SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Few studies have statistically investigated reduced CD20 expression in B-cell lymphoma after rituximab therapy and genomic mutation of CD20 associated with reduction. We examined CD20-positive rate in follicular lymphoma (FL) and diffuse large B-cell lymphoma (DLBCL) by flow cytometry (FCM) and immunohistochemical staining (IHS), comparing 138 cases after rituximab therapy with 360 initial, not yet treated cases. Sequence analysis of exons 3 to 8 of CD20 was performed on 22 cases with low CD20-positive rate after rituximab treatment. The results showed a statistical correlation between CD20-positive rate in FCM and IHS. By FCM, the CD20-positive rate among post-rituximab cases was significantly lower than among initial cases in DLBCL, non-germinal center origin B-cell type (average values [avg] 57.8 and 87.9, respectively) (P < 0.0001), FL2 (avg, 93.9; 103.2) (P = 0.0083), and FL3A (avg, 90.6; 100.7) (P = 0.033). Stratified analyses of post-rituximab cases showed significantly lower CD20-positive rate in cases that were resistant at the start of the treatment and cases with progressive disease during rituximab therapy before biopsy. Sequence analysis showed silent mutation of exon 4 (632 C/T) in seven cases, although this number was not statistically significant. These results suggest the influence of B-lymphoma subtype and a therapeutic effect before biopsy on CD20 expression at relapse and contribute to a better therapeutic approach for relapse cases after rituximab therapy. (Cancer Sci, doi: 10.1111/j.1349-7006.2012.02307.x, 2012)

Rituximab is a chimeric monoclonal antibody recognizing the CD20 antigen on B-lymphocytes.[1] Additions of rituximab to both conventional chemotherapy and rituximab monotherapy have achieved high response rates against B-cell non-Hodgkin lymphoma (B-NHL).[2, 3] The cytotoxic effects of rituximab, due to antibody-dependent cellular cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC), and direct induction of apoptosis, require the binding of rituximab to CD20 antigen.[4] Low CD20 expression was shown to reduce tumor lysis and induce resistance to rituximab in vitro analyses.[5, 6] In cases of diffuse large B-cell lymphoma (DLBCL), reduced CD20 expression was reported to be associated with reduced survival.[7] These studies have verified the importance of CD20 expression for the therapeutic effect of rituximab that was expected from knowing its mechanism of action.

Several studies have been carried out regarding reduced CD20 expression in B-cell lymphoma cells after exposure to rituximab. Takei et al.[8] showed that lymphoma cells with reduced CD20 expression were selected after impregnation of rituximab and complement in rituximab-resistant B-lymphoma cell lines. D'Auria et al.[9] reported that CD20 expression in four patients diminished after rituximab treatment in eight patients with chronic lymphocytic leukemia. Similar results were observed in five of 19 patients with B-cell lymphoma by Hiraga et al.[10] It is suggested from these results that rituximab therapy might decrease CD20 expression in B-cell lymphoma cells. However, no research, to our knowledge, has statistically investigated loss of CD20 expression in B-cell lymphoma cells after rituximab treatment or the degree of that reduction.

Some studies have recently investigated the possibility of an association between genetic mutation and reduced CD20 expression after rituximab treatment. Hiraga et al.[10] reported that when genomic sequence analysis of exons 3 to 8 of CD20 was conducted in five patients with reduced CD20 expression after rituximab therapy of B-cell lymphoma, two patients had point mutations, including serine 97 to phenylalanine (S97F; TCC[RIGHTWARDS ARROW]TTC) and valine 247 to isoleucine (V247I; GTT[RIGHTWARDS ARROW]ATT). CD20 expression was found to be significantly lower in cases with C-terminal deletion mutations among cases of B-cell lymphoma resistant to rituximab.[11] However, reproducible and specific mutations in CD20 associated with reduced CD20 expression have never been detected in B-cell lymphoma. This issue is still controversial because of insufficient investigation.

