Hematological Oncology Hematol Oncol (2015) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/hon.2243
Original Research Article
A retrospective international study on primary extranodal marginal zone lymphoma of the lung (BALT lymphoma) on behalf of International Extranodal Lymphoma Study Group (IELSG) Simona Sammassimo1*, Giancarlo Pruneri2, Giovanna Andreola1, Juan Montoro1, Sara Steffanoni1, Grzegorz S. Nowakowski3, Sara Gandini4, Mara Negri1, Thomas M. Habermann3, Markus Raderer5, Zhi-Ming Li6, Pier Luigi Zinzani7, Patrick Adam8, Emanuele Zucca9 and Giovanni Martinelli1 1
Haematoncology Division, European Institute of Oncology, Milan, Italy Pathology Division, European Institute of Oncology, Milan, Italy 3 Mayo Clinic, Rochester, MN, USA 4 Epidemiology and Biostatistics Department, European Institute of Oncology, Milan, Italy 5 Medical University, Vienna, Austria 6 Sun Yat-Sen University Cancer Center, Guangzhou, China 7 Policlinico Sant’Orsola, Bologna, Italy 8 Pathologisches Institut der Universität, Würzburg, Germany 9 IOSI, Bellinzona, Switzerland 2
*Correspondence to: Simona Sammassimo, Haematoncology Division, European Institute of Oncology, Via Ripamonti 435, Milan, Italy. E-mail: [email protected]
Received 26 February 2015 Revised 21 May 2015 Accepted 31 May 2015
Abstract Primary lymphoma of the lung is a rare entity. Clinical features, optimal treatment, role of surgery and outcomes are not well defined, and the follow-up is variable in published data. Clinical data of 205 patients who were confirmed to have bronchus mucosa-associated lymphoid tissue lymphoma from December 1986 to December 2011 in 17 different centres worldwide were evaluated. Fifty-five per cent of the patients were female. The median age at diagnosis was 62 (range 28–88) years. Only 9% had a history of exposure to toxic substances, while about 45% of the patients had a history of smoking. Ten per cent of the patients had autoimmune disease at presentation, and 19% patients had a reported preexisting lung disease. Treatment modalities included surgery alone in 63 patients (30%), radiotherapy in 3 (2%), antibiotics in 1 (1%) and systemic treatment in 128 (62%). Patients receiving a local approach, mainly surgical resection, experienced significantly improved progression-free survival (p = 0.003) versus those receiving a systemic treatment. There were no other significant differences among treatment modalities. The survival data confirm the indolent nature of the disease. Local therapy (surgery or radiotherapy) results in long-term disease-free survival for patients with localized disease. Systemic treatment, including alkylating-containing regimens, can be reserved to patients in relapse after incomplete surgical excision or for patients with advanced disease. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: BALT lymphoma; marginal zone lymphoma; Achromobacter (Alcaligenes) xylosoxidans
Introduction Primary lymphoma of the lung is a rare entity representing 3.6% of all extranodal lymphomas and 0.4% of all non-Hodgkin’s lymphomas (NHL). It is most frequently of extranodal marginal zone lymphoma of mucosaassociated lymphoid tissue (MALT) type [1]. Pulmonary parenchyma does not contain organized lymphoid tissue physiologically, but some conditions Copyright © 2015 John Wiley & Sons, Ltd.
(such as follicular bronchiolitis, pulmonary inflammatory processes and acute infections) can lead to lymphocytic hyperplasia. The former become bronchus MALT (BALT), histologically distinct from true intrapulmonary lymph node [2]. Additional genetic alterations lead to the transformation of this reactive lymphoid tissue into an NHL [3–5]. BALT lymphoma occurs predominantly, but not exclusively, in patients older than 60 years of age. Previous studies have documented a slight predominance in female
S Sammassimo et al.
