On 2014 Dec 05, Andrea Messori commented:

Treatments for naive patients with follicular non-Hodgkin lymphoma

The study by Rummel et al. [1] is a potentially important advancement in the treatment of naïve patients with indolent non-Hodgkin’s lymphoma. Since, in this trial, the clinical results with bendamustine plus rituximb were very favourable in the subgroup of patients with follicular lymphoma, it can be worthwhile to compare these findings with those observed in other randomized trials aimed at naïve patients with this disease condition. According to the systematic review by Schulz and co-workers [2], two randomized trials [3,4] have previously been conducted in treatment-naive patients with follicular lymphoma. Another trial (the PRIMA trial[5]) has been published thereafter.

If one employs the end-point of progression-free survival at 2 years, these randomized trials have reported the following results (n/N): a) Hiddeman et al. [3]: R-CHOP, 167/223 (74.9%) vs CHOP, 103/205 (50.2%); b) Marcus et al. [4]: R-CVP, 137/159 (86.2%) vs CVP, 81/162 (50%); c) Rummel et al [1] : bendamustine+ rituximab, 93/139 (66.9%) vs R-CHOP, 63/140 (45%); d) PRIMA trial [5] (subgroup of 1,019 patients who completed the induction treatment): i) First data set (unadjusted)= survival data determined from randomization until last date of the follow-up: induction+rituximab maintenance, 418/505 (82.8%) vs induction+no rituximab maintenance, 345/513 (67.3%); ii) Second data set with empirical adjustment (see FOOTNOTE) = survival data determined from registration until last date of the follow-up: induction+rituximab maintenance, 418/529 (79%) vs induction+no rituximab maintenance, 345/537 (64.2%). With the exception of the first data set of the PRIMA trial, these results selectively refer to treatment-naïve patients with follicular lymphoma. The numerators and denominators of the crude rates reported above were estimated from the Kaplan-Meier curves according to graphical methods [6,7]. In each trial, the following sources of information were used: Hiddeman et al. [3]: Figure 2, Panel A; Marcus et al. [4]: Figure 1; Rummel et al. [1]: Figure 3, Panel A; PRIMA trial [5]: Figure 2, Panel A.

If one analyses the above pair-wise comparisons according to the end-point of achievement of progression-free survival at 2 years, the following values of odds-ratio (OR) can be obtained (random-effect model, Open Meta-analysts software): R-CHOP vs CHOP: 2.953 (95%CI: 1.964 to 4.442); R-CVP vs CVP, OR=6.227 (95%CI: 3.609 to 10.744); bendamustine+rituximab vs R-CHOP, OR=2.47 (95%CI: 1.521 to 4.015); R-CHOP+R-maintanance vs R-CHOP (adjusted data set): 2.303 (95%CI: 1.741 to 3.047).

Interestingly enough, all of these comparisons show a significantly better effectiveness of the first treatment as compared to the second. In this framework, a Bayesian meta-analysis aimed at assessing these data could be useful to synthesise the comparative effectiveness of these treatments.

FOOTNOTE: In the PRIMA trial [5], after a total of 1,202 treatment-naive patients were enrolled and registered, randomization to either rituximab maintenance or no maintenance was applied to the subgroup of 1,019 patients who completed the induction treatment. To make the PFS data of this trial (that refer to the randomised patients) comparable to those of the other trials reported above (that refer to treatment-naïve patients), the rates of progression (or, likewise, the rates of PFS) in the PRIMA trial need to be subjected to an empirical adjustment (increase in the numerator of the progression rate or increase in the denominator of the PFS rate applied to both the maintenance and the no-maintenance arms). This adjustment can be based on further information reported on the same patients in a more recent analysis [8].

In their Figures 1 and 2, Ghesquières and co-workers [8] have provided useful details to compare the PFS data between the subgroup of registered patients (data shown in Figure 1, panels A and B) and the subgroups of randomised patients (data shown in Figure 2, panels AI, AII, BI, and BII). For this purpose, the two following comparisons can be made: i) Comparison A between Panel A of Figure 1 vs. Panels AI and AII of Figure 2; ii) Comparison B between Panel B of Figure 1 vs. Panels BI and BII of Figure 2. Comparison A uses the data of FCGR3A polymorphism, while Comparison B uses the data of FCGR2A polymorphism.

In Comparison A, the crude rate of progression for registered patients (according to Figure 1, panel A) is 171/460 (37.17%) where 171=68+80+23 and 460=177+215+68 while the crude rate of progression for randomised patients (according to Figure 2, panels AI and AII) is 141/397 (35.52%) where 141= 35+45+11+17+26+7 and 397=76+93+27+ 72+98+31. The ratio of these two percent rates of progression (37.17% / 35.52%) is 1.0465; this indicates that, using an acceptable approximation, the rate of progression in registered patients can reasonably be estimated from the rate in randomized patients by applying a 4.65% increase in the number of events found in randomised patients (i.e. by increasing the numerator of the rate without changing the denominator). Likewise, in Comparison B, the crude rate of progression for registered patients (according to Figure 1, panel B) is 168/450 (37.33%) while the same rate for randomised patients (according to Figure 2, panels BI and BII) is 138/387 (35.66%). The ratio of these two percent rates of progression (37.33% / 35.66%) is 1.0468. which is very similar to that estimated from Comparison A.

