Early stage head and neck non-Hodgkin's lymphoma. The effect of tumor burden on prognosis

Non-Hodgkin's Lymphomas of the Tonsil: A Retrospective Analysis of Twenty-Eight Patients with Primary Tonsillary Lymphoma

Aims

To analyze the clinical and therapeutic aspects of patients with primary tonsillary non-Hodgkin's lymphoma.

Methods

Twenty-eight patients with primary tonsillary non-Hodgkin's lymphoma who had been followed in the Hacettepe Oncology Institute between 1974 and 1992 were retrospectively analyzed. Fifteen patients were male, 13 were female. Median age was 55 years.

Results

Constitutional symptoms were present in 10 patients (35.7%). Stages according to the Ann Arbor classification were I and II in 12 and 16 patients, respectively. According to the Rappaport classification, poorly differentiated lymphocytic was the most common pathologic subgroup (42.9%). Grades according to the Working Formulation were low, intermediate and high in 3, 22 and 3 patients, respectively. Twenty-two patients had received chemotherapy. Cyclophosphamide, vincristine and prednisone (CVP), and cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP) were the regimens most commonly employed. Eighteen patients received radiotherapy to Waldeyer's ring and neck. Eight patients achieved remission with chemotherapy plus radiotherapy, 7 patients with chemotherapy alone, and 5 patients with radiotherapy alone. In addition to the 20 patients who achieved complete remission, 3 patients achieved partial remission; the overall response rate was 82.1%. The response rates and survival attained with the combined modality, chemotherapy, or radiotherapy alone were not statistically different (P > 0.05). The median follow-up was 14 months. Overall and disease-free survival at 5 years were 62.6% and 77.6%, respectively. Pathologic grade was the most important prognostic factor influencing overall survival in the Cox multivariate model.

Conclusions

Poorly differentiated lymphocytic lymphomas were the most common pathologic subtype, and pathologic grade was the most important prognostic factor to influence survival in the present study. Although combined modality treatment did not appear to be superior to chemotherapy or radiotherapy alone, a larger number of patients is needed to draw definite conclusions.

Acute 'pathological haematoma' without significant antecedent trauma as an unusual presentation of undiagnosed malignancy

The authors report two cases of malignancy that presented initially with acute haematomas without any history of significant trauma. The first case was a 31-year-old male who presented with a large haematoma in the anterior triangle of the neck following minor trauma during a rugby match. This was shown to be due to haemorrhage into an undiagnosed papillary thyroid tumour. The second case was a 41-year-old male who developed a spontaneous sternocleidomastoid haematoma after laying flagstones and without any history of direct trauma. This was due to haemorrhage into a nodal deposit of non-Hodgkin's lymphoma. These cases highlight that sudden onset haematomas without obvious explanation may be the result of underlying malignancy. In such instances further investigation must be considered and re examination of the patient is essential after the haematoma has resolved.

Marginal Zone B-Cell Lymphoma Among Primary B-Cell Lymphoma of Waldeyer's Ring: Histopathologic and Immunohistochemical Study of 16 Tonsillectomy Specimens

Two subtypes of marginal zone B-cell lymphoma (eg, mucosa-associated lymphoid tissue [MALT] type and splenic type) have been reported in the lymph node. To determine the presence or absence of marginal zone B-cell lymphoma of MALT type and the splenic type among Waldeyer's ring (WR) lymphomas, 16 tonsillectomy specimens were studied. Ten cases (63%) were marginal zone B-cell lymphoma. Among marginal zone B-cell lymphoma, 7 were the MALT type and the remaining 3 cases of marginal zone B-cell lymphoma were the splenic type. Moreover, 4 cases of 7 MALT-type lymphomas contained numerous large cells (diffuse large B-cell lymphoma arising from a low-grade marginal zone B-cell lymphoma of MALT type). The low incidence of primary mucosa-associated lymphoid tissue type lymphoma of WR in previous reports may be because it is difficult to correctly identify the characteristic histologic findings of MALT-type lymphoma because of the small biopsy size.

Lymphomas of the head and neck in the pediatric population

OBJECTIVE

Little is known about the incidence and survival outcomes of pediatric patients with head and neck (HN) lymphomas in Asian populations. This study sought to identify the incidence of HN involvement of pediatric lymphomas and to identify factors prognostic of patient survival.

METHODS

We reviewed the medical records of all children aged 0-14 years with previously untreated lymphomas of HN region and compared patient clinicopathologic characteristics and final outcomes in patients with Hodgkin's disease (HD) and non-Hodgkin's lymphoma (NHL). Potential prognostic factors for patient survival were investigated.

