Blood Cancer Journal www.nature.com/bcj ARTICLE OPEN Prophylaxis with intrathecal or high-dose methotrexate in diffuse large B-cell lymphoma and high risk of CNS relapse Sabela Bobillo1,2,3, Erel Joffe 1, David Sermer1, Patrizia Mondello 1, Paola Ghione1, Philip C. Caron1, Audrey Hamilton1, Paul A. Hamlin1,4, Steven M. Horwitz1,4, Anita Kumar1,4, Matthew J. Matasar1,4, Connie L. Batlevi 1,4, Alison Moskowitz1,4, Ariela Noy 1,4, Collette N. Owens1, M. Lia Palomba 1,4, David Straus1,4, Gottfried von Keudell1,4, Ahmet Dogan 5, ✉ Andrew D. Zelenetz1,4, Venkatraman E. Seshan 6 and Anas Younes 1,4 © The Author(s) 2021 Although methotrexate (MTX) is the most widely used therapy for central nervous system (CNS) prophylaxis in patients with diffuse large B-cell lymphoma (DLBCL), the optimal regimen remains unclear. We examined the efficacy of different prophylactic regimens in 585 patients with newly diagnosed DLBCL and high-risk for CNS relapse, treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) or R-CHOP-like regimens from 2001 to 2017, of whom 295 (50%) received prophylaxis. Intrathecal (IT) MTX was given to 253 (86%) and high-dose MTX (HD-MTX) to 42 (14%). After a median follow-up of 6.8 years, 36 of 585 patients relapsed in the CNS, of whom 14 had received prophylaxis. The CNS relapse risk at 1 year was lower for patients who received prophylaxis than patients who did not: 2% vs. 7.1%. However, the difference became less significant over time (5-year risk 5.6% vs. 7.5%), indicating prophylaxis tended to delay CNS relapse rather than prevent it. Furthermore, the CNS relapse risk was similar in patients who received IT and HD-MTX (5-year risk 5.6% vs. 5.2%). Collectively, our data indicate the benefit of MTX for CNS prophylaxis is transient, highlighting the need for more effective prophylactic regimens. In addition, our results failed to demonstrate a clinical advantage for the HD-MTX regimen. Blood Cancer Journal (2021)11:113 ; https://doi.org/10.1038/s41408-021-00506-3 INTRODUCTION Finally, the combination of cell of origin (COO) determined by gene Diffuse large B-cell lymphoma (DLBCL) is the most common expression profiling (GEP) and CNS-IPI has recently improved the subtype of lymphoma accounting for 30–40% of all non-Hodgkin identification of DLBCL patients with high risk of CNS relapse, lymphomas. Central nervous system (CNS) relapse is an uncom- showing a 2-year CNS relapse rate up to 15% in patients with mon yet often fatal complication with a median overall survival activated B-cell (ABC) phenotype and high CNS-IPI [10]. (OS) of less than 6 months [1]. Overall, the incidence of CNS In high-risk patients, CNS prophylaxis is usually recommended, relapse in patients with DLBCL is around 2%, which is lower than although the optimal regimen remains unclear. Some prospective with other aggressive lymphomas, such as Burkitt lymphoma or and retrospective studies conducted in the rituximab era have acute lymphoblastic leukemia. However, the presence of certain demonstrated the lack of efficacy of intrathecal (IT) methotrexate risk factors might increase the risk of CNS relapse to 15% [2]. (MTX) in patients treated with rituximab, cyclophosphamide, Models have been made to identify patients with high risk of CNS doxorubicin, vincristine, and prednisone (R-CHOP) or similar relapse. The German High-Grade non-Hodgkin Lymphoma Study regimens [2, 10–13]. High-dose intravenous methotrexate (HD- Group (DSHNHL) recently proposed a CNS prognostic model (CNS- MTX) has been postulated as a possibly better option since the IPI) that includes five international prognostic index (IPI) factors and majority of relapses in the rituximab era occur in the brain the involvement of kidney or adrenal glands. This model stratified parenchyma. However, different retrospective studies have shown DLBCL patients into three categories, low (0–1 risk factors), conflicting results regarding its efficacy [14–19]. Finally, prelimin- intermediate (2–3 risk factors), and high risk (4–6 risk factors) with ary results from a multicenter retrospective study showed similar a 2-year rate of CNS relapse of 0.6%, 3.4%, and 10.2%, respectively effectiveness of prophylactic HD-MTX and IT MTX in patients with [3]. However, involvement of certain extranodal sites, such as testes, aggressive non-Hodgkin lymphoma [20]. breast, or bone marrow also confers an increased risk, even with The purpose of this study was to evaluate the efficacy of low CNS-IPI [4–7]. In addition, the presence of MYC translocation different CNS prophylaxis regimens in preventing CNS relapse in together with BCL2 translocation has been also associated with a DLBCL patients with risk factors for CNS recurrence who were higher risk of CNS relapse in several retrospective series [8, 9]. treated with rituximab and chemotherapy in a single institution. 