Chapter 134:  Therapeutic approaches according to REAL/WHO classification

Treatment of Early Stage Indolent Lymphoma

The majority of patients with indolent NHL present with advanced disease. Only 15% to 30% of patients have clinical stage I/II disease and less than 10% have pathologic stage I/II.¹⁸⁰ In the largest study with an exceedingly long follow-up, of 177 patients from Stanford, 44% had stage I and 56% had stage II disease.²⁴⁷ Patients were treated with either involved-field (IF) or extended-field (EF) radiation, with a limited number receiving total lymphoid irradiation (TLI). Survival rates at 10, 15, and 20 years were 64%, 44%, and 35%, respectively. The relapse-free survival at 10, 15, and 20 years was 44%, 40%, and 37%. The failure-free remission rate was significantly better for patients under age 60 and for patients who received radiation to both sides of the diaphragm, although there was no difference in overall survival.

This analysis has been extended to the outcome after relapse.²⁴⁸ Most patients (76%) had stage I/II disease at relapse. Actuarial survival rates 5, 10, 15, and 20 years after relapse were 56%, 35%, 17%, and 17%, respectively. The progression-free survival rate for the entire group at 5 years was 44% and remained at 22% 10, 15, and 20 years following relapse. Late second tumors were noted in 17% of the patients treated with extensive radiation therapy, whereas the incidence of second solid tumors was 6.8% in patients who received IF or EF radiation. Patients treated with EF generally relapsed in contiguous lymphatic sites, in contrast to patients treated with TLI, who generally relapsed in extralymphatic sites. Several other studies have similarly observed that early stage patients treated with radiotherapy had 10-year relapse-free survival rates of 45% to 60%, with overall 10-year survivals of 60% to 80%.^249,250

In low-stage disease, chemotherapy alone has rarely been used due to the radioresponsiveness of these tumors. Several studies have employed local radiotherapy with adjuvant chemotherapy (CVP), and there appears to be no significant advantage to the combined modality over local radiotherapy.^250–253 Monfardini and colleagues reported a 5-year relapse-free survival of 55% and an overall survival of 62% for patients treated with radiotherapy, whereas patients treated with the combined modality therapy had a 5-year relapse-free survival and an overall survival rate of 63% and 93%, respectively. There was no statistically significant difference between these treatment arms. A retrospective study by McLaughlin and colleagues reported that the relapse-free survival was 64% versus 37% for patients receiving CHOP chemotherapy plus radiotherapy or radiotherapy alone, respectively. Whether these improved results are due to the more aggressive regimen used remains to be determined in a prospective randomized study. The benefit of chemotherapy in stage I/II indolent NHL remains uncertain.

The extranodal MZLs often present with localized disease involving the gastrointestinal tract, salivary glands, thyroid, orbit, conjunctiva, breast, and lung.^59,68 The gastric MZLs are managed differently from those involving other sites. Because many cases of gastric MZL appear to be a B-cell clonal expansion in response to Helicobacter pylori, treatment has been directed at the chronic gastritis. Treatment with antibiotics (metronidazole, amoxicillin, clarithromycin) and, often, a proton pump inhibitor, induces the regression of superficial low-grade MZL of the stomach in over 70% of patients. The long-term remission status of these patients remains unclear. For patients with localized disease who progress after antibiotic therapy or who are H. pylori negative, IF radiotherapy, with or without surgical resection, the disease-free survival at 10 years is over 90%. For other sites of extranodal MZL, because these diseases tend to remain localized for long periods of time prior to systemic spread, surgery remains a highly effective approach, often with adjuvant IF radiotherapy. In a retrospective study of 70 patients with stage IE or IIE MZL, most were treated with IF radiation therapy alone.²⁵⁴ The 5-year disease-free and overall survival rates for the entire group were 76% and 96%, respectively. Patients with stomach and thyroid disease had a 5-year disease-free survival of 93%, whereas disease-free survival rates for other sites of involvement were 69% (p = .006). The response rate of patients receiving chemotherapy with alkylating agents for MZL is about 60%.⁶⁵