RESULTS

Of 106 eligible lymphoma patients, 45 (42.5%; 31 boys and 14 girls) showed HN involvement. Overall, NHL (n=37) showed more unusual and aggressive presentations than did HD (n=8) in the head and neck region. Cervical lymphadenopathy was frequently observed (30/45, 66.7%) in these patients. Involvement of extralymphatic head and neck sites was found in 15 of 37 NHL patients (40.5%) but not in any HD patients (p=0.027). Five-year disease-free survival (DFS) of all HN lymphoma patients was 76.0%. On multivariate analysis, advanced stage and absence of complete remission following 3 cycles of chemotherapy were poor prognostic indicators of patient survival (p<0.05).

CONCLUSIONS

The incidence of HN involvement in pediatric lymphomas was 42.5% in the studied population. Stage of the lesion and early response to chemotherapy were independent factors prognostic of patient survival.

Clinical relevance of three subtypes of primary sinonasal lymphoma characterized by immunophenotypic analysis

BACKGROUND

The purpose of this study was to investigate the clinical relevance of subtypes categorized by immunophenotypic analysis in primary sinonasal lymphomas.

METHODS

Eighty patients with localized non-Hodgkin's lymphoma involving the nasal cavity and/or paranasal sinuses were divided into three subtypes on the basis of their immunohistochemical findings: (A) B-cell lymphoma (n = 19), (B) T-cell lymphoma (n = 27), and (C) natural killer (NK)/T-cell lymphoma (n = 34). The clinicopathologic profiles, immunophenotypic data, patterns of treatment failure, and survival data among the three patient groups were retrospectively compared.

RESULTS

The nasal cavity was the predominant site of involvement in T-cell and NK/T-cell lymphoma, whereas sinus involvement without nasal disease was common in B-cell lymphoma. Systemic B symptoms were frequently observed in NK/T-cell lymphoma. Almost all patients with NK/T-cell lymphoma showed a strong association with the Epstein-Barr virus by in situ hybridization studies. Sixty-five patients (81%) patients achieved complete remission after initial treatment, but 36 (55%) of these subsequently experienced treatment failure. Although there were no significant differences in locoregional failure rates among the patients of the three groups, distant failure was far more common in B-cell or NK/T-cell lymphoma than in T-cell lymphoma (p =.005). Most B-cell lymphoma cases showed a predilection for sites of systemic failure in the nodal and extranodal sites below the diaphragm, such as the paraaortic lymph nodes or the gastrointestinal (GI) tract, whereas patients with NK/T-cell lymphoma showed an increased risk of systemic dissemination to the skin, testes, or GI tract, including the development of hemophagocytic syndrome. The 5-year actuarial and disease-free survival rates for all patients were 57% and 51%, respectively. Of the three subtypes of primary sinonasal lymphomas, T-cell lymphoma seemed to carry the most favorable prognosis and NK/T-cell lymphoma the worst. (The 5-year actuarial survival rate was 57% for B-cell lymphoma, 80% for T-cell lymphoma, 37% for NK/T-cell lymphoma; p =.02, log-rank.) By univariate and multivariate analyses, immunophenotype was identified as the most important prognostic factor.

CONCLUSIONS

Our data indicate that the three subtypes of primary sinonasal lymphomas classified by immunohistochemical studies exhibit different clinical profiles, different patterns of failure, and different treatment outcomes. Given these observations, it is concluded that the recognition of these distinct subsets, diagnosed on the basis of immunophenotypic study, is very important and clinically relevant in predicting their potential behavior and prognosis.

Lymphoma of the nasal cavity and paranasal sinuses: treatment and outcome of early-stage disease

The records of 23 patients diagnosed and treated at the Massachusetts General Hospital for extranodal non-Hodgkin's lymphoma of the paranasal sinus and nasal cavity were reviewed. The majority of patients were Ann Arbor stage I and approximately evenly divided in T1 or T2 (n = 10) and T3 or T4 (n = 13). Eight patients had nasal-type NK/T cell and 15 patients had diffuse large B-cell lymphoma (DLBCL). The patients with nasal-type NK/T cell lymphoma predominately involved the nasal cavity (5/8), whereas the DLBCL more often had the paranasal sinuses as the primary site (12/15). All patients received radiation as part of their treatment. Only three patients received chemotherapy as part of their initial treatment for three cycles using a cyclophosphamide, doxorubicin, vincristine, and prednisone-based regimen. By coincidence, the estimated overall survival (OS) and disease-free survival rates for both 5 and 10 years were all the same for all analyses. The OS for the entire group at 10 years was 78%. Significant prognostic factors were Ann Arbor stage IEA versus IIEA ( p = 0.0001) and T stage with (T1 or T2) versus (T3 or T4) (p = 0.0243). Combining Ann Arbor stage and T stage created a highly significant prognostic variable (IEA & [T1 or T2], IEA & [T3 or T4], IIEA & [T1 or T2], IIEA & [T3 or T4]) at p = 0.0001, regardless of site or histology. Patients with local-regional disease appear to be well controlled with radiation alone, but distant failure remains a problem. A combined-modality approach with local-regional radiation and systemic chemotherapy is recommended for these patients.

