Abstract
Background/Aim: Pediatric patients with primary refractory or relapsed B-cell non-Hodgkin lymphoma (B-NHL) have highly unfavorable prognosis. In this study, we retrospectively analyzed outcomes in pediatric B-NHL patients treated in a single center in Poland from 1995 to 2022, with emphasis on therapy results in patients with progression or relapse. Patients and Methods: The primary objectives were a 5-year probability of overall survival (pOS) and a 5-year probability of event-free survival (pEFS). The secondary objectives involved the assessment of prognostic factors. Results: A total of 76 children were eligible for the analysis. The 5-year pOS was 76.7%, and the 5-year pEFS was 72.9%. At diagnosis, elevated lactate dehydrogenase activity, the presence of B symptoms, bone marrow, skeletal or mediastinal involvement, and stage IV disease were associated with inferior outcomes. Nine children experienced progression and four relapse. The 5-year pOS for patients with progression was 38.1%. Two patients treated with hematopoietic stem cell transplantation (HSCT) as part of salvage therapy survived. However, only one out of seven patients who were treated without HSCT survived. The 5-year pOS was 0.0% in patients with relapsed disease. Conclusion: The most significant factor related to outcomes in pediatric B-NHL is therapy response, with a high mortality rate in children with refractory disease and relapse. There is no consensus on the salvage therapy approach; however, HSCT appears to be the optimal choice.
Non-Hodgkin lymphomas in children represent a heterogeneous group of neoplasms that includes over 30 histological subtypes, mostly of B-cell origin (1-3). The B-cell non-Hodgkin lymphomas (B-NHL) diagnosed in children predominantly comprise high-grade tumors, such as Burkitt lymphoma (BL), B-cell lymphoblastic lymphoma (BLL), or diffuse large B-cell lymphoma (DLBCL) (2). In contrast to adults, pediatric B-NHL cases are relatively rare, requiring multicenter collaborations to assess therapy results (1, 4). Analysis of data from large international studies has facilitated the identification of risk factors associated with clinical symptoms, laboratory and imaging test results, and response to therapy (5). Although multicenter studies, including clinical trials, provide valuable information about prognostic factors and therapy results, real-world data validate the feasibility of incorporating them into everyday clinical practice.
In general, outcomes in the pediatric population are better than those in adults, but patients with primary refractory or relapsed disease have a highly unfavorable prognosis (2, 3, 6-8). Moreover, pediatric patients presenting with refractory and relapsed lymphoma represent only 10-15% of children diagnosed with B-NHL. Therefore, recruiting sufficient patients for clinical trials is challenging, and published therapy results are usually based on outcomes from small groups of patients (7, 9-11). While numerous strategies are being investigated as second-line treatment, established standards for salvage therapy are lacking, which emphasizes the need to provide information about therapy results in this subgroup (6-8, 11).
In this study, we retrospectively analyzed outcomes in pediatric B-NHL patients treated at the Department of Pediatric Hematology and Oncology in Poland from 1995 to 2022, with particular emphasis on prognostic factors and therapy results in patients presenting with progression during first-line treatment or relapsed disease.
Patients and Methods
Design of the study. We report the results of a retrospective, single-center analysis of prognostic factors and therapy results in pediatric patients with B-cell non-Hodgkin lymphoma treated in the period from November 1995 to December 2022. The study was approved by the Ethics Committee of Collegium Medicum, Nicolaus Copernicus University in Bydgoszcz.
Patients. The medical records of patients treated at the Department of Pediatric Hematology and Oncology in Bydgoszcz were reviewed. Patients <19 years with newly diagnosed B-cell non-Hodgkin lymphoma were eligible for the study. Children with incomplete data, those treated with experimental or unstandardized therapy, and patients with uncommon histological diagnosis were excluded.
Diagnosis and stratification. Data obtained from medical interviews were analyzed with particular focus on the incidence of B symptoms (unexplained fever over 38°C, unintentional weight loss, night sweats) and the medical history of immunodeficiency or cancer predisposition syndromes.
