Predictive Factors of Relative Dose Intensity Reduction in Clinical Practice for First Line Treatment in Patients with Non-Hodgkin and Hodgkin Lymphoma

The first line treatment regimen differs according to the type of lymphoma. ABVD for Hodgkin’s Lymphomas (HL) and CHOP or R-CHOP for non-Hodgkin’s Lymphomas (NHL) are the most worldwide used treatments. There is good evidence that reduction in relative dose intensity (RDI) results in loss of clinical benefit. The aims of this study were to define predictive risk factors for the RDI reduction and to evaluate patterns of Granulocyte-Colony Stimulating Factors (G-CSF) use in these first line treatment regimens. Retrospective data from three Portuguese centres of 133 patients (pts) with NHL and HL treated with ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine), R-CHOP 21 like (rituximab, cyclophosphamide, vincristine, doxorubicin/mitoxantrone, prednisolone) and CHOP 21 like regimens was collected. Dose Intensity Evaluation Programme (DIEP^®) software was developed to allow calculation of RDI and includes demographic data, clinical and treatment characteristics, chemotherapy dose modifications and delays, haematological toxicities and patterns of use of G-CSF. Three classes were defined for “size of tumoral mass” (< 5 cm, 5–10 cm, >10 cm) and for “G-CSF administration” (primary prophylaxis, secondary prophylaxis and neutropenia/febrile neutropenia treatment). Univariate and multivariate analyses were performed to identify factors related to RDI ≤ 90% of standard. A P-value < 0.05 was considered for statistical significance. Patient distribution according to lymphoma is HL – 37.6%, NHL – 62.4%. All HL pts were treated with ABVD regimen. Among NHL pts, only 9.9% had a T-cell related lymphoma, 43.7% were treated with CHOP 21 like and 56.3% with R-CHOP 21 like. Median RDI was 92.0% in ABVD, 92.4% in CHOP 21 like and 92.3% in R-CHOP like. RDI ≤ 90% was delivered in 37.1% of all pts. According to the univariate analysis there was no statistical significant difference between RDI (cut-off 90%) and treatment regimen, body mass index, body surface area, serum LDH, albumin, hemoglobin, type of lymphoma, Ann Harbor Staging, symptoms, existence of secondary extranodal involvement, organ with primary involvement, IPI, FLIPI and G-CSF administration. A significant reduction of RDI was associated with age ≥ 65 years, female gender, size of tumoral mass (< 5 cm), bone marrow involvement, hospitalization and presence of a primary extranodal lymphoma. In multivariate analysis, the independent predictor of RDI ≤ 90% was hospitalization (P=0.018). Twenty per cent of R-CHOP 21 like pts were hospitalized due to febrile neutropenia (representing all febrile neutropenia hospitalizations in the study). Six per cent of ABVD pts and 12.1% of CHOP like pts were hospitalized due to other reasons.G-CSF was administrated as primary prophylaxis to 3.33% of the pts treated with ABVD regimen, 12.5% of CHOP 21 like pts and 5.71% of the R-CHOP 21 like pts. The proportion of pts receiving G-CSF as secondary prophylaxis was: ABVD - 80%, CHOP 21 like – 75% and R-CHOP like – 60%. Administration of G-CSF due to neutropenia or febrile neutropenia was performed to 93.33% of ABVD regimen pts, 81.25% of CHOP 21 like pts and 85.71% of R-CHOP 21 like pts. Differences in the reduction of RDI in lymphoma pts treated with the three distinct regimens were not significant. However, hospitalizations in general were statistically associated with RDI reduction. In this study, a relation was found between R-CHOP 21 like regimen and hospitalization due to febrile neutropenia. These results may indicate that R-CHOP 21 like regimen pts could be candidates for primary G-CSF prophylaxis.