Treatment of Advanced-Stage Indolent NHL

The long natural history of indolent NHLs and the lack of symptoms in the majority of patients at diagnosis have fostered close observation as an initial approach to some of these patients.^38,39 Moreover spontaneous remissions of longer than a year have been reported from Stanford in 23% of patients, making treatment unnecessary in this subgroup of patients.^38,255,256 In the study from Stanford, in which patients were randomized either to initial therapy or to deferred treatment until the time of symptoms (usually progressive bulky disease), there was no difference in the 4-year actuarial survival between the two groups. For the entire group of indolent lymphoma patients, the median time until therapy was started was 3 years. However, for the three histologic subtypes, the median time until therapy was required differed: 16.5, 48, and 72 months for follicular lymphoma grade II (follicular mixed small cleaved and large-cell), follicular lymphoma grade I (follicular small cleaved cell), and SLL, respectively.²⁵⁷

In general, systemic chemotherapy has been used for the treatment of advanced-stage indolent NHL. However, fractionated TBI has been employed in the treatment of patients with stage III/IV disease, resulting in a high complete response rate (70% to 85%).^257,258 The relapse-free survival rate has been reported to be approximately 25% at 5 years in studies from both Stanford and the National Cancer Institute (NCI).

Indolent NHLs are very sensitive to both single agents and combination chemotherapy.³⁹ The complete response rates of previously untreated patients to single alkylating agents range between 30% and 60%. However, the median duration of complete remission with either single alkyating agents such as cyclophosphamide or with combinations such as CVP is only about 2.5 years. A study from Stanford that compared daily single alkylating agent treatment, CVP, and TBI in patients with follicular lymphoma grade I/II and SLL has similarly reported that there is no significant difference in relapse-free or overall survival with these different regimens.²⁵⁷ The researchers did note that the median time to achieve complete remission was 12 months, 5 months, and 3 months for single-agent CVP, and TBI, respectively. In general, with either single agents or CVP, only 20% to 25% of patients are disease free at 4 years.²⁵⁹ Following relapse, these diseases continue to be sensitive to single agents and CVP; but the median relapse-free survival progressively decreases with each subsequent relapse.

In an attempt to improve relapse-free and overall survival in patients with indolent NHL, more aggressive combination chemotherapy regimens have been used.^260–263 Regimens such as MOPP; bleomycin, Adriamycin, cyclophosphamide, vincristine, and prednisone (BACOP); moderate dose methotrexate, bleomycin, Adriamycin, cyclophosphamide, Oncovin, and dexamethasone (m-BACOD); CHOP, and BCNU, cyclophosphamide, vincristine, prednisone (BCVP) have been observed to give complete remission rates of 35% to 70%, but the median disease-free survival rate remains similar to that seen with CVP, in the range of 1.5 to 3 years.

It has been reported that the response of follicular lymphoma grade II (follicular mixed small cleaved and large-cell) to aggressive therapy is different from follicular lymphoma grade I. Patients with follicular lymphoma grade II treated with combination chemotherapy (c-MOPP, CHOP-bleomycin) had a high complete remission rate, and significant number of patients (>50%) were still in complete remission 5 to 7 years after treatment.^264–266 However, a prospective randomized trial comparing cyclophosphamide-prednisone, c-MOPP, and BCVP therapy failed to demonstrate significant long-term disease-free survival for patients with follicular lymphoma grade II.²⁶⁷ With the difficulty of reproducible grading of these histologies, it unclear whether patients with this histology have a different outcome than patients with follicular lymphoma grade I.

Several randomized trials have looked at the impact of combination chemotherapy in patients with advanced follicular lymphoma.²⁶⁸ When patients were randomized to cyclophosphamide-prednisone, BCVP, or CVPP trials, the 5-year progression-free survival was greater with CVPP, but no differences in survival were noted. When the role of Adriamycin was examined by comparing COP-Bleo to CHOP-Bleo, no differences were seen. Finally, when the results from the use of CHOP-Bleo were compared with those from cyclophosphamide alone in patients with follicular lymphoma, there was a survival advantage for the patients with follicular lymphoma grade II, although there has not been a recent update of this data.