Marginal zone B cell lymphomas of Waldeyer's ring--a report of two tonsillectomy cases resembling histomorphological features of inflammatory lesions

We identified two cases of lymphoma of the mucosa-associated lymphoid tissue (MALT) type in 11 tonsillectomy specimens of primary B cell lymphoma of Waldeyer's ring. Both patients were Japanese females presenting with bilateral enlargement of the palatine tonsils. One had a history of chronic otitis media. In that case, the lesion was characterized by an extrafollicular growth pattern with marginal zone-like arrangement. The tumor was mainly composed of medium-sized cells with round or indented nuclei with scant cytoplasm (centrocyte-like cells). In the other patient, most tumor cells were mature plasma cells, plasmacytoid cells, proplasmacytes, and immunoblasts with scattered centrocyte-like cells. Tropism of tumor cells for the epithelium was noted in both lesions. Primary marginal zone B cell lymphoma of the MALT type arising from Waldeyer's ring has rarely been reported in the literature, causing certain diagnostic problems. Various florid reactive lymphoproliferative disorders, including chronic tonsillitis and infectious mononucleosis, should be differentiated from this type of primary Waldeyer's ring lymphoma.

Neoplastic masquerade syndromes

Masquerade syndromes are classically defined as entities which emulate inflammatory conditions but which are in fact due to a neoplastic process. Careful history and examination in concert with appropriate ancillary investigations and histopathologic evaluation of tissue specimens are required in order to make the correct diagnosis. Many conditions may result in an appearance mimicking an inflammatory condition. The authors review neoplastic conditions which may be considered masquerades. The most common of these is primary intraocular lymphoma or primary central nervous system lymphoma, occurring predominately in older individuals. Diagnostic strategies, therapy, and prognosis are reviewed in detail. Other conditions that can be considered masquerade syndromes are reviewed as well, including lymphomatous and nonlymphomatous conditions, such as melanoma, retinoblastoma, juvenile xanthogranuloma, metastatic lesions, and paraneoplastic syndromes, among others.

Radiotherapy in the treatment of primary central nervous system lymphoma (PCNSL)

The use of radiotherapy alone to treat primary central nervous system lymphoma (PCNSL) does not produce the high local control and survival rates that it does in limited extranodal non-Hodgkin's lymphoma outside the central nervous system (CNS). Even with doses of whole brain radiation therapy (WBRT) to 40+20 Gy boost, the Radiation Therapy Oncology Group (RTOG) reported a local control rate of 39%. Seventy-nine percent of recurrences were in the 60 Gy region. The median survival was 11.6 months. This response to local radiotherapy is quite different from the response of non-CNS Diffuse Large Cell Lymphoma where doses of 30-40 and >40 Gy have a 75-90% local control rate. Neither systemic lymphoma nor PCNSL have a classic radiotherapy dose response. For PCNSL there appears to be a threshold dose that ranges in the literature between 30 and > 50 Gy with a median of 40 Gy. Therefore, when radiotherapy is combined with chemotherapy that crosses the BBB, WBRT and/or boost doses may be able to be decreased, especially in patients achieving a complete response. Promising data from the Centre Leon Berard suggest that this is possible. When such chemotherapy was combined with intrathecal chemotherapy and 20 Gy WBRT, they obtained a 56% actuarial 5 year survival rate. Confirmation of single institution reports of favorable results such as these are needed. Cooperative group and intergroup trials are needed to define optimal therapy.

