OUP user menu

The Role of CD11c Expression in the Diagnosis of Mantle Cell Lymphoma

Teresa S. Kraus MD, Christine N. Sillings MD, Debra F. Saxe PhD, Shiyong Li MD, PhD, David L. Jaye MD
DOI: http://dx.doi.org/10.1309/AJCPOGCI3DAXVUMI 271-277 First published online: 1 August 2010

Abstract

Flow cytometric immunophenotyping (FCI) aids in the differentiation of chronic lymphocytic leukemia (CLL) from mantle cell lymphoma (MCL); however, overlapping phenotypes may occur. CD11c expression has been reported in up to 90% of CLL cases but has rarely been reported in MCL. Whether CD11c can be used to exclude MCL has not been directly addressed. FCI reports were reviewed for 90 MCL cases (44 patients) and 355 CLL/small lymphocytic lymphoma (SLL) cases (158 patients). MCL cases were confirmed by cyclin D1 immunoreactivity and/or t(11;14) detection by karyotyping or fluorescence in situ hybridization. Cases with typical MCL immunophenotypes did not express CD11c. The 2 MCL cases displaying dim CD11c positivity (2 of 44 patients) expressed other markers not typical of MCL. CD11c was detected in 96 (27.0%) of 355 cases of CLL/SLL representing 53 of 158 patients. CD11c expression is rare in MCL and may aid in differentiation of CD5+ B-cell neoplasms, particularly when small samples limit further ancillary testing.

Key Words:
  • Chronic lymphocytic leukemia
  • Mantle cell lymphoma
  • CD11c
  • Flow cytometry

Chronic lymphocytic leukemia (CLL) and mantle cell lymphoma (MCL) are B-cell neoplasms that usually express CD5.1,2 Differentiation of these entities is essential because MCL generally has a more aggressive clinical course35 and is therefore often managed differently from CLL.3,5

The characteristic cytogenetic abnormality in MCL is the t(11;14)(q13;q32), involving the CCND1 and IGH genes. The resulting increase in expression of cyclin D1, encoded by CCND1, promotes the transition from the G1 to the S phase of the cell cycle.6 MCL is definitively diagnosed by demonstration of the t(11;14) by conventional cytogenetics, fluorescence in situ hybridization (FISH), or polymerase chain reaction (PCR) or by demonstration of cyclin D1 expression by immunohistochemical analysis in a CD5+ small B-cell lymphoma.7,8

Flow cytometric immunophenotyping (FCI) aids in the differentiation of CLL from MCL. The classic CLL immunophenotype is CD5+, CD20dim, surface immunoglobulin (sIg)dim, CD23+, and FMC7–,3,4,9,10 whereas MCL is typically CD5+, CD20bright, sIgbright, CD23–, and FMC7+.3,4,9,10 However, ambiguous or overlapping phenotypes have been described.1,5,11,12 The identification of additional markers that are differentially expressed in MCL and CLL would aid in the flow cytometric differentiation of these 2 entities.

Expression of the cellular adhesion molecule CD11c has been reported in 4% to 89% of CLL cases11,13 but has rarely been reported in MCL.11,14 Whether CD11c expression can be used to exclude a diagnosis of MCL has not been directly addressed using current diagnostic standards for MCL.8 Therefore, in this study, we identified a set of MCL cases for which definitive diagnoses were made using recent World Health Organization standards and determined the frequency of CD11c expression in these confirmed cases of MCL.

Materials and Methods

Patients

Patients were evaluated at Emory University Hospital, Atlanta, GA, from January 1999 to September 2008. Cases of MCL were selected from our archives based on the following criteria: (1) A pathologic diagnosis of MCL was rendered and/or provided in the clinical history. (2) FCI was performed in our laboratory. (3) The t(11;14) was detected by karyotyping and/or FISH and/or positive immunohistochemical staining for cyclin D1 was demonstrated at our institution. Cases of CLL/small lymphocytic lymphoma (SLL) were selected based on final pathologic diagnosis. Only CLL/SLL cases for which the absence of the t(11;14) was confirmed by cytogenetics, FISH, or absent cyclin D1 immunoreactivity were included in this study. Initial samples analyzed at our institution from which the diagnosis of MCL or CLL/SLL was determined or confirmed are designated as “initial diagnostic cases” for the purposes of this study.

