The Best Treatment for Diffuse Large B-Cell Lymphoma: A German Perspective

Without therapy, aggressive lymphomas are fatal within a few months. But due to their sensitivity to radiotherapy and chemotherapy, the majority of young patients with aggressive lymphomas can be cured. Risk factor profile according to the International Prognostic Index (IPI)[1] and age is the main determinant for the primary therapeutic strategy. A young patient is clinically defined as a patient who is considered fit for high-dose chemotherapy requiring hematopoietic stemcell support, which is usually the case up to between 60 and 65 years of age. Rather than imposing a fixed cutoff point, the new generation of the German High-Grade Non-Hodgkin's Lymphoma Study Group (DSHNHL) protocols for young and elderly patients have overlapping age boundaries with respect to age-dependent eligibility, with protocols designed for young patients up to 65 years and protocols for elderly patients starting with patients 61 years and older. Thus it is up to the recruiting physician to decide on an individual basis whether a "middle-aged" patient between 61 and 65 years better fits into a protocol designed for young or elderly patients. According to the IPI,[1] four prognostic subgroups each for young and elderly patients can be distinguished: low-risk, low/intermediate-risk, high/ intermediate-risk, and high-risk patients. However, for practical reasons, the DSHNHL groups young low-riskand low/intermediate-risk patients together into the therapeutic group of "young good-prognosis patients" that are distinguished from "young poorprognosis patients" (comprising patients with high/intermediate and high risk) with differential therapeutic strategies for both groups. In contrast to "good-prognosis" as defined above, "limited-stage disease," "early-stage disease," "low-stage disease," and "localized disease" are less well defined and often used exchangeably. A further subdivision of limited-stage disease into three subgroups according to a stage-modified IPI has been proposed.[ 2] However, as will be discussed later, in the era of chemoimmunotherapy, a subdivision of good-prognosis patients into very favorable and less favorable subgroups seems more appropriate (see below). Because we define elderly patients as those not fit for high-dose chemotherapy, this option is by definition not available for the elderly; hence we do not (yet) strategically subdivide elderly patients into further subgroups. We also have no differential approach for elderly patients in stage I (eg, reduce number of chemotherapy cycles) because we did not find a significant difference with respect to prognosis between elderly patients in stage I and stage II, if they were balanced for other risk factors according to the IPI or bulky disease.[3] Similarly, we do not believe that a further subdivision of elderly patients into old or very old patients (> 70, > 75, or > 80 years of age) is justified, because it has not been shown that the prognosis of these patients is significantly different as long as their increasingly frequent comorbidities do not compromise the consequent adherence to the therapeutic regimen. Therefore, we recommend a careful initial and follow-up evaluation of patients > 70 years of age for coexisting morbidity that is perceivable before the commencement of therapy or might evolve thereunder. Based on these considerations, the DSHNHL has developed therapeutic strategies for three groups of patients (see Figure 1):
1. Young good-prognosis patients (age-adjusted IPI = 0,1)
2. Young poor-prognosis patients (age-adjusted IPI ≥ 2) 3. Elderly patients More recent prognostic classifications are based on gene expression profiles as determined by the microarray technique. Several investigatorshave shown that gene expression profiles have prognostic significance independent from the IPI. The value of the gene expression-based prognostication has to be confirmed in prospective trials, which is demanding, because to date fresh biopsy material is the prerequisite for gene expression analysis and the method is available and reliable in only a few laboratories worldwide. Even though analysis of a limited number of genes might have a similar prognostic power[4] and good correlations between the expression of a few relevant genes at the mRNA level and at the protein level (by immunohistology) have been demonstrated,[5] the IPI is still the most widely applicable prognosticator within the context of a large cooperative group with 300 participants like the DSHNHL. Therefore, for the time being, prognostic classification based on clinical parameters determines the therapeutic strategies of the DSHNHL. Young Good-Prognosis Patients Background
