Abstract 3656


Diffuse large B-cell lymphoma (DLBCL) is predominantly diagnosed in the elderly and its incidence is therefore rapidly increasing. Rituximab in combination with cyclophosphamide, doxorubicin, vincristine and prednisone (R-CHOP) is the standard treatment for patients with newly diagnosed DLBCL. This is mainly based on data from carefully controlled randomised trials with strict eligibility criteria. Therefore data reflecting the “real world” in the rituximab era are sparse. We conducted an observational prospective cohort study to identify whether comorbidity is an independent risk factor for overall survival (OS) in patients with newly diagnosed DLBCL treated with R-CHOP with curative intent.


All patients with newly diagnosed DLBCL in Friesland, a Dutch province, are prospectively registered by their clinicians in a population-based registry (HemoBase®) since January 2005. Patients ≥18 years, CD20 positive, Ann Arbor stage II-IV were included if they received ≥ one cycle of R-CHOP.

Clinical variables registered were age, gender, performance status (PS), Ann Arbor stage, presence of B-symptoms, lactate dehydrogenase, bulky disease, number of extranodal sites involved. Comorbidity was scored using the Charlson Comorbidity Index (CCI) [Charlson et. al.; J Chron Dis, 1987]. The score is a summation of the number of comorbidities existing at the time of diagnosis. The group was divided in low CCI (0–1) and high CCI (≥2).

The primary endpoint was OS, defined as start of diagnose of DLBCL until death by any cause.

Cox Proportional hazards model was used to identify significant risk factors. Variables used to calculate International Prognostic Index (IPI) and CCI were entered into the model. Kaplan-Meier curves were evaluated by log-rank test.


Over a period of 7 years 302 patients were newly diagnosed with DLBCL after the start of the HemoBase® registry. Of those patients, 76 received no R-CHOP therapy, 9 had unknown Ann Arbor stage and 61 had Ann Arbor stage I. Therefore data from 156 patients could be retrieved from the population-based registry for further analysis. The median age was 69 years (range 23 – 93). The median observation period for the total population was 24 months (range 0 – 83) and for patients still alive 33 months (range 2 – 83). Twenty-five percent of patients had a PS≥2, 39% had IPI≥3.

The percentage of patients with 8, 7, 6, 5, 4, 3, 2, or 1 R-CHOP cycle(s) given were 49, 3, 32, 1, 3, 4, 2, 6 respectively. R-CHOP21 was given to 74% and R-CHOP14 to 26%. Dose reduction of 20% or more in doxorubicin, cyclophosphamide or vincristine occurred in 26, 19, and 36% of the patients respectively. In 54% of the patients dose reduction occurred in one or more antineoplastic drugs.

Twenty-five patients had a CCI≥2. Most frequently observed comorbidities in the CCI≥2 group were diabetes (36%), peripheral vascular disease (32%), chronic pulmonary disease or second tumor (28%), myocardial infarction (24%), congestive heart failure or cerebrovascular disease (16%). The percentage of patients with a CCI score of 2, 3, 4 or 5 were 64, 28, 4 and 4% respectively. In the CCI≥2 group only 16% of the patients had also a PS≥2.

Three year OS in the total study population was 68%: in patients with CCI<2 71% and in CCI≥2 48%.

As shown in figure 1, the median overall survival was 18 months in the CCI≥2 group and over 80 months in CCI<2 (p=0.01). From multivariate analysis of CCI and IPI variables only CCI≥2 (p=0.01; HR 2.41 (95%CI 1.20 – 4.87) and PS≥2 (p<0.001; HR 3.77 (95%CI 1.91 – 7.45) were identified as significant risk factors for poor overall survival.

Figure 1.

Overall survival in patients with CCI≥2 compared with patients with CCI<2.

Figure 1.

Overall survival in patients with CCI≥2 compared with patients with CCI<2.


In this population-based analysis of all patients with advanced-stage DLBCL treated with R-CHOP in a Dutch cohort over a period of 7 years, comorbidity is an important and independent adverse risk factor. Comorbidity data provide important prognostic information on OS, thereby aiding patient consultation in daily practice. Furthermore, our results demonstrate the importance of population-based research.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.