OUP user menu

Oral and Extraoral Plasmablastic Lymphoma
Similarities and Differences in Clinicopathologic Characteristics

Damien Hansra MD, Naomi Montague MD, PhD, Alexandra Stefanovic MD, Ikechukwu Akunyili MD, Arash Harzand, Yasodha Natkunam MD, PhD, Margarita de la Ossa MD, Gerald E. Byrne MD, Izidore S. Lossos MD
DOI: http://dx.doi.org/10.1309/AJCPJH6KEUSECQLU 710-719 First published online: 1 November 2010

Abstract

Plasmablastic lymphoma (PBL), initially characterized as an aggressive lymphoma arising in the jaw and oral mucosa in HIV-infected patients, was recently reported to occur with extraoral manifestations, heterogeneous histologic findings, and variable association with immunodeficiency states. We reviewed clinical, morphologic, and immunophenotypic features of 13 cases of PBL to determine whether these different subtypes represent distinct morphologic and clinical entities. Two distinct subtypes of PBL were identified and classified as oral and extraoral PBL. The oral PBLs were strongly associated with HIV infection and commonly demonstrated plasmablastic morphologic features without plasmacytic differentiation. Extraoral PBLs tended to occur in patients with underlying non–HIV-related immunosuppression and universally demonstrated plasmacytic differentiation. The patients with oral PBL demonstrated better overall survival compared with patients with extraoral PBL (P = .02). Our findings suggest that PBL with oral and extraoral manifestation represent 2 distinct clinicopathologic entities.

Key Words:
  • Lymphoma
  • Plasmablastic lymphoma
  • Epstein-Barr virus
  • HIV

Large B-cell lymphoma (LBCL) is the most common type of non-Hodgkin lymphoma (NHL), accounting for 30% of cases in Western countries. It encompasses a heterogeneous spectrum of lymphomas with multiple clinical, histologic, and immunophenotypic subtypes. Rare variants with plasmablastic/plasmacytic features and an immunophenotype compatible with terminal B-cell differentiation have been recognized and include plasmablastic lymphoma (PBL), large B-cell lymphoma arising in human herpesvirus-8 (HHV-8) multicentric Castleman disease, and primary effusion lymphoma.1 Most cases of LBCL with plasmablastic morphologic features are seen in the setting of HIV infection. However, a number of cases with similar morphologic features arising in HIV– patients have also been reported.24

PBL is classified as a distinct entity by the World Health Organization and typically occurs in the oral cavity in the clinical setting of HIV infection, accounting for 2.6% of NHLs in this population.5 It is associated with latent Epstein-Barr virus (EBV) infections in the majority (70%) of cases. PBL is a diffuse proliferation of large neoplastic cells with immunoblastic morphologic features and an immunophenotype of terminally differentiated B cells.6 Although PBL most frequently occurs in the oral cavity, it is also encountered in other extranodal, particularly mucosal, sites. Lymph node involvement has also been reported.610

PBLs harbor a cytologic spectrum. Cases with monomorphic cytologic features are more commonly seen in the setting of HIV infection and occur in the oral, nasal, and paranasal area (oral mucosal type). PBL with plasmacytic differentiation tends to occur in extranodal sites7 and may represent a clinically distinct subtype. Only one third of these lymphomas are associated with HIV infection, but almost half of the reported cases are associated with EBV infection. Clinically, PBLs are rapidly progressive tumors associated with poor response to therapy and an average survival time of 14 months.7

To address whether PBLs occurring in oral and extraoral locations represent a continuum of the same disease or constitute distinct entities, we assessed 15 cases diagnosed as PBL in further detail. Herein we present the clinical, pathologic, and immunophenotypic features and differences in outcome among oral and extraoral manifestations of PBL.

Materials and Methods

Cases

The surgical pathology files of Jackson Memorial Hospital, Miami, FL, were reviewed using the key search term “plasmablastic” to identify cases of PBL. Patients with a large cell lymphoma arising in the setting of HHV-8 multicentric Castleman disease and primary effusion lymphoma were not included. On chart review, 2 of the 15 patients were found to have a history of a plasma cell dyscrasia and were excluded from the study. A total of 13 cases were identified and designated as PBL arising in an oral or extraoral location based on the sites involved at the time of initial diagnosis. Patients’ medical records were reviewed for age, time from initial symptoms to diagnosis, sites involved by disease, HIV status, stage, International Prognostic Index (IPI) score, treatment modality, response, relapse site and therapy at the time of relapse, disease-related and treatment complications, time to death, progression of disease, and last follow-up visit. The study was approved by the institutional review board.

