SEARCH

SEARCH BY CITATION

Keywords:

  • CAMPATH-1H;
  • alemtuzumab;
  • chronic lymphocytic leukemia;
  • plasma;
  • soluble CD52;
  • survival

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

The CD52 antigen is a glycoprotein anchored on the cell membrane of mature B and T lymphocytes, monocytes, and eosinophils. Alemtuzumab (CAMPATH-1H; anti-CD52) is currently approved for the treatment of patients with refractory chronic lymphocytic leukemia (CLL). The authors investigated the possibility that CD52 may be shed from cells and, once soluble, may bind to injected alemtuzumab, forming immune complexes.

METHODS

The authors used Western blot analysis, immunoprecipitation, and enzyme-linked immunoadsorbent assay to investigate the presence of soluble CD52 (sCD52) in the plasma specimens of 117 patients with CLL. They also used in vitro mixing experiments to examine the ability of sCD52 to compete with cells and sequester therapeutic alemtuzumab.

RESULTS

The authors detected high levels of sCD52 in the plasma specimens of patients with CLL. sCD52 can compete with cells in vitro for binding to alemtuzumab, and can form complexes in patients receiving alemtuzumab. Plasma levels of sCD52 were found to be correlated (r)with Rai stage (P = 0.0001), β-2-microglobulin (β-2M) levels (P = 0.00002), soluble CD23 levels (r = 0.42, P < 0.001), and immunoglobulin mutation status (P = 0.003). In the multivariate analysis adjusted for β-2M level, patients with sCD52 levels > 2336 nM/L had a nearly 4-fold increase in risk of death. Higher levels of plasma alemtuzumab were achieved when levels of sCD52 were lower.

CONCLUSIONS

These data not only demonstrated that sCD52 was detectable and useful in the staging and monitoring of patients with CLL, but also showed that sCD52 formed immune complexes with alemtuzumab and may influence the efficacy and toxicity of alemtuzumab therapy. Cancer 2004. © 2004 American Cancer Society.

The CD52 antigen is a glycoprotein with a very short mature protein sequence (comprised of only 12 amino acids) but with a large carbohydrate domain.1–3 The C-terminus is anchored on the cell membrane by a glycosylphosphatidylinositol (GPI) lipid.2, 4 CD52 is expressed on the surface of T and B lymphocytes, monocytes/macrophages, eosinophils, and on some early hematopoietic cells.5–8 It is also expressed in the male reproductive tract.3, 9 CD52 is necessary for spermatozoa to preserve normal motility. It is shed into the seminal plasma and then acquired by sperm cells during their passage through the genital tract. Therefore, it is detectable on the surface of epididymal sperm and in the ejaculate, but not on either the spermatogenetic cells or the testicular spermatozoa. The protein core of the sperm and lymphocyte CD52 is identical—both are products of a single-copy gene located on chromosome 1 (1p36).10 However, N-linked carbohydrate side chains and the GPI anchor structure are different. The physiologic role of CD52 on lymphocytes is unclear.

The CAMPATH-1 family of monoclonal antibodies (MoAb) was originally generated by immunizing rats against human T cells.11 Later studies showed that CAMPATH-1 antibodies recognize CD52.1, 10, 12 The immunoglobulin G1 (IgG1) form of CAMPATH-1 was humanized and this agent, alemtuzumab (CAMPATH-1H), was recently approved for the treatment of patients with refractory chronic lymphocytic leukemia (CLL).13 The CAMPATH-1 family of antibodies is also being used in vitro for lymphocyte depletion in recipients of allogeneic bone marrow grafts, and is being investigated as an immunomodulatory therapy in a variety of diseases.14–19

Antibodies against CD52 are believed to initiate the killing of cells through antigen cross-linking.20, 21 As a result of this cross-linkage, several cytokines are released, among them tumor necrosis factor-α, interferon-γ, and interleukin-6.22 CD52 is expressed on the surface of neoplastic lymphocytes in patients with CLL, low-grade lymphomas, and most T-cell malignancies.23 It is also expressed in some patients with myeloid, monocytic, and acute lymphocytic leukemia.24

CD52 is shed in the male reproductive system and the soluble molecules play an important role in preserving the function of the spermatozoa. However, it is not known whether CD52 is shed from hematopoietic cells or is detectable in the circulation of patients with CLL. We investigated the presence of free, soluble CD52 (sCD52) in the serum and plasma specimens of patients with CLL, and correlated the levels of sCD52 with disease characteristics and outcome.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Patients and Samples