The rationale for combined modality therapy stems from the observation that systemic extranodal sites of relapse are common following radiation therapy and that radiation therapy provides excellent local control for low-grade lymphomas. Studies from the NCI and Stanford failed to demonstrate an improved disease-free or overall survival for patients treated with CVP±TBI or CVP±TLI, respectively. The study in stage III follicular lymphoma patients from M. D. Anderson who were treated with CHOP-Bleo and IF radiotherapy reported an 81% complete remission, a 75% 5-year survival, and a 52% relapse-free survival rate for the entire patient population.²⁶⁹ A study from the NCI attempted to address two issues: first, the role of aggressive combined modality therapy and second, whether early institution of aggressive therapy affects the natural history of the disease.²⁷⁰ In this study, patients with stage III/IV follicular and SLL were randomized to either no initial therapy or ProMACE-MOPP followed by TLI. Of 104 patients, 14% were not randomized but required initial treatment, with a 71% complete remission rate; however, only 33% were in unmaintained remission, with a median follow-up of 23 months or more. The patients who were randomized to initial aggressive treatment had a high complete remission rate (78%), and 86% were in first complete remission from 1 to 82 + months. In contrast, 44% of the patients initially randomized to observation have now crossed over, with a lower complete remission rate (43%) than those treated initially with combined modality treatment; 71% of those treated remained in complete remission from 4 to 40 months or more. To date, there is no difference in the overall survival between patients treated initially and those treated after a period of observation, with over 75% of patients alive at 5 years.

Interferon-α is an active agent in relapsed indolent NHL (predominantly follicular NHL). Several prospective randomized trials have examined the role of interferon when added to combination chemotherapy in advanced-stage disease patients.^271–275 These studies have generally reported a significant effect on progression-free survival, with only two trials involving maintenence interferon-_ observing a significant prolongation of survival.

The purine analogs 2-chlorodeoxyadenosine (2-CDA) and fludarabine are active agents in indolent NHL.^276–281 In previously treated patients, 2-CDA induced responses in approximately 40% of patients, with about half experiencing complete remission and the other half partial remission; median duration of the response was 5 months. The response rate to fludarabine ranges from 45% to 65% response rate, with less than 20% experiencing complete remission. In previously untreated patients, the response rate to fludarabine was 65% response rate, with 37% experiencing complete remission.²⁸² Fludarabine has been used in combination with other drugs, including cyclophosphamide (100% response rate, 89% complete remission) and mitoxantrone (91% response rate, 43% complete remission, 63% at 2 progression-free survival years) with encouraging results.^283,284 A randomized trial compared eight cycles of either fludarabine or CVP in previously untreated patients with stages III or IV low-grade NHL.²⁸⁵ Median durations of remission (26 months in both) and overall survival at 5 years (68% vs 60%, respectively) were similar in the two treatment arms. A randomized trial compared fludarabine to CVP in previously treated patients with follicular lymphoma. Although the response rates (62% vs 52%) and 2-year overall survival (70% vs 75%) appeared similar, 2-year progression-free survival improved significantly with fludarabine (32% vs 14%).²⁸⁶

Monoclonal antibody therapies have received much attention in the treatment of indolent NHL.²⁸⁷ The “humanized” anti-CD20 MAb rituximab has a 50% to 60% response rate, with 6% complete remission in patients with previously treated follicular lymphoma.²⁸⁸ In SLL, the response rate is significantly lower, approximately 10%. The median duration of response is about 9 months. The major toxicity of rituximab is largely infusion related, with fevers, rigors, and hypotension usually associated with the first infusion. A Phase II trial studied the safety and efficacy of re-treatment with rituximab in patients with relapsed indolent NHL, all of whom had previously responded to rituximab.²⁸⁹ The overall response rate was 40%, with 11% experiencing complete remissions. The estimated median time to progression after treatment was 18 months. Two Phase II trials have employed rituximab as initial therapy in patients with indolent lymphoma with overall response rates of 54% and 73%.^290,291 The role of rituximab as an initial treatment remains uncertain.

To deliver targeted radiotherapy in the treatment of indolent B-cell NHL, several radioimmunoconjugates have been developed. Bexxar (tositumomab) is a murine anti-CD20 MAb conjugated to ¹³¹I. The overall response rate in peviously treated patients was 65%, with 20% complete remissions.²⁹² The median duration of remission had not been reached by 47 months. In a limited study in previously untreated patients with follicular lymphoma who received Bexxar, the overall response and complete remission rates were 97% and 63%, respectively, and there was a 3-year PFS of 68%.²⁹³ The anti-CD20 antibody conjugated to yittrium-90 Zevalin (ibritumomab tiuxetan, IDEC Y2B8) has an 82% response rate with 26% complete remission and a median duration of response of more than 12 months in patients with relapsed disease.²⁹⁴ In a Phase III study of 143 patients with relapsed or refractory low-grade, follicular, or transformed CD20+ NHL, patients were randomized to receive either rituximab or rituximab plus Zevalin.²⁹⁵ The overall response rates to rituximab or the rituximab/Zevalin combination were 56% and 80%, respectively.