Staging investigations in most cases included history and physical examination; CBC; lactate dehydrogenase; complete metabolic profile; bone marrow biopsy; complete radiologic bone survey; serum protein electrophoresis and immunofixation; computed tomography scan of chest, abdomen, and pelvis; and positron emission tomography scan. Stage was assigned according to the Ann Arbor staging system. The IPI score was assigned whenever possible if all 5 components constituting the score could be identified in the medical records. The Eastern Oncology Cooperative Group scale was used to determine performance status.

Patients were treated according to the decisions of primary oncologists. Response to treatment was determined as complete if there was elimination of all evidence of lymphoma after therapy. Partial response was defined as reduction in more than 50% of measurable tumor size. Stable disease was assigned if there was no change in the size of the tumors or no more than a 25% increase in size. Progression of disease was defined as a growth of more than 25% in size or development of new sites of disease. Overall survival (OS) was defined as the time from disease diagnosis to death of any cause. Progression free-survival was defined as the time from the date of diagnosis to the date of documented disease progression or death of disease or due to treatment toxic effects. Survival curves were constructed according to the method of Kaplan and Meier and compared using the log-rank test. Differences were considered significant if the P value was equal to or less than .05.

Immunohistochemical Analysis, In Situ Hybridization, and Fluorescent In Situ Hybridization

Immunohistochemical analysis was performed on formalin-fixed, paraffin-embedded tissue sections using routine methods. When necessary, antigen retrieval was performed in a pressure cooker at 120°C for 20 minutes before immunostaining. The following monoclonal antibodies were used with a labeled streptavidin-biotin system in an automatic stainer (Autostainer Plus, DAKO, Carpinteria, CA): CD138/syndecan-1 (clone M115, dilution 1:50), MUM1 (clone MUM1P, dilution 1:50), CD20 (clone L26, dilution 1:25), CD79a (clone JCB117, dilution 1:50), κ (clone R10R10-210F3, dilution 1:4,000), λ (clone N10/2, dilution 1:4,000), Ki-67 (clone MIB-1, dilution 1:100), and anaplastic lymphoma kinase (ALK) (clone M7195, dilution 1:50) (DAKO); XBP1 (clone M-186, dilution 1:25; Santa Cruz Biotechnology, Santa Cruz, CA); BCL-1/cyclin D1/PRAD1 (clone SP4, dilution 1:100; LabVision Products, Fremont, CA); CD56 (clone 1B6, dilution 1:50; Leica Microsystems, Bannockburn, IL); and PAX5 (clone 24, dilution 1:100; BD Biosciences, San Jose, CA). Detection of latent EBV infection was assessed by in situ hybridization on paraffin-embedded tissue sections using EBV peptide nucleic acid (PNA) probe/fluorescein with the PNA ISH Detection Kit (DAKO), and HHV-8 infection was assessed with a monoclonal antibody to HHV-8 (clone BB10, dilution 1:50; Cell Marque, Hot Springs, AR) using standard methods. IgA specific for α chains, IgG specific for γ chains, and IgM specific for μ chains were ready to use antibodies used in the Flex EnVision System (DAKO).

Immunohistochemical staining was scored as positive if immunoreactivity was present in greater than 30% of tumor cells with appropriate localization and staining of control tissue and cell types. XBP1 expression was scored at ×500 magnification: Tumors were considered positive if more than 30% of cells demonstrated homogeneous diffuse nuclear staining; cells with cytoplasmic staining were considered negative.11 The Ki-67 proliferation index was determined on areas of non-necrotic tumor at ×100 using categories of about 25%, less than 50%, 50% to 75%, more than 75% to 90%, more than 90%, and about 100%. Expression of EBV-encoded RNA (EBER) was scored as positive if signals localized to nuclei were present.

Fluorescent in situ hybridization (FISH) analysis for MYC/IGH rearrangement [t(8;14)] was performed on available paraffin blocks from 3 patients using a commercially available tricolor dual-fusion FISH probe set (Vysis, Downers Grove, IL).

