sign in sign up
<<
Home , Burkitt lymphoma, Follicular lymphoma, Mantle cell lymphoma

VOLUME 31 NUMBER 32 NOVEMBER 10 2013

JOURNAL OF CLINICAL ORIGINAL REPORT

Lymphoma Occurring During Pregnancy: Antenatal Therapy, Complications, and Maternal Survival in a Multicenter Analysis Andrew M. Evens, Ranjana Advani, Oliver W. Press, Izidore S. Lossos, Julie M. Vose, Francisco J. Hernandez-Ilizaliturri, Barrett K. Robinson, Stavroula Otis, Liat Nadav Dagan, Ramsey Abdallah, Aimee Kroll-Desrosiers, Jessica L. Yarber, Jose Sandoval, Kelley Foyil, Linda M. Parker, Leo I. Gordon, Andrew M. Evens, Tufts University Kristie A. Blum, Christopher R. Flowers, John P. Leonard, Thomas M. Habermann, and Nancy L. Bartlett School of Medicine, Boston; Aimee Kroll-Desrosiers, University of Massa- ABSTRACT chusetts Medical School, Worcester, MA; Ranjana Advani and Stavroula Otis, Purpose Stanford University Medical Center, is the fourth most frequent in pregnancy; however, current clinical practice is Stanford, CA; Oliver W. Press and Linda M. Parker, Fred Hutchinson based largely on small series and case reports. Cancer Research Center, Seattle, WA; Patients and Methods Izidore S. Lossos, Liat Nadav Dagan, In a multicenter retrospective analysis, we examined treatment, complications, and outcomes for and Jose Sandoval, University of Miami (HL) and non-Hodgkin lymphoma (NHL) occurring during pregnancy. School of Medicine, Miami, FL; Julie M. Vose, University of Nebraska Medical Results Center, Omaha, NE; Francisco Among 90 patients (NHL, n ϭ 50; HL, n ϭ 40), median age was 30 years (range, 18 to 44 years) J. Hernandez-Ilizaliturri, Roswell Park and median diagnosis occurred at 24 weeks gestation. Of patients with NHL, 52% had Cancer Institute, Buffalo; Ramsey advanced-stage versus 25% of patients with HL (P ϭ .01). Pregnancy was terminated in six Abdallah and John P. Leonard, Weill Cornell Medical College, New York, NY; patients. Among the other 84 patients, 28 (33%) had therapy deferred to postpartum; these Barrett K. Robinson, Indiana University patients were diagnosed at a median 30 weeks gestation. This compared with 56 patients (67%) School of Medicine, Indianapolis, IN; who received antenatal therapy with median lymphoma diagnosis at 21 weeks (P Ͻ .001); 89% of Jessica L. Yarber and Leo I. Gordon, these patients received combination . The most common preterm complication was Northwestern University Feinberg induction of labor (33%). Gestation went to full term in 56% of patients with delivery occurring at School of Medicine, Chicago, IL; Kristie a median of 37 weeks. There were no differences in maternal complications, perinatal events, or A. Blum, The Ohio State University, Columbus, OH; Christopher R. Flowers, median infant birth weight based on deferred versus antenatal therapy. At 41 months, 3-year Emory University, Atlanta, GA; Thomas progression-free survival (PFS) and overall survival (OS) for NHL were 53% and 82%, respectively, M. Habermann, Mayo Clinic, Rochester, and 85% and 97%, respectively, for HL. On univariate analysis for NHL, radiotherapy predicted MN; and Nancy L. Bartlett, Washington inferior PFS, and increased and poor Eastern Cooperative Oncology Group University School of Medicine, St. performance status (ECOG PS) portended worse OS. For HL patients, nulliparous status and “B” Louis, MO. symptoms predicted inferior PFS. Published online ahead of print at www.jco.org on September 16, 2013. Conclusion Standard (non-antimetabolite) combination chemotherapy administered past the first trimes- Presented at the 11th International ter, as early as 13 weeks gestation, was associated with few complications and expected Conference on Malignant Lymphoma, maternal survival with lymphoma occurring during pregnancy. Lugano, Switzerland, June 15-18, 2011, and the 53rd American Society of Annual Meeting and Expo- J Clin Oncol 31:4132-4139. © 2013 by American Society of Clinical Oncology sition, San Francisco, CA, December 11-13, 2011. There remains a deficiency in the literature of INTRODUCTION Authors’ disclosures of potential con- lymphoma-specific data to guide clinicians and pa- flicts of interest and author contribu- tions are found at the end of this Approximately 3,500 to 4,000 cases of cancer are tients regarding presentation, optimal tim- article. diagnosed during pregnancy each year in the United ing of therapy, maternal complications, perinatal Corresponding author: Andrew Evens, States, which translates to approximately one in ev- events, and fetal and maternal outcomes. DO, MSc, Division of Hematology and ery 1,000 gestations.1-5 Lymphoma is the fourth Furthermore, there is a paucity of data describ- Oncology, Tufts Medical Center, 800 ing the complications and outcomes associated with Washington St, Boston, MA 02111; most frequent cancer that occurs during pregnancy, e-mail: [email protected]. with Hodgkin lymphoma (HL) being more com- chemotherapy for lymphoma given during gesta- 6 © 2013 by American Society of Clinical monly seen compared with non-Hodgkin lym- tion. Lishner et al reported outcomes on 48 patients 1 Oncology phoma (NHL). Due in part to pathologic and with HL during 50 pregnancies; however, only 10 0732-183X/13/3132w-4132w/$20.00 clinical heterogeneity, however, current clinical patients were diagnosed during gestation with six

DOI: 10.1200/JCO.2013.49.8220 practice of treating lymphoma during pregnancy is who received antenatal chemotherapy. Most NHL based largely on case reports and small case series. data consist of case reports or small case series. Ward

4132 © 2013 by American Society of Clinical Oncology Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Lymphoma During Pregnancy and Weiss7 compiled data on 42 patients with NHL during pregnancy ϭ from the literature and reported a composite overall survival (OS) rate Table 1. Characteristics of Patients With Non-Hodgkin Lymphoma (n 50) of 31%; most patients had been managed with steroids, radiotherapy, Characteristic No. % or single-agent chemotherapy. In addition, none of these reports an- Disease ء alyzed prognostic factors that may predict maternal survival. DLBCL 28 56 Here, we report a multicenter collaboration that examined a T-cell lymphoma 10 20 5 10 cohort of 90 patients with an established diagnosis of lymphoma Burkitt’s lymphoma 4 5 during pregnancy. Approximately two thirds of patients received MALT 1 2 antenatal therapy which, in most cases, was combination chemo- Composite (DLBCL and FL) 1 2 therapy. The aims of this analysis were to examine detailed mater- Prior pregnancies nal and fetal characteristics with attention to therapy received in Nulliparous 20 40 the antenatal period, preterm and perinatal complications, and 11326 Ͼ potential disease and treatment-related prognostic factors that pre- 11122 N/A 2 4 dicted maternal survival. “B” symptoms Yes 19 38 No 30 60 PATIENTS AND METHODS N/A 1 2 ECOG PS We conducted a multicenter, retrospective analysis at 11 academic centers of 02938 patients who had a diagnosis of NHL or HL during pregnancy over a 13-year 11020 period (1999-2011). Investigators collected data for all associated consecutive 2612 patients treated at their local institutions. There were 102 patients identified; 336 eight were excluded from the analysis because of diagnosis of lymphoma after N/A 2 4 pregnancy and four were excluded because of incomplete records (ie, no survival data). The final analysis was performed on 90 patients (NHL, n ϭ 50; Yes 25 50 HL, n ϭ 40). The study was approved by the institutional review board at each No 22 44 participating institution. Diagnosis was established by and reviewed by N/A 3 6 a hematopathologist at each institution. Staging and therapy were performed Lactate dehydrogenase at the discretion of the treating physician; the majority of staging was com- Increased 21 42 pleted by magnetic resonance imaging and ultrasonography. Functional im- Normal 25 50 18 aging (eg, [ F]fluorodeoxyglucose-positron emission tomography) was not N/A 4 8 performed in any patient antepartum. It is also important to highlight that all Albumin patients were comanaged with high-risk maternal-fetal medicine consultation. Decreased 28 56 Data on disease characteristics, treatment, obstetric complications, and Normal 17 34 fetal outcomes were examined, including detailed perinatal events such as N/A 5 10 endometritis, postpartum hemorrhage (ie, Ͼ 500 mL for vaginal delivery, involvement Ͼ 1,000 mL for cesarean section), transfusion of blood products, chorioam- Yes 4 8 nionitis, route of delivery (ie, cesarean v vaginal), gestational diabetes, and No 37 74 pre-eclampsia. Data were collected regarding preterm delivery (ie, Ͻ37 weeks) Not performed 9 18 and spontaneous preterm delivery (ie, premature rupture of membranes Extranodal disease (nonmarrow) 56 ءPROM] or preterm labor). For patients who received therapy during preg- Yes† 28] nancy, data were collected regarding treatment received (nonsteroid) and use No 21 42 of growth factors and complications of therapy. N/A 1 2 Statistical Analysis Bulky disease Yes 3 6 Progression-free survival (PFS) was calculated from the time of lym- No 45 90 phoma diagnosis to the time of disease relapse/progression or death. OS was N/A 2 4 measured from the date of lymphoma diagnosis to the date of death. Complete Stage‡ response (CR) was defined as complete resolution of the disease or absence of I1122 any disease-specific findings. PFS and OS were calculated by using the Kaplan- II 13 26 Meier method, and survival differences between the treatment groups were III 5 10 assessed by using the log-rank test. Hazard ratios (HRs) and their 95% CIs IV 15 30 were reported. Univariable analyses were performed by using Cox propor- N/A 1 2 tional hazards models. Factors included in the univariable model are listed in Appendix Table A1 (online only). SAS version 9.2 (SAS Institute, Cary, NC) Abbreviations: DLBCL, diffuse large B-cell lymphoma; ECOG PS, Eastern was used for all statistical analyses. Cooperative Oncology Group performance status; FL, follicular lymphoma; MALT, mucosa-associated lymphoid tissue; N/A, not available. .Included two patients with primary CNS lymphomaء †Thirteen patients had more than one extranodal site: liver, n ϭ 5; , n ϭ RESULTS 4; lung, n ϭ 3; bone, n ϭ 3; vagina, n ϭ 3; soft tissue, n ϭ 3; , n ϭ 3; psoas, n ϭ 2; brain, n ϭ 2; skin, n ϭ 2; , n ϭ 2; epidural, n ϭ 2; nasopharyngeal, n ϭ 2; gastrointestinal, n ϭ 2; thyroid, n ϭ 1; intraocular, n ϭ Disease Characteristics 1; parotid, n ϭ 1; and cardiac, n ϭ1. Baseline characteristics for the 90 patients with NHL (n ϭ 50) ‡One patient had relapsed disease. and HL (n ϭ 40) during pregnancy are depicted in Tables 1 and 2. Median age at diagnosis for all patients with HL was 29 years (range, 18 www.jco.org © 2013 by American Society of Clinical Oncology 4133 Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Evens et al