patients. Most patients are asymptomatic, and the diagnosis is performed through routine chest X-rays. Non-specific pulmonary symptoms (such as cough, chest pain and dyspnea) are the most common in symptomatic patients [6,7]. In gastric MALT lymphoma, the pathogenic role of Helicobacter pylori is well recognized, but in non-gastric MALT lymphoma, the role of infectious agents is less clearly defined. Previous studies suggest a possible pathogenic role of Chlamydia psittaci and Borrelia burgdorferi infections in orbital lymphoma and skin MALT lymphoma, respectively, but the associations have not been confirmed in all studies [8–10]. The role of chronic infections, toxic exposure or underlying autoimmune diseases in BALT lymphoma is unknown. Achromobacter (Alcaligenes) xylosoxidans, a Gramnegative bacterium with low virulence but with high resistance treatment, has been recently detected using a specific nested polymerase chain reaction in a series of pulmonary MALT lymphomas from different European countries [11]. Whether this finding indicates a potential etiopathogenetic role of the bacterium in BALT lymphoma will, however, require further studies. The natural history of BALT lymphomas appears to be slow compared with other low-grade lymphoproliferative disorders, and so, there is no uniform treatment [6,12–16]. The initial treatment has most commonly been surgical resection performed at the time of initial diagnostic biopsy, which results in long remissions in limited stage disease. In more recent years, less invasive diagnostic and therapeutic approaches are emerging. Patients may be managed with a watch-and-wait approach (10). Limited resection, low-toxicity chemotherapy, immunotherapy or immunochemotherapy with rituximab or with low-dose radiotherapy to obtain symptom control has been reported [15,16]. The clinical features, clinical course, optimal treatment, role of surgery and prognostic factors are not well defined, and the follow-up is variable in published data [12,13,17–19]. In this study, we retrospectively reviewed clinical data of 223 BALT lymphoma patients.
Patients and methods This study was reviewed and approved by the human subjects institutional review boards of the participating institutions; patients consented to participate in this study according to local rules. We have retrospectively evaluated clinical data of 223 patients who were diagnosed with BALT lymphoma from December 1986 to December 2011 in 17 different centres worldwide. Inclusion into this study required an initial diagnosis of BALT lymphoma according to World Health Organization criteria [20] and primary site in the lungs, either with single or multifocal lesions (including cases with bilateral presentation), with Copyright © 2015 John Wiley & Sons, Ltd.
or without lymph node involvement, regardless of treatment type with at least 1 year of follow-up. Patients’ charts were reviewed for complete information about clinical characteristics, presenting signs, symptoms, radiological findings, diagnostic procedures, treatment type and outcome. The histological diagnostic standards were detailed in the protocol, and the pathology slides of lymphoma specimens were reviewed centrally by two expert panels, one located in the USA and the other one in Europe. BALT lymphoma staging was performed according to the Ann Arbor system modified by Ferraro et al. [13], summarized in Table 1. Response criteria were defined according to World Health Organization criteria for response [21].
Statistical methodology Time to progression was defined as the time from date of first response until progression. Time to death was defined as the time from diagnosis until death. All patients with no progression or alive at last follow-up date were considered right censored. Overall survival (OS) and progression-free survival (PFS) curves were estimated by the Kaplan–Meier method [22,23]. Being a retrospective data collection, the criteria defining a ‘watch-and-wait’ approach was not specified in detail. For this reason, we excluded 10 patients from the Kaplan–Meier survival curves because, despite not having received a systemic treatment, most of them received a surgical intervention that was unclear as to whether it was only diagnostic. The log-rank tests were used to compare survival time between groups. All statistical tests were two-sided, and p < 0.05 was considered statistically significant. The statistical analyses were performed with the Statistical Analysis System Version 8.2 (SAS Institute, Cary, NC, USA).