In this study, we statistically compared CD20-positive rates in lymphoma cells between initial cases and cases after rituximab treatment by flow cytometry (FCM) and immunohistochemical staining (IHS). The factors associated with the lower CD20-positive rate were also investigated. Sequence analysis of exons 3 to 8 of CD20 in cases with low CD20-positive rates after rituximab treatment was performed to research potential gene mutation.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Tissue samples

We reviewed samples from 138 cases diagnosed as follicular lymphoma (FL) or diffuse large B-cell lymphoma (DLBCL) in 2006 that had received at least one course of rituximab before biopsy and 360 cases initially diagnosed as FL or DLBCL from 2008 to 2010 in the Department of Pathology, Kurume University. Fresh materials were used for FCM. Paraffin-embedded tissues were used for diagnosis and IHS. All cases were reviewed by hematopathologists (OK, ND, and MH) and diagnosed according to the World Health Organization (WHO) classification. Clinical information was obtained by reviewing the patients' medical charts. The use of materials and clinical information was approved by the Research Ethics Committee of Kurume University and was in accordance with the Declaration of Helsinki.

Evaluation methods for determining CD20-positive rates by FCM and IHS

Flow cytometry analysis was performed with a flow cytometer (FACS-Calive, Becton-Dickinson, Mountain View, CA, USA) and the Cell Quest software program (Becton-Dickinson) according to conventional methods described previously.[12] Calculation of positive rates of CD20 expression in lymphoma cells was performed as follows: cells were gated using forward scatter, side scatter, and CD45 expression. The proportion of CD20-positive cells in the gated cell population was divided by the proportion of CD19-positive cells or the proportion of cells positive for the tumor-specific light chain. The antibodies (clones) used for FCM were CD19 (B4-RD1) (Beckman Coulter, CA, USA), CD20 (B-Ly1) (Dakocytomation, Glostrup, Denmark), kappa (rabbit polyclonal) (Dakocytomation), and lambda (rabbit polyclonal) (Dakocytomation).

Immunohistochemical staining was performed on paraffin sections. Assessment of CD20-positive rates was performed using L-26 antibody (Dakocytomation) except in 12 initial cases. If a neoplastic cell was near-circumferentially and moderately to strongly stained compared with positive control non-neoplastic B-cells, the cell was scored as positive. The proportion of CD20-positive cells among the neoplastic cells was used as the positive rate. In these assessments, immunostaining with CD79a antibody (JCB117) (Dakocytomation) was used as necessary. The CD20-positive rates in neoplastic cells in each case were immunohistochemically scored as follows: 0, 0–25% positive; 1, 25–50% positive; 2, 25–75% positive; 3, 75–90% positive; 4, 90–100% positive (Fig. 1).

image

Figure 1. Classification of CD20-positive rate by immunohistochemical staining (IHS) [L-26 antibody; original magnification, ×400]. Panels show examples of cases categorized in each score: (A) score 0, with positive rate less than 25%; (B) score 1, with positive rate from 25% to 50%; (C) score 2, with positive rate from 50% to 75%; (D) score 3, with positive rate from 75% to 90%; (E) score 4, with positive rate more than 90%. When a neoplastic cell was near-circumferentially and moderately to strongly stained, the cell was considered positive.

Download figure to PowerPoint

Evaluations of antigens other than CD20

When the positive rates of antigens other than CD20 were evaluated, FCM and/or IHS were performed. For FCM analysis, antibodies (clones) against CD5 (T1) (Beckman Coulter) and CD10 (J5) (Beckman Coulter) were used. Antibodies against CD5 (4C7) (Leica Microsystems, Wetzler, Germany), CD10 (56C6) (Leica Microsystems), bcl6 (P1F6) (Leica Microsystems), and MUM1 (MUM1p) (Dakocytomation) were used for immunohistochemistry. Except when evaluating CD20, tumor cells were considered positive if more than 30% of the neoplastic cells were positive. Hans' algorithm[13] was used to determine whether DLBCLs were of germinal center origin B-cell type (GC-type) or non-germinal center origin B-cell type (non-GC type) when not staining for CD5.

Sequence analysis

DNA preparation was performed from sections of paraffin-embedded tissues. DNA samples were extracted using a commercial kit (KAPA Express Extract, KAPA BIOSYSTEMS, Boston, MA, USA) following the manufacturer's recommendations.

