Effectiveness and tolerability of radiotherapy for patients with indolent non-Hodgkin’s lymphoma: a monocenter analysis

To analyze the effectiveness and toxicities of radiotherapy in indolent non-Hodgkin’s lymphoma (iNHL) patients treated in our institution. Patients with iNHL treated with radiotherapy between 1999 and 2016 were included. The primary endpoint was progression-free survival (PFS). Secondary endpoints were local control (LC), overall survival (OS) and toxicities. PFS, LC, and OS were analyzed using Kaplan–Meier method. Log-rank test was used to investigate the differences between subgroups. Cox proportional hazard model was used for univariate continuous analysis. Seventy-five patients were identified in our institutional database between 1999 and 2016. Fifty-eight (77.3%) had stage I after Ann-Arbor and 17 patients (22.7%) had stage II. The median follow-up was 87 months (95% CI 72–102 months). Median single dose per fraction was 2.0 Gy (range 1.5–2 Gy) and median total dose was 30.6 Gy (range 16–45 Gy). Radiotherapy was performed in 2D (n = 10; 13.3%), 3D (n = 63; 84.0%) and VMAT (n = 2; 2.7%) techniques, respectively. The median PFS was 14.0 years (95% CI 8.3–19.7 years). The estimated PFS after 5 and 10 years were 73.0% and 65.5% in Kaplan–Meier analysis, respectively. The 5- and 10-year LC were 94.9% and 92.3%, respectively. The 5- and 10-year OS were 88.6% and 73.9%. In univariate analyses of PFS, younger patients (≤ 60 years old) had significantly superior PFS to those older than 60 years old (5-year PFS 81.9% vs. 65.1%, p = 0.021). Dose escalation > 36.0 Gy had no prognostic influence in term of PFS (p = 0.425). Extranodal involvement, stage and histology had no prognostic impact on PFS. Depending on the site of lymphomas, the most common acute side effects were: dermatitis CTCAE° I–II (8.0%), xerostomia CTC° I (8.0%), cataract CTC° I (12.0%) and dry eyes CTC° I–II (14.6%). No adverse event CTC° III was reported. Most acute side effects recovered at 3 to 6 months after radiotherapy except for CTC° I cataract and xerostomia. Local Radiotherapy was highly effective for treatment of early stage iNHL with no serious side effects in our cohort. The most acute CTCAE° I–II side effects recovered 3 to 6 months later. Technique advances seem to have further improved effectiveness and tolerability of radiotherapy. Trial registration: Local ethics committee of Ludwig-Maximilian-University (LMU) Munich approved this retrospective analysis on the May 7th, 2019 (Nr. 19–137).

Statistical analysis. Patient demographics were calculated using descriptive statistics as absolute and relative frequencies. The primary endpoint of this study was progression-free survival (PFS). PFS was a time-to-event endpoint and defined as the interval between the beginning of radiotherapy to the earliest date of progressive disease, relapse, or death resulting from any cause 9 . PFS, LC, and OS were analyzed using Kaplan-Meier method. Log-rank test was performed to investigate the differences between subgroups. Chi-square and Cramer's V were utilized to analyze association of nominal parameters. Cox proportional hazard model was used for univariate continuous analysis. A two tailed p-value of < 0.05 was considered significant. We performed statistical analyses with IBM SPSS Statistics, Version 25 (IBM, Armonk, New York, USA). All methods were carried out in accordance with relevant guidelines and regulations.
Ethics approval. Local ethics committee approved this retrospective analysis on 7th of May 2019 (Nr. .