Results Patient characteristics Two-hundred twenty-three patients were enrolled. Eighteen patients were excluded. The histological diagnosis was not confirmed by the expert panel in eight patients: four had no Table 1. Ann Arbor system modified by Ferraro et al. [13] (8) Stage
IE II 1E II 2E II 2EW III E IV E
Description
Unilateral or bilateral presentation of the lung Lung presentation with hilar lymph node involvement Lung presentation with mediastinal lymph node involvement Lung presentation with chest wall or diaphragm involvement Lung presentation with abdominal lymph node involvement Lung presentation with extra-lymphatic organs or tissue involvement
Hematol Oncol (2015) DOI: 10.1002/hon
Retrospective analysis on BALT lymphoma
evidence of lymphoproliferative disorder; in two cases, the specimen material was insufficient for review; two patients had another histology, diffuse large B-cell lymphoma and follicular lymphoma, respectively. Clinical data of 10 patients lacked essential information on the treatment performed. Therefore, 205 patients were included in the analysis. Table 2 shows the main clinical data at presentation. Fifty-five per cent of the patients were female. The median age at diagnosis was 62 (range 28–88) years. Seventeen patients (17/185; 9%) had a history of exposure to toxic substances such as lead, pesticides or asbestos, while about 45% (88/197) of the patients had a history of smoking. Nineteen patients (19/184; 10%) had autoimmune disease at presentation, including Sjogren’s syndrome (n = 9) and rheumatoid arthritis (n = 2), and 38/202 (19%) patients had a reported pre-existing lung disease such as bronchial asthma, chronic bronchitis or emphysema. The majority of the patients (169/197; 86%) presented at diagnosis with early stage disease (stages I–II), good performance status (performance status 0–1, 192/198; 97%), no constitutional symptoms (170/199; 85%) and low prognostic score (International Prognostic Index 0–2, 187/196; 95%). Fifty-five per cent of patients had no specific respiratory symptoms such as dry cough or dyspnea.
single or multiple pulmonary nodules, masses or mass-like areas of consolidation, localized in the centre or in the periphery of parenchyma was not definable. An 18F-2-fluoro-2-deoxy-D-glucose positron emission tomography scan was available in 52 (26%), and in 40 patients (93%), it showed an increased uptake. Diagnostic procedures were the following: a transthoracic biopsy in 22 patients, a trans-bronchial biopsy in 26 and major surgery in 41 patients. In 90 cases, a single procedure (trans-thoracic biopsy or trans-bronchial biopsy) was not sufficient to make an accurate diagnosis, so patients had undergone additional procedures. In the remaining 26 cases, the diagnostic approach was not reported. A bone marrow biopsy resulted positive in 10 patients out of 144 (7%) in which it was performed.
Treatment Treatments and results are reported in Table 3. Seventyseven patients (38%) did not receive chemotherapy: 63 (30%) patients underwent surgery alone, 3 (2%) patients were treated with radiotherapy, 1 (1%) patient was prescribed antibiotics and 10 (5%) patients did not Table 3. Treatments, response and disease progression after first line treatment
Diagnostic studies
Treatment (n)
All patients underwent thoracic computerized tomography scan or chest X-ray. A common radiologic pattern such as
Local treatment (67) Surgery (63)
Table 2. Main clinical patients’ characteristics Antibiotics (1) Radiotherapy (3)
Characteristic
All Median age at diagnosis (range)
205 62 years (28–88) n (%)
Sex Male/female
91 (45)/114 (55)
Pre-existing risk factors Exposure to toxic substances Smoking Autoimmune disorders Pre-existing lung disease
17/185 (9) 88/197 (45) 19/184 (10) 38/202 (19)
Stage I–II III–IV PS 0–1 2–3
192/198 (97) 6/198 (3)
IPI score 0–2 3–4 Constitutional symptoms Respiratory symptoms
187/196 (95) 9/196 (5) 29/199 (15) 100/183 (55)
Immunotherapy — rituximab (20) Chemotherapy (70)
169/197 (86) 28/197 (14)
PS, performance score; IPI, International Prognostic Index. Copyright © 2015 John Wiley & Sons, Ltd.
Systemic treatment (128) Immunochemotherapy (38)
Watch and wait (10)a
Response
n (%)
CR PR UN SD CR PR
58 (92) 4 (6) 1 (2) 1 (100) 2 (67) 1 (33)
CR PR SD PD UN CR PR SD UN CR PR SD PD UN NE SD UN
20 (53) 13 (34) 1 (3) 2 (5) 2 (5) 4 (20) 4 (20) 10 (50) 2 (10) 31 (44) 25 (36) 8 (11) 3 (4) 1 (1) 2 (3) 8 (80) 2 (20)
PFS (months)
66 68 5 62 33 33
24
37
26
CR, complete remission; PR, partial remission; UN, unavailable; SD, stable disease; PD, progressive disease; NE, not evaluable; PFS, progression-free survival. a Excluded from the Kaplan–Meier analysis (described in the Section on Statistical methodology). Hematol Oncol (2015) DOI: 10.1002/hon
S Sammassimo et al.
receive any treatment. One-hundred twenty-eight patients (62%) received systemic treatment: 70 patients received chemotherapy alone, 38 patients received rituximab in combination with chemotherapy and only 20 patients received rituximab alone. Most of the patients treated with chemotherapy (83/108, 79%) received an alkylatingcontaining regimen, while 23% (25/108) of patients were treated with an anthracycline-containing regimen.
