Clinical Lymphoma, Myeloma and Leukemia

Primary Dural Diffuse Large B-cell Lymphoma: A Comprehensive Review of Survival and Treatment Outcomes

Zachary L. Quinn, MD, Karam Zakharia, MD, Janet L. Schmid, MD, John J. Schmieg, MD, Hana Safah, MD, Nakhle S. Saba, MD

Title: Primary Dural Diffuse Large B-cell Lymphoma: A Comprehensive Review of Survival and Treatment Outcomes

Authors: Zachary L. Quinn, MD1*; Karam Zakharia, MD2*; Janet L. Schmid, MD3; John J. Schmieg, MD3; Hana Safah, MD2; and Nakhle S. Saba, MD2
1 Department of Medicine, Tulane University, New Orleans, LA
2 Section of Hematology and Medical Oncology, Department of Medicine, Tulane University, New Orleans, LA
3 Department of Pathology, Tulane University, New Orleans, LA
*, ZLQ and KZ contributed equally to this manuscript.

Running Title: Survival in PD-DLBCL

Keywords: PD-DLBCL, Dura Mater, methotrexate, CHOP-R, PCNSL

Financial Support: There was no funding for this project

Corresponding Author:
Nakhle Saba, MD
Associate Professor of Clinical Medicine Section of Hematology & Medical Oncology Tulane University School of Medicine
1430 Tulane Ave, code#8578 New Orleans, LA 70112 Office: 504-988-6234
Email: [email protected]

Conflict of Interest: There is no conflict of interest to declare for this project.

Counts:
Abstract word count: 151 Text word count: 2957

Number of figures: 5 Number of tables: 1
Number of supplementary data: 1 Number of references: 35

Abstract

Primary dural diffuse large B-cell lymphoma (PD-DLBCL) is a rare and aggressive B-cell non- Hodgkin lymphoma (NHL) that can present in intracranial or intraspinal locations. While the optimal management is unknown, PD-DLBCL therapy is often mirrored after primary central nervous system lymphoma (PCNSL) therapy and aggressive treatment with a high dose Methotrexate (MTX)-based regimen is frequently used. Our comprehensive, retrospective study of 24 reported cases of PD-DLBCL provide the most complete analysis of this rare disease including data on biology, treatment outcomes and survival. Our findings demonstrate good outcomes following induction treatment with R-CHOP, suggesting that these cases can be treated as DLBCL rather than PCNSL, obviating the need for more aggressive and toxic approaches. The durable responses following R-CHOP also confirm that PD-DLBCL is not protected by the blood brain barrier.
Data in this paper were presented in part at the 2017 American Society of Hematology Annual Meeting, abstract # 4166.

Keywords: PD-DLBCL; Dura Mater; methotrexate; CHOP-R; PCNSL

Introduction

Primary dural lymphoma (PDL) is an uncommon type of extranodal B-cell non-Hodgkin lymphoma (NHL) that arises within the dura mater, incasing either the brain or spinal cord. PDL is rare, representing less than 0.1% of all cases of NHL, and approximately 0.6% of all intracranial tumors.1,2 Most cases present in the 5th or 6th decade of life and occur more frequently in women, with a female to male ratio of 3-4:1.3,4
Diagnosis of PDL is typically made after a patient presents with neurologic sequelae due to mass effect from the neoplasm, mimicking most intracranial processes. Headache, visual disturbances, cranial nerve deficits and seizures are commonly described when an intracranial lesion is discovered; while hemiparesis and radicular pain are most commonly observed with intraspinal disease.5 Magnetic Resonance Imaging (MRI), which is the most sensitive modality for diagnosis, classically shows a single or multiple dural based masses that enhance with gadolinium, displaying a “dural tail” sign.4 The MRI findings in PDL overlap significantly with that of a meningioma, which is the most common alternative preoperative diagnosis. Vasogenic edema out of proportion to tumor mass, restricted diffusion MRI and bone lysis instead of hyperostosis and absence of vascular blush on angiography are all imaging features that may suggest a diagnosis of PDL over meningioma.6,7 However, definitive distinction can only be made after surgical pathology is performed.
Due to its unique clinical presentation that mimics intra-cranial/spinal tumors, PDL is frequently misdiagnosed as primary central nervous system lymphoma (PCNSL). In most cases, however, the histology of PDL is most consistent with an indolent extranodal marginal zone lymphoma (MZL), which can be easily differentiated from the classic large B-cell histology of PCNSL.7 Nevertheless, primary dural diffuse large B-cell lymphoma (PD-DLBCL) can develop and is hard to differentiate from PCNSL. The challenges of making a surgical or pathological diagnosis of tissue location and concern over possible extension into deeper tissue spaces has likely driven some clinicians to treat the entire mass as a PCNSL, when the bulk of the disease does not behave as such.
PD-DLBCL is rarely seen, with only 24 cases reported thus far (including our case). Given the paucity of clinical data, optimal management remains uncertain with no standardized treatment established. Coupling data collected at our institution, with an extensive review of the literature we provide analysis of biologic classification, survival data and recommendations on diagnosis and treatment strategies for PD-DLBCL.