The laboratory test results included in the analysis were a complete blood cell (CBC) count, bone marrow smears, cerebrospinal fluid (CSF) analysis and dehydrogenase lactate (LDH) activity. LDH was considered elevated when >500 U/I. The imaging test results included to the analysis were chest X-rays, lymphatic nodes and abdominal ultrasound, and a computed tomography (CT) scans of the neck, thoracic chest, abdomen, and pelvis. Patients with neurological symptoms underwent additional magnetic resonance imaging (MRI) scans of the central nervous system (CNS).
Histological diagnosis was obtained from a tissue sample. The following lymphoma subtypes were analyzed: Burkitt lymphoma, B-cell lymphoblastic lymphoma, diffuse large B-cell lymphoma with detailed primary mediastinal large B-cell lymphoma (PMLBCL), and an unclassified B-cell non-Hodgkin lymphoma (BNHLU).
Bone marrow disease was defined as ≥5% and <25% of lymphoma cells in myelogram results confirmed by immunophenotyping. When the number of blasts exceeds 25%, mature B-cell acute lymphoblastic leukemia (ALL) was diagnosed. The CNS was defined as involved in the presence of >5 lymphoma cells/μl in the CSF based on cytospin results, intracranial mass confirmed by MRI, or cranial palsy that could not be explained otherwise. Mediastinal involvement was considered in the presence of a tumor mass on X-ray or CT scans. Skeletal involvement was diagnosed in cases where bone lesions were observed on X-ray, provided that the diagnosis was confirmed through histological examination. If the bone lesion was the only manifestation of a suspected NHL, biopsy was performed.
Therapeutic protocols. Patients were treated according to the following therapeutic protocols: LMB 89, NHL-BFM 90, NHL-BFM 95, EURO-LB 02, LMB 2001, B-NHL BFM 04, and Inter-B-NHL-Ritux 2010 (12-16).
Response criteria. Complete remission (CR) was defined as the disappearance of all measurable or evaluable lesions and no lymphoma cells in the BM and CSF. In the presence of a residual mass, patients were considered in CR when large biopsies showed no viable lymphoma cells.
Definitions. Progression was defined as any enlargement of primary lesions, relapse or new, histologically proven lesions during the treatment or up to three months after completion of the therapy. Relapse was defined as the appearance or reappearance of histologically proven lesions three or more months after completion of the therapy. Event was defined as death due to any cause, progression, relapse, or secondary malignancy. Overall survival (OS) was calculated as the time in years from the first day of treatment to death of any cause or last follow-up contact for alive patients. Event-free survival (EFS) was calculated as the time in years from the first day of treatment to the event.
Statistical analysis and language corrections. The primary endpoints of the study were a 5-year probability of overall survival (pOS) and a 5-year probability of event-free survival (pEFS). The pOS and the pEFS were estimated using the Kaplan-Meier method and compared using the log-rank test. The secondary endpoints involved the assessment of prognostic factors in pediatric patients with B-NHL. The evaluation of clinical, laboratory, and imaging findings was conducted through univariate and multivariate analyses. The results were considered statistically significant when p<0.05. Statistical calculations were performed using MedCalc 20.100 (MedCalc Software, Mariakerke, Belgium). Language corrections were made using TRINKA grammar checker on the www.trinka.ai website.
Results
Demographics. A total of 97 patients were diagnosed with B-NHL between November 1994 and December 2022. Eighteen patients were excluded due to missed or insufficient data, two because of uncommon histological diagnoses (one case of mucosa-associated lymphoid tissue lymphoma and one patient with lymphomatoid granulomatosis of CNS due to EBV infection in patient with immunodeficiency), and one because of unstandardized treatment. A total of 76 children were eligible for the analysis.
