We present data on 10 children (<18 years) with R/R cHL treated at the Princess Máxima Center for Pediatric Oncology, Utrecht, The Netherlands. The primary manifestation of cHL was initially treated according to the Euronet PHL-interim or -C2 treatment protocols: 2 OEPA cycles (vincristine, etoposide, prednisolone and doxorubicine) followed by 1 to 4 COPDAC-28 cycles (28 days; cyclophosphamide, vincristine, prednisolone and dacarbazine) or 2 to 4 DECOPDAC-21 cycles (21 days; doxorubicin, etoposide, cyclophosphamide, vincristine, prednisolone and dacarbazine) +/- radiotherapy (25). R/R cHL was confirmed by pathology, when clinically possible, and staging performed by total body MRI scan, FDG-PET-CT scan and abdominal ultrasound. Patients were stratified into 2 groups (standard and low risk) according the EuroNet-PHL guidelines (4) and as published (22, 24). Disease was considered refractory if remission had not been achieved or if progression occurred < 3 months after end-of-treatment. All patients and/or parents or guardians were informed about standard chemotherapy (4) and offered to choose treatment with nivolumab + BV. They provided written informed consent according to national laws and in agreement with the declaration of Helsinki (2013). This study is IRB-approved and registered under a national trial registry number 7744.

All patients were treated as published for the Checkmate 744 trial (22, 24): Patients received 4-6 cycles of nivolumab (3 mg/kg on day 8 of cycle 1 and on day 1 thereafter) + BV (1.8 mg/kg on day 1 of every cycle). If CMR [Deauville score < 4, according to Lugano criteria (23)] was reached after cycle 4, patients from the low risk group received consolidation treatment by involved-field radiotherapy. After achieving CMR patients from the standard risk group were treated with HD-BEAM chemotherapy followed by autologous stem cell reinfusion. Autologous peripheral CD34+ blood stem cells were harvested in standard risk patients after the 3^rd or 4^th cycle of nivolumab + BV according to standard procedures. In patients not in CMR after the 4^th nivolumab + BV cycle treatment was intensified by administering 2 cycles of BV + bendamustine (90 mg/m2 on day 1 and 2). One patient was treated with DHAP (dexamethasone, high-dose cytarabine, cisplatinum) + BV (26) because of progressive disease after BV + bendamustine. Treatment toxicity was assessed according to the National Cancer Institute Common Terminology Criteria (CTC) to evaluate adverse events.

Response to treatment was evaluated by total body MRI scan and FDG-PET-CT after 2 and 4 cycles of nivolumab + BV or after the second cycle of bendamustine + BV according to the Lugano criteria (23). A Deauville score <4 was considered a CMR.

Laboratory tests were performed according to standardized procedures at the laboratories for clinical chemistry at the University Medical Center Utrecht and the Princess Máxima Center. Serum TARC levels were measured by ELISA at the laboratory of the UMC Utrecht.

Data were descriptively analyzed using Stata 18 SE for Windows (StataCorp., College Station, TX, USA). Continuous data were shown as median and range or as mean values and range when useful. Categorical data were shown as frequencies and percentages.

We here report the results of 10 patients treated for R/R cHL at the Princess Máxima Center for Pediatric Oncology ( Table 1 ). The median age at primary cHL diagnosis was 13 years (range: 9-16). Nine children had received first-line cHL therapy according to the EuroNet PHL-C2 treatment protocol and one according to the EuroNet PHL-interim treatment recommendation (25) at our center ( Table 1 ). Two patients had received radiotherapy after completion of chemotherapy. Notably, 5 patients were in complete metabolic remission (CMR) at the end of first line treatment.

Recurrences occurred on average 236 days (range 8-804, median 176 days) after end-of-treatment and were diagnosed during routine follow-up visits in 90% of cases. One case came to attention due to a visit scheduled at the patients’ request after feeling a swelling the neck. In 9/10 patients R/R disease was detected in the initially involved lymph node region. 3/10 patients showed progressive disease during treatment (1 at early and 2 at late response assessment). In 20% of cases the relapse occurred at a site previously treated by radiotherapy. One patient presented with a second recurrence. Nine R/R cHL cases were confirmed by pathology. In 1 patient a biopsy was considered too dangerous due to the bleeding risk because of the anatomical localization. The diagnosis of this specific cHL recurrence was made based on detection of enlarging FDG-PET-positive (Deauville score 5) lymph nodes in the previously afflicted lymph node region, an increase of the erythrocyte sedimentation rate (ESR) from normal to 63 mm/1^st h and serum TARC-elevation to 971 pg/ml.