Cytogenetic and FISH Studies

A subset of patients diagnosed with MCL had documented evidence of the t(11;14) by conventional cytogenetics or FISH studies. Bone marrow and blood samples submitted for karyotyping were prepared from 24- and 48-hour unstimulated cultures following standard methods. Solid tissue samples were disaggregated and cultured by conventional methods. GTG-banded karyotypes were reported in accordance with the International System for Human Cytogenetic Nomenclature.15

FISH studies were performed on fresh tissue or paraffin-embedded tissue using an LSI IGH/CCND1 XT dual-color, dual-fusion probe set (Abbott Molecular, Des Plaines, IL) according to the manufacturer’s instructions. At least 2 technologists scored analyzable interphase cells of each specimen, recording all patterns observed, for a total of 200 nuclei. Results were considered positive when the percentage of cells containing the IGH/CCND1 rearrangement exceeded the established, validated normal cutoff values (>1.5% for cultured cells and >10% for formalin-fixed, paraffin-embedded tissues).

Flow Cytometric Immunophenotyping

Antibodies were used according to the manufacturers’ instructions and purchased from Becton Dickinson (San Jose, CA) unless otherwise noted. The antibody combinations were CD14-fluorescein isothiocyanate (FITC)/CD13-phycoerythrin (PE)/CD34-allophycocyanin (APC)/CD45-peridinin chlorophyll protein (PerCP), HLA-DR-FITC/CD25-PE/CD33-APC/CD45-PerCP, CD3-FITC/CD4-PE/CD8-APC (Pharmingen, San Diego, CA)/CD45-PerCP, CD2-FITC/CD7-PE (Pharmingen)/CD5-APC (Pharmingen)/CD45-PerCP, CD10-FITC/CD19-PE/CD34-APC/CD45-APC, CD36-FITC (Beckman Coulter, Fullerton, CA)/CD117-PE/CD34-APC/CD45-PerCP, CD15-FITC/CD11b-PE/CD34-APC/CD45-PerCP, CD103-FITC (DAKO, Carpinteria, CA)/CD22-PE/CD11c-APC/CD45-PerCP, CD16-FITC (Beckman)/CD56-PE/CD38-APC/CD45-PerCP, FMC7-FITC (Beckman)/CD23-PE/CD10-APC/CD45-PerCP, κ sIg-FITC (DAKO)/CD20-PE/CD19-APC/CD45-PerCP, λ sIg-FITC (DAKO)/CD20-PE/CD19-APC/CD45-PerCP, κ sIg-FITC/λ sIg-PE/CD5-APC/CD20 PerCP, and corresponding isotype controls. Data were acquired using a FACSCalibur cytometer or FACSCanto II cytometer (BD Biosciences, San Jose, CA) and analyzed using CellQuest software, version 3.3 (BD Biosciences).

FCI reports for all cases were reviewed by one of us (C.N.S.), and expression patterns and intensities for CD11c, CD20, CD22, CD23, FMC7, and sIg were recorded for each MCL case. “Dim” staining of a population was defined on dot plots when the population fluorescence intensity shifted above but showed overlap with the isotype control staining. “Bright” staining of a population was defined as a mean fluorescence intensity of staining of the population of interest at least 2 logs above the isotype control. “Partial” staining was recorded when a subset of the cells of interest showed a fluorescence intensity at least a log above that of the isotype control.16,17