Patient characteristics and outcomes are shown in Table I. The study population consisted of 51 boys (male-to-female ratio 2.1:1), aged 1.6 to 18.8 years (median 11.9 years) at the time of diagnosis. The most common histological diagnosis was BL, and the majority of children had advanced stage lymphoma. One patient was diagnosed with immunodeficiency before B-NHL diagnosis, and none were diagnosed with a cancer predisposition syndrome. B-cell non-Hodgkin lymphoma was a secondary malignancy in two patients, both after ALL treatment.
Association of patient characteristics with outcome [probability of overall survival (pOS) and 5-year probability of event-free survival (pEFS)].
Outcomes. The 5-year pOS for the entire cohort was 76.7%, and the 5-year pEFS was 72.9%. Complete remission during the initial treatment was achieved in 67 patients (88.2%). Nine children experienced progression, and the median time from diagnosis to progression was 5.2 months (range=2.3-8.0 months).
All children with progression received second-line treatment based on chemotherapy. In addition to chemotherapy, three children received anti-CD20 monoclonal antibodies (rituximab), one patient received Rituximab followed by autologous hematopoietic stem cell transplantation (auto-HSCT), and another patient received chemotherapy followed by auto-HSCT. One patient with progressed disease, treated with chemotherapy combined with rituximab in first-line treatment in 2019, was eligible for the Individualized Therapy For Relapsed Malignancies in Childhood (INFORM) program (17). He received personalized treatment including ibrutinib according to INFORM feedback, which was administered along with rutiximab, vincristine, idarubicin, carboplatin and ifosfamide (R-VICI). However, the targeted therapy had to be stopped because of the side effect of the drug (severe bleeding from gastrointestinal track) and he subsequently died due to disease progression. The 5-year pOS for patients with progression was 38.1%. However, both of the two patients who were eligible for HSCT in salvage therapy have survived, in contrast to only one out of seven patients who did not undergo HSCT in second-line therapy.
Four patients relapsed (5.3%) with a median time to relapse after diagnosis of 5.5 months (range=3.5-12.1 months). Relapse was manifested in the primary involved sites in three children (BM, CNS, and bones) whereas one patient with primarily affected lymph nodes, mediastinum, and adrenal glands experienced a relapse in the CNS. Among the patients who relapsed, all received rescue chemotherapy. In addition to chemotherapy, one patient was treated with rituximab, and one received a treatment involving chemotherapy and subsequent auto-HSCT. None of the patients who experienced relapse survived.
Overall, 16 children died (21.1%), mostly due to therapy-related complications (n=11; 68.7%), such as infections (n=7) and treatment-induced toxicity (n=4). Four patients died of disease progression, and one died of secondary malignancy (T-cell ALL in a girl initially treated for B-lymphoblastic lymphoma according to the LMB 89 protocol).
Prognostic factors. The incidence of B-symptoms was associated with a significantly lower 5-year pOS. Among B symptoms, the most crucial risk factor for death was unexplained fever (OR=6.8; 95%CI=1.7-27.1; p=0.007). Concerning the LDH level, the 5-year pOS was significantly lower among patients with an elevated LDH activity, who also demonstrated a tendency toward lower 5-year pEFS. Primary lymphoma localization had no significant impact on the 5-year pOS. However, we observed a significantly lower 5-year pEFS among patients with primary mediastinal disease or involvement of sites other than those commonly observed in pediatric BHNL cases (Table I).
According to histological subtypes, children diagnosed with BLL had a six-fold increased risk of death (OR=6.4; 95%CI=1.8-23.1; p=0.005) and an increased risk of event (OR=4.1; 95%CI=1.2-13.9; p=0.024). Stratifying by disease stage, stage IV patients had a significantly higher risk of mortality (OR=4.7; 95%CI=1.4-16.4; p=0.014) and event (OR=3.8; 95%CI=1.2-12.1; p=0.023) compared to other patients. Although there were no statistically significant differences in the 5-year pOS between particular therapeutic protocols, we observed superior outcomes for children treated according to the Inter-B-NHL-Ritux 2010 protocol, particularly in stage III disease with elevated LDH activity and stage IV (5-year pOS 77.8% vs. 58.6%), but the results were not statistically significant (p=0.480). The 5-year pOS and 5-year pEFS in respective histological subtypes and disease stages are shown in Figure 1.