Radiation doses between 36 and 40 Gy suffice to eradicate the malignant clone. If abdominal involvement beyond the diaphragm is not excluded by a diagnostic laparotomy with splenectomy-an approach that has been abandoned due to its significant morbidity and mortality-the results of radiotherapy alone are not satisfactory.[ 6] Disease-free survival rates of 70% to 80% have been observed only in monocentric and mostly retrospective studies in patients with small lymphomas (< 2.5 cm)[7]; radiotherapy alone has been widely abandoned for patients with aggressive lymphoma, except for rare cases with contraindications against chemotherapy. Similarly, the data supporting a combined-modality approach consisting of chemotherapy plus radiotherapy are conflicting. Because neither chemotherapy nor radiotherapy can be given at full dose in such an approach, there is a wide variety of combinations, and the optimal ones remain to be defined. Combinations of three to five cycles of chemotherapy and radiotherapy with 35 to 45 Gy achieved survival rates between 75% to 80%and 5-year survival rates of 80% to 90% in limited-stage disease. However, follow-up studies show that relapse rates are significant even after 5 years if four or fewer cycles of chemotherapy are given, suggesting that abbreviated chemotherapy is unable to eradicate the malignant clone.[8] In the late 1990s, a Southwest Oncology Group (SWOG) study set the benchmark for the treatment of "limited- stage disease."[2] This study had randomized patients in stage I and those in stage II if they had nonbulky (< 10 cm) disease into eight cycles of CHOP (cyclophosphamide, doxorubicin(Drug information on doxorubicin) HCl, vincristine [Oncovin], prednisone(Drug information on prednisone)) or three cycles of CHOP followed by involved- field radiotherapy with 40 to 55 Gy. While an early analysis of this study had shown an advantage for the combined approach, longer follow-up revealed crossing of the event-free and overall survival curves after 7 and 9 years, respectively.[9] Moreover, the LNH-93.1 trial of the French Group d'Etude des Lymphomes de l'Adulte (GELA)[10] that included patients in stage I and II (irrespective of bulky disease) with no risk factors according to the IPI showed the superiority of three cycles of the ACVBP regimen (doxorubicin [Adriamycin], cyclophosphamide(Drug information on cyclophosphamide), vindesine, bleomycin(Drug information on bleomycin), prednisone) followed by a sequential consolidation chemotherapy over the combination of three cycles of CHOP with involved- field radiotherapy with 30 to 40 Gy. These recent results, publishedto date only in abstract form, support the recommendation that patients with early stages of aggressive lymphoma should receive full-cycle chemotherapy, ie, not less than the equivalent of six cycles of a CHOP-like regimen. Radiotherapy in addition to fullcycle chemotherapy was tested in the Eastern Cooperative Oncology Group (ECOG) 1484 study that randomized patients in stage I with mediastinal, retroperitoneal, or bulky disease (> 10 cm) and patients in stage II in complete remission after eight cycles of CHOP chemotherapy into additional involved-field therapy with 30 Gy[11] or observation. There was a borderline significance in favor of the combined approach with respect to disease-free but not overall survival. Limiting additional radiotherapy to areas of primary bulky disease derives support from a small Mexican study in which radiotherapy to initial bulky disease resulted in prolonged relapse-free and overall survival.[12] Finally, there are no data from prospective trials supporting additive radiotherapy to sites of residual masses("iceberg radiotherapy"). Whether a differentiation between residual "scarry" disease from remaining vital lymphoma can be made by fluorodeoxyglucose positron-emission tomography (FDG-PET) with clinical relevance (ie, a better outcome if salvage therapy is started immediately) is the scope of ongoing studies. Similar uncertainty exists with respect to additive radiotherapy to extranodal disease, eg, facial bones. In summary, the data supporting the use of radiotherapy in a combinedmodality approach or additive to fulldose chemotherapy have come under scrutiny, and some cooperative groups like the French GELA have totallyeliminated radiotherapy from their therapeutic armamentarium. The combination of cyclophosphamide, doxorubicin, vincristine, and prednisone-or CHOP-published nearly 30 years ago was the first breakthrough in the treatment of what we call today aggressive lymphomas.