Materials and Methods
Illustrative Case
A 44 year-old male presented to the hospital after a tonic clonic seizure. Computed Tomography (CT) of the head without contrast demonstrated a hyperdense epidural right posterior parietal lesion with associated vasogenic edema. MRI of the brain with gadolinium further defined the lesion as a 4.4 x 2.6 cm extra axial multilobulated enhancing mass within the right parietal convexity with thin connection to masses posterior to bilateral occipital lobes extending across the torcula (Figure 1A). Smaller lobular moieties extended inferiorly, abutting the right temporoparietal region with a representative mass measuring 1.4 x 0.8 cm. Right sided craniotomy with tumor debulking was performed and a grossly thickened dura was noted during the operation. The patient recovered nicely from the procedure and was discharged home on anti-epileptics.

Pathology identified the lesion as DLBCL, germinal center B-cell like (GCB). The malignant B-cells were reactive for CD20, CD79a, PAX5, and FMC7, with germinal center subtype reactive for CD10 and BCL6, negative for MUM1 (Figure 2). Cells were negative for Cyclin-D1, BCL2, CD3, CD5, CD15, CD23, CD43 and EMA. Ki67 was estimated to be 40%. Fluorescence in situ hybridization (FISH) for MYC, BCL2, and BCL6 was normal. Bone marrow biopsy, cerebrospinal fluid (CSF) analysis, as well as staging CT scans of neck, chest, abdomen and pelvis did not demonstrate any other abnormality. The patient was diagnosed with Stage IE PD-DLBCL, of the GCB cell of origin.

Shortly after diagnosis, the patient began treatment with cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP) with rituximab (R), and he was provided intrathecal methotrexate (MTX) with each round of therapy for CNS prophylaxis. His course was complicated by neutropenia following his first round of treatment and he was given colony stimulating factor injections after subsequent treatments. An MRI of the brain following his second round of therapy demonstrated a complete remission at the surgical site. He completed six rounds of chemotherapy and has remained disease free 31 months after diagnosis, the date of his last follow-up.

Search and Study Selection
A comprehensive review of the literature was conducted on pathologically confirmed intracranial and intraspinal cases of PD-DLBCL reported from Jan 1st 1992 to February 20th 2019 in PubMed/Medline and EMBASE (Figure 3). Key words used (single or in combination) included: primary dural lymphoma, PDL, diffuse large B-cell lymphoma, DLBCL, primary diffuse large B- cell lymphoma dura mater and cranial vault lymphoma. Additionally, we performed a manual search through cited references of articles found by our systematic review to identify cases possibly missed in our database search.
Only cases with pathologic confirmation of DLBCL histology where disease was present in dural tissue were included. We included cases of intracranial disease where extension of disease through bone was present if the origin of disease was shown to be in the dura mater. We excluded all other histologies, including marginal zone and follicular lymphoma. We excluded any case of epidural lymphoma and any case where disease penetrated the central nervous system, which we defined as radiographic or pathological parenchymal invasion or presence of disease in cerebrospinal fluid.