Immunohistochemical Analysis

Formalin-fixed, paraffin-embedded tissue sections were immunostained using a rabbit monoclonal anti–cyclin D1 antibody according to the manufacturer’s instructions (SP4; dilution 1:70, NeoMarkers, Fremont, CA). Bound antibody was detected using the DAKO Envision+ dual link system. Sections were counterstained with hematoxylin before mounting for microscopy.18

Results

Patients

To investigate the value of CD11c expression as a basis to exclude MCL, we identified cases of MCL confirmed by current diagnostic criteria from our archives. As shown in Table 1, 44 patients with MCL and 158 with CLL/SLL meeting the criteria described were identified. A male predominance was noted in patients with MCL and with CLL/SLL. The median age in both groups was 65 years. Patients diagnosed with MCL ranged in age from 39 to 91 years and patients diagnosed with CLL/SLL ranged in age from 33 to 88 years. Patients with MCL were followed up for up to 44 months, and patients with CLL/SLL were followed up for up to 106 months.

Sites of Involvement

Samples were obtained from a broad spectrum of anatomic sites, including lymph nodes, bone marrow, blood, spleen, parotid gland, adenoids, colon and rectum, soft tissue, pleura and pleural fluid, and hematopoietic progenitor cells obtained by apheresis (summarized in Table 1). The sites of involvement were similar between groups and similar to those reported in the literature, suggesting that these clinical features do not readily differentiate MCL from CLL.

Confirmatory Laboratory Studies

All MCL cases included in the study had documented evidence of the t(11;14) by conventional karyotyping or FISH studies and/or had positive immunohistochemical results for cyclin D1 on neoplastic cells. In 25 (28%) of 90 FCI cases representing 16 patients, the t(11;14) was identified in the same sample by conventional cytogenetics or FISH. In 13 (14%) of 90 cases representing 9 patients, positivity was displayed by cyclin D1 immunohistochemical studies in neoplastic cells. Two patients had samples that showed positive FISH results and positive cyclin D1 immunohistochemical results. The remaining 19 patients had diagnostic material not included in this study that was positive for one or more of these confirmatory tests.

The absence of the t(11;14) was confirmed in the CLL/SLL cases by cytogenetic studies (89 patients), FISH (35 patients), cytogenetics and FISH (30 patients), or by negative cyclin D1 immunohistochemical results (4 patients). The presence of one or more cytogenetic abnormalities commonly associated with CLL, including del 13q14, trisomy 12, del 11q22-23, and del 17p13,8 was detected in 50 patients.

Immunophenotyping Studies

A total of 90 MCL FCI cases from 44 patients were studied. On average, there were 2 FCI studies per patient (range, 1–10 cases). In 82 studies (91%), the following pattern of expression was shown: CD20+, FMC7+, CD23–, and sIg+. In 22 cases (24%), there was bright expression of CD20 and bright sIg expression characteristic of MCL.

In 6 cases (7%) of MCL, there was positive or dim expression for CD23, each showing coexpression with FMC7. In 2 cases (2%), results were negative for FMC7 expression; however, in each of these cases, the clonal population represented less than 1% of the total population of cells analyzed. Both of these cases were negative for CD23 expression, and one was also negative for CD20 expression.

Dim CD11c expression was detected in 2 (2%) of the total MCL cases (5% of initial diagnostic cases), representing 5% of patients (2/44). Of note, both cases displayed unusual phenotypes for MCL, with one case showing aberrant expression of CD7 Image 1, and the other aberrantly expressing CD56 Image 2. Both displayed bright expression of sIg and positive expression of CD20. The clinicopathologic features of the CD11c+ MCL cases are listed in Table 2. Of note, the case with aberrant CD7 expression was diagnosed as MCL, pleomorphic blastoid variant.

View this table:
Table 1
Image 1

Flow cytometric dot plots of a case of mantle cell lymphoma, pleomorphic blastoid variant with expression of CD7, dim CD11c, and bright CD20 and surface immunoglobulin. The t(11;14) was detected by conventional cytogenetics. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein.