CD20 exons and the flanking intronic sequences were amplified by semi-nested PCR with the oligonucleotide primers listed in Table 1. Polymerase chain reaction amplification was performed with KAPA2G Robust HotStart ReadyMix (KAPA BIOSYSTEMS) for first-round PCR, and AmpliTaq Gold PCR Master Mix (Applied Biosystems, Foster City, CA, USA) for the second round. The first-round PCR conditions were as follows: an initial denaturation at 95°C for 3 min was followed by 40 cycles (95°C for 15 s, 60°C for 5 s and 72°C for 30 s), with a final extension at 72°C for 10 min. The second-round PCR consisted of 40 cycles of 95°C for 30 s, 65°C for 30 s and 72°C for 30 s. The amplified products were evaluated on 3% agarose gels and visualized by ethidium bromide staining under ultraviolet light.

Table 1. Primers for semi-nested polymerase chain reaction (semi-nested PCR) amplification and sequencing of CD20 exons (shown 5′[RIGHTWARDS ARROW]3′)
Name ExonSequence (5′[RIGHTWARDS ARROW]3′)Semi-nested PCR product size
  1. Reference sequence NG_023388.1 (chromosome 11, 60223282-60238225). PCR, polymerase chain reaction.

EX3FFW common3ATGAC ACAAG GTAAG ACTGC CAAA313 bp
EX3R-OUTRV outerTATGT GGCCT ATACC GCATC AGC
EX3R-INRV innerCCTAT ACCGC ATCAG CTTCT GTC
EX4F-OUTFW outer4GGAGC CAGAG CTTCC AACCT TG276 bp
EX4F-INFW innerGCCAG AGCTT CCAAC CTTGT CTT
EX4RRV commonTCTGG CATAT CCCTG TGGAG CC
EX5FFW common5GTGAT CTCCC TCTCT CCTCT ATC231 bp
EX5R-OUTRV outerGCTGA GAGGC TGTGA TACCT TG
EX5R-INRV innerGAGAG GCTGT GATAC CTTGG CT
EX6FFW common6CCCTC CCAGA TTATG TTTTC CAAAG361 bp
EX6R-OUTRV outerTGCTC ATAAT AAACA AGTCA TCCTT C
EX6R-INRV innerTCATA ATAAA CAAGT CATCC TTCCT C
EX7FFW common7GTCAA CAATA ACTTA CTGAA CACCA A287 bp
EX7R-OUTRV outerATCAA GCTCC TAGTT TCAAC AACTC
EX7R-INRV innerGCTCC TAGTT TCAAC AACTC ATCA
EX8FFW common8TTGTG GAGAT TGTTG ACAAA GGTGT353 bp
EX8R-OUTRV outerAGTCA GCATG TCTCT TGGAA GCTA
EX8R-INRV innerCAGCA TGTCT CTTGG AAGCT ATGA

The PCR products were purified using the ExoSAP-IT Clean-Up method (Affymetrix USB, Cleveland, OH, USA). The DNA sequences of the PCR products were determined directly by using the primers listed in Table 1 with the BigDye Terminator version 1.1 Cycle Sequencing Kit (Applied Biosystems) according to the manufacturer's instructions and analyzed on an Applied Biosystems 310 DNA Sequencer (Applied Biosystems). CD20 exon gene sequences were compared with the reference sequence NM_152866.2 for CD20 cDNA. An amino acid sequence number of CD20 was referred to NP_690605.1.

Statistical analysis

For each group, the average positive rates for CD20 by FCM were calculated, and box-and-whisker plots were created. Student's t-test was used to compare these average values. A P-value of <0.05 indicated statistical significance.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Clinical features

The clinical features of 138 cases with at least one course of rituximab before biopsy (post-rituximab) and 360 cases at initial diagnosis (initial cases) are shown in Table 2. The number of males slightly exceeded that of females, and the median age was about 65 in both groups. High-intermediate risk or high risk by the International Prognostic Index (IPI) was present in 34.5% of post-rituximab cases and 55.0% of initial cases, although IPI information was available for only a limited number of cases. Post-rituximab cases had a lower proportion of DLBCL, non-GC type, and a higher proportion of follicular lymphoma – grade 1 (FL1) and grade 2 (FL2) – compared with initial cases.