Data Collection
Data were tabulated regarding the following 23 variables (Table S1): histology, tumor location (cranial vs. spinal), age, gender, additional site of disease detection by either positron emission tomography-computed tomography (PET-CT) scan or CT scan, bone marrow biopsy, cerebrospinal fluid (CSF) cytology, disease stage, local tumor extension, cell of origin (germinal center-like (GCB)-DLBCL vs. activated B-cell like (ABC)-DLBCL), CD20, CD10, CD79a, CD30, BCL2, BCL6, MUM1, Ki-67 (above or below 70%), type of chemotherapy, use of radiation therapy, use of chemo-prophylaxis, progression free survival (PFS), and overall survival (OS). Data were reported as percent and total number of cases with available information of each variable.

Results

Out of 463 screened cases, 24 cases of PD-DLBCL were detected from 20 published sources and one case from our institution.1,2,5-22 Twenty of these cases were intracranial and four were intraspinal. Median age at time of diagnosis was 49 years (range, 14-75), and there was no gender predominance noted (12 females, 12 males) (Table S1). Infection with HIV was not apparent in any case, although specific mention of negative HIV serology was only provided in 10 instances.

Pathologic Data
Analysis of pathology data demonstrated a diverse representation of molecular characteristics as illustrated in Figure 4. When reported, each case was positive for CD20 (19/19). CD10 was positive in four of eight reported cases, CD79a was positive each time it was reported (6 of 6 cases). MUM1 was positive in two of three reported cases and CD30 positive only once in four reported instances. Using Hans criteria when possible to determine cell of origin23, two cases were classified as activated B-cell-like (ABC)-DLBCL, and four cases as germinal-center B-cell-like (GCB)-DLBCL. The Ki-67 fraction was given in ten cases, only three cases demonstrated fractions greater than 70%.24

Clinical Presentation
Either PET-CT or CT imaging was performed in 20 of 24 cases and bone marrow biopsy results were available in 18 of 24 cases (Table S1). Lumbar punctures were performed in 11 cases and in only one instance abnormal cytology was detected (Table S1). Four cases did not

report imaging or bone marrow biopsy results and were unstageable (Table S1). Based upon the Lugano classification, 18 cases were classified as stage IE.25 Two cases were classified as stage IV. In one case, two cranial tumors were separated only by intact skull bone.19 The other case was classified as stage IV with nerve tissue invasion noted on surgical pathology and atypical cells discovered on CSF cytology.18 However, this was only found after saline was flushed into the spinal canal then redrawn following a failed attempt of a regular spinal tap. In addition, the lumbar puncture was performed through the primary dural mass, leading to questionable validity of this result and possible CSF contamination by the operator.18 We elected to consider this case as a PD-DLBCL and not PCNSL is because lesions in PCNSL should be confined to the CNS and the eyes.26 While the site of the original tumor remains undetermined, we favored the possibility that it started in the dura and extended internally to involve the nerve tissue and possibly the CSF. Thus we classified this case as a stage IV PD- DLBCL, with secondary CNS involvement.
Macroscopic extension of the primary tumor into leptomeningeal tissue was described during surgical resection (three cases) or MRI (three cases). With the exception of the case reported by Nakashima et al, pathologic or histologic confirmation of disease extension into deeper tissues was not reported in any other instance (Table S1).

Treatment and Survival Outcomes
Survival data was available in 21 of 24 cases. Using the Kaplan-Meier method, the overall survival (OS) was determined to be 90% at 1 and 5 years (Figure 5A). Tumor location, when comparing intracranial vs. intraspinal disease, did not impact OS (P = 0.48; Figure 5B). Nineteen cases reported both survival data and specified agents used for induction chemotherapy (Table 1). From these 19 cases, 13 patients received treatment with CHOP, 7 of these patients additionally received rituximab (R-CHOP). Six patients received high dose methotrexate (MTX)-based therapy. Interestingly, no difference in overall survival was observed between CHOP vs. high dose MTX-based therapy (P=0.49; Figure 5C). All seven patients who received rituximab in addition to CHOP remained disease free and alive at the time of last follow-up. Consolidation with high dose chemotherapy and stem cell rescue was not reported in any case.
Three progressions of disease following treatment were reported, two following high dose MTX-based therapy and one following induction with CHOP. For both patients who demonstrated recurrence of disease after induction with high dose MTX-based therapy, R- CHOP was subsequently given, resulting in complete remission. One patient showed progression of disease after 4 cycles of high dose MTX, but subsequently had a complete response to R-CHOP and was alive at 34 months. The other patient had a good response to high dose MTX, coupled with reversible osmotic opening of the blood brain barrier (BBB) using a mannitol infusion, but distant disease was detected in the left adrenal 34 months after initial diagnosis. This patient also subsequently achieved complete remission following six cycles of R-CHOP. There was one case of primary refractory disease in a patient who received induction with CHOP, without rituximab, who declined further treatment after 3 rounds of therapy.
Radiation therapy was always given in the treatment of spinal disease (4 of 4 reported cases), but less frequently when treating intracranial disease (5 of 20 reported cases), with a range of 36-4000 cGy and median dose of 3240 cGy. Four patients received whole brain radiation therapy following induction chemotherapy. The use of radiation therapy did not impact OS (data not shown).

