Abstract 4944

Background:

DLBCL is the most common form of aggressive non-Hodgkin lymphoma (NHL). Despite its aggressive nature, the majority of patients are diagnosed and managed in an out-patient setting and only minority of patients require hospital admission, for symptom control or management of associated co-morbid conditions. The outcomes of patients admitted to hospital with newly diagnosed DLBCL, though presumed to be inferior to patients managed in an outpatient setting, due to advanced disease or poorer performance status is presently unknown. The aims of this study were to identify predictors of treatment location (in-patient vs. out-patient), and assess the survival of patients according to treatment location (in-hospital or out-patient).

Methods:

Retrospective chart review over 5 years (January 2005 to December 2009) of newly diagnosed patients with DLBCL in Winnipeg, Canada. These patients were treated either in a teaching hospital, at Health Sciences Center, or in the out-patient setting, at CancerCare Manitoba. Clinical predictors of treatment setting were analyzed using multivariable logistic regression. Survival at one and three years was assessed with Kaplan-Meier statistics.

Results:

We included140 patients (46 in-patients and 96 outpatients). The mean age of the in-patient population was 68.3 (SD 14.2); while for the out-patient group it was 65.2 (SD 15.3). The in-patient group was comprised of 47.8% female, with the outpatient having 52.1%. Fifty percent of the in-patient group came from a rural residence, while only 28.1% of the out-patient group was from a rural residence. Of the in-patient group 21.7% had a favorable IPI (0-2), compared with 70.8% of the out-patient population.

Of the 46 in-patients, 28 (60.9%) received R-CHOP, compared with 69 (71.9%) from the out-patient group. Four (8.7%) in-patients received an alternate form of chemotherapy (e.g. R-CVP), compared with 18 (18.8%) outpatients. Fourteen (30.4%) in-patients received no chemotherapy, compared with 9 (9.4%) in the out-patient group. Patients with an IPI of 3 or higher at diagnosis were significantly more likely to require hospital admission for initial treatment [Odds ratio (OR) = 8.43; (95% CI 2.55–19.30), p-value <0.01]. Patients living in rural setting were more likely to be hospitalized compared to those who resided in Winnipeg [OR = 2.34; (95% CI 0.86–6.46), p-value = 0.04]. Overall survival at one and three year was 50.0% and 38.8% for the in-patient group, compared with 85.3% and 69.3 for the out-patient group (p<0.01). In a subgroup of patients who received R-CHOP, survival for in-patients compared with out-patients was 71.4%% vs. 89.7%% at one year, and 57.7% vs. 76.8% (p=0.03) at three years respectively. Survival of in-patients with a low IPI (0-2) that completed R-CHOP (six cycles) therapy was 100% at one year, compared with 97.7% in the out-patient group (p=0.10). Survival of in-patients with high IPI (≥3) that completed R-CHOP was 68.2% at one year, compared to 66.7% in the out-patient group (p=0.51).

Conclusions:

The overall survival of patients with DLBCL that require hospital admission to receive their first cycle of chemotherapy is inferior to patients who can be treated in the out-patient setting. Observed differences in survival may relate to the decreased administration of chemotherapy among in-patients, which may further relate to patient co-morbidity and functional status. Among patients who receive a full course of chemotherapy, the location of treatment initiation does not appear to impact survival. Factors found to be associated with in-patient treatment initiation include high IPI and rural status. Initial in-patient treatment is not a necessarily associated with poor prognosis if a complete course of chemotherapy can be delivered and to better inform prognosis, further studies are needed to predict which patients will ultimately not be able to tolerate a full course of chemotherapy and thus be at high risk for death.

Disclosures:

Rubinger:Roche Canada: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution