We report on the real world population-based study of patterns of care of DLBCL in Belgium (2013-2015) with a specific focus on the elderly population.


The Belgian Cancer Registry (BCR) collects, processes and analyses data on all new cancers diagnosed in Belgian residents by independent collection of double input: oncological care programs and pathology reports. Coded data were thus obtained of all adult (≥ 20 year) DLBCL diagnosed between 2013 and 2015 (n=1,890).

From pathology reports we extracted immunohistochemistry (IHC) expression of CD10, BCL2, BCL6, IRF4 and MYC proteins, cell of origin (COO) classification estimated by Hans algorithm and BCL2, BCL6 or MYC gene translocation by FISH and/or PCR.

Belgian health insurance data were used to infer DLBCL treatment modalities (chemotherapy schemes based on the combination of reimbursed drugs, HSCT and/or radiotherapy received) as well as to assess comorbidities.

First-line treatments were grouped into 3 main categories: "standard" R-CHOP (≥6 or ≥4 cycles if Ann-Arbor = 1, including R-miniCHOP); other anthracycline containing schemes (other R-CHOP, CHOP, (R-)ACVBP, (R-)CHOP-like, intensified regimens); non-anthracycline treatments (COP and bendamustine schemes, palliative treatments).


COO was available in 63% of cases; KI-67, BCL2, BCL2 & MYC expressions in 58%, 62% and 16% of cases respectively; MYC, BCL2 & MYC gene rearrangements in 11% and 9% of cases. Of the evaluable cases, 49% were double expressor (DE) and 10% were double-hit DLBCL (FISH only performed in ±10%). The negative prognostic markers identified from univariable models were: age, WHO PS, Ann-Arbor stage, non-GCB COO, MYC rearrangement, BCL2 expression, BCL2/MYC DE, comorbidities, other malignancies.

First line systemic treatment was started in 84% of patients (Rituximab-containing in 96%) divided in standard R-CHOP (52%), other anthracycline schemes (39%) and non-anthracycline regimens (10%). The median [IQR] delay from diagnosis to treatment was 19 [9-31] days.

In 18% of treated patients 2nd line therapy was initiated for refractory (10%) or relapsed disease (8%). This contained either a platinum-derivative (with or without cytarabine), HD cytarabine, anthracycline or bendamustine in 54%, 6%, 6 and 3% of cases respectively. Autologous HSCT was performed in 67 patients (BEAM-like conditioning in 88%), allogeneic HSCT in only 4 cases (< 65 yr-old).

Importantly 56% of patients were ≥ 70 yr-old (transplant-ineligible) and 44% ≥ 80 yr-old (unfit for intensive treatments). In patients older than 80 (and +85) years, no systemic treatment was administered in 38% (57%) of which 27% did receive radiotherapy. First line treatment was started in 62% (43%): "standard" R-(mini)CHOP in 22% (11%); other anthracycline schemes in 23% (13%) and non-anthracycline regimens in 17% (19%) (Table 1). In this elderly population treated with R-CHOP, the median number of cycles and interval still remained 6 and 21 days showing feasibility of dose-dense therapy. Second line therapy was started in 6% and no HSCT was performed.

The 2-yr overall survival [95% IC] between the different age categories [20-59], [60-69], [70-79], [80-84] and ≥85 yr-old were 84 [80-87], 74 [69-78], 62 [58-66], 44 [38-49] and 31 [25-36], respectively (Figure 1).

Survival curves of treatments by age categories showed that the older patients [70-84 yr-old] still benefit from the R-(mini)CHOP treatment (Figure 2).

In the multivariable analyses the following variables were associated with a significant hazard ratio: sex, age, WHO PS, respiratory comorbidity, tumor history > 1,5 year after diagnosis, "standard" R-CHOP treatment.


This real world population-based study allows to assess patients usually excluded from clinical trials (elderly population, patients with comorbidities including malignancies (13%). Although limitations due to treatment inference and to the absence of some prognostic markers, this real world analysis of diagnostic-work up and inference of treatment modalities of DLBCL in Belgium showed that the disease characteristics of the elderly population did not seem to be different from the younger population.

The majority (63%) of older patients [70-84 yr-old] are started on first line treatments with curative intent. Our study suggests that a substantial fraction of this elderly population qualifies for standard R-(mini)CHOP treatment and benefits from it.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution