Prognostic Factors and a New Prognostic Index Model for Children and Adolescents with Hodgkin’s Lymphoma Who Underwent Autologous Hematopoietic Stem Cell Transplantation: A Multicenter Study of the Turkish Pediatric Bone Marrow Transplantation Study Group

Objective: The prognostic factors and a new childhood prognostic index after autologous hematopoietic stem cell transplantation (AHSCT) in patients with relapsed/refractory Hodgkin’s lymphoma (HL) were evaluated. Materials and Methods: The prognostic factors of 61 patients who underwent AHSCT between January 1990 and December 2014 were evaluated. In addition, the Age-Adjusted International Prognostic Index and the Childhood International Prognostic Index (CIPI) were evaluated for their impact on prognosis. Results: The median age of the 61 patients was 14.8 years (minimum-maximum: 5-20 years) at the time of AHSCT. There were single relapses in 28 patients, ≥2 relapses in eight patients, and refractory disease in 25 patients. The chemosensitivity/chemorefractory ratio was 36/25. No pretransplant radiotherapy, no remission at the time of transplantation, posttransplant white blood cell count over 10x103/µL, posttransplant positron emission tomography positivity at day 100, and serum albumin of <2.5 g/dL at diagnosis were correlated with progression-free survival. No remission at the time of transplantation, bone marrow positivity at diagnosis, and relapse after AHSCT were significant parameters for overall survival. Conclusion: The major factors affecting the progression-free and overall survival were clearly demonstrated. A CIPI that uses a lactate dehydrogenase level of 500 IU/L worked well for estimating the prognosis. We recommend AHSCT at first complete remission for relapsed cases, and it should also be taken into consideration for patients with high prognostic scores at diagnosis.


Introduction
Children with Hodgkin's lymphoma (HL) have an excellent prognosis, and their survival rates are satisfactory. The survival rates for HL have increased from 81% to 95% for children and adolescents, even those with advanced-stage disease [1]. Approximately 15% of patients cannot be cured and experience relapse after first-line treatment [2]. The relapse rate is increased up to 30% in advanced stages, and most relapses of HL occur within the first 2 years after completing treatment [3,4]. Autologous hematopoietic stem cell transplantation (AHSCT) is recommended for patients with refractory disease during therapy or disease relapse within 1 year after completing therapy [5,6]. However, some patients who undergo AHSCT develop recurrence within 1 year [7].
There has been much research on the prognostic parameters of patients who get worse [8,9,10,11,12]. These prognostic parameters and scores include the International Prognostic Index (IPI) and the Age-Adjusted IPI (AAIPI), which have been proven to show a close relationship between relapse and a poor prognosis [13]. However, most of these parameters and scores are for newly diagnosed and adult patients. Thus, in this multicenter study, we evaluated the effect of various factors and prognostic indexes, at diagnosis and pre-and posttransplant, on the relapse rate and survival of children with HL who undergo AHSCT. In addition, we aimed to create and develop a new international prognostic index that is specific to children with newly diagnosed advanced Hodgkin's disease in order to evaluate the prognosis even after AHSCT.

Materials and Methods
This was a multicenter and retrospective study. Eleven pediatric stem cell transplant centers in different cities around Turkey participated, and the data were recorded retrospectively for 61 patients between January 1990 and December 2014. Three patients had undergone AHSCT before the year 2000, while 58 had undergone AHSCT after that point. Approval for the study was obtained from the local ethics committee. Exclusion criteria for the subjects included lymphomas other than HL and missing data. In four cases of an unknown pathologic subgroup of HL, the histologic slides were re-reviewed.