Results
Patient characteristics. In the initial database screening, 574 lymphoma patients treated with radiotherapy in our department between 1999 and 2016 were identified. After exclusion of aggressive lymphoma, patients with stage III and IV, and follow up less than 3 months, 75 patients with stage I or II iNHL remained for retrospective analysis. Median follow up was 87 months (95% CI 72-102 months). A CONSORT (Consolidated Standards of Reporting Trials) diagram of our cohort is presented in Suppl. Fig. 1.
In order to rule out systemic involvement, PET/CT were performed in 19 patients (25.3%) and 56 patients (74.7%) underwent contrast enhanced whole body computer tomography (CT). We observed an increasing number of PET/CT utilization over the years. Only 5 from 36 patients (14%), who were treated before 2010, underwent PET/CT at the diagnosis. In comparison, PET/CTs were performed in 14 from 39 patients (36%), who were treated in 2010 and afterwards.
We applied Follicular Lymphoma International Prognostic Index (FLIPI) for patients with FL and observed 34 patients (75.6%) with low-risk, 6 patients (13.3%) with intermediate-risk, and 5 patients (11.1%) with unknown FLIPI score due to unknown LDH. We used Marginal Zone Lymphoma of Mucosa-associated Lymphoid Tissue International Prognostic Index (MALT-IPI) for MZL, and it resulted in 15 patients (50.0%) with low-risk, 11 patients (36.7%) with intermediate-risk and 4 patients (13.3%) with unknown risk (due to unknown LDH).
Patients' characteristics are summarized in Table 1.
In 9 patients who received RT in recurrent situation, 6 patients (66.7%) were treated with systemic therapy prior to RT and 3 patients (33.3%) underwent resection prior to RT.
Radiotherapy was performed with a median single dose of 2 Gy (range 1.5-2 Gy) and a median total dose of 30.6 Gy (range 16-45 Gy). Target volume delineation was based on involved-field radiation therapy (IFRT) in 43 patients (57.3%) and involved-site radiation therapy (ISRT) in 32 patients (42.7%). Radiotherapy planning was simulator-based (2D-RT) in 10 cases (13.3%), three dimensional (3D-RT) in 63 cases (84.0%), and volumetric As for radiation volume, 40 patients (53.3%) underwent radiation in extranodal regions, 18 patients (24.0%) in inguinal or femoral lymph node regions, and 10 patients (13.3%) in cervical and supraclavicular lymph node regions. Radiation of other lymph node regions, such as Waldeyer's ring, axillary, paraaortic, mesenteric and iliac lymph node region was applied in 7 patients (9.3%). All patients with stage II disease received RT in one target volume. The summary of treatment parameters are described in Table 2.
Progression-free survival. The Kaplan-Meier estimates of 5-and 10-years PFS were 73.0% and 65.5%, respectively. The median PFS was 14 years (95% CI 8.3-19.7 years, Fig. 1a). In 49 patients, who underwent RT alone, the 5-and 10-years PFS were 81.0% and 67.4%. Different lymphoma subtypes achieved a comparable long-term outcome (5-year PFS for FL 68.8% vs. MZL 79.4%, p = 0.427, Fig. 2a). Dose escalation > 36.0 Gy had no prognostic influence to of PFS than ≤ 36.0 Gy (5-year PFS 65.5% vs. 72.1%, p = 0.425, Fig. 2b). There was no significant difference between patients with nodal and extranodal iNHL in term of PFS (5-year PFS 69.6% vs. 76.0% for nodal and extranodal iNHL, p = 0.541, Fig. 2c). Younger patients (≤ 60 years old) had significantly  Fig. 2d). ISRT was not inferior to IFRT (p = 0.543). Univariate analysis of PFS was summarized in Table 3. We performed subgroup analyses with patients who received RT at the frontline, and precluded four patients, who received systemic therapy prior to RT (rituximab or R-CHOP), or underwent resection or systemic therapy prior to RT. In this subgroup of 62 patients, 5-and 10-years PFS were 76.1% and 64.5%, respectively. The subgroup analyses also demonstrated, that patients who received dose escalation of > 36 Gy had no better PFS than ≤ 36 Gy (p = 0.466). In the further univariate analyses of PFS, we did not find significant prognostic role of histology (FL vs. MZL), age (> 60 years old vs. younger) and nodal vs. extranodal. ISRT was not inferior to IFRT. In 9 patients, who received RT in the recurrent setting, 5-and 10-years PFS were 55.6% and 0% respectively.
Local control and overall survival. Local progress after radiotherapy was reported in 6 patients (8.0%).
The 5-and 10-year LC was 94.4% and 92.3%, respectively, the median was not reached (Fig. 1b). Bases on his-  Sixteen patients (21.3%) were deceased at the time of last follow up. The 5-and 10-year OS were 88.6% and 73.9% respectively, with a median of 19 years (Fig. 1c). Based on different histologies, the 5-and 10-years OS were 90.2% and 77.2% for FL, versus 86.3% and 70.3% for MZL (p = 0.156), respectively. The 5-and 10-years OS in 62 patients, who received RT at the frontline, were 91.6% and 74.5% respectively. In patients, who underwent RT at the recurrent, the 5-and 10-years OS were 63.5% and 0% respectively.

Toxicity.
Depending on the site of lymphoma, the most common acute side effects were: dermatitis CTCAE° I-II (n = 6; 8.0%), xerostomia CTC° I (n = 6; 8.0%), cataract CTC° I (n = 9; 12.0%), and dry eyes CTC° I-II (n = 11; 14.6%). No adverse event CTC° III was reported. Most of acute side effects recovered at 3 to 6 months after radiotherapy except for CTC° I cataract, dermatitis, and xerostomia. The summary of toxicities after radiotherapy is described in Table 4.
Using Chi-square and Cramer's V statistical methods, we analyzed the correlation between toxicities, radiation dose and extranodal involvement. We did not find any significant difference regarding toxicity between > 36.0 and ≤ 36.0 Gy (p = 0.197). There were significant more side effects for extranodal involvements (CTCAE° I-II: 57.5% in extranodal cohort vs 28.6%, p = 0.012).