When only FCI studies at the initial patient encounter are considered, the expression of CD5, CD11c, CD23, FMC7, and sIg fall into 5 distinct patterns Table 3. The predominant pattern, seen in 39 cases (89%), is characterized by expression of CD5, FMC7, and sIg, without detectable CD23 and CD11c. There were 2 cases (5%) that differed from the aforementioned pattern in the coexpression of CD23. Less common patterns, each representing 2% of initial diagnostic cases (1/44), included 1 CD5– case, 1 CD11c+ case, and 1 case that expressed all 5 markers (including CD11c). Cases initially lacking CD11c expression did not become positive in 46 follow-up samples at up to 44 months (median, 22 months).

Image 2

Flow cytometric dot plots of a case of mantle cell lymphoma with expression of dim CD11c, dim CD56, and bright CD20 and surface immunoglobulin. The t(11;14) was detected by fluorescence in situ hybridization. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein.

View this table:
Table 2

A total of 355 FCI reports from 158 patients with CLL/SLL were reviewed. The number of FCI cases averaged 2 per patient (range, 1–15). In 246 total cases (69.3%) and 116 initial diagnostic cases (73.4%), there was the characteristic CLL/SLL phenotype (CD20+, CD23+, sIg+, and FMC7–). In 69 (19.4%) of the total CLL/SLL cases, there were varying degrees of FMC7 expression, ranging from dim to intermediate intensity.

CD11c expression was identified in 96 CLL studies (27.0%) representing 53 (33.5%) of 158 patients. In contrast with the MCL cohort, a subset of patients with CLL demonstrated variably detectable CD11c expression at up to 106 months’ follow-up.

Discussion

CLL and MCL are CD5+ B-cell lymphomas that generally differ significantly in terms of prognosis. CLL tends to follow an indolent course, with median survival ranging from 95 to 293 months depending on immunoglobulin gene mutational status.1,8 In contrast, MCL becomes more clinically aggressive and treatment-refractory over time,19 with a median survival of 36 to 60 months.1,8 Therefore, differentiation of these entities has important prognostic and therapeutic implications.

View this table:
Table 3

FCI is often used to differentiate MCL from CLL and other low-grade B-cell lymphomas. The most common useful markers in their differentiation are CD23 and FMC7.2,3,5,20 MCL is generally CD23– and FMC7+, with CLL showing the opposite pattern. When aberrant phenotypes occur, flow cytometric differentiation of CLL from MCL can be difficult.5 In our study, 85 (23.9%) of the total CLL/SLL FCI studies showed varying degrees of FMC7 expression. The incidence of CD23 expression among our CLL cases was similar to that reported in a review by Wohlschlaeger et al,7 wherein 27 (10.9%) of 247 MCL cases were CD23+. In a review by Schlette et al,12 CD23 expression was identified in up to 45% of MCL cases. Similarly, Palumbo et al5 reported CD23 expression in 6 (43%) of 14 MCL cases.

Other nonspecific phenotypes, such as CD20bright/sIgdim or CD20dim/sIgbright, may also occur.11 Even the “classic” flow cytometric phenotype may be misleading, as Ho et al2 reported that 9 of 28 cases with FCI phenotypes suggestive of MCL lacked t(11;14) by FISH and instead had cytogenetic abnormalities more typical of CLL, including trisomy 12 and deletion 11q22-23.

Because of the potential for immunophenotypic overlap, the diagnosis of MCL must be confirmed by additional tests, such as immunohistochemical staining for cyclin D1 or demonstration of the t(11;14)(q13;q32) by karyotyping, PCR, or FISH. However, immunohistochemical staining for cyclin D1 generally uses fixed tissue sections,5 and results may be falsely negative if the tissue quality is suboptimal.6 In addition, rare cyclin D1–negative cases of MCL have been reported,1,6 some of which were confirmed by gene expression profiling.21 FCI is widely available, relatively easy to perform, and cost-effective; therefore, the identification of additional markers to improve its specificity would be beneficial.