Histologic Transformation

The implications of histologic progression of indolent lymphoma on prognosis are controversial.^39–42,296 Armitage and colleagues compared similarly treated patients who had histologic conversion to DLBCL with de novo DLBCL patients and found no instances of prolonged complete remission in the former group and a significantly shorter median survival (12 vs 40 months).⁴⁰ The median survival in a large series of patients with follicular NHL undergoing histologic conversion was 11 months.⁴³ An update from Stanford suggests that patients who were never treated had a better prognosis, as did patients with limited disease.⁴⁴ Although the median survival for the entire group was only 22 months, patients who achieved a complete remission after conversion had an actuarial survival of 75% at 5 years. As detailed below, a subgroup of patients with a history of indolent NHL who transform to a more aggressive histology may be cured by high-dose therapy and autologous bone marrow transplantation (ABMT).^297,298

Treatment of Early-Stage Aggressive Lymphoma

Less than 20% of patients with diffuse large-cell lymphoma have truly localized disease. Diffuse large-cell lymphomas with nonbulky stage I (I or IE) or limited stage II (II or IIE) disease have been treated with radiotherapy alone with variable results.^249,299 Dosages in the range of 4,500 to 5,000 cGy appear to be necessary to maximize local control. Approximately 30% of clinically staged patients experience long term disease-free survival with radiotherapy alone. A number of Phase II studies have demonstrated that patients who received local IF radiation therapy followed by adjuvant chemotherapy did significantly better than patients who received radiation therapy alone.^300–302 To address the role of radiotherapy in localized diffuse aggressive NHL, a randomized trial of patients with localized DLBCL compared eight cycles of CHOP to three cycles of CHOP plus IF radiotherapy.²⁰⁹ Patients treated with three cycles of CHOP plus radiation therapy had a significantly better 5-year progression-free survival and overall survival than patients treated with eight cycles of CHOP (77% vs 64% for PFS, 82% vs 72% for overall survival). Overall life-threatening toxicity and cardiac toxicity were significantly higher in the patients receiving CHOP alone. The benefit of attenuated chemotherapy was largely in patients over the age of 60. Another randomized trial compared eight courses of CHOP with or without IF radiotherapy in patients with previously untreated bulky or extranodal stage I or stage II diffuse aggressive NHL. The 10-year overall survival was similar in the two treatment arms (64% vs 60%).³⁰³

Treatment of Advanced-Stage Aggressive Lymphoma

Based on the success of MOPP in the treatment of Hodgkin's disease, the earliest combination chemotherapeutic regimens developed specifically for large-cell lymphoma were C-MOPP and CHOP (Table 134-8).^304,305 These regimens led to approximately 50% of patients achieving complete remission and 35% enjoying long-term disease-free survival (Table 134-9).³⁰⁵ Additional first-generation regimens were developed, including BACOP and COMLA.^306,307 Although early reports suggested that complete remission rates were higher, most studies have demonstrated no significant improvement of outcome with the use of these regimens over C-MOPP or CHOP (see Table 134-9).

The second generation of intensive regimens (BACOD, m-BACOD, and ProMACE-MOPP) attempted to add noncrossresistant agents to build upon C-MOPP and BACOP.^308–310 A series of more aggressive third-line regimens was tested. All of these regimens (ProMACE-CytaBOM, COP-BLAM-III, MACOP-B) attempted to add more active agents in a shorter period of time.^311–313 These regimens initially reported greater than 80% complete remission rates and greater than 60% long-term disease-free survival (see Table 134-9).