Discussion
In the present study, we report that radiotherapy provided excellent LC, PFS and OS for iNHL, with no difference between FL and MZL, as the two major subgroups. These results were in accordance with those from several preceding studies [11][12][13][14][15][16][17][18][19][20] . However, PFS decreased from 73.0 at 5 years to 65.5% at 10 years. Similar results were also reported by some previously published studies [11][12][13][14][15][16][17][18] . The discrepancy between LC and PFS was mainly caused by distant progression outside of radiation volume, which was observed in 15 patients (20.0%) in our cohort with a median time of 87 months after RT.
The issue of distant relapses after radiotherapy has raised the question whether adding systemic therapy might help to improve PFS. A German phase II MIR trial combined anti-CD20 antibody rituximab with involved-field radiotherapy (IFRT, 30-40 Gy) 21 . IFRT combined with rituximab was well tolerated and 5-years PFS was 78%, www.nature.com/scientificreports/ which was slightly higher than our results 22 . The potential improvement of PFS by adding rituximab to IFRT (30-40 Gy) was also confirmed by Ruella et al. 23 . However, this observational multicenter study showed no OS difference between RT alone and RT combined with rituximab. A more recent randomized trial, comparing IFRT alone with IFRT plus immunochemotherapy (rituximab, cyclophosphamide, vincristine, and prednisolone), reported significantly better 10-years PFS in the arm of combined therapy (59% vs 41%), albeit at cost of grade III or IV acute toxicities (65% in the combined arm) 24 . In contrast, we did not observe any grade III or higher toxicities in the present study, in which the majority was treated with RT alone. The advance in diagnostic imaging enabled better localization of involved lymph nodes and radiation volumes. Supported by the development of more sophisticated RT technique (e.g. 3D, IMRT, VMAT, instead of 2D) it was possible to further reduce RT volumes from IFRT to ISRT. In our cohort, target volume delineation were Table 3. Univariate analysis of PFS.  www.nature.com/scientificreports/ performed according to IFRT (57.3%) and ISRT (42.7%) strategies respectively. In our univariate analysis, ISRT was not inferior to IFRT in term of PFS. This stands in accordance with a retrospective analysis of more than 200 patients, which showed the non-inferiority of ISRT and the most common pattern of failure in IF-and ISRT groups was distant recurrence 25 . In the current study, extranodal involvement was not associated with inferior PFS, this stands in line with a recent analysis from König et al. 26 . Our analysis showed that radiation dose higher than 36.0 Gy did not result in any benefit of clinical outcomes. This finding was in line with a randomized phase III trial comparing 40-45 Gy with 24 Gy, which showed no difference in overall response rate, LC, PFS, or OS 27 . Another low dose radiation therapy (LDRT) study investigated further dose de-escalation and compared 24 Gy with 4 Gy in patients with indolent lymphoma 28 . However, the group of 4 Gy was inferior to 24 Gy in term of PFS. Thus, the authors concluded that 24 Gy should be considered as standard dose for definitive radiotherapy of iNHL and LDRT of 4 Gy remained an useful alternative, especially for palliative care 28,29 . However, in the era with innovation of targeted therapy and more sensitive functional PET imaging, further reduction of radiation dose is still a striving issue with the purpose of saving radiation-induced side effects. A prospective study (GAZAI) is now ongoing to examine the effectiveness of combined low-dose radiotherapy with CD20-antibody Obinutuzumab for stage I/II follicular lymphoma. For non-responder, evaluated in FDG-PET, a second radiotherapy with 36 Gy was applied for salvage treatment. The results are eagerly awaited which may help further optimizing treatment for early stages iNHL.
Because of its indolent nature, "watch and wait" strategy has also been performed in patients with early stage iNHL (Ann Arbor I and II). A retrospective analysis of 41 selected patients showed estimated survival at 5 and 10 years of 97% and 85% after deferred therapy 30 . Therapy was not initiated mostly because of physician choice, large radiation field, advanced age, and concern about toxicity 30 . Another large retrospective cohort study of National Cancer Data Base with almost 36,000 patients reported a decline in the use of RT in patients with early stage follicular lymphoma from 37 in 1999 to 24% in 2012. In contrast to the Stanford series, this study showed an improved OS after RT 8 .
Regarding toxicity, we could show in the present study that RT of indolent lymphoma was well tolerated with no serious adverse events (≥ CTCAE° III). Extranodal involvement was associated with significant increased acute toxicities; this was attributable to the fact that the majority of extranodal involvements in our cohort were orbita lymphoma and conjunctivas, in which conjunctiva as a relative radiation sensible organ nearby oftenreceived relevant radiation dose and developed conjunctivitis. The most recent multicenter retrospective study from Brady et al. analyzed definitive radiotherapy for localized follicular lymphoma and presented similar results as the present study 31 . There were no significant adverse effects after the relative low doses as well as limited radiation fields 31 .
With its excellent effectiveness and low toxicities, our results support the pivotal role of RT in curative treatment of early stage iNHL, being in line with the recommendation in international guidelines and the results of other studies 5,6,12,32 . Despite the long follow-up of our study allowing reliable analysis of survival rates, major limitations of our study remain its retrospective nature and the limited number of patients.

Conclusion
Local Radiotherapy was highly effective for treatment of early stage iNHL with no serious side effects in our cohort. The most acute CTCAE° I-II side effects recovered 3 to 6 months later. Technique advances seem to have further improved effectiveness and tolerability of radiotherapy.

Data availability
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. 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 Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.