The 5-year probability of overall survival and 5-year probability of event-free survival for consecutive disease stages and histological subtypes. A) The 5-year probability of overall survival for disease stages. B) The 5-year probability of event-free survival for disease stages. C) The 5-year probability of overall survival for respective histological subtypes. D) The 5-year probability of event-free survival for respective histological subtypes.
The most important prognostic factor was the response to therapy, with a significantly higher mortality in children experiencing progression (OR=6.0; 95%CI=1.2-31.1; p=0.032; 5-year pOS 79.9% vs. 38,1%; p=0.023). Relapse was associated with the highest risk of death (OR=45.8; 95%CI=2.3-909.9; p=0.012; 5-year pOS 82.7% vs. 0.0%; p<0.001) in the analyzed cohort.
None of the factors that increased the risk of the event remained statistically significant in the multivariate analysis (Table II), while the incidence of progression during first-line treatment (p=0.037) and relapse (p=0.001) were the only independent risk factors of death (Table III).
The multivariate analysis of prognostic factors for event.
The multivariate analysis of prognostic factors for mortality.
Discussion
In this study, we evaluated the outcomes of pediatric patients with B-NHL with a particular focus on prognostic factors and therapy response. The majority of risk factors shown in the analysis have already been incorporated into therapeutic protocols, whereas others have not reached significance in larger studies as part of the stratification system (5, 12-16, 18). However, only the incidence of progression or relapse remains an independent risk factor in multivariable analysis, highlighting the importance of response to therapy in patient outcomes.
At diagnosis, elevated LDH activity, the presence of B symptoms, BM, skeletal or mediastinal involvement, and stage IV disease were associated with inferior outcomes. Stage of the disease is a well-known prognostic factor in pediatric B-NHL (19). Although therapy outcomes in stage III lymphoma within the analyzed cohort were not statistically different from those in stages I and II, stage IV disease was associated with over a fourfold higher mortality. The consistently poor therapy outcomes observed in this subgroup in multicenter studies resulted in the introduction of anti-CD20 monoclonal antibodies (rituximab) for treating patients from risk groups (1, 16, 20, 21). This therapeutic approach led to an improvement in the 3-year EFS (93.9% in the chemotherapy and rituximab group versus 82.3% in the chemotherapy group) and 3-year OS (93.9% versus 82.3%), as demonstrated in a phase III clinical trial (16). In the analyzed cohort, we also noted improved pEFS and pOS among patients from high-risk groups who received treatment according to the Inter-B-NHL-Ritux 2010 protocol. Nevertheless, the overall outcomes were inferior compared with those observed in the clinical trials, primarily due to high mortality among patients experiencing disease progression or relapse.
Response to therapy remains the most crucial prognostic factor in pediatric B-NHL. Of the four patients who experienced relapse in the analyzed cohort, all died, whereas in the subgroup with disease progression, the 5-year pOS was lower than 40%. Despite significant progress in the therapy of pediatric B-NHL, outcomes in children with refractory or relapsed disease remain unfavorable, with overall survival rates ranging between 15% and 40% (6-8, 11). Moreover, the small number of patients experiencing relapse has limited studies on salvage therapy (6-8). The largest analysis to date comes from the network of the International Berlin-Frankfurt Muenster Group and the European Inter-Group for Childhood and Adolescent with Non-Hodgkin Lymphoma, where the authors analyzed data from 639 children with relapsed NHL (8). The study highlights the role of HSCT in salvage therapy, indicating that OS for patients who underwent HSCT reached 50%, whereas OS was below 10% in the group treated without HSCT (8). Similar observations were made in our study, where two out of three patients (two with progression and one with relapse) treated with HSCT in second-line therapy survived, compared with only 10% of patients treated without HSCT.