Surgery plus chemotherapy or chemotherapy alone for primary intermediate- and high-grade gastric non-Hodgkin's lymphoma: the Royal Marsden Hospital experience

Primary gastric lymphomas (PGL) have traditionally been treated with surgery followed by chemotherapy or radiotherapy. Surgery was thought to improve staging, optimise local disease control and reduce risk of perforation or bleeding, but recent studies question its role. In this study, patients with intermediate- or high-grade PGL who received chemotherapy from 1985 to 1996 at the Royal Marsden Hospital were identified using a prospectively accrued database. A total of 37 patients (6 with low-grade mucosa-associated lymphoid tissue lymphoma (MALT-L), 9 with high-grade MALT-L, 20 with diffuse large B-cell (DLBC) lymphoma and 2 other histologies), 17 of whom had localised disease, were treated with either surgery plus chemotherapy or chemotherapy alone. 5-year overall survival for localised and advanced PGL was 94 and 50%, respectively, with no differences between the two treatments over a 53 months median follow-up. No perforations or serious bleeding occurred. Surgery is associated with important morbidity and we detected no benefit of surgery prior to chemotherapy in this limited series of patients.

Combined therapy for angioimmunoproliferative lesions

43 patients with a diagnosis of angioimmunoproliferative lesions (AIL) entered onto a prospective clinical trial to evaluate the use of combined therapy as a primary therapeutic approach. Patients were treated initially with involved field radiotherapy 40-55 Gy (40 patients received 45 Gy) followed by six cycles of chemotherapy which consisted of CEOP-Bleo (cyclophosphamide, epirubin, vincristine, prednisone and bleomycin). Complete response was achieved in 41 cases (95%). At a median follow-up of 40 months, 40 patients (91%) remain in first complete remission. 2 patients died during radiotherapy secondary to sepsis and tumour progression. Treatment was well tolerated. The treatment of AIL remains controversial. Our results show that combined therapy appears to be the best therapeutic approach in patients with this type of malignant lymphoma. More studies are necessary to define the role of combined therapy in patients with AIL.

Treatment of localized non-Hodgkin's lymphomas of the head and neck

BACKGROUND

Localized non-Hodgkin's lymphomas of the head and neck are generally treated with radiotherapy with or without chemotherapy, although the results of treatment of localized non-Hodgkin's lymphomas with of treatment of localized non-Hodgkin's lymphomas with chemotherapy alone appear to be favorable. It is unclear if and when combined modality therapy should be used.

METHODS

The authors reviewed the records of 53 patients with Stage I or II non-Hodgkin's lymphoma of the head and neck, who were treated with radiotherapy alone (13 patients), chemotherapy according to the cyclophosphamide, doxorubicin, vincristine, prednisone- (CHOP) regimen (27 patients), or a combination of both treatments (13 patients).

RESULTS

A complete remission was achieved in 43 (81%) patients. The 5-year survival for all patients was 78%. A significant difference (P = 0.03) in 5-year relapse-free survival was observed between Stages I and II disease, of 92 and 60%, respectively. Extensive tumor was a significantly poor prognostic factor (P = 0.04) with a 5-year relapse-free survival of 52 versus 84% for patients with nonextensive lymphoma. Eight relapses occurred; in five patients, a local relapse was the first presentation. Although salvage radiotherapy was successful in these five patients, a distant relapse developed in three. No relapses were observed in previously irradiated areas.

CONCLUSIONS

Our results suggest that radiotherapy alone is the appropriate treatment for nonextensive Stage I intermediate grade non-Hodgkin's lymphoma of the head and neck. For extensive Stage I or II non-Hodgkin's lymphomas, chemotherapy is preferable. The value of combined modality therapy remains unclear.

Malignant lymphoma of Waldeyer's ring. A histological and immunohistochemical study

The histopathological and immunohistological features of non-Hodgkin's lymphoma limited to the Waldeyer's ring were studied in 22 Japanese patients using a panel of T- and B-cell markers on paraffin-embedded sections. All cases showed a diffuse growth pattern. Twenty cases were B-cell lymphomas and two were T-cell lymphomas. In contrast to the primary malignant lymphomas of the nasal cavity and paranasal sinuses, in which T-cell neoplasms are more frequently seen, the majority of the primary Waldeyer's ring lymphomas were B-cell neoplasms. Sixteen of the 20 cases of B-cell lymphoma were centroblastic lymphomas, and the monomorphic variant comprised the majority of these; the other three B-cell lymphomas were immunocytomas. Two of the T-cell lymphomas showed morphological features of angiocentric lymphomas.

