Non-Hodgkin Lymphoma

Diffuse large B-cell lymphoma

Clinical presentation

DLBCL makes up about one-third of the cases of NHL and is classified as a mature peripheral B-cell neoplasm by WHO. The clinical presentation is variable, but generally patients present with either peripheral lymphadenopathy (neck, axillae) or enlarged nodes in the mediastinum, the mesenteric region, or the retroperitoneum. These sites predict symptoms, which may include chest pain; facial swelling and suffusion of the eyelids (superior vena cava [SVC] syndrome from mediastinal disease); abdominal discomfort, ascites (mesenteric), or back pain; or renal obstruction (retroperitoneal presentations). More than 30% of patients present with disease in extranodal sites, such as the GI tract (including Waldeyer's ring), skin, bone marrow, sinuses, genitourinary tract, thyroid, and CNS. B symptoms, consisting of fever, sweats, and weight loss, are more common in DLBCL than in the indolent lymphomas and occur in about 30% of patients. The median age at presentation is 60 years.

Once the diagnosis is clearly established, staging studies are carried out to determine treatment and define parameters for follow-up. Generally, imaging studies of the chest, abdomen, and pelvis are obtained, and CT scans provide the most accurate anatomic information. Recently, functional imaging using PET scans (which have largely replaced gallium scans) has shown promise as a means of distinguishing between residual scar and active disease after treatment. Further, some investigators have shown that early response by PET scan (after 2 to 3 cycles) is a good prognostic indicator. In addition to CT and PET scans, bone marrow aspirate and biopsy, serum LDH level, and serum beta2-microglobulin level have been described as important predictors of outcome.

Pathology/Immunology

The diagnosis should be made by incisional or excisional biopsy of an available lymph node, with adequate tissue for immunologic studies, such as flow cytometry or immunohistochemistry (IHC), to identify the characteristic B-cell clonality (kappa or lambda restriction). In many cases of DLBCL, CD10 is present, indicating a germinal center origin. The CD20 antigen is present in almost all cases. In addition, markers for bcl-2 and bcl-6 offer prognostic information and are part of most diagnostic evaluations of DLBCL. The use of FNA or core biopsy should be discouraged and is acceptable only when tissue cannot be safely obtained by other means and only if flow cytometry is used to help classify the disease and distinguish it from epithelial malignancies that can masquerade as lymphoma.

Prognostic factors

Clinical predictors of response have been identified and are now widely used to help design therapeutic plans and clinical trials. These predictors include patient age (< or > age 60), performance status (0, 1 vs 2–4), number of extranodal sites (more than two), Ann Arbor stage (I or II vs III or IV), and serum LDH level (> normal; Table 6). Older patients, higher stage, poorer performance status, higher number of extranodal sites, and higher LDH level all predict a worse outcome, and this model has been validated in more than 3,000 patients. These parameters have been called the IPI; this index is used to plan therapy and clinical trials in the United States and abroad and may be used to predict survival (Table 12).

More recently, genomics have been used to help predict outcome based on molecular signature (see subsection on Molecular profiling). This molecular system provides prognostic information independent of the IPI. Several investigators using similar statistical methodologies have yielded comparable results, and recently these analyses have been extended to other lymphomas.

Treatment

Prior to the 1970s, most patients with stage I/II large cell lymphoma (intermediate grade in the Working Formulation) were treated with irradiation alone, with overall cure rates of 40% to 50%. Patients with pathologically favorable stage I/II disease had even better outcomes, but relapse rates, even in these patients, were still 20% to 30%. Pathologic staging, therefore, selected a group suitable for irradiation alone. This approach is no longer appropriate, in view of the success of combined chemotherapy and irradiation in clinically staged patients.

TABLE 12
Outcome according to International Prognostic Index (IPI) factors in 365 DLBCL patients treated with R-CHOP in British Columbia

Coiffier et al found that the addition of rituximab(Drug information on rituximab) improved results in elderly patients with DLBCL, and recent data confirm these observations for younger patients as well. For patients with clinical stage I or II disease (by the Ann Arbor criteria), most studies suggest that chemotherapy (CHOP, and most would add rituximab) for 3 to 4 cycles followed by localized radiation therapy is preferred. Excellent local and systemic tumor control is obtained with combined-modality therapy.

In an ECOG phase III trial, Horning et al showed that 8 cycles of CHOP and irradiation produced a 10-year disease-free survival rate of 57%, compared with 46% with CHOP alone (P = .04). Overall survival was 64% vs 60%, respectively (P = .23), and time to disease progression was 73% vs 63%, respectively (P = .07).