Discussion

Here we present the most comprehensive analysis of PD-DLBCL to date, including key clinical and biological features that aim to guide the diagnosis and management of this rare entity. Our analysis of 24 distinct cases of PD-DLBCL, suggests that PD-DLBCL could be managed as DLBCL, sparing the patient unnecessary aggressive therapy for PCNSL.
We evaluated the OS in 21 patients, all of whom had a pathologic diagnosis of PD- DLBCL. When intracranial and intraspinal cases of PD-DLBCL were combined, we observed OS at 5 years to be 90%. This is consistent with previously reported survival rates for PDL as a whole and is similar to that of stage IE DLBCL, but contrasts with PCNSL outcomes, where median OS is as low as 26 months, further demonstrating an apparent distinction between the two diagnoses.1,27,28
Notwithstanding that only 4 cases of intra-spinal cases were identified, it does not appear that site of origin affects survivorship. When cases of intracranial PD-DLBCL were compared to intra-spinal cases, we observed no statistically significant difference in survival. Previous arguments have been made that spinal and intracranial PDL may be separate diseases, citing the predilection for intracranial PDL to be of MZL subtype and spinal to be predominantly DLBCL.29 However, evidence of equivalent survival and treatment outcomes for PD-DLBCL at each site suggests a single diagnosis regardless of site of origin.