Plasma samples from patients with CLL were collected at the time of presentation to the University of Texas M. D. Anderson Cancer Center (Houston, TX) and other sites participating in the CLL Research Consortium (CRC; University of California, San Diego, CA) via the CRC Tissue Core. All samples were collected according to the individual institution's institutional research board-approved clinical research protocols. Written informed consent was obtained from all patients before their participation in the investigational program. Plasma samples were collected in ethylenediaminetetraacetic acid tubes. The diagnosis of CLL was established from the results of morphologic, immunologic, and molecular evaluations. Samples were immunophenotyped as CD19, CD5, CD20, CD23, CD11C, CD22, FMC7, CD79b, CD3, CD4, CD8, CD52, and light chain (kappa and lambda). All samples were positive for CD19, CD5, CD52, and CD23. Cytogenetic studies were also performed on the majority of samples, as was molecular analysis to detect rearrangement of Ig genes, BCL-1 and BCL-2 genes, and T-cell receptor genes.

Western Blot Analysis of Plasma and Cellular CD52

Protein samples were extracted from cells as described previously.25 After removing albumin on a column, 5 μL of total plasma samples from patients and 200 μg of cellular protein were subjected to electrophoresis on 12.5% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) gels. The protein sample was transferred to nitrocellulose membranes using standard techniques. Membranes were blocked with 5% nonfat milk in phosphate-buffered saline (PBS) containing 0.1% Tween 20 (Sigma Chemical Co; St. Louis, MO) and 0.01% sodium azide for 6–8 hours at room temperature. The blots were incubated overnight at 4 °C with 1 μg of anti-CD52 (CAMPATH-IM) antibodies and PBS containing 2.5% nonfat milk, 2.5% bovine serum albumin (BSA), and 0.01% Tween 20. The membranes then were washed with PBS containing 0.01% Tween 20. The blots were incubated with diluted rabbit anti-rat IgM linked to horseradish peroxidase (Jackson ImmunoResearch Laboratories, West Grove, PA) and PBS containing 1% nonfat milk and 0.1% Tween 20 for 1.5 hours. Active bands were developed using an electrochemiluminescence detection system (Amersham, Arlington Heights, IL).

Immunoprecipitation of CD52–Alemtuzumab Complexes

Rabbit anti-rat Ig was attached to Sepharose beads using a Pierce (Rockford, IL) immunoprecipitation kit, as recommended by the manufacturer. The anti-rat Ig antibodies were absorbed against the human Ig and were specific for rats. Furthermore, unlike the anti-idiotype, these antibodies showed higher affinity binding to the residual rat sequence in alemtuzumab. Plasma samples from patients with CLL, taken before and after treatment with alemtuzumab, were incubated with these columns overnight at 4 °C. The immune complexes were then resolved by 12.5% SDS–PAGE and transferred to membranes. The membranes were probed first with CAMPATH-IM, after which these filters were stripped and the membranes were reprobed with anti-rat IgG.

Competition Experiments and Plasma Blocking of Alemtuzumab Binding to CLL Cells

We performed several mixing studies to test the ability of circulating sCD52 to bind to alemtuzumab and prevent it reaching leukemic cells. We added 1 μg alemtuzumab to 1 million CLL cells freshly isolated from patients with CLL and then used flow cytometry and phycoerythrin-labeled anti-CD52 antibodies to observe the effect of alemtuzumab on the CD52 binding sites on the cell surface. We used the anti-Fc fragment of the rat Ig (312-035-049; Jackson ImmunoResearch Laboratories) to detect alemtuzumab on the surface of cells, and anti-CD19 antibody to detect B cells. The binding of alemtuzumab to CLL cells was also analyzed after cells were mixed with 500 μL of the plasma sample from the same patient or with a similar amount of buffered saline. The plasma sample was heated for 30 minutes at 56 °C to inactivate the complement pathway and to reduce cell death.