1 Department of Medicine, Lymphoma Service, Memorial Sloan Kettering Cancer Center, New York, NY, USA. 2Department of Hematology, Vall d’Hebron Institute of Oncology (VHIO), Barcelona, Spain. 3Department of Medicine, Universitat Autonoma de Barcelona, Barcelona, Spain. 4Weill Cornell Department of Medicine, Weill Cornell Medical College, New York, NY, USA. 5Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, NY, USA. 6Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, USA. ✉email: anas.younes@astrazeneca.com Received: 7 January 2021 Revised: 29 April 2021 Accepted: 10 May 2021 S. Bobillo et al. 2 from relapse until death of any cause. Patients’ characteristics were DLBCL 2001-2017 compared using Fisher exact test for categorical variables and n=2308 Kruskal–Wallis test for continuous variables. Kaplan-Meier analysis was implemented to report survival estimates across patients and logrank test to compare groups. We applied a non-parametric analysis of competing Excluded risks and used Gray’s test to compare the cumulative incidence of CNS relapse between the different groups, with death and systemic relapse as - HIV posive n=42 competing events. Confidence intervals for risk ratio (RR) (ratio of - CNS disease at dx n= 34 cumulative incidence rate at fixed time-points) were obtained using - R-CODOX/M-IVAC n=4 bootstrap resampling. All statistical analyses were performed in R v3.6. - Primary mediasnal n= 168 - Follow-up less 12 months n= 40 - Previously treated indolent lymphoma n=18 RESULTS Patient characteristics We identified 2308 patients treated with R-CHOP or R-CHOP-like regimens from 2001 to 2017 at MSKCC. We excluded 306 patients N=2002 for the following reasons: HIV positive, n = 42, CNS disease at diagnosis, n = 34; primary mediastinal lymphoma, n = 168; pre- viously treated indolent lymphoma, n = 18; treatment with R- CODOX/M-IVAC, n = 4 and follow-up shorter than 12 months, n = CNS high risk CNS low risk 40. Finally, 2002 patients were considered and 585 (29%) were n=585 n=1417 classified as HR-CNS relapse and were included in the study (Fig. 1). Median age was 68 years (range 21–91) and 301 (51%) patients were male. The high-risk extranodal sites were: BM, n = 118 (20%), kidney/adrenal, n = 106 (18%); testes, n = 51 (9%) and breast, n = IT MTX n=253 47 (8%). One hundred and fifty-eight patients (27%) had more 1234567890();,: than two extranodal sites involved. By CNS-IPI, patients were classified as low, 10%; intermediate, 22%; and high risk, 68%. Eighteen (7%) patients had MYC and BCL2 rearrangements. COO HD-MTX n=42 determined by Hans algorithm [21] was non-germinal center (non- GCB) in 229 (39%) patients, germinal center (GCB) in 234 (40%) patients and missing in 122 (21%). No prophylaxis n=290 CNS prophylaxis Two hundred and ninety (50%) patients received at least one Fig. 1 Consort diagram. DLBCL diffuse large B cell lymphoma, CNS administration of CNS prophylaxis: IT MTX and/or cytarabine, n = central nervous system, IT MTX intrathecal methotrexate, HD-MTX 253 (87%); or HD-MTX, n = 42 (13%). Among the 42 patients who high-dose methotrexate. received HD-MTX, 11 received concomitant IT prophylaxis. Patients’ characteristics by prophylactic regimen are shown in Table 1. Patients who received prophylaxis (IT or HD-MTX) were PATIENTS AND METHODS younger (p < 0.001) and presented with a better performance Patients status (ECOG < 2) (p = 0.002) compared with patients who did not We retrospectively reviewed the records of all newly diagnosed patients receive prophylaxis. Patients with two or more extranodal sites with DLBCL at Memorial Sloan Kettering Cancer Center (MSKCC) from 2001 involvement were more likely to receive CNS prophylaxis (p < to 2017, treated with frontline R-CHOP or R-CHOP-like regimens. In all 0.001). cases, the pathology at diagnosis was confirmed by expert