Results

Pathologic Findings

We identified 13 cases of PBL clinically classified as PBL of the oral mucosa (6 patients) and extraoral PBL (7 patients) in our records from 2000 to 2009. Histologic sections were reviewed and discussed over a multiheaded microscope by 3 pathologists (N.M., M.O., and G.E.B.) to evaluate the cytologic, morphologic, and immunophenotypic features of the tumors. Of the 6 cases in the oral mucosa, 5 exhibited monomorphic plasmablastic cytologic features. These tumors were composed of a relativity uniform population of large cells with scant cytoplasm, a nucleus with dispersed or fine reticular chromatin pattern with minimal or no chromatin clumping, and 1 or more large nucleoli Image 1A. Necrosis, frequent mitotic figures, and numerous apoptotic bodies with abundant tingible body macrophages resulting in a “starry-sky” appearance were present in 4 of 5 tumors Image 1B. These tumors were EBER+ with a Ki-67 index of approximately 100%. One tumor differed and demonstrated few apoptotic bodies despite the presence of abundant mitosis. This particular tumor was EBER– in a HIV– patient and expressed a lower Ki-67 index compared with the other 4 cases with monomorphic plasmablastic cytologic features.

Image 1

Morphologic features of plasmablastic lymphoma (PBL) in oral and extraoral sites. PBL in the oral mucosa exhibits monomorphic plasmablastic cytologic features. A, These tumors were composed of a relativity uniform population of large cells with a moderate amount of cytoplasm and a nucleus with dispersed or fine reticular chromatin with minimal or no chromatin clumping and 1 or more large nucleoli (H&E, ×1,000). B, Frequent mitotic figures and numerous apoptotic bodies with abundant tingible body macrophages resulting in a “starry-sky” appearance were present in 4 of 5 tumors (H&E, ×200). PBL in extraoral sites exhibited plasmacytic differentiation. C, These tumors were composed of large cells with relatively increased cytoplasm and an eccentric nucleus with clumped chromatin and relatively smaller nucleoli. These large cells were admixed with variable numbers of smaller cells with plasmacytic differentiation resulting in a varied cytologic appearance (H&E, ×1,000). D, Starry-sky appearance was not present in any of these tumors (H&E, ×200).

Plasmacytic differentiation was exhibited in 7 cases with extraoral and 1 case with oral cavity manifestations. The large cells in these tumors possessed relatively increased cytoplasm, an eccentric nucleus with clumped chromatin, and relatively smaller nucleoli Image 1C. These larger cells were admixed with variable numbers of smaller cells with plasmacytic differentiation resulting in a varied cytologic appearance. Similar observations were reported by Colomo et al.7 Necrosis or a starry-sky appearance was not present in any of these tumors Image 1D. Differences in the cytologic features exhibited by tumors with monomorphic plasmablastic cytologic features vs plasmacytic differentiation are highlighted in Image 2. Cells with monomorphic cytologic features (Image 2A) are generally larger with less cytoplasm. Rare to few cells exhibit plasmacytic differentiation. In contrast, cells constituting tumors with plasmacytic differentiation (Images 2B–2D) are generally smaller with more abundant cytoplasm. In addition, these tumors were observed to contain cells with a cytologic appearance approaching that of mature plasma cells.

Image 2

H&E-stained sections comparing the cytologic features of monomorphic plasmablastic tumor cells with tumor cells with plasmacytic differentiation. A, Cells with monomorphic cytologic features are generally larger with less cytoplasm. Nuclei are often eccentric, and large central nucleoli are common. Rare to few cells exhibiting features of plasmacytic differentiation are present (×1,000). B, C, and D, Cells comprising tumors with plasmacytic differentiation differ with a slightly smaller relative size and relatively more abundant cytoplasm. The nuclei are conspicuously eccentric and are, in general, smaller with coarser chromatin. In addition, cells with a cytologic appearance approaching mature plasma cells are admixed within these tumors (×1,000).