Participants' Characteristics
Sixty-one children from 11 centers, who underwent AHSCT for refractory or recurrent HL, were included and retrospectively analyzed. They showed some significant index values at diagnosis, pretransplant, and posttransplant for survival and event status after AHSCT.
Age, sex, concomitant diseases such as immune deficiency syndromes, and risk factors were analyzed. At diagnosis, risk factors were based on age, sex, primary tumor localization, number of relapses, time to relapse from diagnosis until completing the treatment, response to salvage chemotherapy, stage of disease, Karnofsky/Lansky score, pathologic HL type, bulky tumor, spleen involvement, extranodal involvement, bone or bone marrow involvement, B symptoms, hemoglobin, white blood cell (WBC) count, lymphocytes, monocytes, mean platelet volume (MPV), ferritin, albumin, lactate dehydrogenase (LDH), and sedimentation rate. Pretransplant risk factors were based on positron emission tomography-computed tomography (PET-CT) positivity, number of chemotherapy regimens, radiotherapy (RT), conditioning regimen, and disease status at the time of transplantation. After transplant, the risk factors were based on lymphocyte counts on posttransplant days 15 and 100, WBC counts, neutrophils, monocytes, platelets, MPV, neutrophil/ lymphocyte ratio (NLR), platelet/lymphocyte ratio (PLR), LDH, and positivity of PET-CT on posttransplant day 100.

Description of Features
Chemosensitivity: Susceptibility to the action of the chemotherapy protocol with a complete or very good response.
Staging: HL was graded using the Cotswold modification of the Ann Arbor staging system. Bulky disease is defined as the transplantation, posttransplant white blood cell count over 10x10 3 / µL, posttransplant positron emission tomography positivity at day 100, and serum albumin of <2.5 g/dL at diagnosis were correlated with progression-free survival. No remission at the time of transplantation, bone marrow positivity at diagnosis, and relapse after AHSCT were significant parameters for overall survival.

Conclusion:
The major factors affecting the progression-free and overall survival were clearly demonstrated. A CIPI that uses a lactate dehydrogenase level of 500 IU/L worked well for estimating the prognosis. We recommend AHSCT at first complete remission for relapsed cases, and it should also be taken into consideration for patients with high prognostic scores at diagnosis.
Time to relapse: Refractory disease was defined as occurring within 3 months after completion of therapy or during therapy. Early relapse was defined as disease recurring within 3-12 months, and late relapse was defined as disease occurring more than 12 months from the end of therapy [5].

Karnofsky/Lansky a Status
The Karnofsky/Lansky performance status is used to determine the functional status of the patient and is essential for all outcome-based analyses. The Karnofsky scale is designed for patients aged 16 years and older, and the Lansky scale is designed for those under 16 years. The Karnofsky/Lansky scores range from 100 to 0, with 100 indicating 'perfect' health and 0 representing death [15].

Childhood International Prognostic Index
The original AAIPI incorporates serum LDH levels, Eastern Cooperative Oncology Group (ECOG) performance status, and Ann Arbor clinical stage at diagnosis. Based on these factors, patients are divided into four risk groups: 0, low risk; 1, lowintermediate risk; 2, high-intermediate risk; and 3, high risk [13]. We adapted the AAIPI for children according to an LDH level of 500 IU/L instead of 250 IU/L because of its high prognostic predictively in childhood HL, and we used the Cotswold modification of the Ann Arbor clinical stage at diagnosis with an ECOG modification score according to the pretransplant Karnofsky/Lansky performance score [15] and the Childhood International Prognostic Index (CIPI).

Types of Outcome Measures
The definitions used as survival terms were as follows: 1) overall survival (OS) was calculated from the start of the treatment until death from any cause; 2) progression-free survival (PFS) was the achievement of stable disease without signs of progression, calculated from the day of transplant to the date of the next relapse or from the date of randomization for post-complete remission (CR) questions; and 3) event-free survival (EFS) was calculated from the date of the start of treatment to the date of the first event (failure to achieve CR, relapse, or death from any cause).