hematopathol- The median administrations of IT prophylaxis were 4 (range ogists at MSKCC according to the World Health Organization classification 1–9), and the majority of patients received MTX alone. The dose of of hematopoietic and lymphoid tumors. Patients with primary mediastinal IT MTX and IT cytarabine was 12 and 70 mg, respectively. Patients B cell lymphoma, HIV positive or known CNS disease at diagnosis were had a median of two cycles of HD-MTX (range 1–6), at a median excluded. Patients with history of indolent lymphoma previously treated with chemotherapy were also excluded. The Hans algorithm [21] was used dose of 3500 mg/m2 (range 2000–3500 mg/m2). Overall, 23 to classify patients as germinal center B-cell like phenotype (GCB) or non- patients (55%) had HD-MTX at the end of R-CHOP-like treatment germinal center B-cell like (non-GCB). High risk for CNS relapse (HR-CNS) and 19 (45%) patients during R-CHOP-like treatment. Patients was defined by high-CNS-IPI (4–6 risk factors) or low or intermediate CNS- receiving HD-MTX were admitted to the hospital and received IPI along with testicular, breast, kidney, adrenal glands, and/or bone leucovorin rescue starting 24 h after MTX. Overall, 6 out of 42 marrow involvement. Patients with MYC and BCL2 rearrangement were patients (14%) developed acute renal injury (grade 3 in all cases) also included in HR-CNS group. Patients with low or intermediate CNS-IPI related to HD-MTX; two patients at the end of chemotherapy and paraspinal masses, sinus, orbit or skull involvement were not treatment and four during systemic chemotherapy treatment. All considered HR-CNS. patients recovered completely and no patient required dialysis. CNS prophylaxis was administered based on physician’s preference. CNS relapse was diagnosed by the presence of radiological findings, detection None of the patients received further HD-MTX after developing of lymphoma cells in the CSF and/or by brain biopsy. This study was renal injury, two patients transitioned to IT MTX, one patient approved by the institutional reviewed board at MSKCC. received a single administration of high-dose cytarabine, and three patients did not receive further prophylactic treatment. Statistical analysis Competing risk analysis was used to analyze time to CNS relapse, with CNS relapse and the effect of prophylaxis on CNS relapse systemic non-CNS relapse and death without relapse as competing events. After a median follow-up of 6.8 years, 36 out of 585 patients Time to event was calculated from the date of diagnosis to the date of one considered HR-CNS, relapsed in the CNS with a 5-year risk of 6.5%. of the three events, relapse in the CNS, systemic relapse, or death without Fourteen (39%) patients had received prophylaxis: 12 IT and 2 HD- relapse whichever occurs first. OS from relapse was defined as the time MTX. The risk of CNS relapse at 5 years for patients who received Blood Cancer Journal (2021)11:113 S. Bobillo et al. 3 Table 1. Patient’s characteristics. Variables IT methotrexate HD-methotrexate No prophylaxis p value n (%) n (%) n (%) Number (n) 253 (43%) 42 (7%) 290 (50%) Median age (range) 64 (21–86) 63 (27–81) 72 (24–91) <0.001 Male 142 (56) 24 (57) 135 (47) 0.07 ECOG 0–1 160 (63) 27 (64) 142 (49) 0.002 ≥2 93 (37) 15 (38) 148 (51) Stage I–II 30 (12) 4 (10) 30 (10) 0.81 III–IV 223 (88) 38 (90) 260 (90) Serum LDH Above normal 198 (78) 28 (67) 207 (71) 0.14 Missing 5 (2) 0 17 (6) CNS-IPI riska 0 10 (4) 3 (7) 12 (4) 1 16 (6) 2 (5) 18 (6) 2 28 (11) 4 (10) 17 (6) 3 41 (16) 4 (9) 31 (11) 4 107 (43) 16 (38) 157 (54) 5 41 (16) 12 (29) 49 (17) 6 10 (4) 1 (2) 6 (2) a CNS-IPI risk groups Low 0–1 26 (10) 5 (12) 30 (10) 0.009 Intermediate 2–3 69 (27) 8 (19) 48 (17) High 4–6 158 (63) 29 (69) 212 (73) High risk site Testis 39 (15) 8 (19) 4 (1) Breast 15 (6) 5 (12) 27 (9) Kidney/adrenal glands 50 (20) 12 (29) 44 (15) Bone marrow 63 (25) 6 (14) 49 (17) Extranodal sites >2 sites 87 (36) 14 (35) 57 (20) <0.001 Double-hit 18 (7) 5 (12) 11 (4) 0.05 Treatment R-CHOP 143 (57) 32 (76) 220 (76) <0.001 R-EPOCH 43 (17) 7 (17) 40 (14) R-CHOP/RICE 67 (26) 3 (7) 30 (10) Cell of origin Germinal center 104 (41) 14 (33) 116 (40) 0.15 Non-germinal center 104 (41) 23 (55) 102 (35) Missing 45 (18) 5 (12) 72 (25) IT intrathecal, MTX methotrexate, HD high-dose, ECOG Eastern Cooperative Oncology Group, LDH lactate