[13] In several studies, five to nine cycles of CHOP achieved complete remission rates of 50% to 70% and 5-year survival rates of 30% to 50%. In the 1980s, several intensified modifications of CHOP, based on model calculations of Goldie and Coldman and the concept of dose intensity developed by Hryniuk, achieved complete response rates up to 90% and 5-year overall survival rates up to 85% in phase II trials. However, in the pivotal trial by the American Intergroup, the intensified m-BACOD, ProMACE-CytaBOM, and MACOP-B regimens* were not superior to CHOP with respect to complete response rates, event-free, or overall survival,[14] but proved to be more toxic. Retrospective subgroup analyses also did not find an advantage of the primary use of high-dose chemotherapy over conventional therapy in the group of young good-prognosis patients.[15,16] However, the fact that intensified conventional chemotherapy regimens have their role in this group of patients has been shown by the NHL-B1 trial of the DSHNHL. This trial tested, in a 2 * 2 factorial design, whether the addition of etoposide(Drug information on etoposide) to the CHOP regimen (CHOEP-21)[17] and/or the reduction of treatment intervals from 3 to 2 weeks (CHOP-14, CHOEP-14) improves the outcome of young goodprognosis patients (defined as those with normal pretreatment serum lactate dehydrogenase [LDH]). CHOEP was significantly better than CHOP with respect to the primary end point of event-free survival, while the reduction of treatment intervals from 3 to 2 weeks resulted in a significantly better overall survival.[ 17] When the three intensified regimens CHOP-14, CHOEP-21, and CHOEP-14 were compared with the standard CHOP-21 regimen, CHOEP- 21 improved event-free survival while CHOEP-14 improved event-free sur-vival, complete remission rates, and overall survival over baseline CHOP- 21 (Table 1). Therefore, CHOEP-14 is our preferred chemotherapy regimen for young good-prognosis patients. The MInT (MabThera International Trial Group) trial addressed the role of the monoclonal anti-CD20 antibody rituximab(Drug information on rituximab) (Rituxan) in young goodprognosis patients. A total of 824 patients with age adjusted IPI = 0,1 (excluding patients with stage I nonbulky disease) from 18 countries with stages II to IV disease and stage I with bulky disease were randomized into six cycles of a country-specific CHOPlike regimen (CHEMO: CHOP-21, CHOEP-21, MACOP-B, PMitCEBO [prednisone, mitoxantrone(Drug information on mitoxantrone), cyclophosphamide, etoposide, bleomycin, vincristine]) or the same regimen plus rituximab given on day 1 of each chemotherapy cycle (R-CHEMO). Patients with bulky disease received additional radiotherapy to the respective areas. After a median observation time of nearly 2 years, the addition of rituximab increased event-free survival from 61% to 80% (P = .000000007) and overall survival from 86% to 95% (P = .0002) (see Table 2).[18] Conclusions
The results of the MInT trial obtained with a combination of six cycles of CHOP-like chemotherapy with rituximab are superior to any other report published to date for the group of young good-prognosis patients. Therefore, we consider six cycles of a combination of a CHOP-like regimen with rituximab the best therapy for young patients with good-prognosis aggressive lymphoma and consider it the standard and reference approach novel concepts will have to be compared to. Additional radiotherapy (eg, to sites of initial bulky disease or extranodal involvement) should not be given outside clinical trials. Perspectives
A multivariate analysis of the MInT results[18] revealed, in the treatment arm with rituximab, the presence of bulky disease and one risk factor according to the IPI as highly significant prognostic factors. Patients with IPI = 0 and no bulky disease after R-CHEMO represent a very favorable subgroup with a 2-year event-free survival of 90%, while patients with ageadjusted IPI = 1 and/or bulky disease represent a less favorable subgroup with only 77% 2-year event-free survival. The respective figures for 2-year overall survival are 90% and 97%, respectively. Notably, the results in the very favorable group were achieved with six cycles of a CHOPlike chemotherapy with rituximab andwithout radiotherapy, while in the less favorable group bulky disease was a negative prognostic factor despite radiotherapy given to such areas. Thus, in the era of combined rituximab CHOP-like chemotherapy, two novel therapeutic subgroups have evolved for young good-prognosis patients (Table 2). In the very favorable subgroup (IPI = 0, no bulk) further improvement will be difficult to achieve and demonstrate, but the results in the less favorable subgroup definitely warrant further improvement. The results in the very favorable subgroup suggest that a considerable proportion of patients are overtreated. We therefore have designed a trial that compares six cycles of a CHOP-like regimen combinedwith six cycles rituximab with four cycles of the same CHOP-like regimen combined with six applications of rituximab. For the less favorable group, we aim at improving the results further. This will be attempted in a randomized trial that compares six cycles of CHOEP-21 with rituximab with a dose-dense regimen of six cycles of CHOEP-14 with rituximab (Figure 2). Young Poor-Prognosis Patients Background