Further supporting the notion that PD-DLBCL should be treated separately from PCNSL, was the finding that induction treatment with MTX-based therapy did not appear to offer a survival benefit over CHOP, nor was there mention of any patient requiring consolidation therapy with high-dose chemotherapy and stem cell rescue. The ability to use a far more tolerable and less toxic treatment regimen and achieve similar outcomes best supports the notion that PD-DLBCL should be treated as DLBCL and not PCNSL. Furthermore, several of the cases pre-dated the use of rituximab as an adjunct to CHOP therapy, and it is worth noting that all patients who received rituximab achieved a durable remission and were alive at time of respective case publication.
The observation that survival data suggests efficacy of R-CHOP in PD-DLBCL is not surprising when considering nervous system anatomy. One of the main challenges in the pharmacologic treatment for brain diseases is the presence of the Blood Brain Barrier (BBB).30 The BBB has three main structural units: endothelial cells and their basement membrane adjoined by tight cell to cell junctions, pericytes embedded within the basement membrane and astrocytes end feet.31 With endothelial cells that lack the usual fenestrations and acquired tight junctions, the BBB has highly selective semipermeable features, limiting the entrance of large water-soluble molecules including most chemotherapeutic drugs.30 Accordingly, the determination of the targeted anatomical site in respect to the BBB seems to influence the preferred therapeutic plan. The dura mater, derived from mesodermic tissue and containing fenestrated capillaries, is a connective tissue layer that encapsulates the central nervous system, but remains outside of the BBB.31,32 In contrast, the leptomeninges along with brain parenchyma, are derived from neural crest cells of ectodermic origin.33 Consequentially the blood supply to the leptomeninges is within the BBB. In comparison to a lymphoma found in the sub-arachnoid space or brain parenchyma, a lymphoma arising in the dura mater and extending externally toward the skull and scalp remains outside of the BBB and thus does not require high-dose chemotherapy to overcome the protective effect of this barrier.30,31 Interestingly, in the six cases where extension into the leptomeninges was suggested either by imaging or during surgery, there was no incidence of death, regardless of whether the patient was treated with high dose MTX (2 cases), or CHOP without brain chemo-prophylaxis (3 cases, Table S1).
Although proximity would suggest epidural and dural lymphoma of the spine to be similar diseases, review of the anatomy suggests otherwise and led to our decision to exclude epidural disease from our data set. The epidural space surrounds the dural sac and it extends from the foramen magnum superiorly to the sacral hiatus inferiorly.34 It contains loose areolar connective tissue, fat, lymphatics, arteries, an extensive plexus of veins, and the spinal nerve roots as they exit the dural sac.34 This heterogenous mixture of tissues starkly contrasts the composition of dural tissue, which most notably lacks lymphatic tissue, when considering the pathogenesis of lymphoma.6 Of the 22 cases of epidural lymphoma we excluded, we observed a higher incidence of death (6/22 cases), suggesting a more aggressive disease than PD-DLBCL (data not shown).
The decision whether to offer adjunctive radiation therapy and/or CNS prophylaxis remains less clear from our review. Radiation was more commonly used in the treatment of spinal disease in comparison to intracranial location. In our analysis, every patient with spinal disease presented with cord compression symptoms, and subsequently received radiation therapy (4/4 cases); however, only 5 of 20 patients with intracranial disease received radiation therapy. Whole brain radiation was given four times. Most commonly, concern over cognitive deterioration from radiation in the setting of an anticipation of a curative response to frontline therapy was discussed as reason for avoiding use of this potentially toxic modality.6
CNS chemo-prophylaxis was rarely reported as a provided treatment in our review. At our institution intra-thecal methotrexate was given, however in only 5 other cases of PD-DLBCL was chemo-prophylaxis provided. Systemic high dose MTX was given in one case, while intra- thecal administration of either cytarabine, MTX or both agents in combination was given in four instances following induction treatment with CHOP (+/-R). Unfortunately, given the paucity of data, the efficacy of chemo-prophylaxis could not be concluded. However, due to the close proximity of the tumor to the CNS and possible deeper tissue extension in some cases, sampling the CSF and adding a prophylactic agent such as intra-thecal methotrexate with each R-CHOP, or systemic high dose MTX alternating with every other cycle of R-CHOP, is a reasonable choice.
Finally, our collected biological characteristics of PD-DLBCL did not follow a pattern, which might aid to distinguish this disease from PCNSL. While most cases PCNSL are of ABC- DLBCL origin, both cell of origin sub-types are represented in PD-DLBCL notwithstanding the small number of cases available for our analysis.35 Overall, we observed a less aggressive proliferation pattern as reported by Ki-67 index; with only 3 out of 10 cases reporting fractions greater than 70%, which may help to explain the encouraging treatment outcomes.

Conclusion

We concede that determination of an optimal treatment strategy cannot be concluded from such a small, retrospective analysis. However, given the rarity of diagnoses and the paucity of available data, prospective trials likely will never be feasible. We offer the first attempt to quantify outcomes for this disease in hopes of aiding clinical decision making in future cases.

Acknowledgements

ZLQ, KZ, and NSS performed the data collection, analysis and writing of the manuscript. JLS and JJS confirmed the pathologic diagnoses of our case. HS and NSS assisted in clinical decision making of our case. All authors read and approved the final manuscript. We also acknowledge contributions of Miss Lei Yin and Dr. Theresa Brown who offered assistance with statistical evaluation of our early data, and review of the final manuscript, respectively.