dehydrogenase, CNS-IPI central nervous system international prognostic index, R-CHOP rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone, R-EPOCH rituximab, etoposide, cyclophosphamide, doxorubicin, vincristine, prednisone, RICE rituximab, ifosfamide, carboplatin, etoposide. a Includes patients who were missing baseline LDH but were grouped regardless of its value. IT, HD-MTX, or no prophylaxis was 5.5%; 5% and 7.5% (p = 0.34), The median time to relapse since first diagnosis was 9 months respectively (Fig. 2). The risk of CNS relapse by prophylaxis was (range 6–110 months). Patients who received prophylaxis, either IT similar when excluding patients who received concomitant HD- or HD-MTX, relapsed later than patients who did not receive MTX and IT (Supplementary Fig. 1). CNS relapse risk was similar prophylaxis, with a median time to relapse of 19 months (range among patients who received R-CHOP, R-EPOCH, and R-CHOP/ 7–55 months) vs. 8 months (range 6–110 months), respectively. RICE (p = 0.12) (Supplementary Fig. 2). The 5-year CNS relapse risk The risk of relapse at 1 year was lower for patients who received for patients considered low-risk CNS and not included in the study prophylaxis compared to patients who did not receive prophylaxis was 1.1%. 2% vs. 7.1%, RR 0.29 (95% CI; 0.08,0.66). However, over time, the Blood Cancer Journal (2021)11:113 S. Bobillo et al. 4 Table 2. Site of CNS relapse by CNS prophylaxis. Site of CNS relapse CNS prophylaxis Leptomeninges Parenchyma Both regimen IT 5 (42%) 3 (25%) 4 (33%) HD-MTX 1 (50%) 1 (50%) – No prophylaxis 5 (23%) 13 (57%) 4 (18%) CNS central nervous system, IT intrathecal, MTX methotrexate, HD high- dose. Fig. 2 Cumulative incidence of CNS relapse rate by prophylactic strategy. IT MTX intrathecal methotrexate, HD-MTX high-dose methotrexate. Fig. 4 Overall survival of patients with CNS relapse vs. other relapse. Fig. 3 Risk ratio of CNS relapse with and without prophylaxis over time. risk of CNS relapse became closer among prophylaxis and no- prophylaxis groups, with a 3-year risk of 3.8% vs. 7.5% (RR 0.51, CI 95%; 0.22, 1.04) and a 5-year risk of 5.6% vs. 7.5% (RR 0.76, CI 95%; 0.35,1.50), respectively (Fig. 3). At the time of CNS relapse, 26 (72%) patients presented with disease confined to the CNS and 10 (28%) patients had a concomitant systemic and CNS relapse. The brain parenchyma was the most common site of relapse (47%), followed by Fig. 5 Cumulative incidence of CNS relapse by cell of origin. GCB leptomeninges (30%) and both sites (23%). Sites of relapse by germinal center B-cell phenotype, non-GCB non germinal center B- prophylactic treatment are detailed in Table 2. cell phenotype. Patients with CNS relapse (n = 36) presented worse outcomes than patients with systemic relapse without CNS involvement DISCUSSION (n = 145) with a median OS of 4.9 months vs. 17.1 months, To our knowledge, this is one of the largest series analyzing the respectively (p = 0.003) (Fig. 4). role of CNS prophylaxis exclusively in patients with high risk of The COO determined by the Hans algorithm was available in 463 CNS relapse in the modern era. According to the CNS-IPI model, out of 585 patients, of whom 229 (49%) had a non-GCB phenotype we considered high-risk for CNS relapse patients with 4 to 6 risk and 234 (51%) patients had a GCB phenotype. Overall, 127 (55%) factors. In addition, we included patients with involvement of patients in the non-GCB group and 108 (46%) in the GCB group certain extranodal sites traditionally associated with an increased received prophylaxis. Patients with non-GCB phenotype had a risk of CNS relapse, such as testes, breast, kidney, and bone higher risk of CNS relapse compared with patients with GCB subtype marrow [3–6], regardless the CNS-IPI; making our results more with a 5-year risk of 9.9% vs. 4.5%, respectively (p = 0.03) (Fig. 5). informative. We also included double-hit lymphomas that have Finally, we performed a sub-analysis excluding patients with been associated with a high risk of CNS relapse, although low or intermediate CNS-IPI and bone marrow involvement (n = estimates vary widely among different studies [2, 8, 9]. Notably, 66), since the association between bone marrow and CNS relapse 50% of the patients included in our study did not receive CNS is controversial. We observed a 5-year CNS relapse risk of 