Non-Hodgkin's lymphoma of the brain: can high dose, large volume radiation therapy improve survival? Report on a prospective trial by the Radiation Therapy Oncology Group (RTOG): RTOG 8315

Between 1983 and 1987 the Radiation Therapy Oncology Group conducted a prospective phase II study to evaluate survival in primary non-Hodgkin's lymphoma of the brain treated with whole brain irradiation to 40 Gy and a 20 Gy boost to tumor plus a 2 cm margin. Forty-one patients are reported. Full follow-up is available on 35/41 who have died. Six are alive at 8.8-67.2 months from start of radiation therapy with a median followup of 53.9 months. Overall median survival was 11.6 months from start of radiation therapy and 12.2 months from diagnosis, with 48% surviving 1 year and 28% surviving 2 years. Karnofsky Performance Status and age were significant prognostic factors. Patients with a Karnofsky Performance Status of 70-100 had a median survival of 21.1 months compared to 5.6 months for patients with a status of 40-60 (p less than .001). Fourteen patients less than 60 years of age had a median survival of 23.1 months, while 27 patients greater than or equal to 60 years of age had a median survival of 7.6 months (log-rank p = .001). Disease recurred in the brain in 25/41 (61%) of the patients, (21/41 in the brain only and 4/41 in the brain plus distant metastases). Despite high dose and large volume irradiation, primary Central Nervous System lymphoma still exhibits excessive mortality, especially in older patients. This paradox of the relative radioresistance of primary Central Nervous System lymphoma remains unresolved.

Prognostic factors in patients with early stage non-Hodgkin's lymphomas of the head and neck treated with definitive irradiation

Between 1974 and 1989, 58 patients with clinical Stages I and II non-Hodgkin's lymphomas of the head and neck were treated with radiation at the Fox Chase Cancer Center. Forty-one treated with radiotherapy alone form the basis for this retrospective analysis of outcome and prognostic factors. With a mean radiation dose of 4400 cGy, the 5-year actuarial local control rate is 92%. Only one patient failed within an irradiated field. The 5-year actuarial survival and relapse-free survival rates are 85% and 54%, respectively. In a univariate analysis, poor survival was significantly correlated with involvement of Waldeyer's ring, postoperative tumor size greater than 3 cm, and greater than two involved lymph nodes and extranodal sites (p less than 0.02). No such correlations were seen for stage, histologic grade, the presence of extranodal disease, or any of the other parameters that were examined. Relapse free survival was significantly correlated only with the total of the number of involved nodes and extranodal sites. Patients with one or two involved nodes and sites had a 68% chance of remaining disease-free at 5 years compared to 0% for patients with greater than two (p = .02). Again, significant trends were not seen for the other parameters analyzed. These data demonstrate excellent local control, survival, and relapse-free survival using radiation alone with doses of 3000-5000 cGy. In our group of clinically staged patients preselected for treatment with radiation alone, the total of the number of involved nodes and extranodal sites, involvement of Waldeyer's ring, and tumor size after resection correlated strongly with relapse-free survival and overall survival. In patients with early stage non-Hodgkin's lymphomas of the head and neck, initial management with external beam radiotherapy should be considered in particular for those with one or two involved nodes and extranodal sites that are less than 3 cm following resection and that do not involve Waldeyer's ring.

Primary lymphoma of the mandible: report of a case

A case involving primary lymphoma of the mandible is presented. The difficulties in diagnosis of this particular pathology are highlighted. The features of primary lymphoma of the mandible are described based on a review of the literature. Malignant lymphoma must be considered in the differential diagnosis of unexplained dental pain, swelling, ulceration, and radiographic evidence of rarefaction.

Primary nasal-paranasal oropharyngeal lymphoma in the pediatric age group

Nasal-paranasal oropharyngeal (NPOP) non-Hodgkin's lymphoma (NHL) is a disease of the very young (median age, 5 years) and of the aging adult (median age, 50-60 years). Of a total of 208 pediatric patients with NHL studied, 20 (9.6%) had primary NPOP. Sixty percent of the patients had Stage I and II disease. Primary sites were maxillary sinus in eight patients; tonsils in eight; posterior pharynx in two; mandible in one; and orbit in one patient. Histologically, the disease is different than that of the adults since most patients had B-cell lymphomas of the diffuse undifferentiated type (Rappaport) or small cell non-cleaved types (Lukes-Collins, Kiel, and Working Formulation). None of these patients had gastrointestinal involvement. All patients were treated with the LSA2-L2 regimen and radiation therapy was given to primary unresectable tumors and regional metastases. The lymphoma event-free survival was 75%, with a median observation period of 99+ months. In staging systems that refer mostly to amount of disease outside of the primary (such as ours, Murphy's, and the Ann Arbor staging systems) stage did not correlate well with disease-free survival. In the TNM staging of 1977, a staging system that refers to size of primary tumor as well as regional and systemic disease, stage correlated better with prognosis and survival. In our staging system, eight of 12 patients (66.7%) with Stage I and II disease; four of four with Stage III; two of two with Stage IVA; and zero of two with Stage IVB survived. In the TNM staging system, three of three patients with Stage II and III disease and 12 of 18 patients (67%) with Stage IV disease survived. All recurrences occurred early suggesting that early intensification of chemotherapy may produce better results.