We subsume young patients with an age-adjusted IPI = 2 or 3 into this group. For these young poor-prognosis patients with aggressive lymphomas CHOP-21 is formally still the standard conventional chemotherapy regimen, because approaches shown to be superior to CHOP-21 in young good-prognosis patients, such asCHOEP-14[17] or the combination of a CHOP-like regimen with rituximab,[ 18] have not been tested in a randomized fashion in young poorprognosis patients with aggressive lymphomas. However, CHOEP-14 is the only chemotherapy regimen that has been shown in a randomized trial to be superior to CHOP-21 and well tolerated in young patients. Moreover, the majority of young high-risk patients present with elevated pretreatment LDH and/or bulky disease, two clinical parameters that profited most from dose densification in the NHL-B1 and NHLB2 trials, respectively.[3,17] We therefore consider CHOEP-14 the best conventional chemotherapy regimen for young poor-prognosis patients. The French ACVBP is similar to CHOEP-14: it represents another dose-dense (biweekly) regimen with possibly even higher dose intensity due to higher doses of some cytotoxic drugs. But the most efficaciousdrugs, anthracycline and cyclophosphamide, are given only four times, resulting in a lower total dose of these drugs than in eight cycles of CHOEP- 14. Whether this is compensated for by the intensive and somewhat lengthy consolidation program that follows ACVBP can only be answered by a prospective randomized trial. In contrast to relapsing aggressive lymphomas, where high-dose chemotherapy with hematopoietic stem cell support was superior to conventional chemotherapy,[19] the results of highdose chemotherapy in the primary treatment of aggressive lymphomas are contradictory. A summary of randomized trials addressing this question is shown in Table 3. Most trials did not find differences between the two treatment strategies, some favored the high-dose approach only in retrospective subgroup analyses, and in one trial the high-dose approach was even inferior to the conventional chemotherapy.[ 15,16,20-27] Several factors may be responsible for the failure of high-dose concepts to improve the outcome of young patients in these trials. First, the cytotoxic drugs used for the myeloablative therapy are not the most active in aggressive lymphomas. Second, in many studies high-dose chemotherapy was given instead of one to three cycles of conventional chemotherapy, but not in addition to a full-cycle conventional chemotherapy, resulting in a compromised total dose. Third, the dose intensity of many high-dose strategies is low due to prolonged treatment intervals. Fourth, up to 40% of the patients randomized to highdose chemotherapy did not receive it, often due to progressive disease before high-dose chemotherapy. In general, the results of these trials are difficult to interpret because the conventional and high-dose arms in these trials differ in more than the dose of chemotherapy. For example, Gianni's approach of "sequential highdose chemotherapy"[21] uses highdose methotrexate(Drug information on methotrexate) and the repeated application of single maximum-dosed cytotoxic drugs. The GOELAMS (Groupe Ouest d'Etude des Léucemies et Autres Maladies du Sang) trial,[27] which com-pared eight cycles of CHOP-21 with two cycles of a variant of biweekly CHOEP, followed by high-dose methotrexate plus cytarabine(Drug information on cytarabine) (Ara-C) and a myeloablative BEAM (carmustine [BiCNU], etoposide, cytarabine [Ara- C], melphalan(Drug information on melphalan) [Alkeran]) regimen with autologous stem cell support in patients with low-intermediate and high-intermediate risk according to the IPI, found a significant advantage for the intensified concept with respect to event-free survival for the entire study population and overall survival after 5 years for the subpopulation with intermediate/high risk only. Despite of the suggestive title of the publication in the New England Journal of Medicine, it is not clear whether the high-dose BEAM or the dose-dense chemotherapy given early in the experimental arm are responsible for the superiority of this approach over eight cycles of CHOP- 21. Notably, patients in the experimental arm of the GOELAMS study had received all their therapy by day 64, when patients in the CHOP-21 arm had received only 50% of their CHOP chemotherapy. Like the dose-dense regimen CHOEP-14, the monoclonal anti- CD20 antibody rituximab has never been tested in a randomized trial in young poor-prognosis patients with aggressive lymphomas, and such trials would be warranted. Indeed, the DSHNHL started to test the role of rituximab in this population