Because of these studies, the Southwest Oncology Group (SWOG) undertook a Phase III study comparing CHOP, m-BACOD, MACOP-B, and ProMACE-CytaBOM.³¹⁴ The completed study of 897 eligible patients with intermediate or high-grade NHL found no difference in outcome at a median follow-up of 31 months for CHOP, m-BACOD, ProMACE-CytaBom, or MACOP-B. At 4 years the patients alive without disease were those treated with CHOP (36.4%), m-BACOD (34.4%), ProMACE-CytaBOM (45.1%), MACOP-B (38.8%). Fatal toxicities were 1%, 5%, 4%, and 6%, respectively.³¹⁴ There have been additional randomized trials, including: comparisons of MACOP-B and ProMACE-MOPP, escalated versus standard doses of doxorubin in the BACOP regimen, and CHOP versus m-BACOD; these have clearly demonstrated that more intensive regimens offer no improvement in remission rate or disease-free or overall survival but have documented increased toxicity.^315–317

In the context of the IPI for diffuse aggressive lymphomas, new therapeutic programs for the treatment of DLBCL have been developed.²⁰⁸ For those patients with the best prognosis, present therapeutic regimens are reasonably effective. In contrast, for those patients who fall into the worst prognostic subgroups, present aggressive combination regimens are inadequate. More aggressive regimens in which the doses of cyclophosphamide and Adriamycin are escalated with hematopoietic growth factor support have shown impressive complete remission rates and durable remissions in Phase I/II studies. However, to date, no survival benefit of a doseescalated regimen has been demonstrated in a randomized study. Similarly, a number of studies have looked at the role of consolidative high-dose therapy and stem cell support for high-risk patients in first remission.

For the myriad of peripheral T-cell lymphomas, similar treatment approaches have been taken for patients with localized and advanced-stage disease. When patients are stratified by the IPI, the failure-free survival and overall survival is generally lower for patients with aggressive T-cell lymphomas than for patients with B-cell disease.

Treatment of Aggressive Lymphoma in Older Patients

In the analysis that led to the development of the IPI, the outcome for patients over age 60 was significantly different from that of those age 60 and less. The clinical characteristics of patients above or below age 60 were similar, and the complete remission rates were similar or slightly lower than those seen in patients 60 or under. The relapse-free and overall survival rates for patients over age 60 in the low and lowintermediate risk groups were significantly less than the rates in patients 60 or under. The reasons for these differences may be related to less intensive treatment as well as the presence of comorbid diseases.³¹⁸ In some studies, increased deaths unrelated to lymphoma or its treatment have been reported, whereas in others, an increased treatment-related mortality has been reported, often related to poor performance status.^319,320

Several attempts have been made to modify the chemotherapy for elderly patients.³²¹ In a randomized trial comparing weekly CHOP with a standard dose, the complete remission rate and PFS were similar, whereas the overall survival was better with standard-dose CHOP.³²² In studies comparing anthracycline-containing regimens and regimens without anthracycline, the myelosuppression has been higher, although treatment-related mortality is not significantly different. More importantly, the complete remission rate, time to treatment failure, and 5-year survival were better for those receiving the anthracycline-containing combination. Substitution of mitoxantrone for Adriamycin has resulted in lower complete remission and 3-year survival rates.^323,324 The group Etudes Lymphomas Adults trial in elderly patients (median age 69, range 60 to 80 years) with previously untreated advanced diffuse large B-cell NHL randomized patients to receive either eight cycles of CHOP or CHOP plus rituximab.³²⁵ There was no significant difference in toxicity between the two arms, and the event-free and overall survival rates at 18 months for the R-CHOP and CHOP arms were 62% versus 43% (p = .0001) and 73% versus 61% (p = .006), respectively.

Follicular Large-Cell Lymphoma

Follicular large-cell lymphoma (follicular lymphoma grade III) is clinically similar to diffuse large-cell lymphoma.^45,46,326 It is generally agreed that all patients should be treated with aggressive combination chemotherapy, which results in complete remission rates of approximately 80% and 50% and overall survival of approximately 70% and 60% for stage I/II and III/IV patients, respectively.⁴⁵ A study of 100 patients with follicular large-cell lymphoma from a single institution supported this strategy, suggesting prolonged disease-free survival and an overall favorable prognosis.⁴⁶

MCL

The median survival of patients with MCL is 3 to 4 years. The treatment of MCL generally involves alkylating agents, in response to which 30% to 50% of patients will have a complete remission, with a median duration of 1 to 3 years.^71,72,327 A randomized trial comparing CVP to CHOP demonstrated no benefit from the addition of doxorubicin in terms of overall survival. Generally, single alkylating agents offer results similar to combination chemotherapy. Newer agents such as fludarabine and rituximab are giving response rates in the range of 30%, with remission duration similar to alkylating agents.^328–330 The results of high-dose therapy and autologous bone marrow transplantation have been generally disappointing in both previously untreated and relapsed patients. The Hyper-CVAD regimen, with escalated doses of cyclophosphamide, high-dose methotrexate, and cytarabine has a very high response rate (38% complete remission).³³¹ In this report, all patients went on to either autologous or allogeneic stem cell transplantation. The results for previously untreated patients were encouraging, with 3-year event-free survival of 72%. A recent study of CHOP plus rituximab showed a high overall response rate of 96% with complete remission/complete remission unconfirmed of 48%. However, the median PFS was 16.6 months, which is similar to that seen with CHOP alone.³³² Further study into new induction regimens as well as the role of allogeneic transplantation may have an impact on the outcome of these patients.