g/dL] and 10.8 g/dL [range, 9.4 to 11.8 g/dL], respectively). Patients ϭ Table 2. Characteristics of Patients with Hodgkin Lymphoma (n 40) with NHL more commonly had stage III to IV disease compared with Characteristic No. % patients with HL (54% v 25%; P ϭ .01). Of patients with stage I to II Disease histology disease, 17% of patients with NHL and 30% of patients with HL had 35 88 “B” symptoms and/or bulky disease. Approximately 85% of patients Not otherwise specified 3 7 had bone marrow evaluation; only 8% of patients with NHL and 3% Mixed cellularity 2 5 of patients with HL had involvement. Extranodal disease was fre- Prior pregnancies Nulliparous 7 18 quently identified in patients with NHL, with 26% of patients having 11845more than one extranodal site with several atypical sites seen (Table 1). Ͼ 11025Of the 26 patients with DLBCL (non-CNS), the median International N/A 5 12 Prognostic Index (IPI) score was 1 (range, 0 to 4); 13 of these patients “B” symptoms had stage III to IV disease with a median IPI score of 3 (range, 1 to 4). Yes 9 23 No 27 67 Gestation Data N/A 4 10 ECOG PS For the entire cohort, the diagnosis of lymphoma occurred at a 02870median of 24 weeks gestation (range, 10 to 38 weeks). There were no 1718differences based on NHL versus HL (25 and 23 weeks; P ϭ .49) or by 2-4 0 0 NHL subtype (data not shown). Two patients had a pre-existing N/A 5 12 diagnosis of follicular lymphoma that was also present during gesta- Anemia tion. Among all other patients, 11 (12%) of 88 were diagnosed in the Yes 26 65 No 10 25 first trimester, 44 (50%) of 88 in the second trimester, and 33 (38%) of N/A 4 10 88 in the third trimester, with no differences based on lymphoma type. Albumin All but one patient with NHL and one patient with HL had newly Decreased 24 60 diagnosed disease. The patient with HL was a 31-year-old female who Normal 11 28 presented with relapsed disease during pregnancy following an 8-year N/A 5 12 remission, and the patient with NHL was a 36-year-old female who Bone marrow involvement relapsed during pregnancy 21 years after treatment for anaplastic Yes 1 3 No 32 80 large-cell lymphoma (ALK-positive). Not performed 5 12 Pregnancy was terminated in six patients (three with NHL and N/A 2 5 three with HL) to enable immediate chemotherapy (five in first tri- Extranodal disease (nonmarrow) mester [at 5, 8, 10, 10, and 12 weeks] and one patient early in second ء ء Yes 2 5 trimester who warranted high-dose ). Among the other No 37 92 84 patients, 28 (33%) had therapy deferred until postpartum (15 N/A 1 3 Bulky disease [32%] of 47 with NHL and 13 [35%] of 37 with HL). Those 28 patients Yes 2 5 were diagnosed with lymphoma at a median of 30 weeks gestation No 38 95 (range, 12 to 38 weeks) compared with 56 (67%) patients who re- Stage† ceived antenatal therapy were diagnosed with lymphoma at a median I25gestation of 22 weeks (range, 6 to 32 weeks; P Ͻ .001; NHL: 22 weeks II 28 70 [range, 10 to 32 weeks]; HL: 20 weeks [range, 6 to 30 weeks]; P ϭ .35). III 8 20 Specific histologic subtypes for patients with NHL who received ante- IV 2 5 natal therapy are described in the footnote of Table 3. Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; N/A, not available. One patient with lung involvement and one patient with lung and Treatmentء liver involvement. Antenatal treatment was initiated in 56 patients at a median of 25 †One patient had relapsed disease. weeks gestation (range, 13 to 37 weeks) with no difference based on lymphoma subtype (data not shown). Further, therapy was initiated in the second trimester for 37 (66%) of 56 of patients. Treatment by to 40 years) versus 32 years (range, 19 to 44 years) for patients with lymphoma subtype and gestational period is delineated in Table 3. NHL (Pϭ.13). Median age for B-cell NHL was 29.5 years (range, 20 to The only patient with follicular lymphoma who received antenatal 44 years) versus 34 years (range, 19 to 37 years) for T-cell lymphoma therapy was a 28-year-old diagnosed at 19 weeks gestation with stage (TCL; P ϭ .09). The most common NHL diagnosis was diffuse large IVXE disease with leukemic phase and a 14- ϫ 11-cm pelvic mass B-cell lymphoma (DLBCL), which constituted 56% of all NHLs and causing epidural compression. Four of the five patients with NHL and 73% of B-cell NHLs. TCL subtypes included anaplastic large-cell lym- all of the four patients with HL who received antenatal radiotherapy phoma (n ϭ 7; four anaplastic lymphoma kinase (ALK) –positive and (range, 25.2 to 30.0 Gy) had supradiaphragmatic stage I to II disease. three ALK-negative), natural killer TCL (n ϭ 2), and peripheral TCL For supportive therapy, 14 patients with NHL and seven with HL not otherwise specified (n ϭ 1). received antenatal myeloid growth factor, and three patients with Hemoglobin was decreased in approximately half the patients NHL and one with HL received erythropoietin. Restaging was done with NHL or HL at diagnosis (median 9.8 g/dL [range, 8.2 to 11.6 postpartum after completion of all intended first-line therapy in all

4134 © 2013 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Lymphoma During Pregnancy

Table 3. Therapy Based on Lymphoma Type and Week Therapy Table 4. Maternal and Fetal Complications Based on Lymphoma Type and Started (n ϭ 56) Therapy (n ϭ 72) (HL (n ϭ 24) NHL (n ϭ 41) HL (n ϭ 31 ء(NHL (n ϭ 32