5.2%, prophylaxis. These patients tended to be older and had worse 5.3%, and 7.5% for patients receiving IT and HD-MTX and no performance status than those who receive prophylaxis. Similar prophylaxis; respectively. Due to the low number of events in this results were found in recent large retrospective studies [13, 19]. group, we could not perform a sub-analysis to investigate the risk Consistent with recently published studies in the rituximab era, of CNS relapse in patients with bone marrow involvement. we observed a risk of CNS relapse of 6.5% at 5 years in the HR-CNS Blood Cancer Journal (2021)11:113 S. Bobillo et al. 5 group [13, 15, 20, 22, 23]. Furthermore, our results are similar to a Although in our series the risk of CNS relapse was similar recent real-world study from Sweden including 4205 patients with regardless the use of prophylaxis, we found that within the first DLBCL, that reported a 2-year CNS risk of 8% in the high-risk year from diagnosis the risk was higher in patients who did not patients [23]. Other investigators reported a slightly higher receive prophylaxis compared to patients who received IT or HD- incidence of CNS relapse, with a 2-year risk of 10.2% [10]. The MTX with a risk of 7.1% vs. 2% vs. 2.4%, respectively. However, lower incidence of our study is likely due to the fact that we over time, the risk became similar between groups, raising up to frequently perform brain magnetic resonance imaging in addition 5.6% and 5.2% at 5 years in the HD-MTX and IT groups, to a diagnostic lumbar puncture. Therefore, patients with active respectively. These findings were not observed in previous studies CNS involvement at diagnosis are promptly identified and analyzing the role of CNS prophylaxis [13, 22]. However, the excluded from the prophylactic strategy. median follow-up in these studies was around 2.5 years, so late Whether CNS prophylaxis is useful in preventing CNS relapse has relapses might have been underestimated. Our evidence suggests been addressed in several studies over the past years, including the CNS prophylaxis might help to partially control undetected CNS most widely used strategy IT MTX and more recently intravenous disease present at diagnosis delaying the occurrence of CNS methotrexate. The majority of the retrospective studies and post relapse, rather than preventing it. Furthermore, the use of CNS hoc analysis from prospective trials showed the lack of efficacy of IT prophylaxis might not prevent from late CNS relapse in the HR- prophylaxis in the rituximab era [10–13, 22]. In this regard, Kumar CNS population. et al. using the prospectively collected National Comprehensive Finally, we noticed that the presence of non-GCB phenotype Cancer Network database, analyzed the risk of CNS relapse in 989 determined by immunohistochemistry confers an increased risk of newly diagnosed DLBCL, 117 (11.8%) of whom had received CNS relapse in the HR-CNS population. These results are in prophylaxis, mostly IT MTX. They described no benefit from accordance with a recent publication analyzing the impact of COO prophylaxis in the whole population and also in high-risk patients, by GEP in 1418 patients from the GOYA phase 3 trial. In this study, though patients with involvement of kidney or adrenal glands or the authors found that ABC or unclassified phenotype was an patients with double-hit lymphomas were not included in the high- independent risk factor for CNS relapse, and patients with high- risk group [22]. More recently, a large retrospective study including CNS-IPI together with an ABC phenotype had a 2-year CNS relapse 690 patients older than 70 years of whom 271 had high-risk CNS-IPI, rate of 15% [10]. Interestingly, although we used immunohisto- also described a similar rate of CNS relapse regardless the use of IT chemistry techniques to determine the COO phenotype, we MTX prophylaxis [HR 1.34 (95% CI, 0.46–3.86)] [13]. Our series observed similar results, suggesting that in the absence of included only high-risk patients from all ages who received mostly molecular analysis, COO assessed by immunohistochemistry along IT MTX as prophylaxis (43%). In line with previously reported, we did with other risk factors could be useful to identify high-risk not observe a clear benefit from IT MTX on preventing CNS relapse, patients. especially after the first year of