Immunohistochemical characterization of the 13 tumors is summarized in Table 1. All tumors expressed a plasma cell phenotype with expression of CD138 Image 3A and/or MUM1 Image 3B. XBP1, a transcription factor required for terminal differentiation of B lymphocytes into plasma cells, was uniformly expressed in tumors in the oral cavity and in about one half of extraoral cases. None of the tumors expressed PAX5, and 1 expressed CD20. CD79a expression was uniformly absent in tumors with monomorphic cytologic features and expressed in about one half of the tumors with plasmacytic differentiation. Monotypic light chain expression was variable. IgA and IgG immunoglobulin heavy chains were expressed in 3 and 1 of the tested tumors, respectively. All cases expressing IgA exhibited plasmacytic differentiation. CD56 was expressed by 2 PBL tumors. This antigen is frequently expressed in plasma cell myeloma and rarely in extramedullary plasmacytoma and has been reported in cases of PBL.6,7,12 Because a subset of diffuse large B-cell lymphomas (ALK+) can exhibit morphologic features similar to those of PBL,1 we analyzed ALK expression in our cases. None of the tumors expressed ALK. Similarly, none of the tumors expressed BCL1 (cyclin D1), which can be found in extramedullary multiple myeloma.13

Overall, 6 of 13 tumors were EBER+. In HIV+ patients, 5 of 7 tumors were EBER+. EBER was expressed in 5 of the 6 tumors in the oral mucosa, including 4 of 5 tumors with monomorphic cytologic features Image 3C. Only 1 of 7 extraoral cases was EBER+. The Ki-67 proliferation index Image 3D was, on average, higher in EBER+ tumors compared with EBER– tumors.

FISH for MYC/IGH rearrangement [t(8;14)] was performed in 3 cases with residual available paraffin tissue. MYC translocation was detected in the 1 case in the oral cavity with plasmacytic differentiation. This patient was HIV+. No MYC translocation was detected in 2 extraoral PBL tumors with plasmacytic differentiation. One patient was HIV+ and the other, HIV–.

Clinical Manifestations

The main clinical characteristics of all patients are summarized in Table 2. The median age was 40 years (range, 4–52 years) with no marked difference between subtypes. There was an overall slight male predominance (8:5, ratio 1.6:1). Of 6 cases with PBL of the oral mucosa, 5 were HIV+. In contrast, extraoral PBL usually occurred in HIV– patients (5 of 7). HIV infection was detected around the time of lymphoma diagnosis in 3 patients with PBL of the oral mucosa, while a long history of HIV infection was present in the remaining 5 patients, ranging from 6 to 16 years before lymphoma diagnosis. At the time of lymphoma diagnosis, the CD4 count ranged between 57 and 458/μL. Only 2 patients received highly active antiretroviral therapy before the lymphoma diagnosis.

View this table:
Table 1

Patients with extraoral PBL had varied sites of occurrence that included skin, lymph nodes, nasopharynx, terminal ileum, and cecum. Two patients with extraoral PBL had a history of multivisceral transplantation and had undergone long-term immunosuppression before the lymphoma diagnosis. Only 1 patient with extraoral PBL had stage I disease. All remaining patients for whom complete staging documentation was available (n = 11) had stage III or IV disease. The bone marrow and central nervous system were involved at the time of diagnosis in 2 patients each, and both were observed in patients with extraoral PBL. Four patients had a low or low-intermediate IPI (0–2), and 6 patients had a high-intermediate or high-risk IPI (3–4). IPI information was not available for 3 patients. Of the 5 patients with PBL of the oral mucosa and known IPI scores, 3 had low-low-intermediate IPI, while 4 of the 5 patients with extraoral PBL and known IPI scores had high-intermediate-high IPI.

Image 3

Immunohistochemical phenotype of plasmablastic lymphoma. All tumors expressed post–germinal center B-cell antigens CD138 (A, ×200) and MUM1 (B, ×200). Epstein-Barr virus–encoded RNA was expressed in 4 of 5 tumors with monomorphic cytologic features and in 2 of 8 tumors exhibiting plasmacytic differentiation (C, ×200). The Ki-67 proliferation index was high (>90%) in the majority of tumors (D, ×200).