Week Therapy No. of No. of Antenatal No Antenatal No Started Therapy Patients Therapy Patients Therapy Therapy Therapy Therapy (n ϭ 28) (n ϭ 13) (n ϭ 20) (n ϭ 11) Second trimester 13-17 R-CHOP 3 ABVD 4 Complication No. % No. % No. % No. % CHOP 1 RT 2 Gestational DM 2 7 0 0 1 9 Modified hyper-CVAD† 1 AVD 1 Pre-eclampsia 3 11 2 15 0 0 18-22 R-CHOP 4 ABVD 4 IOL 8 29 4 31 7 35 5 45 CHOP 1 RT 1 PROM 5 18 3 23 0 1 9 38 5 39 9 40 6 32 17 ءEPOCH 1 Preterm delivery 23-27 R-CHOP 3 ABVD 4 Cesarean section 11 39 7 54 4 20 2 18 CHOP 2 Postpartum CHOP ϩ RT 2 hemorrhage† 1 4 1 7‡ 2 10 0 CHOP/ ϩ RT 1 NICU admission 5 18 3 23 0 0 Modified 1 Fetal demise CODOX-M/IVACϩ () 1 4 0 0 0 RT 1 Malformations 1 4 0 0 0 Third trimester Low gestational 28-30 CHOP 3 ABVD 2 age§ 9 41 1 9 2 10 2 29 R-CHOP 2 AVD 2 Abbreviations: DM, diabetes mellitus; HL, Hodgkin lymphoma; IOL, induction R-CHOP ϩ RT with 1RT1 of labor; NHL, non-Hodgkin lymphoma; NICU, neonatal intensive care unit; modified CODOX-M/ PROM, premature rupture of membranes. For this single category, the denominators for patients with NHL whoء IVACϩ 31-33 R-CHOP 2 ABVD 3 received therapy v no therapy were n ϭ 32 and n ϭ 15, respectively; the Modified ESHAP 1 denominators for patients with HL who received therapy v no therapy were n ϭ 24 and n ϭ 13, respectively. In addition, the incidence of spontaneous 34-38 and steroids 1 — preterm birth (defined as delivery occurring at Ͻ 37 weeks, patients whose membranes ruptured, and/or patients who had spontaneous onset of preterm Abbreviations: ABVD, , bleomycin, vinblastine, dacarbazine; AVD, labor) was six of 23 for patients with NHL v six of 14 for patients with HL. doxorubicin, vinblastine, dacarbazine; CHOP, , doxorubicin, †Between 500 and 1,000 mL (associated with delivery). , ; CODOX-M/IVAC, cyclophosphamide, doxorubicin, ‡Not including one patient with HELLP syndrome ͓hemolysis, elevated liver high-dose methotrexate/, etoposide, high-dose ; EPOCH, enzymes, and low platelet count͔ associated with Ͻ 500 mL of blood loss. etoposide, vincristine, doxorubicin, cyclophosphamide, prednisone; ESHAP, §Defined as below the 10th percentile by gestational age and sex. Data etoposide, methylprednisolone, cisplatin, cytarabine; HL, Hodgkin lym- known for children of 33 patients with NHL (22 were treated during phoma; hyper-CVAD, cyclophosphamide, vincristine, doxorubicine, dexa- gestation and 11 deferred therapy) and 27 patients with HL (20 were methasone, followed by methotrexate and cytarabine; NHL, non-Hodgkin treated during gestation and seven deferred therapy). lymphoma; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincris- tine, prednisone; RT, radiotherapy. for patients who received antenatal therapy were diffuse largeء B-cell lymphoma (n ϭ 21), T-cell lymphoma (n ϭ 7), Burkitt’s lymphoma (n ϭ 3), and follicular lymphoma (n ϭ 1); histologies for patients who had therapy deferred to postpartum were diffuse large B-cell lymphoma (n ϭ 5), follicular lymphoma (n ϭ 4), T-cell lymphoma (n ϭ 3), Burkitt’s lymphoma (n ϭ 2), and section in 33% of patients; more patients with NHL versus HL had lymphoma (n ϭ 1). cesarean section (44% v 19%, respectively; P ϭ .007). Timing of †Modified hyper-CVAD regimen.8 ϩWithout high-dose methotrexate. delivery was available in 83 of 84 patients; 44% had preterm deliveries with no difference based on NHL versus HL (51% v 36%; P ϭ .19).

Fetal Outcomes cases, this included functional imaging. The overall response rate The median gestation at delivery was 37 weeks (range, 31 to 40 (ORR) for the 56 patients who received antenatal therapy was 82% weeks). Notably, gestation went to full term in 47 (56%) of 84 patients (CR rate, 64%); ORR for patients with NHL was 71% (CR, 50%) with no differences between lymphoma types or whether antenatal versus 96% (CR, 83%) for patients with HL (ORR, P ϭ .03; CR, therapy was given (Table 4). There was one instance of fetal demise P ϭ .013). (miscarriage) that occurred at 19 weeks gestation in a 34-year-old patient with double-hit NHL (ie, concurrent c- and BCL-2 rear- Preterm and Perinatal Complications rangements) following one cycle of R-CHOP (rituximab combined Obstetrical information was available for 72 of 84 patients in with cyclophosphamide, doxorubicin, vincristine, and prednisone). whom pregnancy had not been terminated. Among these, preterm The median birth weight of infants was 2,668 g (range, 1,005 to 3,628 complications included induction of labor in 33% of patients (Table g) with no difference based on receipt of antenatal chemotherapy 4). Perinatal events included PROM in 13% and pre-eclampsia in 7%. versus deferred therapy (2,670 v 2,665 g, respectively; P ϭ .74); how- There were no episodes of chorioamnionitis or endometritis. Further- ever, there was a trend for infants to be small for gestational age if their more, there were no differences in complications detected among mothers received antenatal versus deferred therapy (41% v 9%, re- patientswhoreceivedantenataltreatment(radiotherapyand/orchem- spectively; Pϭ.09). Of gestations for which information was available, otherapy) versus deferred therapy. There were trends identified for an eight (11%) of 72 infants required admission to the neonatal intensive increased incidence of PROM (P ϭ .07) and pre-eclampsia (P ϭ .06) care unit (median stay, 12 days; range, 3 to 40 days) with no differences in patients with NHL versus HL. The route of delivery was cesarean based on antenatal therapy versus deferred therapy (Table 4). All www.jco.org © 2013 by American Society of Clinical Oncology 4135 Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Evens et al infants who were admitted to the neonatal intensive care unit were Maternal Outcomes born to patients with NHL. Microcephaly was reported in one infant At a median follow-up of 41 months (range, 6 to 147 months), following four antenatal cycles of CHOP [cyclophosphamide, doxo- the 3-year PFS and OS for all patients with NHL and HL were 53% and rubicin, vincristine, and prednisone] for the mother’s DLBCL (treat- 82%, and 85% and 97%, respectively (Fig 1). Three-year PFS and OS ment started at 28 weeks gestation; delivery was at 38 weeks). Grade 1 for all patients with DLBCL were 55% and 79%, respectively, pelviectasis was seen in one infant whose mother was treated with four compared with 37% and 90%, respectively, for patients with T-cell antenatal cycles of R-CHOP for DLBCL at 21 weeks gestation and NHL (P ϭ .43 and P ϭ .60, respectively). There were eight deaths delivered at 34 weeks due to pre-eclampsia. There were no other related to NHL, including five DLBCL, one peripheral TCL not oth- malformations detected. erwise specified, one Burkitt’s lymphoma, and one double-hit NHL.