in the randomized Mega-CHOEP trial, where both the Mega-CHOEP (see below) arm and the conventionally dosed arm that consists of eight cycles of CHOEP-14 are compared in a 2 * 2 factorial design both with and without rituximab. However, after the results of the MInT trial[18] became available, the study participants decided to stop the rituximab randomization, so that rituximab is now given in both arms. Whether this was a wise decision is questioned by a recent analysis of the phase II Mega-CHOEP trials. This early analysis revealed no significant difference between Mega- CHOEP given with or without rituximab in young poor-prognosis patients (abstract submitted). The overview on primary highdosechemotherapy in aggressive lymphoma confirms that the statement of the Lyon 1997 consensus conference is still valid in 2005: "There is no justification of high-dose chemotherapy in the primary treatment of aggressive lymphomas outside clinical trials."[28] This should hold even more true in the era of combined chemoimmunotherapy with rituximab that "equalizes" differences of efficacy of different chemotherapy regimens (abstract submitted). Nevertheless, we believe that novel approaches to high-dose chemotherapy are still warranted and justified for young poor-prognosis patients. Such novel approaches must be tested in randomized trials that address the question of primary high-dose chemotherapy in a noncontaminated way. A noncontaminated testing of the value of high-dose chemotherapy is possible if the high-dose chemotherapy regimen and the conventional arm use the same cytotoxic drugs, which should include those with the highest efficacy in aggressive lymphomas-in particular alkylating agents and anthracyclines. In the high-dose arm these drugs should be given at the maximum tolerated doses and the high-dose regimen should be given in addition to, and not as a substitute for, a fully dosed conventional chemotherapy. In the Mega-CHOEP trial of the DSHNHL patients are randomized into a first cycle of conventionally escalated CHOEP followed by three cycles of high-dose CHOEP, each necessitating stem cell support. With doses of 280 mg/m² doxorubicin, > 18,000 mg/m² cyclophosphamide, and > 2,000 mg/m² etoposide given at 21-day intervals, the Mega-CHOEP protocol achieves a very high dose intensity for those drugs that have been shown to be the most active in aggressive lymphomas.[29] The true value of this aggressive and toxic approach is currently being tested in a randomized fashion against eight cycles of (dose-dense) CHOEP-14. Conclusions
The currently available data make it evident that in the group of young poor-prognosis patients there is no formal proof that any of the morerecent approaches that have been successful in elderly patients or young good-prognosis patients can also improve the outcome of young poorprognosis patients. In light of the experiences in young low-risk patients, where the interval reduction from 3 to 2 weeks and the addition of etoposide (CHOEP-14) improved complete response, event-free survival, and overall survival rates, and the addition of rituximab further improved the outcome of these patients, it is problematic to randomize young poorprognosis patients into a control arm with CHOP-21. Due to the low toxicity of rituximab and its efficacy in young good-prognosis patients, it is difficult not to give this antibody to young poor-prognosis patients, even though its efficacy has not been demonstrated in this subpopulation and is less pronounced in elderly high-risk than elderly low-risk patients.[30] Because there is no standard therapy for young poor-prognosis patients, we recommend treating all of these patients within randomized prospective trials that compare dose-dense regimens (eg, CHOEP-14) with a maximally escalated high-dose regimen containing the most efficacious drugs for the treatment of aggressive lymphomas (eg, Mega-CHOEP, see Figure 3). For young poor-prognosis patients who are not candidates for or who or are not willing to participate in clinical trials, we believe that a full-cycle dose-dense regimen (eight cycles of CHOEP-14 or four cycles of ABCVP) in combination with rituximab is the best choice. Elderly Patients General Treatment Considerations