Lymphoblastic, BL, and BLLs

Lymphoblastic lymphoma in adults has been treated with some success, although results are inferior to those seen in children. A 56% disease-free survival at 3 years has been reported in patients treated with CHOP plus high-dose methotrexate, l-asparaginase, and intrathecal methotrexate.³³³ When treated with regimens initially developed for childhood ALL (for example LSA₂L₂), approximately 40% of adults are reported to survive to 5 years. In a randomized study of autologous stem cell transplantation in first remission, there was a trend for improved relapse-free survival but no improvement in overall survival when compared with conventional therapy.³³⁴ Hyper-CVAD has been studied at MD Anderson Cancer Center, with very high complete response rates.³³⁵

BL and BLLs in HIV-1-negative adults have been similarly treated with regimens designed for the pediatric population.^336–339 In selected series, survivals of 50% to 70% have been reported, although for patients with bone marrow or central nervous system disease, the long-term survival ranges from 0% to 30%. An analysis of 65 adults treated on groupe Etudes Lymphomas Adults pediatric protocol reported a complete response rate of 89%, with a 3-year survival of 74%.³³⁹ Using an ALL-like regimen, the Cancer and Acute Leukemia Group B has reported results from 75 adults with small noncleaved cell NHL or the L3 ALL variant.³⁴⁰ Complete remission was attained in 80% of cases, with 52% alive and in continuous complete remission at a median follow-up of 5.1 years. Future prospective randomized trials will assess the efficacy of these regimens in BL and BLLs in adults.

Conventional Salvage Therapy

The vast majority of patients with relapsed or refractory NHL receive limited benefit from conventional salvage regimens. Patients with primary refractory disease who do not achieve a complete remission with initial therapy rarely achieve a complete remission when treated with a non-crossresistant conventional regimen. Following relapse from a first complete remission, although a subset of patients achieves complete remission with standard dose therapy, subsequent remissions are generally not durable, and long-term disease free survivors are rare. In a report of approximately 400 patients with relapsed NHL treated with a variety of salvage regimens, only 3% were in continuous complete remission at 2 years.³⁴¹

In an attempt to improve the disease-free survival rate for patients with relapsed or refractory disease, a number of regimens have been developed with noncrossresistant drugs at higher doses. These regimens have included single agent regimens as well as combinations of cisplatin/carboplatin, etoposide, ifosfamide, and cytosine arabinoside.^343–345 The response rates for these salvage regimens have ranged from 20% to 77%, with upwards of 30% complete remissions reported. The duration of these responses and overall survival of these patients are generally short. Another approach for salvage therapy has been the use of continuous infusions of cyclophosphamide, etoposide, and doxorubicin. The overall responses are in the range of 40% to 80%, with approximately 20% complete remissions.^346,347 Patients who relapse within 12 months of diagnosis of aggressive NHL have a significantly lower response rate to salvage therapy (DHAP) than patients who relapse 12 months or more after diagnosis. (40 vs 69%).³⁴⁸ Although a higher response is seen in indolent NHLs, the response rate of patients with relapsed aggressive NHL to single agent rituximab is 31% (22% partial remission, 9% complete remission).³³⁰

Syngeneic and Allogeneic Bone Marrow Transplantation

Only a limited number of patients with NHL have undergone syngeneic bone marrow transplantation.³⁴⁹ In the largest series (8 patients), 7 achieved complete remission and 4 patients were reported to remain in clinical complete remission from 12 to 126 months after syngeneic bone marrow transplantation.