Outcomes The median time between diagnosis and start of treatment was 37 days (range 0–2279). With a median follow-up of 70 months (range 15–185), 66/205 patients (32%) experienced progression/relapse after first line treatment (median time to progression was 19 months, range 0–112) as reported in Table 3. Figures 1 and 2 show Kaplan–Meier curves of PFS and OS, by stage, of the entire population. Patients receiving a local approach, mainly surgical resection, experienced significantly better PFS versus those receiving a systemic treatment (Figure 3). Furthermore, PFS analysis showed a trend in favour of patients treated with immunotherapy and immuno-chemotherapy versus those receiving chemotherapy alone (Figure 4). PFS analysis according to different types of treatment is described in Figure 5; patients undergoing antibiotics and radiotherapy were excluded from the analysis because of the very small sample size. There was no difference in OS between patients receiving local versus systemic treatment, receiving anthracyclinecontaining versus alkylating-containing regimens and receiving immunotherapy or immunochemotherapy versus chemotherapy alone. Notably, patients with more advanced stages usually undergo different treatments and have a different prognosis; 28 patients (14%) in our series were in stage III/IV.
Figure 2. Overall survival by stage
Figure 3. Progression-free survival by system treatment (CHT Yes) versus local treatment (CHT No)
Figure 4. Progression-free survival by immunotherapy or immuno-chemotherapy (Immunot. Yes) versus chemotherapy alone (Immunot. No)
Figure 1. Progression-free survival by stage Copyright © 2015 John Wiley & Sons, Ltd.
To exclude a possible bias on the outcome due to those patients, we carried out a sensitivity analysis including only stage I/II patients: as shown in Figures 1 and 2 in the Supporting information, PFS according to different treatments was comparable with the entire population. Hematol Oncol (2015) DOI: 10.1002/hon
Retrospective analysis on BALT lymphoma
Figure 5. Progression-free survival by surgery immuno-chemotherapy, chemotherapy and rituximab alone
Discussion Primary lymphoma of the lung is a rare entity representing 3.6% of all extranodal lymphomas and 0.4% of all NHL. MALT type is the most frequent NHL histology that primarily involves the lung [1,24–26]. To our knowledge, this report on 205 BALT lymphomas actually represents the largest series in the literature. Our data confirm the generally indolent presentation, clinical course of the disease, the median age of 62 years and the slight female predominance as described in other series [6,16,24,27]. About 50% of patients were symptomatic, and they usually reported non-specific pulmonary complaints such as cough, dyspnea and chest pain, while a minority of patients (15%) experienced constitutional symptoms such as fever, night sweats and weight loss. As mentioned in the preceding text, 14% of the patients were in stage III/IV, and therefore, we cannot exclude that lung was not the primary site of the disease. MALT lymphomas usually arise in organs normally devoid of lymphoid tissue, such as stomach and salivary glands [28,29]. Chronic antigenic stimulation, triggered by autoimmune or infectious processes, may lead to lymphocytic hyperplasia and then MALT. Native pulmonary parenchyma normally does not contain organized lymphoid tissue, but conditions that include follicular bronchiolitis, pulmonary inflammatory processes, smoking and acute infections may lead to BALT [30,31]. Additional genetic alterations cause the transformation of this reactive lymphoid tissue into NHL. However, no single predisposing inflammatory stimulus appears to be present in all or even most of pulmonary BALT lymphomas. Some authors suggest a possible correlation between pulmonary lymphomas and pre-existing autoimmune disorders, such as Sjogren’s syndrome [7,16] and rheumatoid arthritis [32]. Wöhrer has indeed reported a higher incidence of Copyright © 2015 John Wiley & Sons, Ltd.
