SEARCH

SEARCH BY CITATION

Keywords:

  • extranodal lymphoma;
  • angiotropic lymphoma;
  • intravascular lymphoma;
  • cutaneous lymphoma;
  • central nervous system lymphomas

Summary

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Despite its recognition as a distinct, extremely rare entity, no large studies of intravascular lymphoma (IVL) have been reported. The clinico-pathological characteristics of 38 human immunodeficiency virus-negative patients with IVL diagnosed in Western countries were reviewed to better delineate clinical presentation, clinical variants, natural history and optimal therapy. The IVL is an aggressive and usually disseminated disease (Ann Arbor stage IV in 68% of cases) that predominantly affects elderly patients (median age 70 years, range: 34–90; male:female ratio 0·9), resulting in poor Eastern Cooperative Oncology Group Performance Status (ECOG-PS >1 in 61%), B symptoms (55%), anaemia (63%) and high serum lactate dehydrogenase level (86%). The brain and skin are the most common sites of disease. In contrast to previous reports, hepatosplenic involvement (26%) and bone marrow infiltration (32%) were found to be common features in IVL, while nodal disease was confirmed as rare (11% of cases). Patients with disease limited to the skin (‘cutaneous variant’; 26% of cases) were invariably females with a normal platelet count, and exhibited a significantly better outcome than the remaining patients, which deserves further investigation. Overall survival was usually poor; however, the early use of intensive therapies could improve outcome in young patients with unfavourable features. ECOG-PS >1, ‘cutaneous variant’, stage I and chemotherapy use were independently associated with improved survival.

Intravascular lymphoma (IVL) or ‘angiotropic lymphoma’ (formerly known as ‘malignant angioendotheliomatosis’Bhawan, 1987) is a rare entity characterized by exclusive or predominant growth of neoplastic cells within the lumina of blood vessels (Bhawan et al, 1985; Wrotnowski et al, 1985; Carroll et al, 1986). This disorder has been recently recognized as a subtype of diffuse large B-cell lymphoma in the World Health Organization (WHO) Classification (Gatter & Warnke, 2001), although rare forms with a T-cell phenotype do exist (Chen et al, 1998). The understanding of IVL is very limited considering that literature on this malignancy is almost exclusively represented by case reports, cumulative reviews and occasional studies that do not exceed 10–15 patients (Stroup et al, 1990; Glass et al, 1993; DiGiuseppe et al, 1994).

Most clinical and biological properties of IVL are largely unknown. The heterogeneity of clinical presentation and the lack of diagnostic algorithms may explain why approximately half of IVL cases are diagnosed only after autopsy (Domizio et al, 1989) and many of the antemortem diagnoses are made incidentally in biopsies performed for different reasons. Accordingly, clinical syndromes and predictors of survival have not been defined in IVL. The course of this malignancy is generally rapidly progressive and ultimately fatal, with the exception of some patients that achieve durable remission after chemotherapy and rare cases of untreated long-term survivors (DiGiuseppe et al, 1994; Bogomolski-Yahalom et al, 1998).

This report describes presenting symptoms, clinical variants, course, prognostic factors, therapeutic management and outcome of the largest reported series (n = 38) of IVL diagnosed in Western Countries, including either in vivo and postmortem diagnosed cases. The peculiar clinical features and behaviour of an hitherto not described ‘cutaneous variant’ of IVL are analysed.

Study group

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Twenty-two centres affiliated with the International Extranodal Lymphoma Study Group (IELSG) provided data on patient characteristics, diagnosis, sites of disease, stage, treatment and outcome of 38 human immunodeficiency virus (HIV)-negative patients with an in vivo or postmortem pathological diagnosis of IVL between 1985 and 2003. WHO diagnostic criteria were used (Gatter & Warnke, 2001). Diagnosis was established in vivo in 30 (79%) patients and postmortem in eight (21%); it was incidental in four patients referred for surgical resection of benign prostate hyperplasia (n = 2), a cervical polyp or renal cancer. Staging work-up included physical examination, complete biochemical profile, whole-body computerized tomography (CT) scan, and bone marrow aspirate and biopsy. Disease stage was defined according to the Ann Arbor staging system (Carbone et al, 1971). Some procedures, such as CT or magnetic resonance imaging (MRI) of the brain, cerebrospinal fluid (CSF) cytology examination, gastroscopy, abdominal ultrasound or hysteroscopy, were indicated according to presenting features in individual cases. The study conformed to the tenets of the Declaration of Helsinki. Some preliminary results have been published (Ferreri et al, 2002a,b).

Statistical considerations

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Correlations among variables were assessed by Spearman test. Survival curves were generated by the Kaplan–Meier method. Overall survival (OS) was calculated from the date of pathological diagnosis to death or to the last date of follow-up, while event-free survival (EFS) was calculated from the first day of treatment to relapse, progression or death, or to the last date of follow-up. Impact on survival of clinical and therapeutic variables was evaluated using the log-rank test. The independent prognostic value of variables was analysed using the Cox model. All the probability values were two-sided, with an overall significance level of 0·05. There was no correction for multiple comparisons. Analyses were carried out using the statistica 4·0 statistical package for Windows (Statsoft Inc., 1993, Tulsa, OK, USA).

Clinical presentation

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

The median age of the 38 patients was 70 years (range: 34–90), with a male/female ratio of 0·9. IVL was associated with a previous or concomitant malignancy in six cases (16%): four patients had a previous neoplasia [colon cancer, prostate cancer, breast cancer and marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT)-type of the stomach] diagnosed 4–5 years before IVL; IVL was concomitant to other tumours in two patients (renal cancer and diffuse large B-cell lymphoma of salivary gland).

Patients usually presented non-specific symptoms with a remarkable deterioration in the Eastern Cooperative Oncology Group (ECOG) performance status (PS), which was ≥1 in 36 cases (95%). Twenty-one patients (55%) had systemic symptoms (Table I), mainly fever, which was present in 17 cases (45%). Fever was associated with other B symptoms in 10 cases, while four patients experienced weight loss without fever. B symptoms were the only presenting symptom in nine patients (24%).