Allogeneic stem cell transplantation has received increasing attention for patients with relapsed and refractory NHL.^350–352 The majority of patients had aggressive lymphomas, although increasingly, patients with indolent lymphomas are being considered for allogeneic transplant. Nearly all patients had relapsed disease, and many were resistant to conventional dose therapy. The majority of patients treated achieved a complete remission in response to high-dose ablative therapy, and the relapse rate was generally less than 20%. However, the treatment-associated mortality in these studies ranged from 20% to as high as 50%. The probability of long term disease-free survival is between 20% and 40%. In a prospective comparative trial of allogeneic or autologous bone marrow transplantation, there was no significant difference in event-free survival between the two sources of marrow; however, the probability of relapse was significantly greater in the autologous patients.³⁵² Similar results have been seen in patients with indolent NHL. In an effort to reduce the morbidity and mortality associated with graft-versus-host disease (GVHD), several groups have employed T cell depletion for its prevention.^353,354 The incidence of significant (grade 2-4) GVHD in these studies is about 20%, and it is largely grade 2. The treatment-related mortality is less than 20% and the disease-free and overall survival between 50% and 60%. Studies of other approaches to improving the results of allogeneic transplants, such as non-myeloablative conditioning and donor lymphocyte infusions, are ongoing. In one study, 23 heavily pretreated patients underwent matched allogeneic transplantation after low-intensity conditioning with a fludarabine-based treatment program. Transplant-associated mortality was 30%, and estimated disease-free survival at 3 years was 40%.³⁵⁵

Autologous Stem Cell Transplantation for Relapsed Aggressive NHL

The disease sensitivity at the time of autologous stem cell transplantation (ASCT) has remained the most significant prognostic variable for predicting treatment outcome.^356–359 Several large series have shown that patients who undergo ASCT when the disease is resistant to the initial induction therapy have less than 10% probability of disease-free survival. Although many patients die of progressive lymphoma, in some studies the treatment-related mortality has been higher in this patient population (20% to 30%). Those patients in sensitive relapse have a 30% to 60% probability of long-term disease-free survival. In contrast, 10% to 20% of patients with resistant disease are long-term survivors.

The question as to how ABMT compares to conventional salvage therapy for relapsed aggressive NHL was addressed by a multicenter trial known as the PARMA trial.³⁶⁰ Patients with relapsed aggressive NHL (largely DLBCL) received two cycles of cisplatin, ara-C, and solumedrol (DHAP) and, if responsive, were randomized to either continued DHAP for four additional cycles or high-dose chemotherapy BCNV, etoposide, ara-c, cyclophosphamide and ABMT. With median follow-up in excess of 5 years, patients randomized to the high-dose arm had superior event-free survival (46% vs 12%) and overall survival (53% vs 32%). In a subsequent analysis of prognostic factors for the patients in this study, patients with IPI greater than zero at the time of relapse, prior to re-induction with two cycles of DHAP, had a significantly better overall survival when treated with ABMT.³⁶¹ Furthermore, IPI at relapse was highly correlated with overall survival in patients treated with DHAP but not those who underwent ABMT.

Analogous to relapsed highly aggressive lymphomas in the pediatric population, adult lymphoblastic lymphoma has been reported to be a disease in which high-dose therapy has a role. A retrospective analysis of 109 patients undergoing bone marrow transplantation in second complete remission resulted in a 31% rate of actuarial overall survival at 6 years.³⁶² Similar to other lymphomas, patients with resistant disease at the time of transplantation did relatively poorly, with 15% overall survival at 6 years. For patients with BL and BLL lymphoma, adults who underwent ASCT with sensitive disease had 3-year overall survival of 37%, whereas for chemoresistant patients, the overall survival was 7%.³⁶³

The majority of relapses are seen in the first 2 to 3 years after ASCT. Approximately two-thirds of patients relapse in previous sites of disease; nearly one-third relapse in entirely new sites. It is not possible to determine whether relapse results from endogenous tumor cells or re-infused cells contaminating the stem cell product (marrow or peripheral blood stem cells). Because most patients relapse in previous sites of bulk disease, re-infused tumor cells may not contribute to relapse or, alternatively, may selectively localize back to an affected site. The contribution of re-infused lymphoma cells to relapse is supported by gene marking studies as well as studies of minimal residual disease in the stem cell product.^364,365 In studies by Gribben and colleagues after ex vivo marrow treatment, PCR amplification of BCL2 rearrangement detected residual lymphoma cells in the purged marrow in 50% of patients.²⁰⁶ The disease-free survival rate was significantly better in the patients who had no residual detectable bcl-2+ cells in the purged marrow compared with those in whom cells were detectable after purging by PCR.