Soluble CD52 Enzyme-Linked Immunoadsorbent Assay

A standard enzyme-linked immunoadsorbent assay (ELISA) was modified to detect sCD52 in the plasma sample. Briefly, 96-well polystyrene microtiter plates were coated with CAMPATH-1M antibodies. The plates were then washed 6 times with PBS containing 0.01% Tween 20, blocked with 2% BSA in PBS containing 0.01% Tween 20 for 1–3 hours at 37 °C, and then washed in PBS containing 0.01% Tween 20. An aliquot of 100 μL of patient plasma sample was added and incubated for 3 hours at room temperature, then washed 8 times with PBS containing 0.01% Tween 20. sCD52 was detected using the horseradish peroxidase-labeled humanized anti-CD52 (alemtuzumab) diluted 1:400 in 2% BSA and 0.01% Tween 20. The wells were then washed 6 times with PBS containing 0.01% Tween 20. Substrate (100 mL) was added to develop the color, and the plates were incubated for 15–30 minutes with constant shaking. The reaction then was stopped by adding 50 μL of 2 M hydrochloric acid (HCl), and plates were analyzed for sCD52 at a wavelength of 450 nanometers (nm). Serial dilutions of purified CD52 from cells collected from a patient with CLL were used to generate a standard curve.

Measurement of Alemtuzumab

Affinity-purified rabbit anti-rat IgG (whole molecule) absorbed with human IgG to prevent binding to human IgG (1 μg/100 μL in 0.05 M carbonate buffer, pH 9.4) was added to flat-button 96-well microtiter plates (Fisher Scientific International, Houston, TX) overnight at 4 °C in a cold room. Plates were washed 5 times with PBS–Tween 20 (PBS containing 0.01% Tween 20; Fisher Scientific International) and blocked using 300 μL BSA (2%) in PBS–Tween 20 for 1.5 hours at 37 °C. Plates then were washed 6 times with PBS–Tween 20 and 100 μL of a 1:20,000 dilution of each patient sample in 2% BSA–PBS was added in duplicate. Plates were incubated for 1.2 hours at room temperature with slow shaking, after which they were washed at least 8 times with PBS–Tween 20 and incubated with 100 μL of a 1:50,000 dilution of peroxidase-conjugated, affinity-purified rabbit anti-human-Fc in PBS–Tween 20 containing 15% fetal calf serum for 1.5 hours at room temperature with constant shaking. Eight more washes with PBS–Tween 20 were performed, after which 100 μL of substrate (3,3′-5,5′-tetramethylbenzidine; Dako Corporation, Carpinteria, CA) was added to each well. After 4–8 minutes, the reaction was stopped with 50 μL of 2 N HCl. Plates were read at 450 nm, and a log reading of samples against control was calculated. Samples were tested in BSA solution or in plasma specimens from normal individuals.

Immunoglobulin VH Mutation Status Analysis

Mutation status was evaluated following the procedure of Hamblin et al.26. Briefly, the RNA sample was reversed transcribed using oligo(dT) after which the VH gene was amplified using polymerase chain reaction (PCR) with a mixture of 5′ primers specific for each of the leader sequences of the VH1–VH6 families and 3′ primer for the germline JH region. The amplification product was isolated from a gel and sequenced using the 3′ primer and standard automated sequencing as recommended by the manufacturer (Applied Biosystem, Foster City, CA). The sequence was aligned to the V-gene database using two websites for Ig sequence (http://www.ncbi.nlm.nih.gov/igblast/ and http://imgt.cines.fr). Homology of ≥ 98% was considered to be unmutated and < 98% homology was considered to be mutated.

Statistical Analysis

The Wilcoxon and Kruskal–Wallis tests were used to compare CD52 levels between groups defined by categoric variables. Spearman rank correlation coefficients were computed for CD52 and all continuous covariates.

Univariable Cox proportional hazards regression models were used to evaluate the predictive effect of various covariates on survival. Covariates of interest included patient age, gender, and Rai stage; status of CD38/CD19; hemoglobin levels; platelet count; leukocyte count; lymphocyte count; and status of β-2-microglobulin (β-2M) and CD52. Spearman correlation coefficients were computed to assess the level of intercorrelation or degree of confounding between covariates. Martingale residual plots were used to assess the appropriate functional form of the predictors, as well as a goodness-of-fit test. In addition, cut-off points for covariates were determined by inspection of the residual plots and further assessed in the Cox proportional hazards regression, as appropriate.