Table I.  Presenting symptoms.
 Exclusive symptomsAssociated symptoms
  1. *The 10 patients with disease exclusively limited to the skin constituted the subgroup of patients with the ‘cutaneous variant’ of intravascular lymphoma (IVL).

  2. Values in parentheses expressed as percentage.

Systemic symptoms (n = 21; 55%)9 (24)12 (32)
 Fever3 (8)4 (11)
 Fever + weight loss ± night sweats5 (13)4 (11)
 Fever + night sweats0 (0)1 (3)
 Weight loss1 (3)3 (8)
Cutaneous lesions10 (26)*5 (13)
Neurological symptoms5 (13)8 (21)
Pain1 (3)7 (18)
Fatigue0 (0)6 (16)
Gastrointestinal symptoms0 (0)2 (5)
Urinary symptoms1 (3)2 (5)
Cardiac dysfunction0 (0)2 (5)
Oedema0 (0)2 (5)
Dyspnoea0 (0)1 (3)

Cutaneous lesions were the dominant presenting feature in 15 (39%) cases (Table I). Lesions encompassed a widespread morphology and distribution, including painful indurate erythematous eruption, poorly circumscribed violaceous plaques, swelling overlying skin ‘peau d'orange’, cellulitis, large solitary plaques, painful blue-red palpable nodular discolorations, tumour, ulcerated nodules, small red palpable spots, and erythematous and desquamative plaques. Lesions were more commonly situated in upper arms, thighs and legs, lower abdomen, breast and submammary region. Cutaneous lesions were single in four cases and multiple in 11. They were the only site of involvement in 10 cases, but were associated with other symptoms, mostly neurological and B symptoms, in five cases. Standard lymphoma staging procedures demonstrated further sites of disease in all these five patients, with the involved organs being liver, central nervous system (CNS), marrow, and/or spleen. The 10 patients whose disease was exclusively limited to the skin constituted the subgroup of patients with a ‘cutaneous variant’ of IVL.

Thirteen patients (34%) presented with neurological symptoms at diagnosis (five of them were diagnosed at autopsy). These symptoms were extremely heterogeneous, including sensory and motor deficits, meningoradiculitis, paresthesias, hypostenia, aphasia, dysarthria, hemiparesis, seizures, myoclonus, transient visual loss, vertigo, sensory neuropathy and altered conscious state. They were the unique presenting symptoms in five cases (Table I). Neuroimaging confirmed CNS involvement in four of the eight patients with neurological symptoms and in vivo diagnosis of IVL; in two cases CNS staging was negative although neoplastic infiltration was detected at autopsy a few weeks thereafter. Neuroimaging disclosed the presence of brain lesions in two additional patients without neurological symptoms at diagnosis.

Pain, mostly associated with cutaneous or abdominal (renal or adrenal) masses, and fatigue were two common presenting symptoms, affecting more than one-third of patients (Table I).

Histopathological features

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

All cases showed large lymphoid cells within vessel lumina (Fig 1). Neoplastic lymphocytes showed large nuclei with one or more nucleoli and scant cytoplasm. All cases but one (T cell) shared a B-cell immunophenotype. Concomitant extravascular infiltrates of neoplastic lymphocytes were observed in four (11%) cases. A more accurate analysis of histopathological and immunophenotypic features will be performed in a forthcoming study.

image

Figure 1. Intravascular lymphoma of the intestinal wall (haematoxylin and eosin). Large neoplastic cells (open arrows) were detected in a blood vessel lumen (endothelial cells; black arrows).

Download figure to PowerPoint

Stage and sites of disease

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Among the 30 patients with an in vivo diagnosis of IVL, the Ann Arbor disease stage was IE in 12 (40%) cases and IV in 18 (60%). Stage IV disease was present in 10 patients with systemic disease after staging completion, in seven patients with multiple cutaneous lesions and negative staging at other sites, and in one patient with bone marrow infiltration as the sole site of disease (Table II). After diagnostic biopsy, standard staging work-up for lymphomas detected the involvement of bone marrow (n = 8), CNS (n = 4), spleen (n = 7), liver (n = 6), stomach (n = 1), prostate (n = 1), adrenal gland (n = 1) and/or retroperitoneal lymph nodes (n = 2). The limitations of standard lymphoma staging procedures in IVL patients were observed in two cases of early death, where a short-time period occurred between diagnosis and autopsy. In fact, these cases were considered as having stage I disease after conventional staging procedures, while the autopsy, performed 2–7 weeks later, showed the presence of neoplastic cells in the vessels of several additional organs. Disseminated disease (stage IV) was reported in all autopsied cases.

Table II.  Sites of disease.
Involved organsOne organMultiple organs
  1. *According to the Ann Arbor staging system, multiple cutaneous lesions and bone marrow infiltration were considered as stage-IV disease.

  2. CNS, central nervous system.

  3. Values in parentheses expressed as percentage.

Skin
 Single lesion3 (8)1 (3)
 Multiple lesions7 (18)*4 (11)
CNS2 (5)13 (34)
Liver 10 (26)
Spleen 10 (26)
Bone marrow1 (3)*11 (29)
Peripheral blood smear 2 (5)
Lymph nodes 4 (11)
Endocrine glands 6 (16)
Lung1 (3)6 (16)
Heart 4 (11)
Gastrointestinal tract 3 (8)
Gallbladder1 (3) 
Prostate2 (5)4 (11)
Kidney2 (5)6 (16)
Urinary bladder 3 (8)
Uterus2 (5)1 (3)

The most commonly involved organs were the skin, CNS, bone marrow, liver and spleen (Table II). Cutaneous involvement was observed in 15 (39%) patients, in association with multiple organs in five. CNS lesions were detected in 15 (39%) patients, mainly within the context of multiorgan infiltration. Simultaneous liver, spleen and marrow involvement was observed in eight patients (21%). Overall, bone marrow biopsies were positive in 12 cases (32%); in one of them, bone marrow was the sole site of disease (stage IV). Peripheral blood smears were positive in two cases (5%) that were included in the subgroup with bone marrow infiltration. Lymph nodes were usually spared; two patients presented enlarged retroperitoneal lymph nodes at CT scan, which were not biopsied, and two additional patients had multiple retroperitoneal lymphadenopathy at autopsy. Involvement of lung, endocrine glands, kidney and prostate was common (Table II).