Table 2. Clinical features of post-Rituximab and initial cases
CharacteristicsPost-RituximabInitial
  1. CHOP, CHOP regimen including doxorubicin, cyclophosphamide, vincristine and prednisone; CR(u), complete response or complete response uncertain; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; GC-type, germinal center origin B-cell type; H, high; H-int, high-intermediate; IPI, international prognostic index; non-GC type, non-germinal center origin B-cell type; PD, progressive disease; PR, partial response; R, rituximab; THP-COP, THP-COP regimen including cyclophosphamide, pirarubicin, vincristine, and prednisone.

Sex: Male/Female (unknown)80/58 (0)191/168 (1)
Age: (years) median [range]65 [43–90]66 [14–92]
IPI, H-int or H34.5% (29/84)55.0% (137/249)
Diagnosis
DLBCL, GC type20.3% (28/138)20.8% (75/360)
DLBCL, non-GC type15.9% (22/138)31.1% (112/360)
DLBCL, CD5 positive9.4% (13/138)8.1% (29/360)
FL, grade 18.7% (12/138)5.0% (18/360)
FL, grade 232.6% (45/138)19.2% (69/360)
FL, grade 3A10.9% (15/138)12.5% (45/360)
FL, grade 3B2.2% (3/138)3.3% (12/360)
Main treatment
R-CHOP71.7% (99/138)
R-THP-COP20.3% (28/138)
Rituximab, only2.2% (3/138)
Others including Rituximab5.8% (8/138)
Periods from initial diagnosis: median [range]29 [1–175] months
Number of Rituximab treatments: median [range]6 [1–16]
Number of relapses
0 (resistant to initial rituximab therapy)6.5% (9/138)
1st71.0% (98/138)
2nd15.2% (21/138)
3rd or more4.3% (6/138)
unknown2.9% (4/138)
Therapeutic effect of Rituximab treatment before biopsy
CR(u)71.0% (98/138)
PR17.4% (24/138)
PD8.0% (11/138)
unknown3.6% (5/138)

Most post-rituximab cases (92.0%) had mainly received the R-CHOP regimen, including doxorubicin, cyclophosphamide, vincristine, and prednisone, or the R-THP-COP regimen, including cyclophosphamide, pirarubicin, vincristine, and prednisone. The median period from initial diagnosis to biopsy was 29 months and ranged from 1 to 175 months. The median number of administrations of rituximab was 6 and ranged from 1 to 16. More than half of cases (71.0%) showed initial relapse at the time of biopsy. Among treatments that involved rituximab before biopsy, 71.0% of cases achieved complete response or were complete response uncertain (CR[u]), while 17.4% of cases showed partial response (PR) and 8.0% revealed progressive disease (PD).

Comparison of CD20-positive rate between FCM and IHS

Figure 2 indicates the relationships of CD20-positive rates in DLBCL and FL between FCM and IHS. The average CD20-positive rates by FCM analysis for each score determined by IHS were as follows: 0, 40; 1, 54.9; 2, 77.0; 3, 92.3; 4, 99.9. These average values were significantly different from each other in all combinations except for between score 0 and 1, which was slightly below the significance threshold (P = 0.0501). These results suggest a correlation between the results of FCM and IHS.

image

Figure 2. Comparison of CD20-positive rate between flow cytometry (FCM) and immunohistochemical staining (IHS) in diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL). Black dots indicate values determined by FCM. Green bars indicate average values (avg). Box-and-whisker plots are shown in red. These results suggest a correlation between results of FCM and IHS.

Download figure to PowerPoint

Comparison of CD20-positive rates between initial cases and post-rituximab cases

Figure 3 shows a comparison of CD20-positive rates by FCM in all cases and each subtype between initial cases and post-rituximab cases. In all cases and in each subtype, except for CD5-positive DLBCL, the CD20-positive rate in post-rituximab cases was lower than that in initial cases. There were significant differences in all cases (average value, 94.1 and 85.9; P = 0.0010) and all subtypes of DLBCL – non-GC type (average value, 87.9 and 57.8; P < 0.0001), FL2 (average value, 103.2 and 93.9; P = 0.0083), and FL3A (average value, 100.7 and 90.6; P = 0.033) – while cases of DLBCL, GC type, showed the same tendency without significance (average value, 93.5 and 83.6; P = 0.0785).

image

Figure 3. Comparison of CD20-positive rate by flow cytometry (FCM) in all cases and each subtype between initial cases and post-rituximab cases. Average values and box-and-whisker plots are indicated in the same way as in Figure 2. DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; GC, germinal center origin.