The covariates that appeared to be statistically significant (P < 0.05 in the univariate analyses) then were used in the multivariable models. The model coefficients obtained from the Cox models were transformed by exponentiation and interpreted as the relative odds of the different levels of the covariate compared with the reference level. Finally, survival distribution curves were estimated using Kaplan–Meier methods and compared by the log-rank test. All P values were two sided, with P < 0.05 regarded as being statistically significant. All analyses were performed using S-PLUS (Insightful Corporation, Seattle, WA).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Detection of Soluble CD52 and Soluble CD52–Alemtuzumab Complexes in Plasma Samples

To determine whether sCD52 is present in the plasma, we used Western blot analysis to confirm the presence of CD52 in plasma samples from normal individuals and from patients with CLL. As shown in Figure 1, protein extract from the cells of patients with CLL showed the expected 14–20 kilodalton (kD) CD52 glycoproteins, as detected using the CAMPATH-1G MoAb. Cell lysates from Raji cells and HL60 cell lines showed no detectable levels of CD52. Plasma samples from the patients with CLL showed CD52 bands corresponding to those observed in cells. The multiple bands represent variation in the glycosylation of the CD52 protein. However, a difference in the relative intensity of the bands between cells and plasma samples was observed. In addition, the intensity of the individual bands in the plasma samples varied between patients, reflecting differences in glycosylation between patients. We also detected CD52 in the plasma samples of normal individuals, although at very low levels, compared with patients with CLL (Fig. 1).

thumbnail image

Figure 1. Western blot analysis demonstrating the detection of CD52 in cell lysates from three patients with chronic lymphocytic leukemia (CLL) (C1, C2, and C3), but not in Raji (RA) or HL60 (HL) cell lines (upper panel). The lower panel shows the detection of CD52 in cells (C) and plasma (P) samples from a patient with CLL, cells (NC) and plasma (NP) samples from a normal individual, and plasma samples from patients with CLL. The multiple bands of CD52 represent different levels of glycosylation.

Download figure to PowerPoint

We used immunoprecipitation techniques to demonstrate the formation of CD52–alemtuzumab immune complexes in patients treated with alemtuzumab. Antibodies against the rat Ig were used to precipitate alemtuzumab. Unlike anti-idiotype, the anti-rat antibodies were absorbed against human Ig as they do not cross-react with human Ig. As we have previously reported, the antibodies detect the residual rat sequence in alemtuzumab.27 Figure 2 shows the expected precipitation of alemtuzumab, but only after the initiation of therapy, confirming the specificity of the antibodies. When the same filter was stripped and reprobed with anti-CD52 MoAb (CAMPATH-1M), the expected 14–20 kD CD52 protein was detected as a coprecipitate of alemtuzumab, but only from samples collected after alemtuzumab therapy was initiated (Fig. 2).

thumbnail image

Figure 2. Detection of soluble CD52 (sCD52)–alemtuzumab complexes in three patients with chronic lymphocytic leukemia (CLL) who were treated with alemtuzumab. Samples were collected before therapy (0) or at three points (1–3) of alemtuzumab therapy. The alemtuzumab antibodies first were precipitated (lower panel), after which the filter was stripped and reprobed with anti-CD52, demonstrating the coimmunoprecipitation of sCD52. The cell lysate from a CLL sample (C) was run directly on the gel as a positive control.

Download figure to PowerPoint

Based on these findings, we developed ELISAs to detect sCD52. The median level of sCD52 detected in 25 normal individuals was 73 nM/L (range, 31.7–196.9 nM/L), and significantly higher levels were detected in the plasma samples of patients with CLL (median, 709 nM/L; range, 62.4–8615.1 nM/L; Fig. 3A). As shown in Figure 1, sCD52 levels detected by ELISA correlated with those detected by Western blot analysis. The bands from Western blot analysis were scanned and quantified. Equal amounts of plasma from all samples were loaded onto the gel. To verify the linearity of the ELISA, we analyzed dilutions of plasma and cell lysates from a patient with CLL with a high level of plasma sCD52. There was complete correlation between the dilutions and the levels detected by the ELISA (data not shown).

thumbnail image

Figure 3. (A) Significantly higher levels of soluble CD52 (sCD52) were found in patients with chronic lymphocytic leukemia (CLL) compared with healthy individuals (P < 0.0001, Wilcoxon signed rank test). (B) Representative example showing sCD52 levels in the supernatant fluid from cultured cells of a patient with CLL. There was no significant increase noted in sCD52 with time when cells were cultured alone. In contrast, there was significant increase in sCD52 levels noted when cells were cultured with a shedding agent (phorbol 12-myristate 13-acetate [PMA]).