autoimmune diseases in his series of 185 extragastric MALT lymphomas, but only 17.3% of them were BALT lymphoma not specifying the actual incidence in such disease. However, in our series, we did not observe a high incidence of immune disorders (
Original Research Article
A retrospective international study on primary extranodal marginal zone lymphoma of the lung (BALT lymphoma) on behalf of International Extranodal Lymphoma Study Group (IELSG) Simona Sammassimo1*, Giancarlo Pruneri2, Giovanna Andreola1, Juan Montoro1, Sara Steffanoni1, Grzegorz S. Nowakowski3, Sara Gandini4, Mara Negri1, Thomas M. Habermann3, Markus Raderer5, Zhi-Ming Li6, Pier Luigi Zinzani7, Patrick Adam8, Emanuele Zucca9 and Giovanni Martinelli1 1
Haematoncology Division, European Institute of Oncology, Milan, Italy Pathology Division, European Institute of Oncology, Milan, Italy 3 Mayo Clinic, Rochester, MN, USA 4 Epidemiology and Biostatistics Department, European Institute of Oncology, Milan, Italy 5 Medical University, Vienna, Austria 6 Sun Yat-Sen University Cancer Center, Guangzhou, China 7 Policlinico Sant’Orsola, Bologna, Italy 8 Pathologisches Institut der Universität, Würzburg, Germany 9 IOSI, Bellinzona, Switzerland 2
*Correspondence to: Simona Sammassimo, Haematoncology Division, European Institute of Oncology, Via Ripamonti 435, Milan, Italy. E-mail: [email protected]
Received 26 February 2015 Revised 21 May 2015 Accepted 31 May 2015
Abstract Primary lymphoma of the lung is a rare entity. Clinical features, optimal treatment, role of surgery and outcomes are not well defined, and the follow-up is variable in published data. Clinical data of 205 patients who were confirmed to have bronchus mucosa-associated lymphoid tissue lymphoma from December 1986 to December 2011 in 17 different centres worldwide were evaluated. Fifty-five per cent of the patients were female. The median age at diagnosis was 62 (range 28–88) years. Only 9% had a history of exposure to toxic substances, while about 45% of the patients had a history of smoking. Ten per cent of the patients had autoimmune disease at presentation, and 19% patients had a reported preexisting lung disease. Treatment modalities included surgery alone in 63 patients (30%), radiotherapy in 3 (2%), antibiotics in 1 (1%) and systemic treatment in 128 (62%). Patients receiving a local approach, mainly surgical resection, experienced significantly improved progression-free survival (p = 0.003) versus those receiving a systemic treatment. There were no other significant differences among treatment modalities. The survival data confirm the indolent nature of the disease. Local therapy (surgery or radiotherapy) results in long-term disease-free survival for patients with localized disease. Systemic treatment, including alkylating-containing regimens, can be reserved to patients in relapse after incomplete surgical excision or for patients with advanced disease. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: BALT lymphoma; marginal zone lymphoma; Achromobacter (Alcaligenes) xylosoxidans
Introduction Primary lymphoma of the lung is a rare entity representing 3.6% of all extranodal lymphomas and 0.4% of all non-Hodgkin’s lymphomas (NHL). It is most frequently of extranodal marginal zone lymphoma of mucosaassociated lymphoid tissue (MALT) type [1]. Pulmonary parenchyma does not contain organized lymphoid tissue physiologically, but some conditions Copyright © 2015 John Wiley & Sons, Ltd.
(such as follicular bronchiolitis, pulmonary inflammatory processes and acute infections) can lead to lymphocytic hyperplasia. The former become bronchus MALT (BALT), histologically distinct from true intrapulmonary lymph node [2]. Additional genetic alterations lead to the transformation of this reactive lymphoid tissue into an NHL [3–5]. BALT lymphoma occurs predominantly, but not exclusively, in patients older than 60 years of age. Previous studies have documented a slight predominance in female
S Sammassimo et al.
patients. Most patients are asymptomatic, and the diagnosis is performed through routine chest X-rays. Non-specific pulmonary symptoms (such as cough, chest pain and dyspnea) are the most common in symptomatic patients [6,7]. In gastric MALT lymphoma, the pathogenic role of Helicobacter pylori is well recognized, but in non-gastric MALT lymphoma, the role of infectious agents is less clearly defined. Previous studies suggest a possible pathogenic role of Chlamydia psittaci and Borrelia burgdorferi infections in orbital lymphoma and skin MALT lymphoma, respectively, but the associations have not been confirmed in all studies [8–10]. The role of chronic infections, toxic exposure or underlying autoimmune diseases in BALT lymphoma is unknown. Achromobacter (Alcaligenes) xylosoxidans, a Gramnegative bacterium with low virulence but with high resistance treatment, has been recently detected using a specific nested polymerase chain reaction in a series of pulmonary MALT lymphomas from different European countries [11]. Whether this finding indicates a potential etiopathogenetic role of the bacterium in BALT lymphoma will, however, require further studies. The natural history of BALT lymphomas appears to be slow compared with other low-grade lymphoproliferative disorders, and so, there is no uniform treatment [6,12–16]. The initial treatment has most commonly been surgical resection performed at the time of initial diagnostic biopsy, which results in long remissions in limited stage disease. In more recent years, less invasive diagnostic and therapeutic approaches are emerging. Patients may be managed with a watch-and-wait approach (10). Limited resection, low-toxicity chemotherapy, immunotherapy or immunochemotherapy with rituximab or with low-dose radiotherapy to obtain symptom control has been reported [15,16]. The clinical features, clinical course, optimal treatment, role of surgery and prognostic factors are not well defined, and the follow-up is variable in published data [12,13,17–19]. In this study, we retrospectively reviewed clinical data of 223 BALT lymphoma patients.
Patients and methods This study was reviewed and approved by the human subjects institutional review boards of the participating institutions; patients consented to participate in this study according to local rules. We have retrospectively evaluated clinical data of 223 patients who were diagnosed with BALT lymphoma from December 1986 to December 2011 in 17 different centres worldwide. Inclusion into this study required an initial diagnosis of BALT lymphoma according to World Health Organization criteria [20] and primary site in the lungs, either with single or multifocal lesions (including cases with bilateral presentation), with Copyright © 2015 John Wiley & Sons, Ltd.
or without lymph node involvement, regardless of treatment type with at least 1 year of follow-up. Patients’ charts were reviewed for complete information about clinical characteristics, presenting signs, symptoms, radiological findings, diagnostic procedures, treatment type and outcome. The histological diagnostic standards were detailed in the protocol, and the pathology slides of lymphoma specimens were reviewed centrally by two expert panels, one located in the USA and the other one in Europe. BALT lymphoma staging was performed according to the Ann Arbor system modified by Ferraro et al. [13], summarized in Table 1. Response criteria were defined according to World Health Organization criteria for response [21].
Statistical methodology Time to progression was defined as the time from date of first response until progression. Time to death was defined as the time from diagnosis until death. All patients with no progression or alive at last follow-up date were considered right censored. Overall survival (OS) and progression-free survival (PFS) curves were estimated by the Kaplan–Meier method [22,23]. Being a retrospective data collection, the criteria defining a ‘watch-and-wait’ approach was not specified in detail. For this reason, we excluded 10 patients from the Kaplan–Meier survival curves because, despite not having received a systemic treatment, most of them received a surgical intervention that was unclear as to whether it was only diagnostic. The log-rank tests were used to compare survival time between groups. All statistical tests were two-sided, and p < 0.05 was considered statistically significant. The statistical analyses were performed with the Statistical Analysis System Version 8.2 (SAS Institute, Cary, NC, USA).