View this table:
Table 4

CD11c (αx integrin) is a member of the superfamily of glycoproteins that mediate cell-cell and cell-matrix interaction22 and is expressed on monocytes, neutrophils, and some lymphocyte subsets.23 Antibodies to detect CD11c are routinely used in clinical FCI. Expression of CD11c has been reported in 4%11 to 90%13 of CLL cases. In our study, CD11c was detected in 27.0% of CLL FCI cases representing 33.5% of patients with CLL/SLL (53/158). These results are similar to those reported by Angelopoulou et al,24 who reported CD11c expression in 7 (22%) of 32 patients with CLL. In other studies, however, the incidence of CD11c expression in CLL was greater than 80%.13,20 In their review of the literature, Angelopoulou et al24 cited 3 series (89 cases) in which CD11c expression was reported in 20% or fewer CLL cases (range, 6%–20%) and 6 series (376 cases) in which CD11c expression was present in at least 37% of CLL cases (range, 37%–90%).

In contrast, CD11c expression seems to be rare in MCL, occurring in 2 of 44 patients in our study (representing 2 of 90 FCI studies) and in 7 of 130 MCL cases reported in the literature (5.4%), as cited in Table 4.11,14,20,22,2429 The presence of cyclin D1 immunoreactivity and/or detection of t(11;14) by karyotyping, FISH, PCR, or Southern blot was confirmed in 43 of these MCL cases. However, results of such confirmatory studies were not reported for the 7 CD11c+ cases. Thus, it is possible that some of these cases would not be diagnosed as MCL using more recent diagnostic criteria.

In our population, approximately 5% (2/44) of initial diagnostic cases of MCL expressed CD11c (2% of total MCL cases). When all analyzed CLL and MCL cases are considered, MCL cases account for 2 of 98 cases that express CD11c. Thus, in our series, the likelihood that a CD5+/CD11c+ small B-cell lymphoma was MCL was approximately 2%.

To our knowledge, our study presents the largest cohort of MCL confirmed by cyclin D1 immunoreactivity or detection of t(11;14) in which CD11c expression was specifically examined. In our series, CD11c expression was not detected in MCL cases that showed an otherwise typical MCL phenotype (0/34 patients). Although the CD11c+ MCL cases exhibited some immunophenotypic features that were typical of MCL, including expression of CD5 and bright expression of CD20 and sIg, both displayed aberrant markers not typical of MCL or CLL.

It is worth noting that cases of cyclin D1–negative MCL have been reported.30,31 Fu et al30 described 6 cases of MCL that lacked the t(11;14) by FISH and failed to express cyclin D1 protein. The morphologic features and gene expression profiles in these cases were otherwise consistent with MCL.30

Rare cases of CD5+ marginal zone lymphoma (MZL) have also been described.32,33 In a 2003 case report and literature review, Batstone et al33 identified 14 CD5+ MZL cases. Such cases could potentially be misdiagnosed as CLL or MCL based on their immunophenotype. The possibility that some of the cases in our series that were classified as CD11c+ CLL actually represented CD5+ MZL or cyclin D1–negative MCL cannot be completely ruled out; however, these entities are rare and unlikely to have contributed significantly to our study population.

The main finding in our study is that when CD11c expression is detected by FCI in a CD5+ B-cell neoplasm, the diagnosis of cyclin D1– and/or t(11;14)-positive MCL is highly unlikely. In combination with other known diagnostic criteria, the detection of CD11c in a CD5+ clonal B-cell population supports the diagnosis of CLL. Thus, the inclusion of CD11c in routine FCI panels is a simple tool for evaluating CD5+ B-cell neoplasms, particularly when further ancillary testing is precluded by limited samples.

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.
  32. 32.
  33. 33.
View Abstract