These two re-analyses (i.e. Comparisons A and B) consistently indicate that the rates of progression in registered patients can be estimated with a very good approximation from the rates found in randomized patients by applying a 4.7% adjustment to the survival data (i.e. an increase in the numerator of the progression rate or an increase in the denominator of the PFS rate).

References

1. Rummel et al. Bendamustine plus rituximab versus CHOP plus rituximab as first-line treatment for patients with indolent and mantle-cell lymphomas: an open-label, multicentre, randomised, phase 3 non-inferiority trial. Lancet. 2013 Apr 6;381(9873):1203-10.

2. Schulz et al. Chemotherapy plus Rituximab versus chemotherapy alone for B-cell non-Hodgkin's lymphoma. Cochrane Database Syst Rev. 2007 Oct 17;(4):CD003805.

3. Hiddemann et al. Frontline therapy with rituximab added to the combination of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) significantly improves the outcome for patients with advanced-stage follicular lymphoma compared with therapy with CHOP alone: results of a prospective randomized study of the German Low-Grade Lymphoma Study Group. Blood 2005;106(12):3725–3732.

4. Marcus et al. CVP chemotherapy plus rituximab compared with CVP as first-line treatment for advanced follicular lymphoma. Blood 2005;105(4):1417–1423.

5. Salles et al. Rituximab maintenance for 2 years in patients with high tumour burden follicular lymphoma responding to rituximab plus chemotherapy (PRIMA): a phase 3, randomised controlled trial. Lancet. 2011 Jan 1;377(9759):42-51. doi: 10.1016/S0140-6736(10)62175-7.

6. Messori et aL. Survival meta-analysis of individual patient data and survival meta-analysis of published (aggregate) data. Clin Drug Invest 2000;20(5):309-316.

7. Tierney et al. Practical methods for incorporating summary time-to-event data into meta-analysis. Trials. 2007 Jun 7;8:16.

8. Ghesquières et al. Clinical outcome of patients with follicular lymphoma receiving chemoimmunotherapy in the PRIMA study is not affected by FCGR3A and FCGR2A polymorphisms. Blood. 2012 Sep 27;120(13):2650-7.

On 2014 Dec 05, Andrea Messori commented:

Treatments for naive patients with follicular non-Hodgkin lymphoma

The study by Rummel et al. [1] is a potentially important advancement in the treatment of naïve patients with indolent non-Hodgkin’s lymphoma. Since, in this trial, the clinical results with bendamustine plus rituximb were very favourable in the subgroup of patients with follicular lymphoma, it can be worthwhile to compare these findings with those observed in other randomized trials aimed at naïve patients with this disease condition. According to the systematic review by Schulz and co-workers [2], two randomized trials [3,4] have previously been conducted in treatment-naive patients with follicular lymphoma. Another trial (the PRIMA trial[5]) has been published thereafter.

If one employs the end-point of progression-free survival at 2 years, these randomized trials have reported the following results (n/N): a) Hiddeman et al. [3]: R-CHOP, 167/223 (74.9%) vs CHOP, 103/205 (50.2%); b) Marcus et al. [4]: R-CVP, 137/159 (86.2%) vs CVP, 81/162 (50%); c) Rummel et al [1] : bendamustine+ rituximab, 93/139 (66.9%) vs R-CHOP, 63/140 (45%); d) PRIMA trial [5] (subgroup of 1,019 patients who completed the induction treatment): i) First data set (unadjusted)= survival data determined from randomization until last date of the follow-up: induction+rituximab maintenance, 418/505 (82.8%) vs induction+no rituximab maintenance, 345/513 (67.3%); ii) Second data set with empirical adjustment (see FOOTNOTE) = survival data determined from registration until last date of the follow-up: induction+rituximab maintenance, 418/529 (79%) vs induction+no rituximab maintenance, 345/537 (64.2%). With the exception of the first data set of the PRIMA trial, these results selectively refer to treatment-naïve patients with follicular lymphoma. The numerators and denominators of the crude rates reported above were estimated from the Kaplan-Meier curves according to graphical methods [6,7]. In each trial, the following sources of information were used: Hiddeman et al. [3]: Figure 2, Panel A; Marcus et al. [4]: Figure 1; Rummel et al. [1]: Figure 3, Panel A; PRIMA trial [5]: Figure 2, Panel A.

If one analyses the above pair-wise comparisons according to the end-point of achievement of progression-free survival at 2 years, the following values of odds-ratio (OR) can be obtained (random-effect model, Open Meta-analysts software): R-CHOP vs CHOP: 2.953 (95%CI: 1.964 to 4.442); R-CVP vs CVP, OR=6.227 (95%CI: 3.609 to 10.744); bendamustine+rituximab vs R-CHOP, OR=2.47 (95%CI: 1.521 to 4.015); R-CHOP+R-maintanance vs R-CHOP (adjusted data set): 2.303 (95%CI: 1.741 to 3.047).

