International Journal of
Radiation Oncology * Biology * Physics
Volume 83, Issue 1 , Pages 284-291, 1 May 2012
Primary bone lymphoma (PBL), either in adults or children is a rare presentation of non-Hodgkin lymphoma, accounting for less than 1% of all malignant lymphomas, for about 5% of all primary malignant bone tumors, and for 45% of all extranodal non-Hodgkin lymphomas. PBL was first described as a distinct clinical entity by Parker and Jackson in 1939 (5), and defined in the 2002 World Health Organization (WHO) classification of tumors of soft tissue and bone, as a single skeletal tumor with, or without regional lymph node involvement, or multiple bone lesions, without visceral or lymph node involvement
Almost 90% of PBL patients present with diffuse large B-cell lymphoma histological subtype, which may have a better prognosis than that of the less common T-cell lymphoma subtype The most commonly affected parts of the skeleton are within the metaphysis and diaphysis of the long bones. The clinical characteristics are nonspecific, making a proper diagnosis difficult at the outset. Pain, swelling, and pathologic fractures are the most common presenting symptoms.
Local radiotherapy (RT) was established as the standard treatment in the 1960s with a local relapse rate of 1020%, but with a distant relapse rate of about 50% and a 5-year survival rate ranging between 55% and 65% 11, 12, The 5-year survival rate has been improved to about 7090% with the addition of chemoradiotherapy (CXRT) in early-stage disease
The role of RT was recently challenged as chemotherapy alone appeared to be quite effective, especially with the development of new agents such as rituximab. The purpose of our Rare Cancer Network (RCN; http://www.rarecancer.net) study was to collect substantial information from a large number of patients to more properly define the disease profile, therapeutic approach, and outcome and prognostic factors of this disease.
Thirteen Rare Cancer Network (RCN) institutions enrolled 116 consecutive patients with PBL treated between 1987 and 2008 in this study. Eighty-seven patients underwent chemoradiotherapy (CXRT) without (78) or with (9) surgery, 15 radiotherapy (RT) without (13) or with (2) surgery, and 14 chemotherapy (CXT) without (9) or with (5) surgery. Median RT dose was 40 Gy (range, 460). The median number of CXT cycles was six (range, 28). Median follow-up was 41 months (range, 6242).
The overall response rate at the end of treatment was 91% (complete response [CR] 74%, partial response [PR] 17%). Local recurrence or progression was observed in 12 (10%) patients and systemic recurrence in 17 (15%). The 5-year overall survival (OS), lymphoma-specific survival (LSS), and local control (LC) were 76%, 78%, and 92%, respectively. In univariate analyses (logrank test), favorable prognostic factors for OS and LSS were International Prognostic Index (IPI) score ≤1 (p = 0.009), high-grade histology (p = 0.04), CXRT (p = 0.05), CXT (p = 0.0004), CR (p < 0.0001), and RT dose >40 Gy (p = 0.005). For LC, only CR and Stage I were favorable factors. In multivariate analysis, IPI score, RT dose, CR, and CXT were independently influencing the outcome (OS and LSS). CR was the only predicting factor for LC.
This large multicenter retrospective study confirms the good prognosis of early-stage PBL treated with combined CXRT. An adequate dose of RT and complete CXT regime were associated with better outcome.
To our knowledge, the current study from 13 institutions of the RCN is the second largest report on early stage PBL.
It has demonstrated relatively similar patient characteristics compared with other published series. A male predominance (male to female: 1.43:1) was found, and median age was 51 years, compared with 3060 years in other series. Pain was the most common presenting symptom, followed by a mass or swelling. The most common sites of pathological fracture were usually located in the long bones, similar to other reports . As in other reports most of our patients presented with Stage IE (ratio between Stage IE and IIE: 4:1).
