You are here
Home > Blog > Internal Medicine > Sinonasal B‐cell lymphomas: A Cohort Study on Progression and Recurrence

Sinonasal B‐cell lymphomas: A Cohort Study on Progression and Recurrence

Welcome to NewMedicalTerms, your ultimate guide to the explanations and definitions of the majority of medical terms you may see.

Feel free to use our complete database with charts, tables and examples.

Sinonasal B‐cell lymphomas: A Cohort Study on Progression and Recurrence


Sinonasal lymphoma is a rare malignancy of rhinologic origin affecting the sinonasal region and surrounding structures. It is a type of non-Hodgkin lymphoma (NHL) comprising around 1.5% of all lymphomas and usually involves males aged between 40 and 70 years. Literature reviews show that about 42 – 94% of all sinonasal B-cell lymphomas (SNBCL) in Caucasian populations are the subtype diffuse large B-cell lymphoma (DLBCL). DLBCL is aggressive cancer where patients most often present with rapidly growing neoplasms in nodal or extranodal sites.

The B-cell malignant non-Hodgkin’s lymphoma of the nasal cavity has rarely been reported in the previous literature. This nationwide cohort study aims to determine the subtype distribution, clinical profile, disease behavior, and prognosis of sinonasal B-cell lymphomas. It also emphasizes the prognosis and recurrence pattern of primary sinonasal diffuse large B-cell lymphoma (PSDLBCL).


Study design, participants, data, and setting

The study included all adult patients who were ≥ 18 years of age with a registered diagnosis of B-cell lymphoma in the nasal cavity or front, maxillary, ethmoid, and sphenoid sinuses. Patients were identified from the Danish Registry of Pathology from 1980 to 2018 using Systemized Nomenclature of Medicine codes. Patient data were retrieved from National archives and electronic healthcare records. The Danish National Lymphoma Registry (LYFO) and the Cause of Death Register were used to identify missing data.

Formalin‐fixed, paraffin‐embedded samples were reviewed using 2017 WHO classification. The immunophenotype and cytogenetics of DLBCL, high‐grade B‐cell lymphoma (HGBCL) with or without double‐hit, or Burkitt lymphoma (BL) were validated using CD3, CD20, BCL2, BCL6, and cMYC with the addition of Epstein‐Barr virus‐encoded RNA in cases of BL. Samples positive for MYC rearrangement were further analyzed for BCL2 and BCL6 rearrangements. All samples of plasmacytoma were collected and validated.

Patient and lymphoma-related characteristics including age at diagnosis, sex, symptoms, serum lactate dehydrogenase levels, location of relapse and dissemination, treatment and response, cause of death, and systemic involvement were collected. Age-adjusted International Prognostic Index was calculated after extracting WHO/ECOG performance status. Stage, extension, and laterality outcomes were determined using clinical data, radiological reports, or diagnostic imaging.

Overall survival (OS), Progression-Free Survival (PFS), and cause‐specific absolute risk (AR) were calculated using the Kaplan‐Meier and Aalen Johansen estimators. Death from other causes was considered a competing risk. Curves were compared using the Gray’s or log-rank test. Non‐parametric additive excess hazards model was used to compare mortality rates between the cohort and the general population matched by age, calendar year, and sex relative survival ratio (RSR). RSR curves were compared using the proportional excess hazards model. Hazard ratios (HRs) were calculated using the Cox proportional hazards model. To ensure balance in follow-up time, the regression model was based on a maximum of 10 years or until the last day of follow-up: 31 December 2019.


Inclusion and exclusion

Of the 295 patients with SNBCL and 39 with extraosseous plasmacytoma, 205 with B-cell lymphoma and 22 with plasmacytoma were included.

Lymphoma subtype classification and incidence

Of the 10 histologically distinct B-cell lymphoma subtypes identified, DLBCL was predominant (n = 163; 80%). Aside from DLBCL, HGBCL, high-grade B-cell lymphoma (NOS); HGBCL (double hit) (HGBCLDH), high‐grade B‐cell lymphoma, HGBCL (double hit); plasmablastic lymphoma (PBL) were identified as primary lymphomas.  The national incidence of SNBCL was 8/year/5,800,000 people and that of NHL was 22.9/year/100,000 people. i.e. SNBCL constituted 0.61% of all non-Hodgkin lymphoma (NHLs) in Denmark.

Patient characteristics

The median age at diagnosis was 72; 62% of patients were male (n = 126). Performance Status scores were evenly distributed between subtypes, with 84% (n =153) having a score of zero or one. In 34 cases, serum LDH at diagnosis was above the upper limit.