Laboratory findings

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Increased serum lactate dehydrogenase (LDH) and β2-microglobulin levels were observed in 86% and 82% of assessed patients respectively (Table III). Anaemia (<12 g/dl) was the most frequent cytopenia (63%). Leucopoenia (<4 × 109/l) or thrombocytopenia (<150 × 109/l) did not occur without anaemia. Eight of the 11 (73%) patients with thrombocytopenia had concomitant bone marrow infiltration and hepatosplenic involvement. All but one patient with marrow infiltration had anaemia or thrombocytopenia; patients with hepatic and/or splenic involvement (n = 12) frequently had anaemia (n = 8), leucopoenia (n = 4) or thrombocytopenia (n = 7).

Table III.  Laboratory findings in the entire series (n = 38) and among patients with the ‘cutaneous variant’ of IVL (n = 10).
FeatureEntire series‘Cutaneous variant’
N*Cases (%)N*Cases (%)
  1. *Assessable patients.

  2. LDH, lactic dehydrogenase; ESR, erythrocyte sedimentation rate; IVL, intravascular lymphoma.

Anaemia (<12 g/dl)3824 (63)105 (50)
Leucopoenia (<4 × 109/l)389 (24)100 (0)
Thrombocytopenia (<150 × 109/l)3811 (29)100 (0)
High serum LDH2925 (86)86 (75)
High serum β2-microglobulin119 (82)21 (50)
Elevated ESR3716 (43)94 (44)
Hypoalbuminaemia (<36 g/l)336 (18)92 (22)
Monoclonal serum component365 (14)90 (0)

An elevated erythrocyte sedimentation rate (ESR) was present in 43% of cases. A monoclonal serum component was reported in five cases (14%): IgA (n = 2), IgG (n = 2) and IgM. Altered hepatic, renal or thyroid functional tests were observed in six cases (16%). In these cases, clinical staging or autopsy demonstrated lymphomatous infiltration of liver, kidneys or thyroid gland.

Laboratory findings in patients with the ‘cutaneous variant’ were similar to those of the other patients: elevated serum LDH in six (75%) of eight assessed patients, anaemia in five (50%) patients and raised ESR in four (44%) of nine assessed patients. Remarkably, patients with the ‘cutaneous variant’ did not have leucopoenia, thrombocytopenia or monoclonal components.

Correlations among variables

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Interestingly, the ‘cutaneous variant’ of IVL was significantly correlated with gender, leucocyte and platelet counts and PS (Table IV). All patients with ‘cutaneous variant’ were females (P = 0·0004) and had normal leucocyte (P = 0·03) and platelet (P = 0·009) counts; all but two of them had an ECOG-PS ≤1 (P = 0·001). These patients were younger than the remainder of the series (median age: 59 vs. 72 years; P = 0·09), and were less associated with B symptoms (30% vs. 65%; P = 0·04) and bone marrow infiltration (0% vs. 43%; P = 0·01).

Table IV.  Correlations among variables.
 AlbuminaemiaMonoclonal componentESRPlatelets countsLeucocytes countsHaemoglobin levelMarrow infiltrationβ2-microglobulinLDH serum levelsB symptomsStageCutaneous variantPSSex
  1. S, significant correlation; Σ, near significant correlation (P = 0·05–0·075); PS, Performance Status; LDH, lactic dehydrogenase; ESR, erythrocyte sedimentation rate.

AgeΣΣΣ
SexSSΣ 
PSSS  
Cutaneous variantSSSSS   
StageSS    
B symptomsSSS     
Serum LDHS      
β2-microglobulinS       
Marrow infiltrationSΣS        
Haemoglobin levelΣΣSS         
Leucocyte count          
Platelets countS           
ESRS            
Monoclonal component             

Significantly positive correlations among some variables suggesting a more aggressive disease, such as advanced stage, high LDH serum level, B symptoms and anaemia, were observed (Table IV).

Outcome

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Therapeutic management of 30 patients with in vivo diagnosis is summarized in Table V. Twenty-two patients were treated with chemotherapy, which was anthracycline-based in 19 cases. Response after chemotherapy was complete (CR) in 10 cases and partial (PR) in three (ORR = 59%); seven patients showed progressive disease during therapy and two died of toxicity. Two women (43 and 52 years old), one with a single endometrial lesion and one with multiorgan disease, were treated with anthracycline-based chemotherapy [MACOP-B (methotrexate, adriamycin, cyclophosphamide, oncovin, prednisone, bleomycin) and CHOEP (cyclophosphamide, adriamycin, oncovin, prednisone, etoposide) regimens] followed by consolidation high-dose chemotherapy supported by autologous stem cell transplantation (ASCT); both are alive and relapse-free at 19 and 71 months from diagnosis.

Table V.  Therapeutic management of 30 patients with in vivo diagnosed IVL.
TreatmentNumber of patientsType of disease
  1. *Followed by high-dose chemotherapy supported by ASCT in two patients.

  2. †Three patients with CNS involvement were also treated with intrathecal chemotherapy.

  3. LTF, lost to follow-up; IVL, intravascular lymphoma; PS, Performance Status; ASCT, autologous stem cell transplantation; CNS, central nervous system.