Interestingly enough, all of these comparisons show a significantly better effectiveness of the first treatment as compared to the second. In this framework, a Bayesian meta-analysis aimed at assessing these data could be useful to synthesise the comparative effectiveness of these treatments.

FOOTNOTE: In the PRIMA trial [5], after a total of 1,202 treatment-naive patients were enrolled and registered, randomization to either rituximab maintenance or no maintenance was applied to the subgroup of 1,019 patients who completed the induction treatment. To make the PFS data of this trial (that refer to the randomised patients) comparable to those of the other trials reported above (that refer to treatment-naïve patients), the rates of progression (or, likewise, the rates of PFS) in the PRIMA trial need to be subjected to an empirical adjustment (increase in the numerator of the progression rate or increase in the denominator of the PFS rate applied to both the maintenance and the no-maintenance arms). This adjustment can be based on further information reported on the same patients in a more recent analysis [8].

In their Figures 1 and 2, Ghesquières and co-workers [8] have provided useful details to compare the PFS data between the subgroup of registered patients (data shown in Figure 1, panels A and B) and the subgroups of randomised patients (data shown in Figure 2, panels AI, AII, BI, and BII). For this purpose, the two following comparisons can be made: i) Comparison A between Panel A of Figure 1 vs. Panels AI and AII of Figure 2; ii) Comparison B between Panel B of Figure 1 vs. Panels BI and BII of Figure 2. Comparison A uses the data of FCGR3A polymorphism, while Comparison B uses the data of FCGR2A polymorphism.

In Comparison A, the crude rate of progression for registered patients (according to Figure 1, panel A) is 171/460 (37.17%) where 171=68+80+23 and 460=177+215+68 while the crude rate of progression for randomised patients (according to Figure 2, panels AI and AII) is 141/397 (35.52%) where 141= 35+45+11+17+26+7 and 397=76+93+27+ 72+98+31. The ratio of these two percent rates of progression (37.17% / 35.52%) is 1.0465; this indicates that, using an acceptable approximation, the rate of progression in registered patients can reasonably be estimated from the rate in randomized patients by applying a 4.65% increase in the number of events found in randomised patients (i.e. by increasing the numerator of the rate without changing the denominator). Likewise, in Comparison B, the crude rate of progression for registered patients (according to Figure 1, panel B) is 168/450 (37.33%) while the same rate for randomised patients (according to Figure 2, panels BI and BII) is 138/387 (35.66%). The ratio of these two percent rates of progression (37.33% / 35.66%) is 1.0468. which is very similar to that estimated from Comparison A.

These two re-analyses (i.e. Comparisons A and B) consistently indicate that the rates of progression in registered patients can be estimated with a very good approximation from the rates found in randomized patients by applying a 4.7% adjustment to the survival data (i.e. an increase in the numerator of the progression rate or an increase in the denominator of the PFS rate).

References

1. Rummel et al. Bendamustine plus rituximab versus CHOP plus rituximab as first-line treatment for patients with indolent and mantle-cell lymphomas: an open-label, multicentre, randomised, phase 3 non-inferiority trial. Lancet. 2013 Apr 6;381(9873):1203-10.

2. Schulz et al. Chemotherapy plus Rituximab versus chemotherapy alone for B-cell non-Hodgkin's lymphoma. Cochrane Database Syst Rev. 2007 Oct 17;(4):CD003805.

3. Hiddemann et al. Frontline therapy with rituximab added to the combination of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) significantly improves the outcome for patients with advanced-stage follicular lymphoma compared with therapy with CHOP alone: results of a prospective randomized study of the German Low-Grade Lymphoma Study Group. Blood 2005;106(12):3725–3732.

4. Marcus et al. CVP chemotherapy plus rituximab compared with CVP as first-line treatment for advanced follicular lymphoma. Blood 2005;105(4):1417–1423.

5. Salles et al. Rituximab maintenance for 2 years in patients with high tumour burden follicular lymphoma responding to rituximab plus chemotherapy (PRIMA): a phase 3, randomised controlled trial. Lancet. 2011 Jan 1;377(9759):42-51. doi: 10.1016/S0140-6736(10)62175-7.

6. Messori et aL. Survival meta-analysis of individual patient data and survival meta-analysis of published (aggregate) data. Clin Drug Invest 2000;20(5):309-316.

7. Tierney et al. Practical methods for incorporating summary time-to-event data into meta-analysis. Trials. 2007 Jun 7;8:16.

8. Ghesquières et al. Clinical outcome of patients with follicular lymphoma receiving chemoimmunotherapy in the PRIMA study is not affected by FCGR3A and FCGR2A polymorphisms. Blood. 2012 Sep 27;120(13):2650-7.