Download figure to PowerPoint

Stratified analysis of CD20-positive rate in post-rituximab cases of DLBCL and FL

Figure 4 indicates relationships in post-rituximab cases of DLBCL and FL between CD20-positive rate by FCM and periods from initial diagnosis, frequency of rituximab treatment, number of relapses, and therapeutic effects before biopsy. No significant difference was found in CD20-positive rates between different times from initial diagnosis or different numbers of rituximab treatment rounds. In the analysis of number of relapses, the average CD20-positive rates were 60.9 in cases resistant to initial rituximab therapy, 90.1 in 1st relapse cases, and 81.2 in cases of two or more relapses. Significant differences were observed between rituximab-resistant cases and 1st relapse cases (P = 0.0071), while only a non-significant difference was observed between rituximab-resistant cases and cases of two or more relapses (P = 0.0880). Significant differences were also found between CR(u) and PD (P = 0.0016) and between PR and PD (P = 0.0147) in the analysis of therapeutic effect of rituximab before biopsy. The average CD20-positive rates in these cases were 89.5 in CR(u), 85.7 in PR, and 58.3 in PD.

image

Figure 4. Stratified analysis of CD20-positive rate in post-rituximab cases of diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL). Average values and box-and-whisker plots are indicated in the same way as in Figure 2. CR (u), complete response or complete response uncertain; PR, partial response; PD, progressive disease.

Download figure to PowerPoint

Sequence analysis of post-rituximab cases with low CD20-positive rates

Among the post-rituximab cases, sequence analysis was conducted on 22 that had CD20-positive rates less than 50% in FCM analysis and/ or IHS. Exons 3 to 8 of CD20 were sequenced in each case, as exons 1 and 2 are in the 5′ non-coding region. Those cases and the results of sequence analysis are summarized in Table 3. Of 22 cases, seven cases had silent mutations (632 C/T) in exon 4. Exon 6 was not amplified by PCR in two cases. As a reference, sequence analysis of exon 4 of CD20 was also conducted on 15 post-rituximab cases that had CD20-positive rates more than 90% in both FCM analysis and IHS. Three cases showed the same silent mutation (632 C/T) in exon 4. This silent mutation was statistically unrelated to decreased CD20-positive rate.

Table 3. Post-Rituximab cases with low CD20-positive rate and sequence analysis
SexAgeDiagnosisMain treatmentPeriods from initial diagnosis (months)Number of Rituximab treatmentsNumber of relapsesTherapeutic effect of Rituximab before biopsyCD20-positive rate by FCM (%)CD20 positive-rate by IHS [score]Sequence analysis
  1. CHOP, CHOP regimen including doxorubicin, cyclophosphamide, vincristine and prednisone; CR, complete response; CRu, complete response uncertain; DLBCL, diffuse largeB-cell lymphoma; FCM, flow cytometry; FL, follicular lymphoma; GC-type, germinal center origin B-cell type; IHS, immunohistochemical stain; non-GC type, non-germinal center origin B-cell type; PD, progressive disease; PR, partial response; R, rituximab; THP-COP, THP-COP regimen including cyclophosphamide, pirarubicin, vincristine and prednisone.