ASCT for Relapsed Indolent NHL

In contrast to those with relapsed aggressive NHL, fewer patients with indolent NHL have undergone high-dose therapy and ASCT.^366–369 The Dana-Farber Cancer Institute has treated 153 patients with a history of follicular NHL in sensitive relapse or incomplete first remission with cyclophosphamide/TBI conditioning and anti-B-cell MAb-treated ABMT between 1985 and 1995.³⁷⁰ The disease-free survival and overall survival 8 years after ABMT were 42% and 66%, respectively. The survival from diagnosis for the entire group of patients is 69% at 12 years.

St. Bartholomew's Hospital has similarly treated 64 relapsed indolent lymphoma patients with anti-B1 MAb-purged ABMT.³⁶⁹ These patients were treated with the same cyclophosphamide/TBI conditioning regimen. Following high-dose therapy, the treatment-related mortality has been very low, and 35 patients have remained in CCR from 1+ to 8+ years. This study performed a retrospective analysis of patients undergoing ABMT in second remission and compared them to patients treated with conventional therapy. The patients undergoing ABMT had a significantly better disease-free survival compared with those receiving standard therapy. However, there was no difference in overall survival between the two groups of patients.

Because bone marrow involvement is so frequent in these diseases, the number of patients receiving unpurged bone marrow is limited. The University of Nebraska reported 13 patients with indolent follicular lymphoma undergoing ABMT with unmanipulated marrow.^366,371 The 4-year failure-free and overall survival in these patients was 62% and 76%, respectively. The University of Nebraska has reported on 100 patients treated (included were 26 patients treated with one previous regimen), with 4-year failure-free and overall survivals of 44% and 65%, respectively.³⁷¹ In this series, there was no statistically significant difference in the outcome between patients receiving unpurged marrow and those receiving PBSC. Similar results with the use of PBSCs have been reported by Bastion and coworkers.³⁷²

Histologic conversion from indolent to an aggressive histology has been associated with a poor prognosis. There have been several reports of autotransplantation in patients with chemosensitive disease and a good performance status after histologic transformation for follicular NHL. In the series from Dana-Farber Cancer Institute in which 21 patients have undergone anti-B cell-purged ABMT for transformed follicular lymphoma, the Kaplan-Meier estimate of the percentage of patients alive and disease free at 5 years is 46%, with follow-up from 12 to 120+ months.²⁹⁷ When pathology results at relapse were available, it was found that all patients recurred with DLBCL. Similar results were reported by St. Bartholomew's Hospital and the European Bone Marrow Transplant Registry.^298,373 Aggressive therapy with ASCT is a reasonable treatment option for selected patients.

ASCT for NHL in First Remission

In an attempt to improve the prognosis for patients with a poor outlook following induction therapy, several studies have examined the role of high-dose therapy and ASCT in first remission/partial remission for patients with aggressive NHL.^374–376 The disease-free survival at 3 years in most of these studies is 70% to 80%. The European Bone Marrow Transplant Group has reported a multicenter study of patients, stratified by histology, transplanted in first complete remission.³⁴¹ The disease-free survival in these cases is approximately 50% for Burkitt lymphoma, 60% for lymphoblastic lymphoma, and 80% for other aggressive NHLs.³⁷⁷ All of these studies suggest that early high-dose therapy may be highly effective in patients who are at high risk of relapse following conventional therapy. However, randomized trials that have examined the role of ABMT in first remission versus conventional therapy for patients with adverse prognostic factors at diagnosis have left this question largely unanswered. In four randomized trials, a significant difference in event-free survival and freedom from progression in favor of high-dose therapy and stem cell transplantation was seen in only one study.^378–382 However, in a retrospective analysis of one of the randomized trials, in which patients were stratified by IPI at diagnosis, a benefit for consolidative ABMT was seen for high-intermediate and high-risk patients who acheived a complete remission with induction treatment. It remains uncertain whether patients with aggressive NHL who attain only a partial remission with induction therapy have an improved prognosis with high-dose therapy and stem cell support. One randomized trial failed to show a benefit.³⁸³ ASCT is clearly not the complete answer for curing patients with relapsed or high-risk lymphoma. Development of new strategies, particularly toward resistant cells and minimal residual disease, may make an additional impact after high-dose therapy.