Prognostic factors in head and neck non-Hodgkin's lymphoma with special reference to serum lactic dehydrogenase and serum copper

An analysis of prognostic variables was performed in a retrospective study of 121 patients with Ann Arbor stage I-II head and neck non-Hodgkin's lymphoma admitted from 1973 to 1988. The overall actuarial 5-year survival rate was 58.8% and the minimum follow-up 15 months. Nine clinical and laboratory parameters were studied, including serum lactic dehydrogenase (LDH) and serum copper (SCL), and subjected to univariate and multivariate analyses. In univariate analysis, histology and LDH were found to be significant prognostic variables. Evaluation by Cox's multivariate proportional hazard model revealed histology, SCL and sex to be of prognostic significance.

Prognostic factors of stage I and II non-Hodgkin's lymphomas of the head and neck: the value of the working formulation and need for chemotherapy

From 1966 through 1984, 142 patients with early stage non-Hodgkin's lymphoma of the head and neck were treated. Histologic slides were reviewed and reclassified according to Rappaport's classification and the Working Formulation. Patients were clinically staged; 82 Stage I, and 60 Stage II. Five-year freedom from relapse was 53% for Stage I and 48% for Stage II patients with no statistically significant difference. No patients with favorable histologies of Rappaport's classification or low grade malignancy of the Working Formulation relapsed. By univariate analyses, stage, primary site, the Working Formulation, Rappaport's classification, extent of radiation field, radiation dose, tumor bulkiness and addition of multiple-agent chemotherapy seemed to be prognostic factors for predicting relapse. Multivariate regression analyses (MVA) showed primary site and multiple-agent chemotherapy were independent prognostic factors. Tumor bulkiness provided marginal prognostic significance. Waldeyer's ring lymphomas fared better than nodal or extralymphatic lymphomas; it seems necessary to report Waldeyer's ring lymphomas independently from nodal or extralymphatic lymphomas. Patients receiving more than 3 chemotherapeutic agents had better prognosis than those receiving 3 or less agents, or no chemotherapy. Pathologic grade was another prognostic factor when the Working Formulation was used in MVA. When Rappaport's classification was used, division of unfavorable histologies into histiocytic and non-histiocytic groups provided only marginal significance in MVA. We conclude that Stage I-II non-Hodgkin's lymphomas of the head and neck with favorable histologies of Rappaport's classification or low grade malignancy of the Working Formulation be treated with radiation therapy only; and patients with other histologies should be treated with multiple-agent chemotherapy and radiation therapy.

Immunohistological characterization of malignant lymphomas of the Waldeyer's ring other than the nasopharynx

Among extranodal lymphomas, the Waldeyer's ring is the second most frequently involved site after the gastrointestinal tract. Fresh tissue from 23 consecutive cases of malignant lymphoma of the faucial tonsil, palate and base of tongue were studied histologically and with a panel of 25 monoclonal antibodies. Twenty cases were primary Waldeyer's ring lymphoma, and all were found to express B-cell phenotype. Most cases were classified as diffuse centroblastic lymphoma, polymorphic subtype, in which there were immunoblast-like, centrocyte-like and/or multilobated centroblasts. All except one case expressed all three B-cell lineage antigens CD19, CD20 and CD22, but they showed inconsistent expression of the B-cell antigens CD9 and CD24. Four cases lacked surface immunoglobulin. Six cases expressed interleukin-2 receptor, suggesting that they were composed of highly activated B-cells. Three cases represented relapse in the tonsil or tongue in patients with known malignant lymphoma in other sites; one case expressed T-cell and two cases B-cell phenotype (both of which also expressed interleukin-2 receptor). The clinical features and immunohistological findings suggest that Waldeyer's ring lymphomas, other than those of the nasopharynx, share some of the characteristics of 'mucosa-associated lymphoid tissue' lymphomas. In contrast, nasopharyngeal lymphomas are more related to nasal lymphomas, and are almost exclusively peripheral T-cell neoplasms.