M65DLBCL, non-GCR-CHOP1180PR3.10Non-amplified exon 6
F43FL 2RUnknownUnknownUnknownUnknown3.90Silent mutation in exon 4 (632 C/T)
M65DLBCL, non-GCR-CHOP10842CR21.61Silent mutation in exon 4 (632 C/T)
M57FL 2R-CHOP1561CR0.32Intact CD20
M75DLBCL, CD5+R-CHOP36Unknown1CR19.94Non-amplified exon 6
F53DLBCL, non-GCR-CHOP1361CR10.01Silent mutation in exon 4 (632 C/T)
M70DLBCL, non-GCR-CHOPUnknown62CR44.22Intact CD20
F48DLBCL, GCOthers60Unknown2PD0.91Intact CD20
F71DLBCL, non-GCR-CHOP781PD7.32Intact CD20
M66DLBCL, non-GCR-CHOP1992PR17.62Intact CD20
F86DLBCL, non-GCR-THP-COP1661CR14.60Intact CD20
M44DLBCL, GCR-THP-COP1261CRu70.81Silent mutation in exon 4 (632 C/T)
F59DLBCL, GCOthersUnknownUnknown0PD0.50Intact CD20
F69DLBCL, GCR-CHOP1361CR72.41Intact CD20
F90DLBCL, GCR-THP-COP761PR28.32Silent mutation in exon 4 (632 C/T)
M69DLBCL, GCR-CHOP48Unknown1CR152.71Silent mutation in exon 4 (632 C/T)
F43DLBCL, non-GCR-CHOP1292CR7.51Intact CD20
F54DLBCL, GCR-CHOP2192CR76.21Intact CD20
M70DLBCL, non-GCR-THP-COP4082CR39.04Intact CD20
F56DLBCL, GCR-CHOP1461CR1.00Intact CD20
M81DLBCL, non-GCR110PD32.20Intact CD20
F57FL 3AR-CHOP84Unknown1PD10.52Silent mutation in exon 4 (632 C/T)

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

This study, for the first time, has found a statistical correlation in CD20-positive rate in B-cell lymphoma between FCM and IHS. Significantly lower CD20-positive rates were observed in post-rituximab cases of DLBCL non-GC type, FL2 and FL3A compared with those in initial cases. In the rituximab-resistant cases, CD20-positive rates were significantly lower than those in relapse cases. In the analysis of the therapeutic effect of rituximab before biopsy, cases with PD had significantly lower CD20-positive rates than those with CR(u) or PR. Sequence analysis did not reveal specific gene mutations associated with low CD20-positive rates.

This study statistically confirmed the hypothesis of previous studies on DLBCL non-GC type, FL2, and FL3A, that treatments that include rituximab might reduce CD20 expression in B-cell lymphoma, although CD20-positive rates before and after rituximab treatment were not compared in the same patients in this study. Increased numbers of cases might allow detection of a significant difference in cases of DLBCL, GC type. A greater number of cases is also necessary for detailed investigation of FL3B.

CD20-positive rates in post-rituximab cases were not observed to be significantly lower in cases of FL1. In these cases, the mechanisms associated with relapse may not include selection of CD20-negative lymphoma. Cases of FL1 have shown stronger expression of bcl2 than those of FL2 and FL3.[14] Lymphoma cells of FL1 damaged by rituximab therapy might possibly undergo less apoptosis because of the anti-apoptotic function of bcl2.[15] This function could cause the survival of B-lymphoma cells and preserve CD20 expression at the time of relapse. The cases of CD5-positive DLBCL showed the opposite result to those of other subtypes, without significance, although the reason remains unknown. There have been no studies investigating the mechanism for reduced CD20 expression of each subtype of B-cell lymphoma. The result of this study should be confirmed by detailed analysis for each subtype.

Miyagi et al.[16] showed that loss of CD20 was unrelated to the interval of biopsies, clinical response, and frequency of rituximab administration. This study demonstrated that the positive rate for CD20 was significantly lower in cases resistant to initial rituximab treatment and cases with PD during retuximab treatment before biopsy. These results suggest the possibility that cases showing reduced expression after rituximab therapy were initially resistant to rituximab and that CD20 expression in these cases decreased after the administration, although those cases might have had low CD20 expression before rituximab therapy, and once the cases showed therapeutic effect – as CR(u) or PR – after rituximab treatment, CD20 expression would not be reduced at the time of relapse. The relapse of those cases could be caused by a mechanism other than reduced CD20 expression, although the mechanism remains unknown.