Statistical Analysis
Statistical analyses were performed using SPSS 15.0. Descriptive analyses were presented using medians or mean ± standard deviation for variables. The Kaplan-Meier method and logrank tests were used in the analysis. The risk factors described above were analyzed as prognostic factors for the survival rate with Cox regression analysis. Variables with values of p<0.05 were shown in the univariate analysis and were included in the multivariate analysis model. A p-value of less than 0.05 was considered to show a statistically significant result.
Forty-two (68.8%) patients underwent RT before AHSCT. At the time of the study, 44 of 61 patients were alive, 15 were dead, and two were lost to follow-up. The causes of death were progressive disease in nine patients, infection in two patients, and other reasons except for relapse in four patients, including posttransplant lymphoproliferative disease in one case, acute myeloid leukemia progression in another, anaplastic large cell lymphoma as a secondary cancer in another, and transplantrelated pulmonary hemorrhage in the last. The nonrelapse mortality rate was 40% (infection in two cases, other causes except for relapse in four cases, and lost to follow-up in two cases). The mortality rate 100 days after AHSCT was 6.5% (4 of 61 cases). The transplant-related mortality rate 100 days after AHSCT was 3.2% (2 of 61 cases; infection in one case and pulmonary hemorrhage in the other). The median relapse time was 11 months (minimum-maximum: 1-105). Types of relapse were refractory disease in 33 (54.1%) patients, early relapse in 12 (19.7%), and late relapse in 16 (26.2%). The PFS and OS rates for type of relapse were as follows: the 3-year PFS rates were 93.8% for late relapse, 91.7% for early relapse, and 38.7% for refractory disease (p<0.001); the OS rates were 100% for late relapse, 83.3% for early relapse (death due to other causes except for relapse and infection), and 57.6% for refractory disease (p=0.003). The 3-year survival rates were 57.6% for refractory disease and 88.9% for patients who responded (p=0.001). Threeyear PFS rates were 38.7% for refractory disease and 94.4% for patients who responded (p<0.001).

Features of Treatment
All of the patients underwent AHSCT. The disease status at the time of transplantation was as follows: 36 patients were in CR (CR2=28 patients, ≥CR3=8 patients) and 25 patients with refractory disease were not in remission. The 3-year PFS rates were 77.3% for patients with remission and 49.3% for patients without remission (p=0.007), while the OS rates were 91.4% for patients with remission and 59.1% for patients without remission (p=0.01).
The BEAM regimen was used in 44 (72.1%) patients, and other regimens were used in 17 (27.9%). Cyclophosphamide, etoposide, and BCNU in one case; busulphan and melphalan in five cases; busulphan, cyclophosphamide, and etoposide in four cases; fludarabine and busulphan in one case; total body irradiation, cyclophosphamide, and etoposide in three cases; CCNU, ifosfamide, and etoposide in two cases; and CCNU, cyclophosphamide, and etoposide in one case were used as other conditioning regimens. PET positivity and negativity were 37% and 63%, respectively, before AHSCT; however, these rates were 23% and 77% after AHSCT on day 100. Forty-two children (68.8%) received RT before AHSCT.

Treatment Results
The median (mean ± standard deviation, minimum-maximum) level of lymphocyte counts at day 15 after AHSCT was 0.45x10 3  With regard to the CIPI scores, the 3-year PFS rates were 100% for a score of 0, 61.3% for a score of 1, and 60.0% for a score of 2 (p=0.36), while the 3-year OS rates were 80.0% for a score of 0 (death due to infection), 93.3% for a score of 1, and 75.0% for a score of 2 (p=0.46) (Figure 1).

