The definition and role of bulky disease in young patients (ie, aged 18–60 years) with good-prognosis diffuse large-B-cell lymphoma (DLBCL), who have been treated with CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisolone)-like chemotherapy with or without rituximab, remain controversial. We aimed to assess the effect of maximum tumour diameter (MTD) in these patients.

Patients from the MInT (Mabthera International Trial Group) study were eligible. We analysed event-free (EFS) and overall survival (OS) after CHOP-like chemotherapy with or without rituximab, according to MTD, by Martingale residual analyses and Cox regression models. Radiotherapy was given to sites of primary bulky disease according to national standards, and to primary extranodal disease at physician discretion. The primary endpoint was EFS and the secondary endpoint was OS. Analyses were by intention to treat. This trial is registered with ClinicalTrials.gov, number NCT 00064116.

Of the 824 patients enrolled in the MInT study, the informed-consent form of one patient was missing, leaving 823 patients evaluable for intention-to-treat analysis. Data on MTD of involved sites were available for 802 patients. Martingale residual analysis showed an adverse prognostic effect of MTD on EFS and OS, which increased linearly. In a multivariable analysis with MTD as a linear regression variable, the effect of MTD was significant after CHOP-like treatment alone for EFS (hazard ratio 1·090 [95% CI 1·051–1·130], p<0·0001) and OS (1·119 [1·057–1·184], p=0·0001), and after CHOP-like treatment and rituximab for OS (1·089 [1·003–1·183], p=0·043), but not for EFS (1·044 [0·991–1·099], p=0·103). For CHOP-like treatment alone, 3-year EFS ranged from 78·2% (MTD <5·0 cm, 95% CI 68·3–85·4) to 41·3% (MTD ≥10·0 cm, 31·8–50·4). For CHOP-like treatment and rituximab, 3-year EFS ranged from 83·2% (MTD <5·0 cm, 72·8–89·9) to 72·7% (MTD ≥10·0 cm, 63·8–79·7). With CHOP-like treatment alone, 3-year OS decreased from 92·9% (MTD <5·0 cm, 84·9–96·8) to 73·5% (MTD ≥10·0 cm, 63·9–81·0); for CHOP-like treatment and rituximab, 3-year OS decreased from 98·0% (MTD <5·0 cm, 92·2–99·5) to 85·2% (MTD ≥10·0 cm, 77·0–90·6). For CHOP-like treatment, any cut-off point between 5·0 cm and 10·0 cm separated two populations with a significant EFS difference (p<0·0001 for all log-rank tests) and OS difference (p≤0·003 for all log-rank tests). For CHOP-like treatment and rituximab, only a cut-off point of 10·0 cm separated two populations with a significant EFS difference (log-rank p=0·047), but any cut-off point of 6·0 cm or more separated two populations with a significant OS difference (log-rank p values 0·0009–0·037).

Rituximab decreased, but did not eliminate the adverse prognostic effect of MTD in young patients with good-prognosis DLBCL. Due to the linear prognostic effect of MTD on outcome, arbitrary cut-off points for bulky disease can be set between 5·0 cm and 10·0 cm, depending on clinical considerations. Based on this study, a cut-off point of 10·0 cm might be a suitable margin in the rituximab era to delineate those patients with bulky disease.

Roche, Basel, Switzerland (M39045).