Results Patient characteristics Two-hundred twenty-three patients were enrolled. Eighteen patients were excluded. The histological diagnosis was not confirmed by the expert panel in eight patients: four had no Table 1. Ann Arbor system modified by Ferraro et al. [13] (8) Stage
IE II 1E II 2E II 2EW III E IV E
Description
Unilateral or bilateral presentation of the lung Lung presentation with hilar lymph node involvement Lung presentation with mediastinal lymph node involvement Lung presentation with chest wall or diaphragm involvement Lung presentation with abdominal lymph node involvement Lung presentation with extra-lymphatic organs or tissue involvement
Hematol Oncol (2015) DOI: 10.1002/hon
Retrospective analysis on BALT lymphoma
evidence of lymphoproliferative disorder; in two cases, the specimen material was insufficient for review; two patients had another histology, diffuse large B-cell lymphoma and follicular lymphoma, respectively. Clinical data of 10 patients lacked essential information on the treatment performed. Therefore, 205 patients were included in the analysis. Table 2 shows the main clinical data at presentation. Fifty-five per cent of the patients were female. The median age at diagnosis was 62 (range 28–88) years. Seventeen patients (17/185; 9%) had a history of exposure to toxic substances such as lead, pesticides or asbestos, while about 45% (88/197) of the patients had a history of smoking. Nineteen patients (19/184; 10%) had autoimmune disease at presentation, including Sjogren’s syndrome (n = 9) and rheumatoid arthritis (n = 2), and 38/202 (19%) patients had a reported pre-existing lung disease such as bronchial asthma, chronic bronchitis or emphysema. The majority of the patients (169/197; 86%) presented at diagnosis with early stage disease (stages I–II), good performance status (performance status 0–1, 192/198; 97%), no constitutional symptoms (170/199; 85%) and low prognostic score (International Prognostic Index 0–2, 187/196; 95%). Fifty-five per cent of patients had no specific respiratory symptoms such as dry cough or dyspnea.
single or multiple pulmonary nodules, masses or mass-like areas of consolidation, localized in the centre or in the periphery of parenchyma was not definable. An 18F-2-fluoro-2-deoxy-D-glucose positron emission tomography scan was available in 52 (26%), and in 40 patients (93%), it showed an increased uptake. Diagnostic procedures were the following: a transthoracic biopsy in 22 patients, a trans-bronchial biopsy in 26 and major surgery in 41 patients. In 90 cases, a single procedure (trans-thoracic biopsy or trans-bronchial biopsy) was not sufficient to make an accurate diagnosis, so patients had undergone additional procedures. In the remaining 26 cases, the diagnostic approach was not reported. A bone marrow biopsy resulted positive in 10 patients out of 144 (7%) in which it was performed.
Treatment Treatments and results are reported in Table 3. Seventyseven patients (38%) did not receive chemotherapy: 63 (30%) patients underwent surgery alone, 3 (2%) patients were treated with radiotherapy, 1 (1%) patient was prescribed antibiotics and 10 (5%) patients did not Table 3. Treatments, response and disease progression after first line treatment
Diagnostic studies
Treatment (n)
All patients underwent thoracic computerized tomography scan or chest X-ray. A common radiologic pattern such as
Local treatment (67) Surgery (63)
Table 2. Main clinical patients’ characteristics Antibiotics (1) Radiotherapy (3)
Characteristic
All Median age at diagnosis (range)
205 62 years (28–88) n (%)
Sex Male/female
91 (45)/114 (55)
Pre-existing risk factors Exposure to toxic substances Smoking Autoimmune disorders Pre-existing lung disease
17/185 (9) 88/197 (45) 19/184 (10) 38/202 (19)
Stage I–II III–IV PS 0–1 2–3
192/198 (97) 6/198 (3)
IPI score 0–2 3–4 Constitutional symptoms Respiratory symptoms
187/196 (95) 9/196 (5) 29/199 (15) 100/183 (55)
Immunotherapy — rituximab (20) Chemotherapy (70)
169/197 (86) 28/197 (14)
PS, performance score; IPI, International Prognostic Index. Copyright © 2015 John Wiley & Sons, Ltd.
Systemic treatment (128) Immunochemotherapy (38)
Watch and wait (10)a
Response
n (%)
CR PR UN SD CR PR
58 (92) 4 (6) 1 (2) 1 (100) 2 (67) 1 (33)
CR PR SD PD UN CR PR SD UN CR PR SD PD UN NE SD UN
20 (53) 13 (34) 1 (3) 2 (5) 2 (5) 4 (20) 4 (20) 10 (50) 2 (10) 31 (44) 25 (36) 8 (11) 3 (4) 1 (1) 2 (3) 8 (80) 2 (20)
PFS (months)
66 68 5 62 33 33
24
37
26
CR, complete remission; PR, partial remission; UN, unavailable; SD, stable disease; PD, progressive disease; NE, not evaluable; PFS, progression-free survival. a Excluded from the Kaplan–Meier analysis (described in the Section on Statistical methodology). Hematol Oncol (2015) DOI: 10.1002/hon
S Sammassimo et al.
receive any treatment. One-hundred twenty-eight patients (62%) received systemic treatment: 70 patients received chemotherapy alone, 38 patients received rituximab in combination with chemotherapy and only 20 patients received rituximab alone. Most of the patients treated with chemotherapy (83/108, 79%) received an alkylatingcontaining regimen, while 23% (25/108) of patients were treated with an anthracycline-containing regimen.