Discussion
Several prognostic factors affect survival in HL, such as extranodal disease at time of relapse, mediastinal mass at time of transplant, advanced stage at relapse, primary refractory disease, and a positive PET scan prior to AHSCT [2,16,17,18,19]. We found that no pretransplant RT, a posttransplant WBC count of >10x10 3 /µL, posttransplant PET positivity at day 100, serum albumin of <2.5 g/dL at diagnosis, no remission at the time of transplantation, bone marrow positivity at diagnosis, and relapse after AHSCT were significant parameters related to events after AHSCT and OS. In addition, CIPI was significant in estimating survival after AHSCT in HL.
RT is an important treatment method for HL. Although there have been several studies on discarding RT in the treatment of HL at earlier stages, its removal has been proven to increase the risk of relapse at later stages [20]. Some centers also tend to give RT after transplantation in order to not waste time before the transplant and to avoid recurrence of the disease. Our results clearly showed that there was a 13.2-fold increased risk of relapse in patients who underwent RT after transplant, but there was no impact on OS. Lowering the tumor volume with RT before AHSCT may improve the prognosis.
The NLR and PLR were reported to indicate disease severity and prognosis in patients with various diseases [21]. However, there is no relationship between pretransplant and posttransplant NLR, PLR, lymphocyte, monocyte, and MPV levels in patients who survive. There is also no relationship between pre-and posttransplant neutrophil, lymphocyte, NLR, PLR, and MPV levels and survival. However, on posttransplant day 100, a WBC count of over 10x10 3 /µL was found to increase the relapse risk 32.8-fold, but it had no impact on OS.
Several prognostic indexes have been used to evaluate the survival of patients with lymphoma. The IPI and the AAIPI are the two of the leading methods [13]. Low scores are correlated with a high survival rate and high scores with a low survival rate. Satwani et al. [22] reported that patients with relapsed/ refractory HL in the high-risk group according to prognostic models for PFS have a high progression risk after AHSCT. We used the CIPI and it worked well for estimating survival. The survival rates of our patients were 100%, 59.3%, and 58.3% for events after AHSCT and 80%, 91.7%, and 80% for OS for scores of 0, 1, and 2, respectively. According to our study, the risk of progression increased with the increase of the prognostic score, and this result was compatible with previous studies.
Some pediatric oncology centers may prefer to treat their patients with chemotherapy without AHSCT. However, the outcomes of patients who underwent salvage chemotherapy alone were not found to be satisfactory [23]. Studies about the proper timing of AHSCT are limited in pediatric populations. Stoneham  .0 g/L), male sex, rapidity of response to initial treatment with chemotherapy, fluorodeoxyglucose-PET avidity after two cycles, low serum albumin (<4 g/dL), and low lymphocyte count (<0.6x10 3 /µL or <8% of WBC count) were reported as prognostic factors in previous studies [8,9,10,11,12,30]. Relapsed patients with HL who had localized disease that was treated with chemotherapy alone and/or low-dose involved-field radiation therapy consolidation, and whose relapse occurred ≥12 months after completing therapy, have better survival with intensive conventional chemotherapy [8]. Extranodal disease at relapse, mediastinal mass at time of transplant, advanced stage at relapse, primary refractory disease, and a positive PET scan prior to AHSCT were significant factors in post-AHSCT events [2,16,17,18,19]. Claviez et al. reported that the most important predictors for disease control following AHSCT were time to relapse and disease status at transplantation [5]. We found by multivariate analysis that a WBC count of >10x10 3 /µL at 100 days after AHSCT, no RT before AHSCT, no remission after AHSCT, PET-CT positivity at 100 days after AHSCT, and serum albumin of <2.5 g/dL were significant factors for PFS, and that no remission after AHSCT, relapse after AHSCT, and bone marrow positivity at diagnosis were significant factors for OS.
In conclusion, the major factors affecting the prognosis of children with relapsed/refractory HL seem to be tumor load and chemosensitivity. Treatments that significantly decrease the tumor volume before AHSCT may improve the survival rate, as we saw a benefit with RT on EFS when performed before AHSCT. In addition, the survival rates of patients in remission before AHSCT were twice as high as those of patients without remission. AHSCT had a significant benefit on OS, but the timing must be investigated in larger studies. We also suggest different treatment approaches for patients with high IPI and/or CIPI scores to improve EFS and PFS. We suggest that a CIPI that uses an LDH level of 500 IU/L is more useful in childhood. A serum albumin status of <2.5 g/dL at diagnosis had a significant effect on PFS, pointing to the study of the immunologic profile of the patients, and the treatment schedule may be redesigned with this immunologic profile. In addition, the characteristics that showed significance in a univariate but not a multivariate analysis appear to have an influence as well, and might show a stronger correlation in larger trials. Patients with high prognostic factors should be evaluated at diagnosis and may be directed to AHSCT consolidation therapy at the time of the first CR.