A B 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4 (probability) (probability) Overall Survival 0.2 0.2 Progression-Free Survival

0 6 12 18 24 30 36 0 6 12 18 24 30 36 Time (months) Time (months) C D 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4 (probability) (probability) Overall Survival 0.2 0.2 Progression-Free Survival

0 6 12 18 24 30 36 0 6 12 18 24 30 36 Time (months) Time (months) E F 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4 (probability) (probability) Overall Survival 0.2 0.2 Progression-Free Survival

0 6 12 18 24 30 36 0 6 12 18 24 30 36 Time (months) Time (months)

Fig 1. Survival. The 3-year (A) progression-free survival (PFS) and (B) overall survival (OS) in the 50 patients with a diagnosis of non-Hodgkin lymphoma during pregnancy were 53% and 82%, respectively. Kaplan-Meier curves comparing (C) PFS and (D) OS for patients with diffuse large B-cell lymphoma; 3-year PFS and OS were 55% and 79%, respectively. Kaplan-Meier curves of (E) PFS and (F) OS for 40 patients with a diagnosis of Hodgkin lymphoma during pregnancy; 3-year PFS and OS were 85% and 97%, respectively.

4136 © 2013 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Lymphoma During Pregnancy

Three of these eight patients had received antenatal chemotherapy, frequently than NHL.1-4,9 This likely reflects the younger typical age of and the other four patients had therapy deferred to postpartum. onset of HL compared with NHL. Here, we report data on 90 patients with lymphoma from 11 academic centers who were diagnosed/man- Univariate Analysis for Survival aged during pregnancy. Approximately two thirds of patients received Multiple variables were analyzed (Appendix Table A1) in uni- antenatal therapy, most commonly combination chemotherapy. To variable analysis to predict outcomes. The only variable that predicted the best of our knowledge, this is the largest series of HL or NHL for increased risk of disease progression in the NHL cohort was the during pregnancy in which the majority of patients received standard receipt of radiotherapy (HR, 5.19; 95% CI, 1.76 to 15.34; P ϭ .003). All combination chemotherapy. patients with NHL received radiotherapy during pregnancy. Poor Only one quarter of patients with HL and just over half the Eastern Cooperative Oncology Group performance status (ECOG PS, patients with NHL in our series had advanced-stage disease. A likely 2 to 4: HR, 5.51; 95% CI, 1.36 to 22.39; P ϭ .001) and increased lactate contributing factor to these low rates of advanced-stage disease was the dehydrogenase (HR, 9.93; 95% CI, 1.22 to 80.87; P ϭ .03) predicted absence of functional imaging as well as less frequent use of other for inferior OS. For patients with DLBCL, radiotherapy predicted commonly used staging modalities (eg, bone marrow biopsy). Despite inferior PFS (HR, 7.72; 95% CI, 1.72 to 34.55; P ϭ .008) and poor the nonstandard and suboptimal staging performed, we identified a ECOG PS predicted worse OS (HR, 2.33; 95% CI, 1.40 to 5.10; P ϭ high rate of extranodal involvement in patients with NHL for whom .04). For patients with HL, multiparous status predicted improved several atypical sites were noted (eg, vaginal and ovary). This has been PFS (HR, 0.07; 95% CI, 0.01 to 0.66; Pϭ.01), although the presence of reported in prior series10-12 and may relate to increased blood flow to at diagnosis predicted inferior PFS (HR, 10.78; 95% CI, reproductive organs during gestation or possibly due to expression of 1.12 to 103.75; P ϭ .04); there were no variables in HL that were gestational hormone receptors. predictive of OS. Table 5 depicts the corresponding 3-year Kaplan and The decision to use chemotherapy during gestation is individ- Meier survival rates for prognostic factors identified on univari- ualized, with the risk of antenatal chemotherapy weighed against ate analysis. the potential adverse effect of delaying curative therapy. As ex- pected, the treatment approaches here were heterogeneous, al- though a consistent finding for all patients was the use of antenatal DISCUSSION therapy for patients diagnosed earlier in gestation. The median gestational age at diagnosis of lymphoma for patients who received Lymphoma diagnosis during pregnancy has an approximate preva- antenatal therapy was 22 weeks. Furthermore, 89% of these pa- lence of 1 in every 5,000 to 6,000 gestations, with HL occurring more tients received combination chemotherapy, with the most com- mon regimens being standard for the particular lymphoma subtype. A concern regarding use of antenatal chemotherapy is the