Clinical Outcome

Initial therapy was recorded for 11 patients. Therapy was not standardized and was given at the discretion of the treating physicians (Table 2). One HIV+ patient with PBL of the oral mucosa refused any treatment (case 2), and 1 patient was treated at another institution (case 13). The most commonly used chemotherapy regimens included etoposide, vincristine, doxorubicin, cyclophosphamide, and prednisone (EPOCH) in 4 patients; hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone in 3 patients; and methotrexate with zidovudine alone or alternating with other chemotherapy agents in 4 patients. Patients with PBL of the oral mucosa more frequently received EPOCH and methotrexate with zidovudine regimens, which are commonly used for the treatment of HIV-associated lymphomas. However, there was no marked difference in treatment among patients whose tumors were morphologically classified as PBL with monomorphic plasmablastic cytologic features vs PBL with plasmacytic differentiation.

Complete response was achieved in 4 patients, 3 patients with PBL of the oral mucosa and 1 with extraoral PBL. Morphologically, 2 of these tumors exhibited monomorphic plasmablastic cytologic features, and 2 exhibited plasmacytic differentiation. Five patients had disease progression during treatment. Of 11 treated patients, 7 have died and 3 are alive, all in continuous complete remission with a median follow-up of alive patients of 15 months (range, 13–17 months). One patient discontinued follow-up at 6 months. The median survival of all patients was 6 months (range, 2–17 months) Figure 1A. There was a statistical difference in OS between patients with oral and extraoral PBL (P = .02) Figure 1B. The median survival of patients with PBL with monomorphic plasmablastic cytologic features was not reached (range, 5–15 months), while a shorter median survival was observed in patients with PBL with plasmacytic differentiation (4 months; range, 2–17 months); however, this difference was not statistically significant (data not shown).

Discussion

PBL was initially characterized as an aggressive subtype of NHL most frequently arising in the jaw and oral cavity in patients with HIV infection, accounting for 2.6% to 3% of all NHLs in this patient population.5,14 However, some reports described patients with PBL with extraoral manifestations, heterogeneity in histologic findings, and variable association with immunodeficiency states.7,8,10,1518

The 13 patients we characterized in this study confirm the observation that PBL represents a heterogeneous group of NHLs with distinct clinical features, pathologic characteristics, and outcome. Clinically, we found that oral and extraoral manifestations of PBL occurred in the presence and absence of HIV infection and were associated with distinct patient outcomes. There were also differences in histopathologic features between these clinical subtypes; ie, plasmacytic differentiation was prominent in extraoral PBL but was present in only 1 of the oral PBLs. While tumors morphologically classified as PBL with monomorphic plasmablastic cytologic features or with plasmacytic differentiation expressed CD138, MUM1, and XBP1 and in all but 1 case lacked CD20 expression, they differed in expression of Ki-67, which was uniformly high in PBL with monomorphic plasmablastic cytologic features but was variable in PBL with plasmacytic differentiation. In addition, 2 cases of PBL with plasmacytic differentiation expressed CD56, which was not expressed in monomorphic PBL. None of these patients had a history or evidence of multiple myeloma. These findings suggest that expression of CD56 is more in keeping with the expression of a terminally differentiated B-cell phenotype rather than an antecedent diagnosis of plasma cell myeloma. The histologic and immunohistochemical findings, therefore, suggest the presence of pathologically distinct subtypes of PBL that correlate with clinical classification based on the site of initial occurrence and with distinct epidemiology of these lymphomas.

The association between PBL and HIV infection has been firmly established, with highest prevalence (80%) in the oral/monomorphic plasmablastic subtype and 30% to 50% in the extraoral/plasmacytic subtype in a large series of 50 patients studied by Colomo et al.7 Similarly, in our series, 5 of 6 cases of PBL of the oral mucosa were associated with HIV infection, and all but 1 of these had monomorphic plasmablastic cytologic features without plasmacytic differentiation. In contrast, most of the extraoral PBL tumors occurred in HIV– patients and uniformly exhibited plasmacytic differentiation.

A number of clinicopathologic reports have implicated a potential role for latent infection with EBV in the pathogenesis of PBL, with positivity ranging from 15% to 80% by in situ hybridization assay.5,6,8,10,15,19 In our cohort, EBER staining showed an association with clinical classification of these cases as PBL of the oral mucosa (5 of 6 cases), a pathologic finding of monomorphic plasmablastic cytologic features (4 of 5 cases), and also a strong association with HIV infection. Our results are in keeping with findings of other published case series using similar methods and demonstrate that the majority of PBLs of the oral mucosa are associated with HIV and EBV infections. However, HIV infection is less common in extraoral PBL with plasmacytic differentiation, thus suggesting that other predispositions or etiologic factors are likely to be involved in the pathogenesis of this variant.