Chemotherapy22 
 Anthracycline-based*†1939–86 years old, all stages, all clinical variants
 Alkylating-based†3One patient with cutaneous variant; two elderly patients
Other therapies  
Surgery alone5 
 Total resection4Elderly patients, PS = 3, stage IE, chemo refused
 Partial resection1F/74 years old, PS = 4, stage IVE, early death
Radiotherapy1F/49 years old, single cutaneous lesion
Steroids1F/51 years old, multiple cutaneous lesions, LTF
None1F/90 years old, cervix biopsy, PS = 3, early death

Eight patients did not receive chemotherapy; five of them underwent surgical resection of tumour masses of kidney (n = 2), prostate, gallbladder, and skin of the breast, and did not receive further treatment because of the presence of concomitant renal cancer (limited disease), PS ≥3, advanced age, or patient refusal. The patient with concomitant renal cancer died of unrelated causes, while lymphoma-free, at 24 months from diagnosis; the patient with a ulcerated cutaneous node of the breast is alive and relapse-free at 81 months of follow-up; the other three patients died of lymphoma within 5 months of diagnosis. The remaining three patients who did not receive chemotherapy were a 90-year-old woman who had concomitant end-stage renal failure; a 51-year-old woman with multiple cutaneous lesions, who was lost to follow-up after 1 month of steroid therapy; and a 49-year-old woman with a single cutaneous lesion who was treated with radiotherapy alone and was alive with no evidence of disease (NED) at 14 years from diagnosis.

Overall, at the completion of first-line treatment, 17 patients achieved an objective response (13 CR and four PR), nine patients experienced progression (PD) and four died of toxicity (TD). Among responders, eight subsequently experienced relapse, which invariably involved extranodal organs and, mostly, the primary site of disease. The median time to disease progression for the 30 patients with an in vivo diagnosis was 7 months (range: 1–81+), with a 3-year EFS of 27 ± 8% (SD). All failures but one (66 months) occurred within the first year of follow-up.

Seven of 17 patients who experienced failure (i.e. progressive disease in nine and relapse in eight) received a salvage therapy: re-irradiation of a single cutaneous lesion in one patient with a ‘cutaneous variant’, who was alive and showed NED at 167 months from diagnosis; conventional chemotherapy in four cases (two PR, two PD), and high-dose chemotherapy supported by ASCT in two cases (PD, TD). Duration of the second responses were 7, 8, 9+ and 19 months and 7 and 18 months respectively. Interestingly, a case of spontaneous regression of a cutaneous relapse was observed in a patient who is alive with NED at 10 years.

Eleven patients were alive (nine disease-free), at a median follow-up of 41 months (range: 1–167), with a 3-year OS of 32 ± 9% for patients diagnosed during life and 25 ± 7% for the entire series (Fig 2A). Of the surviving patients, six had the ‘cutaneous variant’, two had stage-I disease and three had stage-IV disease.

image

Figure 2. (A) Overall survival (OS) curve for the entire series. (B) OS curves for patients with in vivo diagnosis of intravascular lymphoma (IVL) grouped according to the clinical variant. Patients with ‘cutaneous variant’ (solid line) survived longer than the rest (dotted line) (P = 0·007). (C) OS curves for patients with an in vivo diagnosis of IVL grouped according to the Performance Status (PS). Patients with PS 0–1 (solid line) survived longer than others (dotted line) (P = 0·005). (D) OS curves for patients with in vivo diagnosis of IVL grouped according to Ann Arbor stage. Patients with stage IE (solid line) survived longer than patients with stage IV disease (dotted line) (P = 0·05).

Download figure to PowerPoint

Prognostic factors

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

Univariate analyses, which were limited to cases with in vivo diagnosis, showed that patients with the ‘cutaneous variant’ had significantly better survival (3-year OS: 56 ± 16% vs. 22 ± 10%; P = 0·007) (Fig 2B). Patients with good PS (ECOG ≤1) had a better survival (3-year OS: 48 ± 14% vs. 20 ± 10%; P = 0·005) (Fig 2C). Patients with stage IE disease also had better survival (3-year OS: 42 ± 14% vs. 24 ± 11%; P = 0·05) (Fig 2D).

Considering the distribution of the International Prognostic Index (IPI) on the entire series (n = 38), the risk of death was low (score 0–1) in six cases, intermediate (score 2–3) in 16, and high (score 4–5) in 16 cases. The IPI analysis, when limited to the 30 cases with in vivo diagnosis, showed a near significant association with survival, with a 3-year OS of 50 ± 17%, 35 ± 14% and 20 ± 13%, respectively, for patients with low (n = 6), intermediate (n = 14) and high (n = 10) risk (P = 0·09).

Multivariate analysis on the entire series confirmed the independent prognostic value of ‘cutaneous variant’, PS, stage of disease and use of chemotherapy (Table VI). Similar results were obtained when analysis was limited to patients with in vivo diagnosed IVL (data not shown). When related variables were substituted with the IPI, clinical variant [P = 0·04; odds ratio (OR), 0·27; 95% confidence interval (CI): 0·08–0·93], IPI (P = 0·003; OR, 2·78; 95% CI: 1·46–5·30) and use of chemotherapy (P = 0·005; OR, 0·24; 95% CI: 0·09–0·63) were independently associated with survival.

Table VI.  Multivariate analysis: impact on overall survival on the entire series.
VariableSubgroupsRR95% CIP-value
  1. Similar results were obtained when analysis was limited to patients with in vivo diagnosed IVL.

  2. IVL, intravascular lymphoma; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; LDH, lactate dehydrogenase; RR, relative risk; CI, confidence interval.

AgeContinuous variable0·990·96–1·030·99
ECOG-PS0–1 vs. 2–45·541·66–18·50·009
Clinical variantOthers vs. cutaneous0·080·01–0·460·008
StageI vs. IV4·721·31–17·10·02
B symptomsNo vs. yes0·360·09–1·430·16
Serum LDHNormal vs. elevated0·990·08–11·70·99
AnaemiaNo vs. yes0·990·25–4·010·99
ChemotherapyNo vs. yes0·340·13–0·920·04

Discussion

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References

This study is the largest to date on clinical presentation forms, therapeutic management, outcome and prognostic factors in IVL patients from Western Countries. An important aspect of our series is the high prevalence of cases with in vivo diagnosis; this feature might reflect per se, in contrast to the past experience, an increasingly better recognition of IVL. Although our conclusions are determined from these representative numbers, they should be interpreted with caution considering the retrospective, multicentre nature of this series, which spans a 17-year period.