The curative intention of treatment is also valid for elderly patients with aggressive lymphomas, although the prognosis worsens with increasing age. Only in cases where a careful examination of the patient and his concomitant diseases indicates unacceptable risks for a full-dose therapy can a palliative treatment approach be justified (Table 4). However, the definitive decision about the treatability of a patient should only be made after a so-called prephase therapy, which according to our suggestions consists of a single injection of 1 mg vincristine and 100 mg prednisone daily for 1 week. This prephase treatment generally leads to a considerable improvement of the patient's performance status and helps in ameliorating the so-called first-cycle effect. The firstcycleeffect is the phenomenon that side effects of a chemotherapy regimen are most pronounced after the first cycle of chemotherapy, which is the reason why most therapy-associated deaths in elderly patients occur after the first cycle. The upper age limit for chemotherapy with curative intention will be defined by the higher comorbidity that is associated with increasing age, rather than by the chronologic age of a patient. As a consequence, a careful and repeated evaluation of preexisting or concomitant diseases evolving under therapy and comorbid conditions is absolutely mandatory in elderly patients (Table 4). Background
The trials of the British National Lymphoma Investigation demonstrated that elderly patients, particularly those > 70 years, cannot be cured by radiotherapy alone,[31] even if they have limited stage I disease. Becausein the GELA trial LNH93-4 patients above 70 years of age with IPI = 0 who received radiotherapy in addition to four cycles of CHOP-21 had a worse overall survival than those who did not,[32] radiotherapy should be used cautiously and not outside clinical trials for elderly patients. As in young patients, six to eight cycles of CHOP given every 3 weeks was the standard chemotherapy regimen for more than 25 years, and was confirmed for the elderly in the randomized Intergroup trial with nearly 900 eligible patients enrolled,[14] nearly half of whom were > 60 years of age. In randomized trials that were particularly designed for elderly patients, a variety of modifications of the CHOP regimen-dose-reduced weekly application of CHOP, 50% dose-reduced first cycle, and dose-modified schedules omitting anthracyclines or substituting doxorubicin with mitoxantrone- came out with disappointing results. Therefore, classic CHOP given every 3 weeks (CHOP-21) remained the gold standard in aggressive lymphoma treatment until the recent publication of the results of three large randomized trials conducted by the German DSHNHL,[3,17] the French GELA,[30] and the American ECOG[33] that had been specifically designed for elderly patients. After the publication of the Intergroup study,[14] the DSHNHL decided to evaluate two different strategies in order to improve the outcome of patients that do not qualify for high-dose chemotherapy. Coapplication of granulocyte colony-stimulating factor (G-CSF, filgrastim(Drug information on filgrastim) [Neupogen]) allows the shortening of treatment intervals of the CHOP regimen from 21 to 14 days (CHOP-14) without adding toxicity, and results in a significant increase of dose density without compromising the total dose given. As an alternative, etoposide, a cytotoxic drug with proven activity in relapsed lymphomas, can be added to CHOP (CHOEP regimen) and results in an escalation of the total dose of cytotoxic drugs. Furthermore, CHOEP-14, in which both strategies (dose escalation and dose densification) are combined, proved to be feasible in a pilot study.[34] In the NHL-B2 trial of the DSHNHL, 689 elderly patients were randomized between 1993 and 2000 to receive six cycles of either a 3-weekly or biweekly CHOP with and without etoposide (CHOP-21, CHOEP-21, CHOP-14, CHOEP-14), respectively. All three intensified regimens (CHOP- 14, CHOEP-21, CHOEP-14) significantly improved outcome compared to CHOP-21 (Table 5), with CHOP- 14 not only inducing the highest remission rate (complete response: 76% vs 60%; P = .001), but also the best 5-year event-free survival (44% vs 33%; P = .003) and 5-year overall survival (53% vs 41%; P < .0001) compared to CHOP-21. In addition, there were no differences in toxicity between CHOP-14 and CHOP-21 with a treatment-related mortality of 2.9% after CHOP-14 vs 3.4% after CHOP-21. These results form the basis to regard CHOP- 14 as the new standard for the treatment of aggressive lymphomas in the elderly in Germany and several other countries. Due to application ofG-CSF in the biweekly regimens the incidence of leukocytopenias was somewhat less after CHOP-14 than after CHOP-21, while CHOEP-21 and CHOEP-14 induced more leukocytopenia and thrombocytopenia as well as nonhematologic toxicities. While the double-intensive CHOEP-14 was well tolerated in the young population of the NHL-B1 trial, it proved to be too toxic in the elderly population of the NHL-B2 trial, where it was associated with frequent treatment delays that resulted in reduced relative dose intensity and had a higher rate of therapy-associated deaths.