immunochemotherapy. In conclusion, our study highlights the need for developing In recent years, the use of HD-MTX administrated mid more effective CNS prophylaxis regimens than MTX. Furthermore, chemotherapy cycles or after completing systemic chemotherapy our data in addition to emerging data from other centers, did not was proposed as an alternative strategy to prevent CNS recurrence. demonstrate clinical advantage for using intravenous HD-MTX The rational is based on the observation that CNS relapses over traditional IT MTX. Since HD-MTX is associated with higher frequently involve the brain parenchyma [14–16]. This approach incidence of adverse events resulting in more delays in the was initially evaluated in the pre-rituximab era by the Groupe administration of R-CHOP cycles, we recommend that benefit/risk d’Etudes des Lymphomes de l’Adulte/Lymphoma Study Associa- assessment should be carefully considered before adopting HD- tion comparing high-dose chemotherapy plus two cycles of HD- MTX for CNS prophylaxis. MTX vs. CHOP without CNS prophylaxis, with significant reduction in the CNS relapse risk of 0.8% vs. 2.7%, respectively [24]. In the modern era, different retrospective series had supported this REFERENCES evidence in patients treated with R-CHOP chemotherapy [14–16], 1. El-Galaly TC, Cheah CY, Bendtsen MD, Nowakowski GS, Kansara R, Savage KJ, et al. Treatment strategies, outcomes and prognostic factors in 291 patients with although two recent retrospective studies have shown no clear secondary CNS involvement by diffuse large B-cell lymphoma. Eur J Cancer. benefit [17, 18]. More recently, preliminary results from a large 2018;93:57–68. retrospective series from Canada including 326 high-risk DLBCL 2. Gleeson M, Counsell N, Cunningham D, Chadwick N, Lawrie A, Hawkes EA, et al. patients, showed the lack of effectiveness of HD-MTX with a CNS Central nervous system relapse of diffuse large B-cell lymphoma in the rituximab relapse risk of 11.2% for patients who received HD-MTX (n = 115) era: results of the UK NCRI R-CHOP-14 versus 21 trial. Ann Oncol. 2017;28:2511–6. vs. 12.2% for patients who did not [19]. Consistent with these 3. Schmitz N, Zeynalova S, Nickelsen M, Kansara R, Villa D, Sehn LH, et al. CNS studies, our data also failed to demonstrate a clinical advantage for international prognostic index: a risk model for CNS relapse in patients with using the more toxic intravenous HD-MTX regimen. More recently, diffuse large B-cell lymphoma treated with R-CHOP. J Clin Oncol. 2016;34:3150–6. a multicenter retrospective study found no difference in the 4. Vitolo U, Chiappella A, Ferreri AJ, Martelli M, Baldi I, Balzarotti M, et al. First-line treatment for primary testicular diffuse large B-cell lymphoma with rituximab- efficacy between intravenous HD-MTX and IT MTX prophylaxis, CHOP, CNS prophylaxis, and contralateral testis irradiation: final results of an with a CNS relapse risk of 5% vs. 7%, respectively [20]. international phase II trial. J Clin Oncol. 2011;29:2766–72. A multicenter retrospective Australian study analyzed 217 high- 5. Hu S, Song Y, Sun X, Su L, Zhang W, Jia J, et al. Primary breast diffuse large B-cell risk patients treated with: (1) R-CHOP plus IT MTX, (2) R-CHOP pus lymphoma in the rituximab era: Therapeutic strategies and patterns of failure. HD-MTX, or (3) dose-intensive chemotherapy plus HD-MTX. They Cancer Sci. 2018;109:3943–52. observed a lower incidence of CNS relapse in patients receiving 6. Hosein PJ, Maragulia JC, Salzberg MP, Press OW, Habermann TM, Vose JM, et al. A HD-MTX compared to IT MTX, with a 3-year CNS relapse rate of multicentre study of primary breast diffuse large B-cell lymphoma in the ritux- 18.4% vs. 6.9% vs. 2.3%, for groups 1, 2, and 3, respectively [16]. imab era. Br J Haematol. 2014;165:358–63. Notably, in this study, only 37% of patients from the IT MTX group 7. Kridel R, Telio D, Villa D, Sehn LH, Gerrie AS, Shenkier T, et al. Diffuse large B-cell lymphoma with testicular involvement: outcome and risk of CNS relapse in the had received rituximab as part of the induction therapy, while rituximab era. Br J Haematol. 2017;176:210–21. nearly all patients in the HD-MTX group received immunochem- 8. Oki Y, Noorani M, Lin P, Davis RE, Neelapu SS, Ma L, et al. Double hit lymphoma: the otherapy [16]. In our study where all patients had received MD Anderson Cancer Center clinical experience. Br J Haematol. 