The overall outcome of patients in this study (5-year OS of 76%, LSS of 78%) was similar to that found in the literature (5-year OS ranging between 70% and 90%)
Univariate analysis for OS in our study revealed that younger age (<50 years) predicted a better outcome, as reported in other series . Normal lactate dehydrogenase level was considered to be a favorable prognostic factor in the report of Beal et al., but in our series we could not confirm this observation. Patients with IPI scores 01 had a markedly better outcome compared with those with an IPI score 24, as previously found by Ramadan et al. but not by Alencar et al.. In contrast to the study of Ostrowski and Horsman e high-grade histology was slightly beneficial compared with low-grade histology for OS and LSS. Previously published papers using Surveillance, Epidemiology, and End Results (SEER) database analysis have reported that patients with local disease had a better survival than those with extensive disease. However, we could not find a significant difference in 5-year OS and LSS between Stages I and II (78% vs. 67%, p = 0.19), which confirms the findings seen in previous series . This might be explained by patient selection. Soft-tissue involvement was observed in 41% of the patients, and patients with extraosseous involvement did not show a significantly worse outcome in 5-year OS, compared with those without extraosseous involvement (62% vs. 85%, p = 0.12). According to some authors, soft-tissue involvement may just reflect an inflammatory process and not real tumor infiltration
RT was established in the 1960s as the treatment of choice with a high local control rate and overall cure rates ranging from 4463% . However, the role of RT alone was challenged over recent years because of the 50% systemic progression rate . Barbieri et al. reported that an RT dose of 40 Gy with a limited RT volume in combination with CXT seemed to be adequate for local control Our study addressed the issue of RT dose and volume. Radiation of the entire bone did not yield a superior outcome compared with partial bone radiation (74% vs. 77%, p = 0.54). However, both univariate and multivariate analyses showed that RT dose >40 Gy was associated with a significantly better 5-year OS and LSS than ≤40 Gy (95% vs. 66%, p = 0.0054, and 95% vs. 69%, p = 0.02 respectively) and a nonsignificant trend toward a better local control (96% vs. 89%, p = 0.32).
CXRT already demonstrated its superiority compared with single therapeutic approaches with 5-year OS between 6090% in recent studies. Interestingly, Alencar recently reported no benefit with CXRT. Although there is no consensus regarding the optimal timing between either RT or chemotherapy, chemotherapy followed by RT was suggested to be the standard approach
With the advent of highly effective chemotherapy, the role of RT has been questioned by some authors. In advanced-stage disease, Ramadan et al. found that patients who received chemotherapy and RT had a worse outcome compared with those who received chemotherapy alone. In the Southwestern Oncology Group 8736 study update, which was on non-PBL lymphoma, there was no difference between CXRT and chemotherapy alone Similar results were reported in some studies, and also in a report on children . Rituximab is now used in association with CHOP or a CHOP-like regimen in the treatment of lymphoma, and studies have demonstrated its positive impact on survival. In our series, the proportion of patients treated with R-CHOP was lower (3:7) than with CHOP alone, and we could not find any significant differences in survival between the two regimens.
Acute side effects were moderate. Leukopenia was the most common early toxicity after chemotherapy. The only reported late toxicity cases involved a limited occurrence of osteonecrosis.
In conclusion, early-stage PBL has a fairly good prognosis. Local control is excellent, and systemic failure occurs infrequently. Young age (<50) and a good IPI score (<2) were positive prognostic factors at diagnosis. The role of chemotherapy is central in the treatment of PBL. Chemotherapy followed by RT is superior for OS and LSS to a sequence of radiotherapy followed by chemotherapy. Although chemotherapy was superior to radiotherapy alone, radiotherapy still plays a role in local control. An RT dose of more than 40 Gy and more than six chemotherapy cycles are associated with a better outcome. Although our results need to be interpreted with caution because of a relatively limited follow-up (41 months) and its retrospective nature, we believe that our findings are important, especially because it is unlikely that a prospective study will be done, given how rare this cancer is.