Table 5. Prognostic Factors With Associated 3-Year Survival Rates From risk of perinatal complications for the fetus and the mother. Univariate Analysis The most commonly identified perinatal events were induction 3-Year PFS 3-Year OS of labor, PROM, and cesarean delivery. Interestingly, there were no No. of differences in events among patients who received antenatal versus Variable Patients % 95% CI P % 95% CI P deferred therapy; this is consistent with recent data in pregnant pa- Non-Hodgkin lymphoma tients with breast cancer.5 The rate of miscarriage identified in the Received RT as part of therapy .001 .88 current data set was 1.1%, consistent with other cancer-related preg- to 0 80 20 to 97 nancy data.4,5 Van Calsteren et al4 showed that preterm labor occurred 0 0 5 ءYes No 43 65 46 to 79 81 64 to 91 in 54% of pregnancies, with the risk being increased in patients who ECOG PS .33 .007 received antenatal cytotoxic therapy. The rate of preterm labor ap- 0-1 39 54 34 to 71 88 70 to 95 peared slightly lower in our study. In addition, we did not identify an 2-3 9 53 18to80 56 20to80 LDH .29 .008 overall difference in median birth weight among patients with NHL or Normal 21 59 32 to 78 95 72 to 99 HL who received antenatal therapy compared with those who had 4,5 Increased 25 48 22 to 70 62 35 to 80 therapy deferred to postpartum. This is in contrast to other series, Hodgkin lymphoma although in the Loibl5 series, the difference was not significant. No. of prior A critical aspect in the management of cancer during pregnancy pregnancies .002 .04 is timing of delivery. The general goal in all cancer-associated gesta- Nulliparous 7 57 17 to 84 80 20 to 97 Ͼ 1 28 96 76 to 99 100 100 to 100 tions should be to continue the pregnancy to full term. This point was 13 Presence of “B” emphasized by Amant et al, who reported recent data on 236 cycles symptoms .009 .08 of chemotherapy administered in 68 pregnancies. When compared No 27 96 75 to 99 100 100 to 100 with chemotherapy for the general population, antenatal chemother- Yes 9 63 23to86 83 27to97 apy was not associated with increased neurologic, cardiac, auditory, NOTE. Three-year progression-free survival (PFS) and overall survival (OS) by general health, or development impairments. The most significant Kaplan-Meier analysis are reported; variables with P Ͻ .10 in at least one category (ie, PFS or OS) are included. factor that predicted for an impairment in cognitive development was Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance preterm delivery. Moreover, for each additional month of gestation, status; LDH, lactate dehydrogenase; RT, radiotherapy. .All patients (n ϭ 5) received antenatal RT; two non-Hodgkin lymphoma the average intelligence quotient was increased by 11.6 pointsء patients had unknown RT status. Our data also support prior findings that antenatal chemothera- py with standard regimens (non–antimetabolite) during the second or www.jco.org © 2013 by American Society of Clinical Oncology 4137 Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Evens et al third trimester does not appear to increase morbidity or mortality for who received antenatal radiotherapy for supradiaphragmatic disease the fetus.5,14 This includes in utero anthracyline exposure, which has without adverse impact on fetal or maternal outcome. been shown to not adversely affect maternal or fetal cardiac In summary, we found that standard chemotherapy regimens for function.13,15-18 We did not identify an increased rate of congenital NHL and HL (without antimetabolites) administered during the sec- malformations, consistent with prior data.4,5 It is important to note ond and third trimester, including as early as 13 weeks gestation in that none of the patients in our series received antenatal therapy some cases, was associated with minimal maternal complications or during the first trimester. Avile´s et al19 reported that administration of fetal detriment. In addition, patients with low-risk clinical scenarios antenatal therapy during the first trimester may be tenable in select (eg, indolent NHL, low tumor burden, and/or late gestational diagno- patients, although most guidelines do not advocate this practice.20,21 sis) had therapy safely deferred to postpartum. It should be under- Limitations of this study are the unintended selection and referral scored that all patients were managed concurrently with high-risk biases as well as the lack of long-term outcomes due to its retrospective maternal-fetal medical consultation. In our experience, this approach nature and nonsystematic treatment or assessments. It is also impor- was associated with maternal and fetal complications that appeared tant to acknowledge that subtle changes in neurocognitive or other consistent with a healthy population and expected lymphoma- developmental abnormalities of children exposed to antenatal chem- related outcomes. otherapy cannot be excluded in our cohort. Avile´s and Neri22 followed 62 children who had been exposed to antenatal chemotherapy. They AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS reported that all children developed normally physically and neuro- OF INTEREST logically when matching school performance and intelligence quo- tient testing with controls. Nonetheless, more research is needed The author(s) indicated no potential conflicts of interest. regarding appropriate chemotherapy dosing and calculation in preg- nancy, safety of supportive care therapy, and long-term study of child AUTHOR CONTRIBUTIONS cognition and development. There is a paucity of available data assessing potential prognostic Conception and design: Andrew M. Evens, Ranjana Advani, Oliver W. factors for survival of patients diagnosed with lymphoma during preg- Press, Izidore S. Lossos, Barrett K. Robinson, Aimee Kroll-Desrosiers, nancy. The survival rates in our study were consistent with the ex- Leo I. Gordon, Thomas M. Habermann, Nancy L. Bartlett pected lymphoma-related outcomes. In univariate analyses of Financial support: Andrew M. Evens Administrative support: Andrew M. Evens, Linda M. Parker prognostic factors for survival, nulliparous status and the presence of Provision of study materials or patients: Andrew M. Evens, Ranjana B symptoms predicted for inferior PFS. The finding of prior preg- Advani, Julie M. Vose, Christopher R. Flowers nancy status affecting PFS is interesting and may be related to hor- Collection and assembly of data: Andrew M. Evens, Ranjana Advani, monal alterations.23,24 In NHL, poor ECOG PS and increased lactate Oliver W. Press, Izidore S. Lossos, Julie M. Vose, Francisco dehydrogenase at diagnosis were associated with inferior OS, and J. Hernandez-Ilizaliturri, Stavroula Otis, Liat Nadav Dagan, Ramsey Abdallah, Jessica L. Yarber, Jose Sandoval, Kelley Foyil, Linda M. Parker, receipt of radiotherapy had an adverse impact on PFS. The latter is an Kristie A. Blum, Christopher R. Flowers, John P. Leonard, Thomas M. interesting finding and is likely a reflection of the systemic nature of Habermann, Nancy L. Bartlett NHL. Although the adverse impact of radiotherapy on PFS did not Data analysis and interpretation: Andrew M. Evens, Ranjana Advani, translate to an OS disadvantage, if treatment is considered in NHL, Oliver W. Press, Izidore S. Lossos, Julie M. Vose, Francisco chemotherapy should likely be the first choice. In contrast, for HL, J. Hernandez-Ilizaliturri, Barrett K. Robinson, Stavroula Otis, Liat Nadav there are data showing that in select patients (ie, those with supradia- Dagan, Ramsey Abdallah, Aimee Kroll-Desrosiers, Jessica L. Yarber, Kelley Foyil, Kristie A. Blum, Christopher R. Flowers, John P. Leonard, phragmatic disease), radiotherapy may be successfully used during Thomas M. Habermann, Nancy L. Bartlett 25 pregnancy with careful shielding of the fetus. Woo et al reported on Manuscript writing: All authors 16 patients with HL diagnosed during pregnancy over a 35-year period Final approval of manuscript: All authors