2014;166:891–901. immunochemotherapy, we did not detect differences in the 9. Petrich AM, Gandhi M, Jovanovic B, Castillo JJ, Rajguru S, Yang DT, et al. Impact of efficacy between IT and HD-MTX, although the number of patients induction regimen and stem cell transplantation on outcomes in double-hit in the HD-MTX group was low. lymphoma: a multicenter retrospective analysis. Blood. 2014;124:2354–61. Blood Cancer Journal (2021)11:113 S. Bobillo et al. 6 10. Klanova M, Sehn LH, Bence-Bruckler I, Cavallo F, Jin J, Martelli M, et al. Integration analyze the data. CLB, PCC, AH, PAH, SMH, AK, MJM, AM, AN, DS, PG, PM, CNO, MLP, of cell of origin into the clinical CNS International Prognostic Index improves CNS DS, GK, ADZ, and AD treated patients and approved the manuscript. relapse prediction in DLBCL. Blood. 2019;133:919–26. 11. Boehme V, Schmitz N, Zeynalova S, Loeffler M, Pfreundschuh M. CNS events in elderly patients with aggressive lymphoma treated with modern chemotherapy COMPETING INTERESTS (CHOP-14) with or without rituximab: an analysis of patients treated in the AD has received personal fees from Roche, Corvus Pharmaceuticals, Physicians’ RICOVER-60 trial of the German High-Grade Non-Hodgkin Lymphoma Study Education Resource, Seattle Genetics, Peerview Institute, Oncology Specialty Group, Group (DSHNHL). Blood. 2009;113:3896–902. Pharmacyclics, Celgene, Novartis, Takeda and research grants from National Cancer 12. Schmitz N, Zeynalova S, Glass B, Kaiser U, Cavallin-Stahl E, Wolf M, et al. CNS Institute and Roche. CLB has received grant funding from Janssen, Novartis, disease in younger patients with aggressive B-cell lymphoma: an analysis of Epizyme, Xynomics, Bayer, BMS, and served as a consultant for Life Sci, GLG, patients treated on the Mabthera International Trial and trials of the German Celgene, Seattle Genetics, Xynomics, and holds honorarium from Dava Oncology. PH High-Grade Non-Hodgkin Lymphoma Study Group. Ann Oncol. 2012;23:1267–73. receives research support from Portola, Novartis/GSK, Molecular Templates, Janssen 13. Eyre TA, Kirkwood AA, Wolf J, Hildyard C, Mercer C, Plaschkes H. et al. Stand-alone Pharmaceuticals, and served as consultant for Karyopharm, Juno, Portola, Celgene, intrathecal central nervous system (CNS) prophylaxis provide unclear benefit in AstraZeneca. SH has received research funding from ADCT therapeutics, Aileron, reducing CNS relapse risk in elderly DLBCL patients treated with R-CHOP and is Forty-Seven, Verastem, Kyowa Hakko Kirin, Millennium Pharmaceuticals Inc, Celgene, associated increased infection-related toxicity. Br J Haematol. 2019;187:185–94. Trillium, Daiichii Sankyo, and consults for Astex, Affimed, Merck Sharp and Dome, 14. Abramson JS, Hellmann M, Barnes JA, Hammerman P, Toomey C, Takvorian T, Kyowa Hakko Kirin Pharma, Corvus Pharmaceuticals Inc, Celgene, Portola et al. Intravenous methotrexate as central nervous system (CNS) prophylaxis is Pharmaceuticals, Takeda Millennium, Innate Pharma, Verastem, Miragen Therapeu- associated with a low risk of CNS recurrence in high-risk patients with diffuse tics Inc, Seattle Genetics, ADCT. AK receives research funding from Abbvie large B-cell lymphoma. Cancer. 2010;116:4283–90. Pharmaceuticals, Adaptive Biotechnologies, Pharmacyclics, Seattle Genetics, and 15. Ferreri AJ, Bruno-Ventre M, Donadoni G, Ponzoni M, Citterio G, Foppoli M, et al. serves on advisory board for Celgene and AstraZeneca. MLP receives honoraria from Risk-tailored CNS prophylaxis in a mono-institutional series of 200 patients with Flagship Ventures, Novartis, Evelo, Seres Therapeutics, Jazz Pharmaceuticals, diffuse large B-cell lymphoma treated in the rituximab era. Br J Haematol. Therakos, Amgen, Merk and consults for Merk and Pharmacyclics. AN receives 2015;168:654–62. honoraria from Janssen, Pharmacyclics, Prime Oncology, and consults for Medscape, 16. Cheah CY, Herbert KE, O'Rourke K, Kennedy GA, George A, Fedele PL, et al. A and serves on advisory board for Janssen, and serves on the speakers’ bureau for multicentre retrospective comparison of central nervous system prophylaxis Prime Oncology, and receives research funding for Rafael Pharma and Pharma- strategies among patients with high-risk diffuse large B-cell lymphoma. Br J cyclics. AM receives research support from Seattle Genetics, Merck, Bristol-Myers Cancer. 