Outcomes The median time between diagnosis and start of treatment was 37 days (range 0–2279). With a median follow-up of 70 months (range 15–185), 66/205 patients (32%) experienced progression/relapse after first line treatment (median time to progression was 19 months, range 0–112) as reported in Table 3. Figures 1 and 2 show Kaplan–Meier curves of PFS and OS, by stage, of the entire population. Patients receiving a local approach, mainly surgical resection, experienced significantly better PFS versus those receiving a systemic treatment (Figure 3). Furthermore, PFS analysis showed a trend in favour of patients treated with immunotherapy and immuno-chemotherapy versus those receiving chemotherapy alone (Figure 4). PFS analysis according to different types of treatment is described in Figure 5; patients undergoing antibiotics and radiotherapy were excluded from the analysis because of the very small sample size. There was no difference in OS between patients receiving local versus systemic treatment, receiving anthracyclinecontaining versus alkylating-containing regimens and receiving immunotherapy or immunochemotherapy versus chemotherapy alone. Notably, patients with more advanced stages usually undergo different treatments and have a different prognosis; 28 patients (14%) in our series were in stage III/IV.
Figure 2. Overall survival by stage
Figure 3. Progression-free survival by system treatment (CHT Yes) versus local treatment (CHT No)
Figure 4. Progression-free survival by immunotherapy or immuno-chemotherapy (Immunot. Yes) versus chemotherapy alone (Immunot. No)
Figure 1. Progression-free survival by stage Copyright © 2015 John Wiley & Sons, Ltd.
To exclude a possible bias on the outcome due to those patients, we carried out a sensitivity analysis including only stage I/II patients: as shown in Figures 1 and 2 in the Supporting information, PFS according to different treatments was comparable with the entire population. Hematol Oncol (2015) DOI: 10.1002/hon
Retrospective analysis on BALT lymphoma
Figure 5. Progression-free survival by surgery immuno-chemotherapy, chemotherapy and rituximab alone
Discussion Primary lymphoma of the lung is a rare entity representing 3.6% of all extranodal lymphomas and 0.4% of all NHL. MALT type is the most frequent NHL histology that primarily involves the lung [1,24–26]. To our knowledge, this report on 205 BALT lymphomas actually represents the largest series in the literature. Our data confirm the generally indolent presentation, clinical course of the disease, the median age of 62 years and the slight female predominance as described in other series [6,16,24,27]. About 50% of patients were symptomatic, and they usually reported non-specific pulmonary complaints such as cough, dyspnea and chest pain, while a minority of patients (15%) experienced constitutional symptoms such as fever, night sweats and weight loss. As mentioned in the preceding text, 14% of the patients were in stage III/IV, and therefore, we cannot exclude that lung was not the primary site of the disease. MALT lymphomas usually arise in organs normally devoid of lymphoid tissue, such as stomach and salivary glands [28,29]. Chronic antigenic stimulation, triggered by autoimmune or infectious processes, may lead to lymphocytic hyperplasia and then MALT. Native pulmonary parenchyma normally does not contain organized lymphoid tissue, but conditions that include follicular bronchiolitis, pulmonary inflammatory processes, smoking and acute infections may lead to BALT [30,31]. Additional genetic alterations cause the transformation of this reactive lymphoid tissue into NHL. However, no single predisposing inflammatory stimulus appears to be present in all or even most of pulmonary BALT lymphomas. Some authors suggest a possible correlation between pulmonary lymphomas and pre-existing autoimmune disorders, such as Sjogren’s syndrome [7,16] and rheumatoid arthritis [32]. Wöhrer has indeed reported a higher incidence of Copyright © 2015 John Wiley & Sons, Ltd.
autoimmune diseases in his series of 185 extragastric MALT lymphomas, but only 17.3% of them were BALT lymphoma not specifying the actual incidence in such disease. However, in our series, we did not observe a high incidence of immune disorders (