According to our cohort, IVL is an aggressive and disseminated malignancy that affects elderly patients, without gender prevalence, resulting in poor PS, B symptoms, anaemia and frequently elevated serum LDH levels. Clinical presentation is extremely heterogeneous, ranging from monosymptomatic forms, such as fever, pain or local symptoms, to the combination of B symptoms and rapidly progressing manifestations of multiorgan failure. B symptoms constituted the dominant presenting feature in a quarter of cases. In three cases, fever, which was remarkably more frequent in our series with respect to other extranodal aggressive lymphomas (45% vs. 25%), lead to histopathological examination of bone marrow, and ultimately the diagnosis of IVL.

The brain and skin were the most frequently involved organs; 68% of our patients had involvement of at least one of these organs. Cutaneous lesions were the main clinical presentation in 39% of our series; their aspect and distribution were greatly heterogeneous, being multiple in 80% of cases. In one-third of cases, cutaneous lesions were associated with other sites of disease, mainly bone marrow, CNS, liver, and spleen, while in the remaining two-thirds, the skin was the sole detectable site of disease (see below ‘cutaneous variant’). Also, neurological symptoms were variable according to the affected region of the CNS (Beristain & Azzarelli, 2002). There are no pathognomonic neuro-radiological findings for IVL: ischaemic foci are the most common presentation pattern and vasculitis is the most common differential diagnosis (Song et al, 2002). Intriguingly, neuroimaging did not detect brain lesions in half of the patients with neurological symptoms and in vivo diagnosis of IVL. Conversely, in two cases, brain lesions were not associated with symptoms, but were detected during staging. Thus, in spite of possible false-negative results, MRI of the CNS should be routinely included in the current staging work-up of IVL. The presence of neoplastic cells in the cerebrospinal fluid (CSF) is rare, usually being present in an advanced phase of disease (Ossege et al, 2000), and in association with increased CSF protein levels, which is actually a more common finding (Wrotnowski et al, 1985; Nakahara et al, 1999; Vieren et al, 1999).

It has been commonly reported that haemolymphoid organs are usually spared in IVL; for example, bone marrow involvement has been rarely reported in IVL patients (Glass et al, 1993; Demirer et al, 1994). In our series, bone marrow involvement was observed in one-third of IVL patients, and was significantly associated with hepatosplenic involvement and pancytopenia. As in one of our cases, the bone marrow may remain the only detectable site of disease, allowing IVL diagnosis, for example, in patients with persistent unexplained fever. Taken together, these features suggest that a bone marrow biopsy should be performed in IVL patients, for both diagnostic and staging purposes. Our observations indicate that the involvement of haemolymphoid organs in IVL patients shows a peculiar pattern; conversely to what is observed in most non-Hodgkin's lymphoma (NHL) cases, lymph node involvement is rare (11% of our cases), while hepatic, splenic and marrow infiltration is present in one-third of cases.

As also confirmed by our findings, pulmonary IVL is characterized by multifocal disease leading to dyspnoea, air trapping and severe pulmonary hypertension, with a rapidly aggressive behaviour (Snyder et al, 1989; Ko et al, 1997; Jang et al, 1998; Walls et al, 1999; Evert et al, 2000; Owa et al, 2000). Endocrine gland involvement is common, with a variable presentation according to the infiltration of one or more organs (Chu et al, 1996; Hanihara et al, 1996; Shanks et al, 1997). Interestingly, IVL can be associated with solid tumours (Rubin et al, 1997; Shanks et al, 1997; Wang et al, 2001). In our series, one case was diagnosed within a renal cell carcinoma, while IVL cells were also detected within an hepatic haemangioma in two cases. The preferential involvement by IVL cells of solid tumour-associated vessels seems to suggest the presence of basic, unknown mechanism(s) involving adhesion molecules. In these cases, reactive or neoplastic endothelia may act as a selective attraction for neoplastic lymphocytes. Consequently, IVL may be a good candidate for the evaluation of lymphocytic migration, traffic and invasiveness of lymphoma cells (Ponzoni et al, 2000).

Some cases of IVL have arisen in patients with a previous history of NHL (Carter et al, 1996; Gabor et al, 1997). In most cases, IVL is concomitant or subsequent to a large B-cell lymphoma (Bhawan et al, 1985; Wick et al, 1986; Lopez-Gil et al, 1992; Glass et al, 1993), but cases of IVL developing after small lymphocytic lymphoma (Glass et al, 1993) or follicular lymphoma (Carter et al, 1996) have also been reported. In two of our cases, IVL was diagnosed in patients previously affected by a gastric B-cell marginal zone lymphoma of MALT-type and a diffuse large B-cell lymphoma of the salivary gland. Evidence of a common clonal origin is lacking. However, this is a very relevant issue considering that extravascular infiltration, which was observed in 11% of our cases, does not exclude the diagnosis of IVL, since the co-existence of both components has been previously described (Ansell et al, 1982; Bhawan et al, 1985; Stroup et al, 1990; DiGiuseppe et al, 1994).

Among the laboratory findings, increased serum levels of LDH and β2-microglobulin were present in almost 90% of cases, while anaemia was present in two thirds of cases. Altered hepatic (Gabor et al, 1997; Owa et al, 2000; Koizumi et al, 2001), renal or endocrine functional tests (Al-Chalabi & Abbott, 1995; Nakahara et al, 1999; Ramus & Booth, 1999) and increased CSF protein levels (Bhawan et al, 1985; Wrotnowski et al, 1985; Stroup et al, 1990) are common features in IVL, and, according to our findings, these function tests are useful tools in staging IVL; abnormal results were invariably associated with organ involvement by lymphoma cells.

The clinical features in our series are remarkably different to those reported in Japanese series (Murase & Nakamura, 1999; Takahashi et al, 1999; Shimazaki et al, 2000). In Japanese patients, IVL is associated with haemophagocytic syndrome, bone marrow involvement, fever, hepato-splenomegaly, and thrombocytopenia in 73–100% of cases, while CNS and cutaneous involvement are uncommon (Murase & Nakamura, 1999). Some Japanese authors claimed these features are diagnostic for IVL-associated haemophagocytic syndrome (named also ‘Asian variant’) (Shimazaki et al, 2000). In the Japanese series, IVL showed a rapidly aggressive behaviour, with a median survival of 2–8 months (Takahashi et al, 1999), which is in line with the previously reported negative prognostic impact of the haemophagocytic syndrome in haematological malignancies (Majluf-Cruz et al, 1998).

Data from the present series suggest that any patient with the diagnosis of IVL should be considered to have disseminated disease, and should be treated accordingly with intensive, combined chemotherapy; with a few exceptions, patients managed without this strategy died early. Only patients with a single, small cutaneous lesion, for who chemotherapy is contraindicated, could be managed with radiotherapy alone with some probability of cure. Anthracycline-based chemotherapy appears the best choice considering that the exclusion of these drugs has been associated with disappointing outcome (Stroup et al, 1990; Kuwabara, 1999; Evert et al, 2000). Fifty-eight percentage of our patients treated with CHOP (cyclophosphamide, adriamycin, oncovin, prednisone) or a CHOP-like regimen achieved an objective response, with a 3-year OS of 32 ± 11%. In the English literature, many cases treated with this combination showed clinical benefit, but follow-up is usually short (DiGiuseppe et al, 1994; Baumann et al, 2000; Savarese et al, 2000). Since, the same strategy has been followed by inexorable progression in many other cases (Ko et al, 1997; Rubin et al, 1997; Takamura et al, 1997; Vieren et al, 1999; Owa et al, 2000), some authorities suggested that chemotherapy benefit could actually reflect a selection bias.

Intensified therapeutic approaches could improve current outcome. High-dose chemotherapy supported by ASCT, an important strategy to intensify treatment against NHL, is however, feasible only in a small proportion of IVL patients considering that their median age is 70 years and PS is usually poor. Worldwide experience with this strategy in IVL is limited, with some encouraging results both as first- (Koizumi et al, 2001; Yamaguchi et al, 2001) and second-line treatment (Rose et al, 1999); however, it is possible that some unsuccessfully treated IVL patients have not been reported. In our series, four patients were treated with this strategy; two women treated with high-dose chemotherapy supported by ASCT as consolidation after MACOP-B or CHOEP were alive and relapse-free after 19 and 71 months; whereas the two patients treated with this strategy as salvage therapy for systemic relapses died early after transplantation, because of toxicity or progression.

The identification of clinical, histopathological and biological features that are useful for distinguishing different risk groups is a relevant issue in the therapeutic management of these rare lymphomas. In the present series, disease limited to the skin at presentation (‘cutaneous variant’), ECOG-PS >1 and stage-I disease were independent favourable predictors of survival. Importantly, as suggested by limitations in standard lymphoma staging, a negative systemic staging not necessarily defines a stage-I disease. Negative staging may be consistent with the presence of a minor tumour burden in an usually disseminated disease, which could be associated with a better prognosis. Many patients with ‘stage-I disease’, good PS and ‘cutaneous variant’ obtained prolonged remission with conventional anthracycline-based chemotherapy. However, survival of IVL patients treated with this approach is disappointing, and more intensive strategies should be investigated in younger patients with unfavourable features.

Patients with IVL limited to the skin at presentation displayed distinctive clinical characteristics and prognostic profile. In fact, all patients with the ‘cutaneous variant’ were females and had normal leucocyte and platelet counts. Almost all patients with ‘cutaneous variant’ had an excellent PS, were younger and less frequently had B symptoms and bone marrow infiltration. Conversely, other laboratory findings-like elevated serum LDH levels, anaemia and an elevated ESR, were observed with the same frequency in patients with ‘cutaneous variant’. From a prognostic standpoint, patients with ‘cutaneous variant’ of IVL survived significantly longer than the others, which was independent of the IPI and all other prognostic variables investigated. However, therapeutic outcome was remarkably different between patients with single and multiple cutaneous lesions. All cases with single cutaneous lesions were long-term survivors, while patients with multiple lesions showed worse outcome. In some cases with single lesions, local strategies were associated with a prolonged remission, both at diagnosis and relapse. Cases with multiple cutaneous lesions were mainly treated with systemic anthracycline-based chemotherapy, with an objective response in 86% of cases. Nevertheless, the majority of these patients relapsed within a year of treatment, and only a few cases were successfully managed with salvage chemotherapy.

The favourable clinical behaviour of the ‘cutaneous variant’ of IVL, mainly in cases with a single lesion, may be simply explained by the easier and earlier diagnosis of cutaneous lesions or, alternatively, by potential biological and behavioural differences with respect to the rest of IVL cases; these still unanswered questions will be addressed in planned forthcoming studies.

References

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Study group
  5. Statistical considerations
  6. Results
  7. Clinical presentation
  8. Histopathological features
  9. Stage and sites of disease
  10. Laboratory findings
  11. Correlations among variables
  12. Outcome
  13. Prognostic factors
  14. Discussion
  15. References
  • Al-Chalabi, A. & Abbott, R.J. (1995) Angiotropic lymphoma in the differential diagnosis of systemic vasculitis. Journal of Neurology Neurosurgery and Psychiatry, 59, 219.
  • Ansell, J., Bhawan, J., Cohen, S., Sullivan, J. & Sherman, D. (1982) Histiocytic lymphoma and malignant angioendotheliomatosis: one disease or two? Cancer, 50, 15061512.
  • Baumann, T.P., Hurwitz, N., Karamitopolou-Diamantis, E., Probst, A., Herrmann, R. & Steck, A.J. (2000) Diagnosis and treatment of intravascular lymphomatosis. Archives of Neurology, 57, 374377.
  • Beristain, X. & Azzarelli, B. (2002) The neurological masquerade of intravascular lymphomatosis. Archives of Neurology, 59, 439443.
  • Bhawan, J. (1987) Angioendotheliomatosis proliferans systemisata: an angiotropic neoplasm of lymphoid origin. Seminars of Diagnostic Pathology, 4, 1827.
  • Bhawan, J., Wolff, S.M., Ucci, A.A. & Bhan, A.K. (1985) Malignant lymphoma and malignant angioendotheliomatosis: one disease. Cancer, 55, 570576.
  • Bogomolski-Yahalom, V., Lossos, I.S., Okun, E., Sherman, Y., Lossos, A. & Polliack, A. (1998) Intravascular lymphomatosis – an indolent or aggressive entity? Leukemia and Lymphoma, 29, 585593.
  • Carbone, P.P., Kaplan, H.S., Musshoff, K., Smithers, D.W. & Tubiana, M. (1971) Report of the Committee on Hodgkin's Disease Staging Classification. Cancer Research, 31, 18601861.
  • Carroll, T.J.J., Schelper, R.L., Goeken, J.A. & Kemp, J.D. (1986) Neoplastic angioendotheliomatosis: immunopathologic and morphologic evidence for intravascular malignant lymphomatosis. American Journal of Clinical Pathology, 85, 169175.
  • Carter, D.K., Batts, K.P., de Groen, P.C. & Kurtin, P.J. (1996) Angiotropic large cell lymphoma (intravascular lymphomatosis) occurring after follicular small cleaved cell lymphoma. Mayo Clinical Proceedings, 71, 869873.
  • Chen, M., Qiu, B., Kong, J. & Chen, J. (1998) Angiotropic T cell lymphoma. Chinese Medical Journal (England), 111, 762764.
  • Chu, P., Costa, J. & Lachman, M.F. (1996) Angiotropic large cell lymphoma presenting as primary adrenal insufficiency. Human Pathology, 27, 209211.
  • Demirer, T., Dail, D.H. & Aboulafia, D.M. (1994) Four varied cases of intravascular lymphomatosis and a literature review. Cancer, 73, 17381745.
  • DiGiuseppe, J.A., Nelson, W.G., Seifter, E.J., Boitnott, J.K. & Mann, R.B. (1994) Intravascular lymphomatosis: a clinicopathologic study of 10 cases and assessment of response to chemotherapy. Journal of Clinical Oncology, 12, 25732579.
  • Domizio, P., Hall, P.A., Cotter, F., Amiel, S., Tucker, J., Besser, G.M. & Levison, D.A. (1989) Angiotropic large cell lymphoma (ALCL): morphological, immunohistochemical and genotypic studies with analysis of previous reports. Hematology and Oncology, 7, 195206.
  • Evert, M., Lehringer-Polzin, M., Mobius, W. & Pfeifer, U. (2000) Angiotropic large-cell lymphoma presenting as pulmonary small vessel occlusive disease. Human Pathology, 31, 879882.
  • Ferreri, A.J.M., Seymour, J.F., Grisanti, S., Rossi, G., Berti, E., Pavlovsky, M.A., Martelli, M., Govi, S., Ungari, M., Ascani, S., Pileri, S., Pedrinis, E., Patriarca, C., Arrigoni, G., Facchetti, F., Cavalli, F. & Ponzoni, M. (2002a) Clinical presentation, management and prognosis of intravascular lymphomatosis (IVL): an ongoing clinico-pathologic study of the I.E.L.S.G. Annals of Oncology, 13(suppl. 2), abstract 131.
  • Ferreri, A.J.M., Seymour, J.F., Willemze, R., Illariucci, F., Grisanti, S., Rossi, G., López-Guillermo, A., Berti, E., Pavlovsky, M.A., Martelli, M., Geerts, M.L., Eriksson, M., Federico, M., Candoni, A., Cortelazzo, S., de Renzo, A., Govi, S., Asioli, S., Pedrinis, E., Ungari, M., Campo, E., Milani, M., Motta, T., Artusi, T., Ascani, S., Pileri, S., Patriarca, C., Arrigoni, G., Facchetti, F., Cavalli, F. & Ponzoni, M. (2002b) The ‘cutaneous variant’ of intravascular lymphomatosis (IVL) is a favourable clinical form in western countries: a clinico-pathologic study on 31 cases. Blood, 100, abstract 3045.
  • Gabor, E.P., Sherwood, T. & Mercola, K.E. (1997) Intravascular lymphomatosis presenting as adult respiratory distress syndrome. American Journal of Hematology, 56, 155160.
  • Gatter, K.C. & Warnke, R.A. (2001) Intravascular large B-cell lymphoma. In: Tumours of Haematopoietic and Lymphoid Tissues (ed. by E.S.Jaffe, N.L.Harris, H.Stein & J.Vardiman), p. 177. IARC Press, Lyon.
  • Glass, J., Hochberg, F.H. & Miller, D.C. (1993) Intravascular lymphomatosis. A systemic disease with neurologic manifestations. Cancer, 71, 31563164.
  • Hanihara, T., Takahashi, T., Shimada, T., Mizuguchi, M. & Yagishita, S. (1996) Parathyroid hormone-related protein-associated hypercalcemia in probable intravascular lymphoma of B-cell type. American Journal of Hematology, 53, 144145.
  • Jang, H.J., Lee, K.S. & Han, J. (1998) Intravascular lymphomatosis of the lung: radiologic findings. Journal of Computerised Assistant Tomography, 22, 427429.
  • Ko, Y.H., Han, J.H., Go, J.H., Kim, D.S., Kwon, O.J., Yang, W.I., Shin, D.H. & Ree, H.J. (1997) Intravascular lymphomatosis: a clinicopathological study of two cases presenting as an interstitial lung disease. Histopathology, 31, 555562.
  • Koizumi, M., Nishimura, M., Yokota, A., Munekata, S., Kobayashi, T. & Saito, Y. (2001) Successful treatment of intravascular malignant lymphomatosis with high-dose chemotherapy and autologous peripheral blood stem cell transplantation. Bone Marrow Transplantation, 27, 11011103.
  • Kuwabara, H. (1999) Intravascular lymphomatosis presenting as bilateral adrenal enlargement and insufficiency. Acta Cytologica, 43, 975976.
  • Lopez-Gil, F., Roura, M., Umbert, I. & Umbert, P. (1992) Malignant proliferative angioendotheliomatosis or angiotropic lymphoma associated with a soft-tissue lymphoma. Journal of the American Academy of Dermatology, 26, 101104.
  • Majluf-Cruz, A., Sosa-Camas, R., Perez-Ramirez, O., Rosas-Cabral, A., Vargas-Vorackova, F. & Labardini-Mendez, J. (1998) Hemophagocytic syndrome associated with hematological neoplasias. Leukemia Research, 22, 893898.
  • Murase, T. & Nakamura, S. (1999) An Asian variant of intravascular lymphomatosis: an updated review of malignant histiocytosis-like B-cell lymphoma. Leukemia and Lymphoma, 33, 459473.
  • Nakahara, T., Saito, T., Muroi, A., Sugiura, Y., Ogata, M., Sugiyama, Y. & Yamamoto, T. (1999) Intravascular lymphomatosis presenting as an ascending cauda equina: conus medullaris syndrome: remission after biweekly CHOP therapy. Journal of Neurology Neurosurgery and Psychiatry, 67, 403406.
  • Ossege, L.M., Postler, E., Pleger, B., Muller, K.M. & Malin, J.P. (2000) Neoplastic cells in the cerebrospinal fluid in intravascular lymphomatosis. Journal of Neurology, 247, 656658.
  • Owa, M., Koyama, J., Asakawa, K., Hikita, H., Kubo, K. & Ikeda, S.I. (2000) Intravascular lymphomatosis presenting as reversible severe pulmonary hypertension. International Journal of Cardiology, 75, 283284.
  • Ponzoni, M., Arrigoni, G., Gould, V.E., Del Curto, B., Maggioni, M., Scapinello, A., Paolino, S., Cassisa, A. & Patriarca, C. (2000) Lack of CD 29 (beta1 integrin) and CD 54 (ICAM-1) adhesion molecules in intravascular lymphomatosis. Human Pathology, 31, 220226.
  • Ramus, M.W. & Booth, J.L. (1999) Pathologic quiz case. Pathologic diagnosis: intravascular lymphomatosis (angiotropic lymphoma). Archives of Pathology and Laboratory Medicine, 123, 743744.
  • Rose, C., Staumont, D. & Jouet, J.P. (1999) Successful autologous bone marrow transplantation in intravascular lymphomatosis. British Journal of Haematology, 105, 313314.
  • Rubin, M.A., Cossman, J., Freter, C.E. & Azumi, N. (1997) Intravascular large cell lymphoma coexisting within hemangiomas of the skin. American Journal of Surgical Pathology, 21, 860864.
  • Savarese, D.M., Zavarin, M., Smyczynski, M.S., Rohrer, M.J. & Hutzler, M.J. (2000) Superior vena cava syndrome secondary to an angiotropic large cell lymphoma. Cancer, 89, 25152520.
  • Shanks, J.H., Harris, M., Howat, A.J. & Freemont, A.J. (1997) Angiotropic lymphoma with endocrine involvement. Histopathology, 31, 161166.
  • Shimazaki, C., Inaba, T. & Nakagawa, M. (2000) B-cell lymphoma-associated hemophagocytic syndrome. Leukemia and Lymphoma, 38, 121130.
  • Snyder, L.S., Harmon, K.R. & Estensen, R.D. (1989) Intravascular lymphomatosis (malignant angioendotheliomatosis) presenting as pulmonary hypertension. Chest, 96, 11991200.
  • Song, D.K., Boulis, N.M., McKeever, P.E. & Quint, D.J. (2002) Angiotropic large cell lymphoma with imaging characteristics of CNS vasculitis. AJNR American Journal of Neuroradiology, 23, 239242.
  • Stroup, R.M., Sheibani, K., Moncada, A., Purdy, L.J. & Battifora, H. (1990) Angiotropic (intravascular) large cell lymphoma. A clinicopathologic study of seven cases with unique clinical presentations. Cancer, 66, 17811788.
  • Takahashi, N., Chubachi, A., Miura, I., Nakamura, S. & Miura, A.B. (1999) Lymphoma-associated hemophagocytic syndrome in Japan. Rinsho Ketsueki. Japanese Journal of Clinical Hematology, 40, 542549.
  • Takamura, K., Nasuhara, Y., Mishina, T., Matsuda, T., Nishimura, M., Kawakami, Y., Fujita, M., Mikuni, C. & Yamashiro, K. (1997) Intravascular lymphomatosis diagnosed by transbronchial lung biopsy. European Respiratory Journal, 10, 955957.
  • Vieren, M., Sciot, R. & Robberecht, W. (1999) Intravascular lymphomatosis of the brain: a diagnostic problem. Clinical Neurology and Neurosurgery, 101, 3336.
  • Walls, J.G., Hong, Y.G., Cox, J.E., McCabe, K.M., O'Brien, K.E., Allerton, J.P. & Derdak, S. (1999) Pulmonary intravascular lymphomatosis: presentation with dyspnea and air trapping. Chest, 115, 12071210.
  • Wang, B.Y., Strauchen, J.A., Rabinowitz, D., Tillem, S.M. & Unger, P.D. (2001) Renal cell carcinoma with intravascular lymphomatosis: a case report of unusual collision tumors with review of the literature. Archives of Pathology and Laboratory Medicine, 125, 12391241.
  • Wick, M.R., Mills, S.E., Scheithauer, B.W., Cooper, P.H., Davitz, M.A. & Parkinson, K. (1986) Reassessment of malignant ‘angioendotheliomatosis’. Evidence in favour of its reclassification as ‘intravascular lymphomatosis’. American Journal of Surgical Pathology, 10, 112123.
  • Wrotnowski, U., Mills, S.E. & Cooper, P.H. (1985) Malignant angioendotheliomatosis. An angiotropic lymphoma? American Journal of Clinical Pathology, 83, 244248.
  • Yamaguchi, M., Kimura, M., Watanabe, Y., Taniguchi, M., Masuya, M., Kageyama, S., Katayama, N., Ohno, T., Kita, K. & Shiku, H. (2001) Successful autologous peripheral blood stem cell transplantation for relapsed intravascular lymphomatosis. Bone Marrow Transplantation, 27, 8991.