[3] Based on promising results in a first trial with rituximab as a single agent in aggressive lymphomas, the GELA conducted the randomized LNH98-5 trial[30] in 399 patients with diffuse large B-cell lymphoma. Patients aged 60 to 80 years (median: 69 years) received eight cycles of CHOP- 21 or the same chemotherapy plus eight infusions of rituximab at 375 mg/m². The combination of CHOP and rituximab significantly improvedthe complete response rate from 63% to 76%, event-free survival from 38% to 57%, as well as overall survival from 57% to 70% after 2 years, compared to classical CHOP-21 alone. There were no differences in toxicity, with a treatment-related mortality of 6% in both arms. In the ECOG 4494 trial of 632 elderly patients, six to eight cycles (depending on response to treatment) of CHOP-21 were compared in a randomized fashion to the same chemotherapy plus five infusions of rituximab.[33] Treatment results confirmed the positive results of the GELA trial for the addition of rituximab to CHOP for time to treatment failure and overall survival, even though this could be demonstrated only after the biometrical rescue measure of a so-called weighted analysis. Moreover, interpretation of data from the ECOG trial is difficult due to a second randomization step for 415 patients achieving a remission (complete or partial) after induction therapy who received either rituximab maintenance therapy for 2 years or observation only, respectively. Results from the second randomization suggest that patients that had already receivedrituximab together with CHOP during induction therapy did not profit from rituximab maintenance therapy. Conclusions
As shown by the results of the NHL-B2 trial of the DSHNHL[3] and the GELA LNH-98.5 trial,[30] both dose densification (ie, reduction of treatment intervals from 3 to 2 weeks)[3] and the addition of rituximab to 3-weekly CHOP-21[30] can improve treatment outcome compared to the former standard CHOP-21 to a similar extent (Table 6). Even though there is no direct comparison between CHOP-14 and R-CHOP-21 available yet, one can assume equal efficacy and toxicity for CHOP-14 and R-CHOP-21. With CHOP-14 and R-CHOP-21 being apparently equally well tolerated and efficacious, due to the lower cost of G-CSF (which is mandatory for CHOP-14) compared to eight applications of rituximab in R-CHOP- 21, we consider CHOP-14 the best and most cost-effective treatment for elderly patients with DLBCL. The economic aspect gains additional support from the fact that 30% of the RCHOP- 21 patients in the GELA trial received G-CSF in addition to eight cycles of rituximab. However, because neither the CHOP-14 nor R-CHOP- 21 results are satisfactory, we encouragetreatment of elderly patients within clinical trials in an attempt at further improvement. Perspectives
Of note, both approaches (CHOP- 14 and R-CHOP-21) that proved to be better than the classical CHOP-21 were successful in elderly patients despite or because of the fact that they did not add any toxicity compared to the former standard CHOP- 21. At the same time, the experience with CHOEP-14, which proved to be worse than CHOP-14 in the elderly population of the NHL-B2 trial, taught us that increasing dose intensity (and hence toxicity) beyond a certain point can be counterproductive in elderly patients, pointing to the direction of future strategies for the improvement of treatment results in the elderly. Since both dose densification in the form of CHOP-14 and the addition of rituximab to 3-weekly CHOP- 21 are not more toxic than CHOP-21 alone, one can expect that the combination of both approaches should be feasible without added side effects, representing a logical and attractive next step in search of a more effective treatment for the elderly. As a consequence, the DSHNHL is currently comparing CHOP-14 with CHOP-14 plus rituximab. Moreover, in a 2 * 2 factorial design the question of the optimal number of chemotherapy cycles (6 vs 8) is addressed. A first planned interim analysis of 500 patients treated in this trial (DSHNHL 1999-1 or RICOVER- 60) showed that protocol adherence is excellent with a median relative dose intensity of > 96% for the cytotoxic drugs cyclophosphamide and doxorubicin in all four of the treatment arms (6* CHOP-14, 8* CHOP-14, 6* R-CHOP-14, 8* R-CHOP-14) in patients up to 80 years of age. An efficacy interim analysis of the first 437 patients after a median follow- up of nearly 2 years showed an event-free survival of > 60% and an overall survival of > 75% for the entire study population. Possible differences between treatment arms did not reach the criteria necessary for early termination of the study. This indicates that a positive effect of rituximab in combination with CHOP-14,if it evolves at all, will be considerably smaller than in the GELA trial where rituximab was combined with eight cycles of CHOP-21. Therefore, trial recruitment will be continued until early 2005 with a scheduled number of more than 1,200 patients; first results should be available by the end of 2005.