6. Lishner M, Zemlickis D, Degendorfer P, et al: 11. Kirkpatrick AW, Bailey DJ, Weizel HA: Bilat- REFERENCES Maternal and foetal outcome following Hodgkin’s eral primary breast lymphoma in pregnancy: A case disease in pregnancy. Br J Cancer 65:114-117, 1992 report and literature review. Can J Surg 39:333-335, 1. Brenner B, Avivi I, Lishner M: Haematologi- 7. Ward FT, Weiss RB: Lymphoma and preg- 1996 cal in pregnancy. Lancet 379:580-587, nancy. Semin Oncol 16:397-409, 1989 12. Magloire LK, Pettker CM, Buhimschi CS, et al: 2012 8. Kenkre VP, Long WL, Eickhoff JC, et al: Burkitt’s lymphoma of the ovary in pregnancy. Ob- 2. Mu¨ller AM, Ihorst G, Mertelsmann R, et al: Maintenance rituximab following induction chemo- stet Gynecol 108:743-745, 2006 Epidemiology of non-Hodgkin’s lymphoma (NHL): for : Long- 13. Amant F, Van Calsteren K, Halaska MJ, et al: Trends, geographic distribution, and etiology. Ann term follow-up of a pilot study from the Wisconsin Long-term cognitive and cardiac outcomes after Hematol 84:1-12, 2005 Oncology Network. Leuk Lymphoma 52:1675-1680, prenatal exposure to chemotherapy in children aged 3. Pereg D, Koren G, Lishner M: The treatment 2011 18 months or older: An observational study. Lancet of Hodgkin’s and non-Hodgkin’s lymphoma in preg- 9. Habermann TM, Witzig TE. Management of Oncol 13:256-264, 2012 nancy. Haematologica 92:1230-1237, 2007 non-Hodgkin lymphoma during pregnancy. In: Armit- 14. Abdel-Hady el-S, Hemida RA, Gamal A, et al: 4. Van Calsteren K, Heyns L, De Smet F, et al: age JO, Mauch PM, Harris NL, et al (eds) Non- Cancer during pregnancy: Perinatal outcome after in Cancer during pregnancy: An analysis of 215 pa- Hodgkin (ed 2). Philadelphia, PA, utero exposure to chemotherapy. Arch Gynecol tients emphasizing the obstetrical and the neonatal Kluwer Health/Lippincott Williams & Wilkins, 2010, Obstet 286:283-286, 2012 outcomes. J Clin Oncol 28:683-689, 2010 pp. 501-506 15. Gziri MM, Debie`ve F, De Catte L, et al: 5. Loibl S, Han SN, von Minckwitz G, et al: Treat- 10. Wang PH, Chao KC, Lin G, et al: Primary Chemotherapy during pregnancy: Effect of anthra- ment of breast cancer during pregnancy: An observa- malignant lymphoma of the cervix in pregnancy: A cyclines on fetal and maternal cardiac function. Acta tional study. Lancet Oncol 13:887-896, 2012 case report. J Reprod Med 44:630-632, 1999 Obstet Gynecol Scand 91:1465-1468, 2012

4138 © 2013 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Lymphoma During Pregnancy