References

1. Lv ZW, Cheng KL, Tian HJ, Han XM. Primary diffuse large B-cell lymphoma of the dura with skull and scalp involvement: A case report and brief review of the literature. Oncol Lett. 2016;11(6):3583-3588.
2. Lee WH, Kim B, Kim MS. A case of primary dural lymphoma: diffuse large B-cell type.
Turk Neurosurg. 2014;24(5):799-803.
3. Yamada SM, Ikawa N, Toyonaga S, Nakabayashi H, Chang Park K, Shimizu K. Primary malignant B-cell-type dural lymphoma: Case report. Surg Neurol. 2006;66(5):539-543; discussion 543.
4. Iwamoto FM, Abrey LE. Primary dural lymphomas: a review. Neurosurg Focus.
2006;21(5):E5.
5. Said R, Rizk S, Dai Q. Clinical challenges of primary diffuse large B-cell lymphoma of the dura: case report and literature review. ISRN Hematol. 2011;2011:945212.
6. Brito AB, Reis F, de Souza CA, Vassallo J, Lima CS. Intracranial primary dural diffuse large B-cell lymphoma successfully treated with chemotherapy. Int J Clin Exp Med. 2014;7(2):456-460.
7. Chhiber SS, Kirmani AR, Ramzan AU, et al. Primary Diffuse Large B-cell Lymphoma of Dura and Cranial Vault: A Short Series and Review of Literature. Neurosurgery Quarterly. 2015;24(Number 3):324-330.
8. Gao Y, Zhong C, Jiang J. Primary diffuse large B-cell lymphoma of the dura mimicking a meningioma with intervening skull bone invasion. J Neurooncol. 2014;120(1):215-217.
9. Berget E, Helgeland L, Lehmann AK, Smievoll AI, Vintermyr OK, Mork SJ. Primary diffuse large B-cell lymphoma of the dura without systemic recurrence four years after diagnosis and successful therapy. Acta Oncol. 2013;52(5):1047-1049.
10. Wang L, Ouayang T, Zhang N, et al. Primary Diffuse Large B-Cell Dural Lymphoma With Bone and Subcutaneous Tissue Involvement Mimicking Meningioma. J Craniofac Surg. 2015;26(6):e492-494.
11. Macnealy MW, Newton HB, McGregor JM, et al. Primary meningeal CNS lymphoma treated with intra-arterial chemotherapy and blood-brain barrier disruption. J Neurooncol. 2008;90(3):329-333.
12. Munakomi S, Bhattarai B, Srinivas B, Cherian I. Case Report: Primary dural based diffuse large B-Cell lymphoma in a 14 year-old boy. F1000Res. 2015;4:78.

13. Narayanan G, Haridas L, Puthusseri J. Primary Non-Hodgkin’s Lymphoma of Cranial Dura Mater. International Journal of Hematology and Oncology 2014 24 (3):208-209.
14. Renard D, Campello C, Beraru O, Bouillot P, Labauge P. Teaching NeuroImages: Primary diffuse large B-cell lymphoma of the cranial vault. Neurology. 2009;73(17):e84- 85.
15. Miranda RN, Glantz LK, Myint MA, et al. Stage IE non-Hodgkin’s lymphoma involving the dura: A clinicopathologic study of five cases. Arch Pathol Lab Med. 1996;120(3):254- 260.
16. Galarza M, Gazzeri R, Elfeky HA, Johnson RR, 2nd. Primary diffuse large B-cell lymphoma of the dura mater and cranial vault. Case report and literature review. Neurosurg Focus. 2006;21(5):E10.
17. Cecen DA, Tatarli N, Turan Suslu H, Ozdogan S, Barisik NO. Primary Dural Spinal Lymphoma Presentation of a Rare Spinal Tumor Case. Case Rep Surg. 2015;2015:639253.
18. Nakashima H, Imagama S, Ito Z, et al. Primary cauda equina lymphoma: case report and literature review. Nagoya J Med Sci. 2014;76(3-4):349-354.
19. Ochiai H, Kawano H, Miyaoka R, Kawano N, Shimao Y, Kawasaki K. Primary diffuse large B-cell lymphomas of the temporoparietal dura mater and scalp without intervening skull bone invasion. Neurol Med Chir (Tokyo). 2010;50(7):595-598.
20. Amaker BH, Ghatak NR, Jebraili SA, Ferreira-Gonzalez A, Kornstein MJ. Primary T-cell- rich B-cell lymphoma masquerading as a meningioma. Arch Pathol Lab Med. 2000;124(11):1700-1703.
21. Chan DYC, Chan DTM, Poon WS, Wong GKC. Primary cranial vault lymphoma. Br J Neurosurg. 2018;32(2):214-215.
22. Raguž M, Mudrovčić Y, Dlaka D, et al. Primary dural lymphoma mimicking meningioma: a clinical and surgical case report. J Surg Case Rep. 2018;2018(8):rjy189.
23. Hans CP, Weisenburger DD, Greiner TC, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood. 2004;103(1):275-282.
24. Broyde A, Boycov O, Strenov Y, Okon E, Shpilberg O, Bairey O. Role and prognostic significance of the Ki-67 index in non-Hodgkin’s lymphoma. Am J Hematol. 2009;84(6):338-343.
25. Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol. 2014;32(27):3059-3068.
26. Abrey LE, Batchelor TT, Ferreri AJ, et al. Report of an international workshop to standardize baseline evaluation and response criteria for primary CNS lymphoma. J Clin Oncol. 2005;23(22):5034-5043.
27. Friedberg JW, Fisher RI. Diffuse large B-cell lymphoma. Hematol Oncol Clin North Am.
2008;22(5):941-952, ix.
28. Mendez JS, Ostrom QT, Gittleman H, et al. The elderly left behind-changes in survival trends of primary central nervous system lymphoma over the past 4 decades. Neuro Oncol. 2018;20(5):687-694.
29. Mneimneh WS, Ashraf MA, Li L, et al. Primary dural lymphoma: a novel concept of heterogeneous disease. Pathol Int. 2013;63(1):68-72.
30. Bhowmik A, Khan R, Ghosh MK. Blood brain barrier: a challenge for effectual therapy of brain tumors. Biomed Res Int. 2015;2015:320941.
31. Ballabh P, Braun A, Nedergaard M. The blood-brain barrier: an overview: structure, regulation, and clinical implications. Neurobiol Dis. 2004;16(1):1-13.
32. Segal MB, Zlokovic BV. Introduction. In: The Blood-Brain Barrier, Amino-Acids and Peptides. Dordrecht: Springer Netherlands; 1990; 1-11.
33. Saunders NR, Liddelow SA, Dziegielewska KM. Barrier mechanisms in the developing brain. Front Pharmacol. 2012;3:46.
34. Rizvi T, Wintermark M, Schiff D. Chapter 58 – Imaging of Epidural Spinal Cord Compression. In: Newton HB, ed. Handbook of Neuro-Oncology Neuroimaging (Second Edition). Academic Press; 2016:723-744.
35. Camilleri-Broet S, Criniere E, Broet P, et al. A uniform activated B-cell-like immunophenotype might explain the poor prognosis of primary central nervous system lymphomas: analysis of 83 cases. Blood. 2006;107(1):190-196.