Indeed, we observed a remarkable variety of underlying medical conditions among the patients with extraoral PBL, including solid organ transplantation with long-term immunosuppression and autoimmune disease. Similar associations with long-term immunosuppression have been described in the literature, including patients with heart and kidney transplants, systemic lupus erythematosus, ulcerative colitis, and chronic lymphocytic leukemia.7,8,20,21

Figure 1

Kaplan-Meier curves of patients with plasmablastic lymphoma (PBL). A, Overall survival of all patients with PBL. B, Overall survival of patients with PBL of the oral mucosa (solid line) and extraoral PBL (interrupted line) (P = .02).

Owing to the low incidence of PBL and lack of clinical trials to guide management decisions, there is no standard treatment approach for this rare subtype of lymphoma. Our observations confirm that both clinical and histologic variants of PBL are characterized by a generally aggressive clinical course, relative resistance to a variety of chemotherapy regimens, and an overall unfavorable outcome. A statistical difference in OS was observed between patients with oral and extraoral disease manifestations. Our results are in line with findings from a large meta-analysis of 98 patients by Rafaniello Raviele et al,22 who reported a prevalence of disease-related deaths of 59.6% during a mean period of 10.4 months from diagnosis for the oral type and 58.6% during a mean period of 6.2 months for the extraoral type. In our cohort, the median OS of monomorphic PBL was not reached compared with 4 months for PBL with plasmacytic differentiation; this difference was not statistically significant. The absence of a statistical difference in OS may be attributed to the small number of cases in each variant.

In our study, there was no difference in survival between HIV+ and HIV– patients. The death rate for HIV-infected patients in the study by Rafaniello Raviele et al22 was 53% during a mean period of 10.4 months from diagnosis. Their data also confirm the observation of prior case series indicating that the prognosis of HIV-associated PBL can be significantly improved by the addition of highly active antiretroviral therapy to chemotherapy.8,23,24 On the other hand, early clinical studies have shown promising results using EBV-directed therapy with high-dose zidovudine, interleukin-2, and ganciclovir in patients with HIV-associated lymphomas.25,26 Further studies are needed to examine whether PBL of the oral mucosa and extraoral PBL exhibit distinct responses to therapy and prognoses and to determine the optimal therapy for patients with PBL.

The MYC translocation was recently reported in 10 cases of PBL.2731 These cases were reported in oral mucosa (1 case) and extraoral tumors (9 cases), all from HIV+ patients. Analysis for the MYC translocation in 3 of our cases with available tissue disclosed its presence in a single case of PBL of the oral mucosa in an HIV+ patient. The MYC translocation was not detected in 2 other extraoral tumors. Compilation of currently available cases may suggest a higher frequency of MYC translocation in HIV+ patients with extraoral PBL and, thus, may contribute to the shorter survival of the patients observed in our study. Additional studies are needed to confirm a higher frequency of MYC translocations in extraoral PBL compared with oral PBL and to examine whether the MYC translocation can be found in PBL tumors in HIV– patients.

PBL represents a heterogeneous subtype of LBCL with 2 distinguishable clinical variants displaying distinct histopathologic features, viral associations, and outcomes. Future investigations based on this subclassification should be designed to help elucidate the specific pathogenetic mechanisms involved in malignant transformation and lead to identification of molecular targets in the treatment of these rare lymphoid neoplasms.

CME/SAM

Upon completion of this activity you will be able to:

  • describe the morphologic features of plasmablastic lymphoma (PBL) presenting in oral and extraoral sites.

  • discuss the utility of immunohistochemistry in the diagnosis of PBL, and its utility for differentiation between PBL presenting in oral and extraoral sites.

  • discuss the utility of the addition of antiviral therapy to chemotherapy for treatment of PBL.

The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module.

The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Questions appear on p 846. Exam is located at www.ascp.org/ajcpcme.

Footnotes

  • Supported by grants RO1 CA109335 and RO1 CA122105 from the US Public Health Service, National Institutes of Health, Bethesda, MD; and the Dwoskin Family Foundation, Miami, FL.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
View Abstract