Miller et al showed that CHOP (3 cycles of CHOP and irradiation) produced a progression-free survival at 5 years of 77%, vs 64% for 8 cycles of CHOP alone (P = .03). Overall survival at 5 years was 82% vs 72%, respectively (P = .02). A recent update of this SWOG study was reported by Miller et al, with an 8.2-year median follow-up. The 5-year estimates for CHOP (3 cycles plus irradiation) vs CHOP (for 8 cycles) remained unchanged. Kaplan-Meier estimates, however, now show overlapping curves at 7 years for failure-free survival and 9 years for overall survival. The treatment advantage for CHOP (for 3 cycles plus irradiation) for the first 7 to 9 years was diminished because of excess late relapses and NHL deaths occurring between 5 and 10 years. Patients with good IPI risk factors had a 5-year overall survival of 94%; patients with one adverse risk factor had an overall survival of 70%; those with three adverse risk factors had a 5-year survival of 50%.

These results were confirmed by a single-arm (doxorubicin-containing chemotherapy) approach followed by IFRT conducted by the British Columbia Cancer Agency. However, two reports from European investigators question the value of consolidation irradiation in early-stage disease. These studies did not use rituximab or FDG-PET staging, and details on the irradiation technique used were not available. The necessity of consolidation radiation therapy after complete response to R-CHOP chemotherapy is now being tested in a randomized study in Germany.

Until further studies define the optimal therapy for stages IA to IIA DLBCL (nonbulky), many investigators consider 3 to 4 cycles of R-CHOP and IFRT the initial treatment of choice. For patients with bulky disease, a minimum of 6 cycles of R-CHOP is typically administered. In Irradiation doses of 30 to 36 Gy, delivered in 1.75 to 1.8 Gy over 3 to 4 weeks after completion of systemic therapy, appear to be adequate. Radiation fields usually include involved lymph node sites or an involved extranodal site and its immediate lymph node drainage areas. Furthermore, the disease should be easily encompassed in a radiation field with acceptable toxicity.

Disease site or potential toxicities may influence the treatment plan:

• Lymphomas of the head and neck may be managed with chemotherapy alone to avoid the acute mucositis and long-term xerostomia associated with radiation therapy fields that are large and include both parotid glands. Alternatively, precise radiation therapy techniques can be employed with intentional sparing of salivary glands, using intensity-modulated radiation therapy (IMRT).

• Fully resected gastric or small intestinal lymphoma may be treated with chemotherapy alone. Patients at high risk of perforation or life-threatening hemorrhage may require surgical resection. Alternatively, chemotherapy followed by local irradiation allows gastric preservation and is preferred in most patients.

For patients with more advanced stage (III or IV) disease, CHOP has been the standard (now with rituximab) for 6 to 8 cycles or 2 cycles beyond remission. Recent data suggest an advantage to "dose-dense" therapy, shortening the interval between cycles from 3 to 2 weeks with growth factor support. More data are needed to validate these results. Many studies now suggest an advantage to the addition of immunotherapy in the form of rituximab, and in almost every study, the combination of rituximab and chemotherapy has improved the response rate and disease-free survival. There appears to be no advantage to maintenance therapy with rituximab in this setting, however, as long as rituximab is included in the induction. Responses are seen in upward of 80% of patients, and approximately 50% to 60% achieve a complete remission. It appears that 50% of these patients (30% to 50% overall) are likely cured.

For patients who either do not have a complete remission or who relapse, alternative therapies are possible, but long-term responses have been seen mostly with autologous or allogeneic SCT. Patients who do not have responsive disease prior to SCT generally do poorly. The IPI has been used to predict outcome for transplantation in DLBCL. The role of autologous SCT for high-risk patients remains open to debate. A randomized clinical trial of early vs delayed high-dose therapy for patients with high- and high-intermediate risk diffuse aggressive lymphoma conducted by the US Intergroup is just completing accrual. If this trial confirms the benefit of early SCT in poor-risk patients with chemosensitive diffuse aggressive NHL, subsequent studies will focus on increasing the number of patients who become eligible for transplant consolidation. Investigational treatments include novel antibodies, radioimmunotherapy, and single-agent chemotherapy drugs. Nonmyeloablative SCT is being evaluated in patients with recurrent or refractory disease.

Some investigators believe that irradiation for stages III and IV (advanced or extensive) DLBCL may be added after the completion of definitive chemotherapy if there is localized residual disease, to improve local tumor control. Irradiation may also be delivered after chemotherapy to areas of initially bulky disease, again to enhance local tumor control. These recommendations are based on the observation that when DLBCL relapses after definitive chemotherapy, it usually does so in initially involved or bulky areas of disease. The benefits and potential side effects of irradiation should be weighed against the use of alternative chemotherapy salvage regimens.