Download figure to PowerPoint

To investigate the possibility that the sCD52 detected was the result of cell degradation or ex vivo shedding, we analyzed the levels of sCD52 in several samples immediately after collection and again 6, 12, and 24 hours after collection. No significant differences between these samples were found. CLL cells from three patients were cultured with and without phorbol 12-myristate 13-acetate (PMA), which was used as a shedding agent. sCD52 was measured in the culture supernatant fluid at 12 hours, 24 hours, 48 hours, 72 hours, 96 hours, 120 hours, and 148 hours. Analysis of variance revealed that there was no significant difference in sCD52 levels among the various supernatant fluids when samples were cultured without PMA (P = 0.20), confirming that sCD52 is not the result of ex vivo cell degradation (Fig. 3B). However, samples cultured with PMA demonstrated significant increases (P = 0.01) in sCD52 with time, supporting the hypothesis that CD52 is shed from cells (Fig. 3B).

Clinical Relevance of Soluble CD52 in Patients with CLL

sCD52 plasma levels were assayed in 117 patients with CLL. Patient characteristics used in the correlation analysis are shown in Table 1. Of the 117 patients, 66 (56%) were previously untreated, 79 (68%) were male, and 44 (38%) had Rai Stage III or IV disease. The median age of these patients was 61 years, and the median leukocyte count was 60.1 × 109/L. The median levels of hemoglobin and β-2M were 12.7 g/L and 3.6 g/dL, respectively.

Table 1. Patient Characteristics
CharacteristicsNo. of patients (%)
  1. β-2M: β-2-microglobulin; HGB: hemoglobin; Ig: immunoglobulin.

Males79 (68)
Rai Stage III–IV44 (38)
No prior treatment66 (56)
Splenomegaly43 (37)
Median age (range)61 (34–84 yrs)
Median leukocyte count (× 109/L)60.1 (1.4–333 leukocytes)
Median HGB level (range)12.7 (4.0–17.8 g/L)
Mutatated IgVH (available on 84 patients)65%
Median β-2M level (range)3.6 (1.3–11.5 g/L)

There were significant differences in sCD52 levels between patients at different Rai stages (P = 0.0008; Kruskal–Wallis test) and Binet stages of disease (P < 0.0001; Kruskal–Wallis test). More importantly, on correlation with the mutation status of the IgVH, which was available for 84 samples, patients with unmutated IgVH were found to have significantly higher levels of sCD52 compared with those with mutated IgVH. (P = 0.003). Complete cytogenetic studies were performed on 74 patients, and sCD52 levels were compared between patients with adverse cytogenetics (11q21 deletion, trisomy 12, or abnormality of chromosome 17) and patients with other karyotypes. Patients with adverse cytogenetics had significantly higher levels of sCD52 (P = 0.0002; Kruskal–Wallis test). Plasma CD52 levels also correlated with the number of sites of lymphadenopathy (Fig. 4). As shown in Figure 4, the level of plasma sCD52 increased with the number of sites of lymphadenopathy (P = 0.002; Kruskal–Wallis test). There was also a positive correlation (r) noted between increasing sCD52 levels and increasing hepatomegaly (r = 0.31, P = 0.0005) and splenomegaly (r = 0.43, P < 0.001). There were negative correlations between sCD52 levels and hemoglobin levels (r = −0.45, P < 0.0001) and platelet counts (r = −0.39, P < 0.0001). sCD52 levels correlated directly with the total leukocyte count (r = 0.52, P < 0.0001), β-2M level (r = 0.41, P = 0.0001), soluble CD23 level (r = 0.42, P < 0.001), and surface expression of CD38 (r = 0.44, P = 0.0006). No correlation between patient age and plasma sCD52 level was evident (r = 0.08, P = 0.4).

thumbnail image

Figure 4. Levels of soluble CD52 correlated with the numbers of sites of lymphadenopathy.