  • All diffuse large B-cell lymphoma (DLBCL): In DLBCL patients (n = 163), lesions were unilateral to the nasal cavity (n = 58; 38%_ or maxillary sinus (n = 52; 32%). Subsequent diagnosis showed widespread extension to other maxillary (38%), ethmoid (38%), and sphenoid sinuses (75%) to the orbit.
  • Of all the patients with PSDLBCL (n = 116), 60% were male and the median age was 74. About 106 patients had a performance status of 0 or 1; 15 patients presented with contiguous spread to cervical lymph nodes. Primary sinonasal diffuse large B‐cell lymphoma accounted for 84% of all primary SNBCLs.
  • Disseminated or prior lymphoma (secondary sinonasal DLBCL) was present in 28 (17%) and 12(8%) cases. They were solely extranodal in 8 cases, purely nodal in 11 cases, and a mixture of both extranodal and nodal in 9 cases. Earlier lymphomas were of extranodal origin in 9 out of 12 relapsed cases. Relapses to the NPS were extranodal in 92% of cases.

42 patients in the cohort presented with other lymphomas. Peripheral blood involvement was found in one patient with mantle cell lymphoma and one with low-grade B-cell (LGBCL) lymphoma.


Nasal congestion (39%) was the most prominent symptom across all subtypes; facial swelling (24%) and pain (23%) were other common symptoms. Epistaxis was common in aggressive lymphomas (13 -25%) and lymphoplasmacytic lymphomas (50%.

Treatment and Outcome

All Subtypes

Out of 205 patients, 147 received CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) or a CHOP‐like regimen as initial treatment. Other chemotherapy regimens were given in 12 cases. Consolidative radiotherapy was utilized for low-stage disease across all subtypes (EI and EII; n = 68), with patients receiving 30–40 Gray over 15 or 20 fractions.

CNS prophylaxis, either high dose methotrexate (73%) cytarabine (8%), methotrexate – cytarabine combo (13%) or triple CNS (6%), was given to 64 patients. DLBCL (75%) patients received CHOP or CHOPlike regimens, and 5% received other regimens. Of the patients treated with chemotherapy or immunochemotherapy, 41% (n=68) received additional radiotherapy and 31% (n = 51) CNS-prophylaxis.

Primary sinonasal diffuse large B‐cell lymphoma (PSDLBCL)

Chemotherapy was given to 93 out of 116 cases; 53 received additional immunotherapy. 83% of cases showed a complete response, 5% partial response, and 4% no response. 6 patients died before response evaluation. Recurrence was observed in 26 patients (20 extranodal; 5 nodal, 1 both nodal and extranodal). Progression was at the following sites: 3 with locoregional progression, 1 to skin and kidney; 1 to skin and testis; and 2 to the CNS. The median time until relapse or until locoregional progression or dissemination was 3.4 years.

The five-year survival rates (median follow up time = 5.7 years) for all PSDLBCLs were: OS 56%; PFS 50%; AR 33% AND RSR 0.67. Five year survival rates for chemotherapy vs immunochemotherapy were: OS 50% vs 73%; p = 0.023; PFS 36% vs 67%, p = 0.023; and AR 42% vs 16%, p = 0.01. Consolidative radiotherapy in addition to immunotherapy did not show any additional benefit. The addition of immunotherapy reduced the death hazard by 72%.

Secondary Diffuse large B‐cell lymphoma

29 cases received chemotherapy; 16 received additional treatments with Rituximab. Overall, 21 cases showed a complete response, 5 had a partial response, 5 showed no effect, and 2 died before evaluation. Recurrences were equal between patients with or without prior lymphoma. 6 recurrences were extranodal, 3 nodal, and 1 was both. Out of the 8 cases that showed progression 1 was locoregional, 1 purely nodal, and 6 with extranodal dissemination. The median time until relapse or locoregional progression/dissemination was 3.1 years.


The findings of this study suggest DLBCL being the dominant subtype; the incidence of SNBCL was 0.14/100000 person/year i.e. 0.61% of call NHLs/year. In most cases, sinonasal symptoms were the most prominent and the first presentation, even in last-stage patients. The lymphoid lesion in the MPS is most likely a primary SNBCL, as two-thirds of the participants had localized disease at the time of diagnosis. Regional lymph node involvement was mainly through contagious spread to cervical lymph nodes. Since none presented retropharyngeal lymph node involvement, persistent cervical lymphadenopathy should be followed by an examination of the NPS, especially if patients exhibit symptoms of rhinosinusitis.

Due to the time span of the data which covered both the pre-and-post rituximab treatment, the authors were able to explore the differences in survival rates between specific treatment groups. The data showed that contemporary treatment improved the prognosis of PSDLBCL. The regression analysis showed that immunotherapy results in a 78% reduction in death hazards. Numerous cohort studies on lymphoma sites and randomized controlled trials on localized lymphoma showed the limited effect of consolidative radiotherapy. The authors found RCTs to be heterogeneous in terms of stage, subtype, and nodal/extranodal involvement. Because PSDLBCL affects the elderly, the efficacy of radiotherapy must be analyzed against its toxicity and the rate of locoregional progression.

Regression analysis showed that besides the apparent predictors such as age >60 and performance status, regional lymph node involvement can increase the hazard of death by a factor of 2.18 (95% CI 0.87–5.12).


The study has several limitations. i) the lack of enough patients to stratify chemotherapy cycles and survival rates ii) lower degree of certainty for the cases before the year 2000 iii) in regression analysis authors could not consider the differences in radiotherapy.

Oncology Related Tools


Latest Research


Similar Articles

Leave a Reply