cell lymphoma and classical lymphoma and classical Hodgkin lymphoma. In: Swerdlow SH, Campo E, Harris NL, et al., editors. Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press 2008. pp 267–268. 2. Grant C, Dunleavy K, Eberle FC, et al. Primary mediastinal large B-cell lymphoma, classic Hodgkin lymphoma presenting in the mediastinum, and mediastinal gray zone lymphoma: What is the oncologist to do? Curr Hematol Malig Rep 2011; 6:157–163. 3. Eberle FC, Salaverria I, Steidl C, et al. Gray zone lymphoma: Chromosomal aberrations with immunophenotypic and clinical correlations. Mod Pathol 2011; 24:1586–1597. 4. Garcia JF, Mollejo M, Fraga M, et al. Large Bcell lymphoma with Hodgkin’s features. Histopathology 2005; 47:101–110. 5. Dunleavy K, Grant C, Eberle FC, et al. Gray zone lymphoma: Better treated like Hodgkin lymphoma or mediastinal large B-cell lym-

doi:10.1002/ajh.24082

6. 7.

8. 9. 10. 11.

12.

were examined, re-reviewed, and diagnosed by institutions with expert hematopathologists and an expert lymphoma pathologist (R.D.G.) reviewed all case reports and was in agreement with diagnosis of GZL. An additional weakness inherent in this retrospective analysis includes nonuniform treatment and follow-up, whereas a major strength is the inclusion of a large number of patients of a unique and relatively uncommon disease treated across multiple institutions in a contemporary era. In summary, we characterized clinical and disease-related features of a large cohort of GZL patients recently treated across multiple institutions. This included complete delineation of a clinical GZL subtype without mediastinal involvement, which was associated with older age, abundance of extranodal disease, advanced stage, and absence of bulk disease. Collectively, relapse rates for MGZL and NMGZL were relatively high; however, OS rates were overall good. Frontline therapy with a rituximab-containing DLBCL regimen was associated with the best outcomes, whereas salvage with HSCT generally appeared effective. In addition, we identified several prognostic factors (i.e., stage and performance status) that predicted markedly divergent survival in GZL patients. Continued pathologic, biologic, and clinical characterization of GZL is needed, and new treatment regimens should be studied, including the integration of rational, novel therapeutic agents (e.g., PD1 inhibitors, brentuximab vedotin, etc), in order to continue to improve outcomes for this unique and interesting lymphoproliferative disease.

䊏 Author Contributions All the authors are responsible for the conception and design of the study, provision of study materials or patients, manuscript writing, and final approval of the manuscript. AME and AK provided financial and administrative support. AME, JAK, LHS, AtK, TF, AiK, RDG, and KAB are responsible for data analysis and interpretation.

䊏 Acknowledgments The authors thank the institutional practitioners and respective data base management teams.

phoma? Curr Hematol Malig Rep 2012; 7:241– 247. Wilson WH, Pittaluga S, Nicolae A, et al. A prospective study of mediastinal gray-zone lymphoma. Blood 2014; 124:1563–1569. 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. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol 2007; 25:579–586. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. JASA 1958; 53: 457–481. Cox D. Regression models and life tables. J R Stat Soc Series B 1972; 34:187–220. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1958; 22:719– 723. Rudiger T, Jaffe ES, Delsol G, et al. Workshop report on Hodgkin’s disease and related diseases (‘grey zone’ lymphoma). Ann Oncol 1998; 9(Suppl 5):S31–S38.

13. Traverse-Glehen A, Pittaluga S, Gaulard P, et al. Mediastinal gray zone lymphoma: The missing link between classic Hodgkin’s lymphoma and mediastinal large B-cell lymphoma. Am J Surg Pathol 2005; 29:1411–1421. 14. Eberle FC, Rodriguez-Canales J, Wei L, et al. Methylation profiling of mediastinal gray zone lymphoma reveals a distinctive signature with elements shared by classical Hodgkin’s lymphoma and primary mediastinal large B-cell lymphoma. Haematologica 2011; 96:558–566. 15. Dunleavy K, Pittaluga S, Tay K, et al. Comparative clinical and biological features of primary mediastinal B-cell lymphoma (PMBL) and mediastinal grey zone lymphoma (MGZL). 2009 ASH Annual Meeting Abstracts. Blood 2009; 114:Abstract 106. 16. Kasamon YL, Jacene HA, Gocke CD, et al. Phase 2 study of rituximab-ABVD in classical Hodgkin lymphoma. Blood 2012; 119:4129– 4132.

American Journal of Hematology, Vol. 90, No. 9, September 2015

783

Gray zone lymphoma with features intermediate between classical Hodgkin lymphoma and diffuse large B-cell lymphoma: characteristics, outcomes, and prognostication among a large multicenter cohort.

Gray zone lymphoma (GZL) with features between classical Hodgkin lymphoma and diffuse large B-cell lymphoma (DLBCL) is a recently recognized entity re...
569KB Sizes 2 Downloads 13 Views

Recommend Documents