16. Van Calsteren K, Berteloot P, Hanssens M, et al: 20. Cardonick E, Iacobucci A: Use of chemother- lactin levels during proestrus and alters estrogen In utero exposure to chemotherapy: Effect on cardiac apy during human pregnancy. Lancet Oncol 5:283- sensitivity in female rats. Endocrinology 147:2575- and neurologic outcome. J Clin Oncol 24:e16-e17, 2006 291, 2004 2582, 2006 17. Avile´s A, Neri N, Nambo MJ: Long-term evalua- 21. Zemlickis D, Lishner M, Degendorfer P, et al: Fetal 24. Lord SJ, Bernstein L, Johnson KA, et al: tion of cardiac function in children who received anthra- outcome after in utero exposure to cancer chemothera- Breast cancer risk and hormone status in cyclines during pregnancy. Ann Oncol 17:286-288, 2006 py. Arch Intern Med 152:573-576, 1992 older women by parity, age of first birth, and breast- 18. Germann N, Goffinet F, Goldwasser F: Anthracy- 22. Avile´s A, Neri N: Hematological malignancies feeding: A case-control study. Cancer Epidemiol clines during pregnancy: Embryo-fetal outcome in 160 and pregnancy: A final report of 84 children who Biomarkers Prev 17:1723-1730, 2008 patients. Ann Oncol 15:146-150, 2004 received chemotherapy in utero. Clin Lymphoma 25. Woo SY, Fuller LM, Cundiff JH, et al: Radio- 19. Avile´s A, Neri N, Nambo MJ: Hematological ma- 2:173-177, 2001 therapy during pregnancy for clinical stages IA-IIA lignancies and pregnancy: Treat or no treat during first 23. Bridges RS, Byrnes EM: Reproductive experi- Hodgkin’s disease. Int J Radiat Oncol Biol Phys trimester. Int J Cancer 131:2678-2683, 2012 ence reduces circulating 17beta-estradiol and pro- 23:407-412, 1992

■■■

ASCOconnection.org—The Professional Networking Site for the Worldwide Oncology Community

● Expand your professional network and online profile ● Post your own content and news ● Respond to commentary by oncology leaders ● Create working groups on a secure site ● Connect and collaborate with colleagues ● Read exclusive articles and interviews ASCOconnection.org columnists include ASCO President Clifford Hudis, MD; ASCO President-Elect Peter Yu, MD; Past Presidents Sandra M. Swain, MD, Michael P. Link, MD, George W. Sledge Jr, MD, and Douglas W. Blayney, MD; Journal of Oncology Practice Editor-in-Chief John V. Cox, DO, MBA; Editor-in-Chief of the ASCO Educational Book, Don S. Dizon, MD; Editor-in-Chief of ASCO’s Cancer.Net Robert S. Miller, MD; and many other leaders in the oncology community. Log in with your ASCO.org member or guest account to participate.

www.jco.org © 2013 by American Society of Clinical Oncology 4139 Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved. Evens et al

Appendix

Table A1. Variables Examined in Univariable Analyses Non-Hodgkin lymphoma (n ϭ 50) Age: as continuous variable and also above and below median (ie, 29 years) No. of prior pregnancies: 0 v Ͼ 1 No. of prior deliveries: 0 v Ͼ 1 Type and No. of prior abortions: 0 v Ͼ 1 Presence of “B” symptoms (yes v no) Body mass index: Յ 25 v Ͼ 25 ECOG PS at diagnosis: continuous variable and 0-1 v 2-4 Hemoglobin at diagnosis (decreased v normal) LDH at diagnosis (increased v normal) Albumin at diagnosis (low v normal) Stage at diagnosis: continuous variable and I-II v III-IV Bone marrow involved (yes v no) Other extranodal sites (yes v no) More than one extranodal site (yes v no) Prognostic score: as continuous variable and 0-2 v 3-5 Diagnosis date: as continuous variable and in groups of 0-12 v 13-27 v 28-40 weeks Date started treatment: as continuous variable and in groups of 13-27 v 28-40 weeks Treatment during pregnancy (yes v no) Rituximab treatment during pregnancy (yes v no) Radiotherapy was part of the treatment plan during pregnancy (yes v no) Response to first treatment: as continuous (ie, CR, PR, SD, PD) and also CR/PR v SD/PD Use of myeloid growth factor (yes v no) Use of epoetin growth factor (yes v no) Preterm delivery at Ͻ 37 weeks (yes, n ϭ 1; no, n ϭ 2) Perinatal outcomes: route of delivery (cesarean, n ϭ 1; vaginal, n ϭ 2) Hodgkin lymphoma (n ϭ 40) Age: as continuous variable and above and below median (ie, 30 years) No. of prior pregnancies: 0 v Ͼ 1 No. of prior deliveries: 0 v Ͼ 1 Type and No. of prior abortions: 0 v Ͼ 1 Presence of B symptoms (yes v no) Body mass index: Յ 25 v Ͼ 25 ECOG PS at diagnosis: continuous variable and 0-1 v 2-4 Hemoglobin at diagnosis (decreased v normal) LDH at diagnosis (increased v normal) Albumin at diagnosis (low v normal) Stage at diagnosis: continuous variable and I-II v III-IV Bone marrow involved (yes v no) Other extranodal sites (yes v no) More than one extranodal site (yes v no) Prognostic score: as continuous variable and 0-2 v 3-7 Diagnosis date: as continuous variable and in groups of 0-12 v 13-27 v 28-40 weeks Date started treatment: as continuous variable and in groups of 13-27 v 28-40 weeks Treatment during pregnancy (yes v no) Radiotherapy was part of the treatment plan during pregnancy (yes v no) Response to first treatment: as continuous (ie, CR, PR, SD, PD) and CR/PR v SD/PD Use of myeloid growth factor (yes v no) Use of epoetin growth factor (yes v no) Preterm delivery (Ͻ 37 weeks) (yes, n ϭ 1; no, n ϭ 2) Perinatal outcomes: route of delivery (cesarean, n ϭ 1; vaginal, n ϭ 2)

Abbreviations: CR, complete response; ECOG PS, Eastern Cooperative Oncology Group performance status; LDH, lactate dehydrogenase; PD, progressive disease; PR, partial response; SD, stable disease.

© 2013 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org by FRANCISCO PEDROSA on December 10, 2013 from 177.19.142.26 Copyright © 2013 American Society of Clinical Oncology. All rights reserved.