In all patients staging was performed with total body MRI and FDG-PET-CT scans. It demonstrated Ann-Arbor stage I/II disease extension in 50% and stage III/IV in 50% of patients. Characteristic “B-symptoms” for cHL such as fever, drenching night sweats or weight loss >10% body weight were absent in 90% of patients treated for R/R disease. One patient with refractory disease had B-symptoms at first presentation. Three patients presented with an ESR >30 mm/1^st hour. Recently, the cytokine TARC (CCL-17) has been characterized as a biomarker for cHL in children (27–29). At diagnosis of R/R cHL serum TARC levels were elevated (>880 pg/ml) in 8/10 patients and hemoglobin levels reduced (<7.4 mmol/L) in 4/10 patients.

There were no emergency presentations due to a large mediastinal mass. All patients had a Karnofsky/Lansky score above 80 and presented without relevant comorbidities. There were 8 patients in the standard and 2 patients in the low-risk treatment group ( Table 2 ).

Treatment of R/R disease was on average initiated 16 days (range: 8-26) after diagnosis. Treatment was given according to the Checkmate 744 trial protocol: The first cycle of immunotherapy was started with the administration of BV followed by infusion of nivolumab 7 days later. Afterwards nivolumab + BV was administered on the same day every 3 weeks in our outpatient clinic (22). Immunotherapy induced a CMR (Deauville score <4) in 5 patients after 2 administrations of nivolumab + BV. Four patients did not achieve CMR after 4 nivolumab + BV cycles. Three patients achieved CMR after 2 and 1 patient after 3 cycles of BV in combination with bendamustine. In 1 patient with a FDG-PET-CT scan suspicious for only a partial metabolic response a CT-guided biopsy was performed that did not detect viable cHL tissue. A second patient showed progressive disease after 2 cycles of BV + bendamustine. Treatment was intensified by 1 course of DHAP + BV (26), which induced CMR ( Figure 1 ), followed by consolidation with HD-BEAM with autologous stem cell reinfusion. Also this patients is still in ongoing remission for 16 months. TARC levels normalized concomitantly with CMR induction ( Figure 2 ). Peripheral stem cell harvest after 2 to 4 cycles of nivolumab + BV was successful in all standard risk patients. On average 5.2 x10⁶ CD34+ cells/kg body weight were harvested.

The treatment was well tolerated and administered in the outpatient clinic ( Table 3 ). We observed a single CTC grade 3 or higher adverse event: an increase of ALT and AST more than five times the upper limit of normal. Both ALT and AST recovered spontaneously and there were no signs of impaired liver function and no suspicion of drug-induced liver injury. Bilirubin levels stayed within the reference range. However, the next nivolumab cycle was delayed, while the possible causes were evaluated. Mild side effects that were reported more than once included fever (n = 3), rash (n = 2), and fatigue (n = 2). One patient developed hypothyroidism attributed to nivolumab treatment, which recovered during follow-up.

Depending on the initial risk group stratification 7/10 patients received standard consolidation with HD-BEAM chemotherapy plus autologous stem cell reinfusion and 3/10 involved-field RT (30,6 to 36 Gy). One patient with 2^nd recurrence had received prior HD-BEAM chemotherapy with autologous stem cell reinfusion as treatment for first recurrence and therefore received radiotherapy as consolation after reaching CMR. In 2/10 patients radiotherapy was administered after completing HD-BEAM chemotherapy with autologous stem cell reinfusion. The indications for additional consolidation radiotherapy in these 2 patients were detection of viable lymphoma after the 4^th course of nivolumab + BV with CMR after bendamustine + BV; and progression after treatment with bendamustine + BV with CMR after one cycle of DHAP + BV. Most importantly, these 2 patients achieved CMR before starting consolidation. The toxicity observed after HD-BEAM was mild. No CTCAE adverse events above grade 3 were reported. Grade 3 CTCAE adverse events included febrile neutropenia (n = 5), weight loss (n = 1), nausea (n = 1), vomiting (n = 1), oral mucositis (n = 1) and isolated ALT increase (n = 1). On average 3.8 x10⁶ CD34+ stem cells/kg body weight were infused. Leukocyte numbers had recovered to >5 x10e6/L after a mean of 11 days (range 11-15) after stem cell reinfusion with neutrophils reaching counts >4 x10e6/L after a mean of 13 days (range 10-16). At present all patients are in ongoing remission with a median follow-up time of 25 months (range: 12-42) since the end-of-treatment.