2014;111:1072–9. Squibb, Incyte, and receives honorarium from Kyowa Hakko Kirin Pharma, Miragen 17. Lee K, Yoon DH, Hong JY, Kim S, Lee K, Kang EH, et al. Systemic HD-MTX for CNS Therapeutics, Takeda Pharmaceuticals, ADC Therapeutics, Seattle Genetics, Cell prophylaxis in high-risk DLBCL patients: a prospectively collected, single-center Medica, Bristol-Myers Squibb, Erytech Pharma. DS consults for InPractice Elselvier, cohort analysis. Int J Hematol. 2019;110:86–94. Seattle Genetics, and is on speaker’s bureau for Medical Crossfire. ADZ consults for 18. Goldschmidt N, Horowitz NA, Heffes V, Darawshy F, Mashiach T, Shaulov A, et al. Genentech/Roche, Gilead, Celgene, Janssen, Amgen, Novartis, Adaptive Biotechnol- Addition of high-dose methotrexate to standard treatment for patients with ogy, Verastem, and serves on advisory board for MorphoSys, Gilead, Genentech, high-risk diffuse large B-cell lymphoma contributes to improved freedom from Abbvie, AstraZeneca, Pharmacyclics, and receives research funding from MEI progression and survival but does not prevent central nervous system relapse. Pharmaceuticals, Roche, Gilead, Beigene. AY received research support from Leuk Lymphoma. 2019;60:1890–8. Janssen, Curis, Merck, BMS, Syndax, Roche, and honorarium from Janssen, Abbvie, 19. Puckrin R, Darsa EIH, Ghosh S, Peters A, Stewart DA. Lack of effectiveness of Merck, Curis, Epizyme, Roche, Takeda, and consulted for Biopath, Xynomics, intravenous high-dose methotrexate for prevention of CNS relapse in patients Epizyme, Roche, Celgene, HCM. AY is currently an employee of AstraZeneca. The with high-risk DLBCL: a retrospective analysis from Alberta, Canada. Blood. remaining authors declare no competing interests. 2020;136(supplem 1):26–7. 20. Orellana-Noia VM, Reed DR, Sen JM, Barlow C, Malecek MK, Kahl BS, et al. CNS prophylaxis during front-line therapy in aggressive non-Hodgkin lymphomas: ADDITIONAL INFORMATION real-world outcomes and practice patterns from 19 US academic institutions. Blood. 2020;136(supplem 1):27–8. Supplementary information The online version contains supplementary material 21. Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie J, Ott G, et al. available at https://doi.org/10.1038/s41408-021-00506-3. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood. 2004;103:275–82. Correspondence and requests for materials should be addressed to A.Y. 22. Kumar A, Vanderplas A, LaCasce AS, Rodriguez MA, Crosby AL, Lepisto E, et al. Lack of benefit of central nervous system prophylaxis for diffuse large B-cell Reprints and permission information is available at http://www.nature.com/ lymphoma in the rituximab era: findings from a large national database. Cancer. reprints 2012;118:2944–51. 23. Harrysson S, Eloranta S, Ekberg S, Enblad G, Jerkeman M, Wahlin BE, et al. Inci- Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims dence of relapsed/refractory diffuse large B-cell lymphoma (DLBCL) including in published maps and institutional affiliations. CNS relapse in a population-based cohort of 4243 patients in Sweden. Blood Cancer J. 2021;11:9. 24. Tilly H, Lepage E, Coiffier B, Blanc M, Herbrecht R, Bosly A, et al. Intensive con- ventional chemotherapy (ACVBP regimen) compared with standard CHOP for poor-prognosis aggressive non-Hodgkin lymphoma. Blood. 2003;102:4284–9. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative ACKNOWLEDGEMENTS Commons license, and indicate if changes were made. The images or other third party This work was supported in part by research funding from Fundación Alfonso Martín material in this article are included in the article’s Creative Commons license, unless Escudero to SB. Data from this manuscript were presented at the 61st Annual indicated otherwise in a credit line to the material. If material is not included in the Meeting of the American Society of Hematology, Orlando, FL, December article’s Creative Commons license and your intended use is not permitted by statutory 7th–10th, 2019. regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. AUTHOR CONTRIBUTIONS SB and AY conceived the project, analyzed the data, and wrote the manuscript. VS analyze the data, wrote the manuscript and provide the figures. DS contributed to © The Author(s) 2021 Blood Cancer Journal (2021)11:113