So what's left to tell you about this trial? I think we can say something about those who had to have fewer than the recommended number of courses or to have their dose reduced.
The mean number of treatment courses administered was 5·2 (range 0–6) in the chemoimmunotherapy group compared with 4·8 (0–6; p=0·006) in the chemotherapy group. 26% did not receive the planned six courses in the chemoimmunotherapy group compared with 34% in the chemotherapy group. The difference was because there were more non-responders withdrawn from the study in those who did not get rituximab. Patients in Binet stages A and B received more treatment courses (mean 5·28 [range 0–6]) than did those in Binet stage C (4·52 [0–6]; p<0·0001).
For any of the three drugs, the planned dose was reduced by more than 10% in 47% of the patients in the chemoimmunotherapy group and 27% of the patients in the chemotherapy group (p<0·0001). 207 of 800 patients had dose reductions during the first to third courses (chemotherapy 19%; chemoimmunotherapy 33%; p<0·0001), and dose reductions occurred in 216 of 800 patients during the fourth to sixth courses (chemotherapy 20% ; chemoimmunotherapy 34% ; p<0·0001). These dose reductions were mostly because of treatment-related haematological toxicity, particularly neutropenia and leucocytopenia (62% in the chemoimmunotherapy group vs 64% in the chemotherapy group). I think this illustrates that neutrophils get swept up in the immune complexes formed between CD20 and anti-CD20 because they have Fc receptors. The production of neutrophils is not being impaired as it is with chemotherapy-induced neutropenia; they are just being used. I am not sure that it is necessary to dose reduce in this instance, but I admit the matter is moot.
The reason that patients with stage C received less drug that patients with stage B is precisely because these patients start out with hematological impairment.
Although an improvement in PFS with the addition of rituximab was noted in all stages, the improvement for Binet stage C was from only months to 40·7 months which did not reach statistical significance (p=0·081). This may have been a statistical artefact because patients with stage C disease who were given chemoimmunotherapy compared with those who were given chemotherapy showed an accumulation of several unfavourable factors (though none were indivdually statistically different; for all p>0·05): Thus age >65 years 35% vs 26%, unmutated IGHV status 54% vs 48%, elevated ZAP70 42% vs 33%, and β2 microglobulin levels greater than 3·5 mg/L 52% vs 44%.
Dose reductions of more than 10% were more common in patients with Binet stage C disease in the chemoimmunotherapy group 49% vs 28%; p=0·001), mostly because of neutropenia and leucocytopenia. Since dose reductions for prevention of haematological toxicity needed to be maintained for the rest of the treatment according to the protocol recommendations, the authors believe that dose reductions required during early courses of chemoimmunotherapy led to persistently lower doses of all three drugs in patients with Binet stage C disease. Since following the first few courses of treatment the cytopenias might remit, there may not be justification for this continued dose reduction and in clinical practise there exists the possibility of continuing subsequent course at full dose.