Download figure to PowerPoint

Using a univariate Cox proportional hazards model with sCD52 as a continuous variable, higher levels of sCD52 were shown to be associated with shorter survival time (P = 0.002). As expected, univariate Cox models confirmed the correlations between shorter survival times in relation to surface CD38 expression (P = 0.012) and poor cytogenetics (P = 0.005), low hemoglobin level (P < 0.0001), soluble CD23 expression (P = 0.0005), low platelet count (P = 0.034), IgVH mutation status (P < 0.001), and high β-2M level (P < 0.001). A trend analysis suggested that higher Rai stages are associated with an increased risk of death (P = 0.008). These data suggest that the cohort of patients in whom sCD52 was measured is representative of patients with CLL in general.

Using the Martingale residual plot for CD52, a cut-off point of 2336 nM/L was used to separate the patients into two groups—one with high expression and one with low expression. The risk of death in patients with sCD52 level > 2336 nM/L was increased more than 6-fold (P = 0.00001; Fig. 5A). In a multivariate analysis incorporating both the β-2M and sCD52 levels, these factors combined were predictive of survival. With adjustment for β-2M level (> 3.5 μg/mL), patients with sCD52 level > 2336 nM/L had an increased risk of death of nearly 4-fold (95% confidence interval, 1.34–11.3; P < 0.0001; Fig. 5B). None of these patients was treated with alemtuzumab.

thumbnail image

Figure 5. (A) Kaplan–Meier curve showing significantly shorter survival for patients with high levels of soluble CD52 (P = 0.00001, log-rank test). (B) Kaplan–Meier survival curves for 4 groups of patients, divided by median β-2-microglobulin level (β-2M; cutoff point for β-2M: 3.6 mg/mL; CD52 cutoff point: 2336 nM/L).

Download figure to PowerPoint

Plasma Soluble CD52 Blocks the Binding of Alemtuzumab to CLL Cells

To test the significance of the levels of sCD52 and their role on alemtuzumab therapy, we performed ex vivo mixing experiments. Plasma sample was added to a mixture of CLL cells and alemtuzumab. The strength of the binding of alemtuzumab to the surface of CLL cells and its ability to mask surface CD52 intensity were evaluated using conventional flow cytometry. The intensity of CD52 was significantly reduced when 1 μL of alemtuzumab was added to 1 million cells without the addition of plasma sample (Fig. 6). Alemtuzumab was detected on the surface of the CLL cells using fluorescein isothiocyante-labeled anti-rat Ig (Fig. 6). The binding of alemtuzumab to cells was substantially blocked by adding 500 μL of the patients' plasma sample, showing that plasma sCD52 competed with (and therefore diminished) binding of alemtuzumab to CD52 on the surface of CLL cells (Fig. 6). Lower levels of competition were observed when plasma samples from normal individuals with lower levels of CD52 were added (Fig. 6).

thumbnail image

Figure 6. Plasma level of soluble CD52 sequesters alemtuzumab. (Upper panel) Analysis of CD52 on the surface of chronic lymphocytic leukemia (CLL) cells before and after adding alemtuzumab alone or with plasma sample. (Lower panel) Detection of alemtuzumab on the surface of CLL cells. The intensity of staining is presented as a mean of fluorescence intensity and is shown on the upper right corner of the histograms.

Download figure to PowerPoint

Among patients with CLL treated with alemtuzumab (30 mg 3 times a week for 1 month) for minimal residual disease, patients who achieved a complete response (CR) had significantly lower levels of sCD52 compared with patients who failed to respond to treatment (P = 0.02; Fig. 7). When serial samples were collected from two patients during therapy and sCD52 and alemtuzumab levels were measured, the patient with lower levels of sCD52 achieved higher levels of alemtuzumab than the patient with high levels of sCD52 (Fig. 8). Patients were given 3 mg, 10 mg, and then 30 mg alemtuzumab during the first week, after which they received 30 mg 3 times weekly. Figure 8A shows sCD52 levels in samples collected before and after alemtuzumab injection on Days 1, 2, 3, 8, and 22, and only after alemtuzumab injection on Days 5, 15, and 29. There was no injection on Day 33. The patient with low sCD52 levels became negative for residual disease by PCR at the end of therapy, whereas the patient with high sCD52 levels remained positive for residual disease at the end of therapy.

thumbnail image

Figure 7. Patients who achieved a complete response (CR) were found to have significantly lower levels of soluble CD52 when compared with nonresponders (NR) (P = 0.02).