Legends

Figure 1: Brain magnetic resonance imaging (T2 Flair) of patient with PD-DLBCL. A. At presentation, 4.4 x 2.6 cm extra-axial multilobulated mass within the right parietal convexity. The calvarial osseous and scalp structures are normal in appearance. B. Following partial resection of the mass (post-operative day 1). C. After completion of six cycles of R-CHOP.

Figure 2: Histology of our PD-DLBCL case (all modified from 400x). A. Hematoxylin Eosin stain. B-F. Immunohistochemical (IHC) staining. CD20, CD10, Bcl-6, MUM-1, and Ki67.

Figure 3: Flow Diagram of Study Selection Process

Figure 4: Biological and Clinical Characteristics Analyzed in PD-DLBCL Cases. Data are tabulated as percent (Y axis) and total number (N) of cases with available information of each variable. GCB and ABC status were determined using the Hans Criteria.

Figure 5: Probabilities of OS. A, Probabilities of OS of 22 patients with PD-DLBCL using the Kaplan-Meier method. B, Probabilities of OS were compared in subgroups dichotomized by disease sites (cranial vs. spinal) of disease location using the log-rank test. C, Probabilities of OS were compared in subgroups dichotomized by choice of induction chemotherapy using the log-rank test.

Table 1: Induction Chemotherapy Agents and Associated Survival. DLBCL, diffuse large B- cell lymphoma; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisone; R, rituximab; MTX, high-dose methotrexate; N/A, not available; ^, Negative bone marrow disease not indicated provided negative PET-CT imaging, #, Dose unspecified.

Table S1: Summary of 22 cases of primary dural diffuse large B-cell lymphoma (PD- DLBCL). S, Surgical; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisone; R, rituximab; MTX, High-Dose Prednisone methotrexate; IT, intra-thecal; Ara-C, cytarabine; MP, methylprednisolone; D, dexamethasone; L, local; WB, whole brain; +, positively expressed; -, negatively expressed; N/A, not available; *, Bone marrow biopsy was not available; ^, Negative bone marrow disease not indicated provided negative PET-CT imaging; #, Dose unspecified.