Currently, international studies are focusing on new therapies with anti-neoplastic mechanisms other than conventional chemotherapy (7, 22-24). In our analysis, one patient with BL was eligible for the INFORM program, which aims to identify molecularly actionable variants in malignant cells for personalized treatment (17). This patient received a personalized recommendation for potentially effective therapy based on the molecular profiling of lymphoma cells after disease progression. One of the potentially effective drugs in this case was ibrutinib, which was administered with the R-VICI regimen. The treatment was based on promising results from the SPARKLE trial, where patients treated with ibrutinib added to R-VICI or R-ICE (rituximab, ifosfamide, carboplatin, etoposide) regimens had an overall response rate of 57.1% in early results from the run-of phase of the clinical trial (25). However, in this case, the therapy had to be discontinued, and the patient subsequently died due to disease progression.
Since 2022, our department has been one of two centers in Poland that are running therapeutic programs using CD19 chimeric antigen T-cells (CAR-T) therapy for pediatric B-cell ALL and B-NHL, but as of now, none of the patients with B-NHL have met the qualifications for this therapy. CD19 CAR-T have demonstrated the ability to induce durable responses in adult B-NHL patients (26, 27). However, questions persist regarding the optimal CD target, and the comparative benefits of CART therapy versus other second-line therapy options, and these aspects require further investigation (22, 28-30).
Because of the retrospective nature of the analysis, several limitations are evident. Some patients were excluded from the analysis due to missing medical data or non-standardized treatment. The analyzed group exhibited heterogeneity, with an unequal distribution of children treated according to different therapeutic protocols. However, despite differences in therapeutic approaches among treatment periods and histological subtypes, we identified significant features associated with inferior outcomes. This analysis not only presents real-world data on therapy outcomes across various therapeutic protocols, but also highlights the issue of nonstandardized salvage therapy approaches. Our findings are consistent with larger analyses and underline the significance of treatment response in determining outcomes.
Conclusion
Early identification of risk factors, preferably at diagnosis, and relevant therapy stratification are the core of current therapy in pediatric B-NHL. The most significant factors are related to therapy response, with a high mortality rate in children with refractory disease and relapse. Due to the limited number of pediatric cases and the variety of second-line therapeutic approaches, conducting randomized studies in this subgroup is challenging. HSCT appears to be the optimal choice as salvage therapy; however, further studies are necessary to identify the best option for bridge therapy, as well as alternative strategies for patients who are not eligible for HSCT.
Acknowledgements
The Authors would like to thank Andrzej Kołtan, Sylwia Kołtan, Rober Dębski, Monika Pogorzała, Krzysztof Czyżewski, Monika Richert-Przygońska, Joanna Cisek, Ewa Demidowicz, Natalia Bartoszewicz, Agnieszka Jatczak-Gaca, Agata Marjańska, Barbara Tejza, Piotr Księżniakiewicz, Anna Dąbrowska, Elżbieta Grześk, Monika Łęcka, Oliwia Grochowska, Marta Mazalon, Eugenia Winogrodzka, Kamila Jaremek, Katarzyna Balci, Karolina Hejnosz and Jagoda Sadlok, all nurses and other team members from Department of Pediatric Hematology and Oncology in Bydgoszcz for the continuous excellent care of children with malignancy.
Footnotes
Authors’ Contributions
Joanna Stankiewicz, Jan Styczyński and Ewa Demidowicz contributed to the study conception and design. Material preparation, data collection and analysis were performed by Joanna Stankiewicz, Anna Jabłońska and Paweł Traichel. The first draft of the manuscript was written by Joanna Stankiewicz and all Authors commented on previous versions of the manuscript. All Authors read and approved the final manuscript.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflicts of Interest
The Authors have no competing interests to declare that are relevant to the content of this article.
- Received July 25, 2024.
- Revision received August 11, 2024.
- Accepted August 17, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
