Survival outcomes in patients with Hodgkin transformation of CLL: a multicenter study

In 5–10% of patients with chronic lymphocytic leukemia (CLL), an aggressive lymphoma develops in what is known as Richter transformation. Most cases of Richter transformation involve the development of diffuse large B-cell lymphoma (DLBCL) arising from the setting of CLL, though in a very small number of cases (<1%), CLL can transform into Hodgkin lymphoma (HL). The expected overall survival (OS) in patients with Hodgkin transformation (HT) appears to be longer compared with transformation to DLBCL, though OS is shorter (2-year OS of 30–40%) in patients with HT when compared with patients with de novo HL (2-year OS of >90%) after treatment with standard chemotherapy regimens such as adriamycin, bleomycin, vinblastine, and dacarbazine (ABVD). The current data are limited and are mostly derived from small, single institution series. Treatment recommendations for HT are largely based on data extrapolated from patients with Ritcher transformation to DLBCL, with some groups recommending aggressive therapy with hematopoietic cell transplantation (HCT) after first complete remission (CR1). However, patients with CLL are typically elderly with comorbidities, and therefore undergoing HCT may not be feasible in this population.

Here we present the key findings from a retrospective study comparing the clinical features, treatment patterns, and clinical outcomes in patients with HT of CLL with historic case series published by Stephens, et al.¹ in Haematologica.

Study design

This was a retrospective, multi-center, cohort study in patients with CLL who developed biopsy-proven HL, diagnosed between 2000 and 2018 at 13 cancer centers in the United States. Clinicobiologic characteristics, treatment type, and survival outcomes for each patient were analyzed, and International Prognostic Score (IPS) for HL and the Richter Scoring System (RSS) were calculated at the time of HT diagnosis in patients with available data.

Outcomes of interest included:

• OS, defined as time of HT diagnosis until time of death.
• Progression free survival (PFS), defined as the time of HT diagnosis until relapse, subsequent therapy, or death.

Results

Baseline characteristics

At CLL diagnosis

A total of 94 patients with CLL HT were included; the median age at CLL diagnosis was 60 years (range, 31–84 years) and 81% of patients were male. Fifteen percent (11/81) of patients were Rai stage 3–4, 67% (25/37) had an unmutated immunoglobulin heavy chain gene, and 15% (9/61) had del(17p).

Regarding initial therapy for HT, 48 patients received an ABVD-based regimen and 14 patients received an adriamycin, vinblastine, and dacarbazine (AVD) based regimen. Of these 62 patients, seven received an anti-CD20 monoclonal antibody and five received a Bruton’s tyrosine kinase inhibitor (BTKi). Ten patients received brentuximab vedotin, seven received rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), six patients received no therapy for HT (due to frailty), and nine received other regimens.

Overall, 25 patients received ibrutinib (n = 24) or acalabrutinib (n = 1) as CLL treatment immediately prior to first HT treatment. Of these 25 patients, 17 discontinued BTKi treatment during the first HT treatment and three of the 17 resumed BTKi upon completion of HT therapy. Nine patients continued BTKi therapy concurrently with the first HT treatment: six received AVD, two received rituximab plus AVD, and one received brentuximab vedotin.

Thirteen patients received only one salvage regimen for HT and the most common salvage regimens included ifosfamide, carboplatin, and etoposide (n = 7) and brentuximab vedotin (n = 6). Subsequent therapy included autologous HCT (n = 7) and allogenic HCT (n = 11). Seven patients received HCT while in CR1.

At time of HT of CLL

The median age at the time of HT was 67 years (range, 38–85 years); the median time from CLL diagnosis to HT was 5.5 years, and seven patients had simultaneous diagnosis of CLL and HL. The median number of treatments for HT per patient was one, with 61% of patients receiving only one line of therapy. Table 1 shows the baseline characteristics of patients at HT of CLL.

Table 1. Baseline characteristics at HT of CLL*

ALC, absolute lymphocyte count, CLL, chronic lymphocytic leukemia; EBV, Epstein Barr virus; HT, Hodgkin transformation; LDH, lactate dehydrogenase; ECOG, Eastern Cooperative Oncology Group; WBC, white blood cell. *Adapted from Stephens, et al.1 †Includes bendamustine, chlorambucil, fludarabine, and pentostatin. ‡Includes fludarabine and pentostatin.
Characteristics	Total (n)	% (unless otherwise stated)
Ann Arbor Stage III/IV	83	87
ECOG Performance Status 0–1	80	68
B-symptoms	89	65
Bulky disease >10 cm	81	11
HT subtype	56	—
Type I HT	—	59
Type II HT	—	41
EBV	72	57
Elevated LDH	78	40
Median (range), WBC × 109/L	73	7.4 (0.6–199.0)
Median (range), ALC × 109/L	72	1.8 (0.1–187.0)
Median hemoglobin (range), g/dL	74	11.0 (4.9–18.6)
Median (range), platelet × 109/L	74	185 (19–539)
Median number of CLL-directed therapies prior to HT	94	2 (0–12)
CLL-directed therapies prior to HT
No therapy	17	18
No cytotoxic therapy†	27	29
Purine analogue-based‡	43	46
Ibrutinib-based	25	27
Acalabrutinib-based	2	2
Venetoclax-based	1	1
Both cytotoxic and ibrutinib-based	15	16
Both purine analogue and ibrutinib-based therapy	12	13

OS at CLL diagnosis

There was no association of OS with Rai stage 1 (p = 0.98) or 2 (p = 0.63), immunoglobulin heavy chain mutational status (p = 0.88), del(13q) (p = 0.16), trisomy 12 (p = 0.70), del(11q) (p = 0.10), or del(17p) (p = 0.67). The estimated 2-year OS was significantly lower in patients receiving CLL-directed therapy prior to HT vs patients not receiving any prior CLL-directed therapy (69% vs 93%; p = 0.02).

Two-year OS was also significantly lower in patients receiving purine-analogue-based CLL-directed therapy prior to HT compared with patients not receiving purine-analogue-based CLL-directed therapy (60% vs 83%; p = 0.009). Epstein Barr virus (EBV) was frequently present in patients treated with purine-analogue-based CLL-directed therapy prior to HT (odds ratio [OR], 3.24; p = 0.02).

OS at HT diagnosis

Of the 25 patients who developed HT following treatment at any time with ibrutinib:

• The median time from initiation of ibrutinib to HT was 15.5 months (range, 1.2-37.7 months).
• Ten patients had never received treatment with a standard chemoimmunotherapy regimen prior to HT.
• The estimated 2-year OS was slightly lower in the patients who received ibrutinib prior to HT compared with patients who did not receive ibrutinib prior to HT (64% vs 73%; p = 0.33).
• Patients treated with ibrutinib prior to HT were less likely to be positive for EBV (OR, 0.52; p = 0.21).

OS was associated with elevated lactate dehydrogenase (LDH), International Prognostic Score of ≥4 vs <4 and RSS of ≤2 vs <2 (Table 2). A one-point increase in RSS resulted in an increased risk of death (hazard ratio [HR], 2.3; p = 0.004).

The risk of death was similar in patients who received an ABVD-based regimen and those who received R-CHOP as first therapy for HT (HR, 1.6; p = 0.48). There was, however, an increased risk of death in patients who received AVD-based (HR, 3.0; p = 0.04), brentuximab vedotin-based (HR, 3.6; p = 0.05) or other regimens (HR, 7.67; p < 0.001) as first therapy for HT compared with those who received an ABVD-based regimen. The increased risk of death remained significant after adjusting for age at time of first HT therapy.

The median OS was 13.2 years in patients who received an ABVD-based regimen. Patients receiving ABVD vs all other regimens were more likely to be <65 years of age (p = 0.05), have a baseline absolute lymphocyte count >0.6 (p = 0.003), and have an IPS of < 4 (p = 0.03). Two-year OS was similar in patients undergoing HCT for HT vs not undergoing HCT for HT in CR1, respectively (67% vs 72%; p = 0.46).

OS after HT diagnosis

The median follow-up was 1.6 years and the median time between HT diagnosis and HT treatment was 15 days. The median OS and PFS were 65 months and 21 months, respectively, and estimated 2-year OS and PFS after HT diagnosis were 72% and 48%, respectively. Five patients died within two months of HT diagnosis.

Table 2. Univariate analysis for OS at HT of CLL*

ALC, absolute lymphocyte count; CLL, chronic lymphocytic leukemia; CI, confidence interval; EBV, Epstein Barr virus; ESR, erythrocyte sedimentation rate; Hgb, hemoglobin; HR, hazard ratio; HT, Hodgkin transformation of CLL; IPS, International Prognostic Score; LDH, lactate dehydrogenase; PS, performance status; RSS, Richter Scoring System. *Adapted from Stephens, et al.1 †Values in bold are statistically significant.
Covariate	n	HR (95% CI)	p value†
HT subtype 2	57	0.67 (0.21–2.19)	0.511
Ann Arbor stage 3–4	83	1.26 (0.38–4.16)	0.710
ECOG Performance Status 3–4	80	1.13 (0.39–3.28)	0.820
B symptoms	89	1.85 (0.79–4.31)	0.154
LDH above normal	78	4.45 (1.78–11.11)	0.001
Hgb, g/dL
≥12	74	0.75 (0.30–1.86)	0.533
≥10.5	74	0.75 (0.30–1.86)	0.533
WBC × 109/L
≥15	73	1.47 (0.56–3.84)	0.430
ALC × 109/L
≥0.6	72	1.25 (0.36–4.26)	0.725
≥4	72	1.97 (0.81–4.76)	0.133
Creatinine ≥1.5, g/dL	72	1.61 (0.57–4.48)	0.366
Albumin ≥4, g/dL	74	0.41 (0.12–1.39)	0.152
ESR ≥50	36	0.69 (0.23–2.06)	0.502
Lymph node ≥10 cm	80	0.64 (0.15–2.69)	0.542
EBV positive	72	1.72 (0.68–4.39)	0.255
IPSS 4+	57	4.81 (1.05–21.99)	0.043
RSS Score 2–4	57	5.74 (1.77–18.54)	0.003

Conclusion

This retrospective study demonstrated that clinical outcomes, including survival, were higher than previous reports in patients with HT of CLL and interestingly comparable to the outcomes of elderly patients with de novo HL. Based on the findings, the authors recommend that patients with HT should be treated with treatment regimens used in patients with de novo HL. Allogenic HCT in CR1 is not recommended for this population based on similar OS in patients receiving and not receiving HCT in CR1. The OS is predicted to be shorter in patients with HT who received CLL-directed therapies—especially purine-based-analogue therapy—elevated LDH, IPS ≥4, and RSS ≥2. Although patients in this analysis survived longer than previously reported patients who transformed to DLBCL, it is not clear whether this was due to underlying disease biology or effectiveness of therapy. The findings from this study should be interpreted with caveats, including the retrospective nature of the analyses, lack of feasibility to centrally review the included samples, selection bias, and limited generalizability to patients treated in community setting (as included patients were treated at tertiary referral cancer center). Further prospective studies defining the optimum management of patients with HT of CLL are therefore needed.