Several recent studies have investigated the association between CD20 expression and rituximab by FCM and/or IHS. This study for the first time revealed a statistical correlation in CD20-positive rates between FCM and IHS. B-ly1 antibody was used in FCM, and L-26 antibody was used in IHS in this study. Rituximab is reported to bind to extracellular proteins of ANPS at positions 170 to 173 and YCYSI at positions 182 to 185 included in exon 5 of the MS4A1 gene.[17, 18] The B-ly1 antibody targets the same amino acid sequence on the extracellular CD20 epitope as rituximab.[18, 19] Antibody L26 recognizes an intracellular epitope on the CD20 antigen that includes sequences encoded by exons 3, 5, 7, or 8.[20] The correlation of expression of these different epitopes of CD20 in this study suggests that CD20 reduction is caused not by a particular gene deletion or mutation but by pre- or posttranscriptional abnormalities. This speculation is consistent with the result of sequence analysis, which showed no specific genomic mutation. Table 3 shows that some cases revealed discrepancy of CD20 expression between FCM and IHS. These discrepancies are also encountered in daily diagnostic process occasionally. The exact reason remains unknown. However, these might be caused by reduced expression of criteria l B-cell markers in lymphoma cells including CD19, light chain and so on, discrepancy of expression between extracellular epitope and intracellular epitope, and difference of detection sensitivity between FCM and IHS. In the present study, several significant differences in CD20 expression were present not by IHS but by FCM. The reason is that the positive rates were calculated to the first decimal place, giving more precise values, in FCM, and those in IHS were calculated using more approximate figures due to the less quantitative visual evaluation. Because of these considerations and the fact that the B-ly1 antibody targeting the same epitope as rituximab is occasionally used in routine diagnosis, FCM with B-ly1 antibody should be used for analysis of the association between CD20 expression and rituximab.

In this study, sequence analysis of exons 3 to 8 in cases with low CD20-positive rates did not show any specific genomic mutation or even mutations reported in previous studies,[10] although C-terminal deletion mutations were not assessed.[11] In the analysis of cell lines, various investigations regarding loss of CD20 expression after exposure to rituximab have been conducted. Czuczman et al.[6] demonstrated that rituximab-resistant cell lines with significant changes in surface CD20 antigen expression had redistribution of CD20 into lipid raft domains (LRDs), decreased CD20 gene expression, and increased CD20 expression after exposure to proteasome inhibitor. They concluded that the acquisition of rituximab resistance was associated with global gene and protein downregulation of the CD20 antigen affecting LRD organization and downstream signaling. Tomita et al.[21] presented dramatically increased CD20 mRNA and protein expression after exposure to trichostatin A, an epigenetic drug that modulates histone acetylation status, in a B-lymphoma cell line established from a patient repeatedly treated with rituximab-containing chemotherapy, suggesting that epigenetic mechanisms might explain the CD20-negative phenotype in the cell lines. Apart from analyses of cell lines, few studies have been performed, although reverse-transcription-PCR by Hiraga et al.[10] showed that CD20 mRNA expression was significantly lower in CD20-negative cells than in CD20-positive cells. More investigation will be necessary to understand the mechanisms of reduced CD20 expression not associated with genomic mutation after rituximab therapy following the results of this study.

In conclusion, this study demonstrated significantly lower CD20 expression after rituximab therapy in cases of FL and DLBCL, although specific genomic mutations were not detected. The reduction was pronounced in cases resistant to initial administration of rituximab and cases refractory to rituximab therapy before biopsy. These results can contribute to a better selection of therapeutic approach for relapse cases after rituximab therapy. More investigation into the mechanisms of reduced CD20 expression not associated with genomic mutation is required to improve therapy for relapse cases of B-cell lymphoma.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

This study was performed using tissue samples submitted from the many hospitals and/or institutes comprising the Kyushu Lymphoma Study Group. We express our appreciation to them. The hospitals and/or institutes comprising the Kyushu Lymphoma Study Group were as follows: Department of Hematology, National Kyushu Cancer Center; Division of Endocrinology and Metabolism, Faculty of Medicine, University of Ryukyu; Department of Hematology, Hamanomachi Hospital; Department of Hematology, Harasanshin General Hospital; Department of Internal Medicine, NTT Nishinippon Kyushu General Hospital; Department of Internal Medicine, Kyushu Kosei-nenkin Hospital; Department of Hematology, Iizuka Hospital; Department of Internal Medicine, Miyazaki Prefectural Hospital; Department of Hematology, Imamura Hon-in Hospital; Fukuoka Higashi Medical Center; National Hospital Organization Miyakonojo Hospital; Department of Internal Medicine, Koga General Hospital; Fukuoka Teishin Hospital; Yatsushiro Social Insurance General Hospital; Naha City Hospital; Chihaya Hospital; Yanagawa Hospital; Nakagami Hospital.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References