Download figure to PowerPoint

thumbnail image

Figure 8. Higher levels of alemtuzumab were detected in the patient with lower levels of soluble CD52 (sCD52). sCD52 (A) and alemtuzumab (B) levels in serial samples collected from 2 patients with chronic lymphocytic leukemia who were treated with alemtuzumab for minimal residual disease (3 mg, 10 mg, and 30 mg for the first week; then 30 mg given 3 times weekly thereafter). Samples were collected before and after alemtuzumab injection on Days 1, 2, 3, 8, and 22, and only after alemtuzumab injection on Days 5, 15, and 29. Open circles with straight line: Patient 1; open circles with dashed line: Patient 2.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Human CD52 (alemtuzumab antigen) is a surface molecule that is abundant on lymphocytes and is a potential target in the therapy of various lymphoproliferative disorders. The data presented in the current study represent the first report of detectable sCD52 levels in the plasma samples of patients with CLL. The data in the current study demonstrate that sCD52 can be used as a tumor marker. It correlates with all the important prognostic markers in CLL, but more important, sCD52 may have significant implication when patients are treated with alemtuzumab. Future clinical trials should evaluate the clinical relevance of sCD52 on outcome as well as on dosing and scheduling of almetuzumab therapy. The increase in sCD52 levels most likely results from active shedding of the molecule from cell surfaces in a manner similar to that reported in the male reproductive system.3 Adding phorbol ester as a shedding agent to cultured CLL cells significantly increased the levels of sCD52 in the supernatant fluid (Fig. 3B). It has been proposed that sCD52 is necessary for preventing the adhesion of sperm to each other and to other cells. Plasma CD52 may have a similar function, i.e., preventing cells from adhering to each other. Regardless of the mechanisms that lead to plasma sCD52, the current data indicate that sCD52 plasma levels reflect the clinical behavior of CLL, namely, that higher levels of sCD52 are associated with more aggressive disease. The level of sCD52 was found to be positively correlated with the number of leukemic cells in circulation, the presence or absence of hepatomegaly and splenomegaly, the extent of lymphadenopathy, the mutation status of the Ig gene, and with the levels of soluble CD23, and β-2M. sCD52 may therefore be useful as a monitor in CLL, regardless of whether the therapy is based on alemtuzumab.

It is important to ascertain whether the presence of sCD52 significantly affects the efficacy or toxicity of alemtuzumab therapy. CD52 is considered to be a highly suitable target for MoAb therapy because of its abundant expression on target cells and its low rate of modulation.28, 29 The possibility that sCD52 binds to therapeutic MoAbs, and thus modifies their access to target antigens, should be investigated, as such binding may significantly affect the behavior of the antibody. The response of patients to alemtuzumab is not uniform, and the causes of this variability have been examined by a number of investigators. Ginaldi et al.30 proposed that this variability may partly depend on differences in the level of CD52 expression on target cells. The current data suggest that patients with high levels of sCD52 may possibly require higher dosages of antibodies to saturate the sCD52 and allow residual antibodies to reach target cells. Alemtuzumab is relatively ineffective at reducing lymphadenopathy in patients with CLL, even while causing rapid, profound, and persistent reductions in the number of circulating lymphocytes.28 The evidence that patients with more enlarged lymph nodes have higher levels of sCD52 may explain this phenomenon. Such patients require a higher dosage. Our demonstration that patients treated with alemtuzumab who achieved a CR had lower levels of sCD52 also supports the concept that levels of sCD52 interfere with alemtuzumab efficacy. Furthermore, it is possible that patients receiving therapy will require lower doses of antibodies as their sCD52 levels decrease and the tumor mass becomes smaller. A reduction in alemtuzumab dosage might help to reduce the acute toxicities associated with drug administration and the severe immunosuppression reported in patients receiving this therapy.31

The detection of sCD52–alemtuzumab complexes in patients treated with alemtuzumab is a potentially significant observation that is applicable to all antibody-based therapies. The formation of these complexes indicates that soluble antigens are able to reduce the amount of antibody available to attach to target cells. In the current study, we have demonstrated that sCD52 is detectable in the circulation of patients with CLL and that the levels of sCD52 correlate with the stage and prognosis of the disease. sCD52–alemtuzumab immune complexes were detected in patients with CLL receiving treatment with alemtuzumab. We believe that levels of sCD52 should be considered when alemtuzumab therapy is planned.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES