Working groups: NCIWG/IWCLL [2]
Parameter
Diagnosis
Lymphocytes (× 109/L)
>5; ≥1 B-cell marker (CD19, CD20, CD23) + CD5
Atypical cells (%) (e.g. prolymphocytes)
<55
Duration of lymphocytosis
At least 3 months
Disease restricted to lymph nodes in absence of cytopenias and lymphocytosis of <5 × 109/L would qualify for the diagnosis of small lymphocytic lymphoma (SLL). Clonal B-cell lymphocytosis of <5 × 109/L in the absence of lymph node involvement, disease-related symptoms, and cytopenias is designated as “monoclonal B lymphocytosis” (MBL). The prevalence of MBL can be 4–5 % [3, 4] among the general population over the age of 40 years. In a prospective study [3], 15 % of subjects with MBL and lymphocytosis developed CLL after a median follow-up of 6.7 years. The absolute lymphocyte count correlated with progression to CLL. See Table 9.2 for initial and pretreatment tests for patients with CLL.
Table 9.2
Initial and pretreatment tests for patients with CLL
Tests/parameters | Diagnosis | Pretreatment | Treatment in clinical trial |
|---|---|---|---|
Complete blood count | Necessary | Necessary | Necessary |
Physical examination | Necessary | Necessary | Necessary |
Immunophenotyping of peripheral blood | Necessary | Necessary | Necessary |
Serum chemistry immunoglobulin | Desirable | Necessary | Necessary |
Chest X-ray | Not needed | Necessary | Necessary |
Marrow aspirate/biopsy | Not needed | Desirable | Desirable |
FISH | Desirable | Necessary | Necessary |
Infectious panel | Desirable | Necessary | Necessary |
IgVH mutation | Desirable | Desirable | Necessary |
ZAP-70 or CD38 expression | Desirable | Desirable | Necessary |
Serum markers β2M Thymidine kinase CD23 | Desirable | Desirable | Desirable |
CT scan | Not needed | Not needed unless clinically indicated | Desirable |
Several other B-cell disorders need to be considered in the differential diagnosis of CLL. Flow cytometric analysis frequently helps in this process (Table 9.3). A flow cytometry-based scoring system has been useful in situations where the diagnosis is not straightforward [5].
Table 9.3
Immunophenotypic analysis in chronic B-cell disorders
Disease | Sig | CD5 | CD23 | FMC7 | CD22 | CD79b |
|---|---|---|---|---|---|---|
CLL | Weak | ++ | ++ | −/+ | Weak/− | Weak/− |
B-PLL | Strong | −/+ | − | ++ | + | ++ |
HCL | Strong | − | − | ++ | ++ | + |
SL VL | Strong | −/+ | −/+ | ++ | ++ | ++ |
FL | Strong | −/+ | −/+ | ++ | ++ | ++ |
MCL | Strong | ++ | −/+ | ++ | ++ | ++ |
Clinical Features
At diagnosis, most patients are older than 70 years, with more than 95 % over 45 years. The diagnosis of CLL is often incidental; routine blood counts may reveal an elevated absolute lymphocyte count (ALC). In symptomatic patients fatigue and infections may be presenting features; B symptoms are rare. A smaller percentage of patients may present with autoimmune hemolytic anemia (AIHA) or autoimmune thrombocytopenia (AIT). Physical examination may reveal cervical, axillary, and/or inguinal lymphadenopathy. Splenomegaly is not uncommon. The majority of CLL patients will develop hypogammaglobulinemia in their disease course [6, 7] and their impaired cellular immunity is evidenced by lack of response to skin testing with tuberculin, candida, and other antigens. These immune defects predispose these patients to recurrent infections [8]. Treatment with purine nucleosides also can cause sustained alterations in the CD4/CD8 ratio [9]. On the other hand, correction of hypogammaglobulinemia is reported after effective therapy of CLL [10].
Staging
The well-described staging systems for CLL include those of Rai [11] and Binet [12] staging (Table 9.4). The original Rai staging defined five stages from 0 to 4; this has been modified [13] to three stages by defining Rai stage 0 as low-risk, stage 1 and 2 together designated as an intermediate-risk group, and joining stage 3 and 4 to form a high-risk group with a median survival of >12.5, 7, and 1.5 years for each risk group, respectively. Similarly in Binet stages A, B, and C median survivals are >10, 6, and 2 years, respectively [12]. Both these clinical staging systems are based on physical examination and routine blood counts and do not require radiological imaging. The value of computed tomography (CT) in early-stage CLL is not established. A report from Spain suggested that an abnormal abdominal CT scan predicted for increased risk of disease progression and the need for therapy in patients with early-stage CLL [14]. The diagnostic workup that is generally recommended to be undertaken in CLL patients at initial presentation and at the time of treatment is listed in Table 9.2.
Table 9.4
Staging of CLL
Rai stage [11] | Modified Rai stage [13] | Description | Binet stage [12] | Description | Median survival |
|---|---|---|---|---|---|
0 | Low risk | Lymphocytosis | A | Two or fewer lymphoid bearing areas | >10 Years |
1 | Intermediate risk | Lymphocytosis and lymphadenopathy | B | Three or more lymphoid bearing areas | 5–7 Years |
2 | Intermediate risk | Lymphocytosis and splenomegaly with/without lymphadenopathy | |||
3 | High risk | Lymphocytosis and anemia (hemoglobin <11 g/dL) | C | Anemia (hemoglobin <10 g/dL) or thrombocytopenia (platelets 100 × 106/dL) | 2–3 Years |
4 | High risk | Lymphocytosis and thrombocytopenia (platelets <100 × 106/dL) |
Workup at Diagnosis
A physical examination that includes examination of the lymph nodes, routine blood counts, and peripheral blood flow cytometry to establish clonal B-cell lymphocytosis is sufficient to establish the diagnosis of CLL. Though the type of bone marrow infiltration (diffuse vs. nondiffuse) may carry prognostic information, the need for a bone marrow aspiration and biopsy may be replaced by new prognostic markers.
Prognostic Factors
Clinical Prognostic Factors
In a randomized trial of chlorambucil versus observation in Binet stage A patients [15, 16], a subgroup of patients (designated “A”) with hemoglobin ≥12 g/dL, lymphocyte count <30 × 109/L, and fewer than 80 % lymphocytes in the bone marrow aspirate was identified and had an overall survival (OS) comparable to an age-matched French population. Similarly, a lymphocyte doubling time of >12 months, Rai stage 0 disease, nondiffuse bone marrow pattern, hemoglobin ≥13 g/dL, and absolute lymphocyte count <30 × 109/L define a group of “smoldering CLL” with an excellent prognosis [17]. Age and response to treatment are also prognostic factors [18]. Women fare better than men and this is independent of stage and age.
Laboratory Parameters
Several serum factors have been identified as prognostic indicators in early-stage CLL. Among patients with early-stage CLL (Binet stage A, Rai stage 0–2) considered to have “smoldering disease” (blood hemoglobin greater than 13.0 g/dL, a low absolute lymphocyte count (<30,000/μL), a lymphocyte doubling time greater than 12 months, and a nondiffuse pattern of lymphoid bone marrow infiltration [17]), serum thymidine kinase (TK) levels >7.0 U/L identify a group with significantly shorter progression-free survival (PFS) compared to those with lower TK levels [19]. Elevated serum β2-microglobulin level is also an adverse prognostic feature [20]. Serum soluble CD23 segregates Binet stage B disease into more or less aggressive forms [21]. High serum LDH levels indicate a poor prognosis [22]. These parameters appear to be surrogate markers of disease burden or cellular turnover.
A series of new prognostic markers has been identified recently. Testing for these is not needed for the diagnosis of CLL but may help clinicians to have an informed discussion with patients regarding prognosis. As most patients with CLL are diagnosed with early-stage disease, these data aid in predicting the likelihood of disease progression. Also, some (although not all) of the prognostic markers are correlated with response to therapy.
Genetic Studies
Using conventional chromosome banding techniques, cytogenetic abnormalities can be detected in 40–50 % of cases of CLL [23]. This technique is hampered by the low mitotic activity of CLL cells; B-cell mitogens may be used to enhance this activity. In addition, metaphases obtained for karyotyping after mitogen stimulation may arise from normal T cells in the sample [24].
Fluorescence in situ hybridization (FISH) using genomic DNA probes has greatly enhanced the ability to detect molecular abnormalities in malignant cells. This technique can detect aberrations in interphase cells. FISH has demonstrated that molecular abnormalities occur in up to 80 % of cases of CLL [25].
13q deletion is the most common genetic aberration found in CLL by FISH (55 %) followed by11q deletion (18 %), 12q trisomy (16 %), and 17p deletion (7 %) [25]. Prior to the use of FISH, trisomy 12 was the most frequently detected chromosomal abnormality in CLL by conventional cytogenetic methods. Structural abnormalities of 13q were often missed by Giemsa banding, presumably because of the small size of the deletion. The prognosis of CLL varies with the chromosomal abnormality. When divided into five prognostic categories: 17p deletion, 11q deletion, 12q trisomy, normal karyotype, and 13q deletion (as sole abnormality), the survival times were 32, 79, 114, 111, and 133 months [25], respectively. Patients with 17p or 11q deletion [26] had more advanced disease with frequent splenomegaly, mediastinal and abdominal lymphadenopathy, as well as more extensive peripheral lymphadenopathy.
The search for tumor suppressor genes in the commonly deleted region of 13q14 led to the discovery that two microRNA genes, mir-15a and mir-16, are located in this region [27] and in majority of patients with CLL (approximately 68 %) these microRNAs are deleted or downregulated. These two microRNAs are inversely linked to the expression of the antiapoptotic Bcl-2 protein at a posttranslational level [28]. Thus loss of these two microRNAs can be linked to apoptosis resistance in CLL cells through upregulation of Bcl-2. Additional targets of mir-15 and 16 include proteins related to cell cycle progression [29]. The 11q22-q23 deletion is associated with loss of the ataxia-telangiectasia mutated (ATM) [30], a gene that is responsible for repair of DNA double-stranded breaks (DNA DSB), activation of cell cycle check points, and inducing apoptosis in response to DNA DSB. ATM also functions directly in the repair of chromosomal DNA DSBs by maintaining DNA ends in repair complexes generated during VDJ gene rearrangement in the lymphocyte receptor assembly [31]. This explains the occurrence of lymphoid malignancies in patients with ataxia-telangiectasia as chromosomal instability in the lymphoid population arises at a time when lymphoid cell receptor diversity is established. The residual ATM allele is mutated in 36 % of CLLs with an 11q deletion [32]. Deletion of 17p13 in CLL always includes loss of tumor suppressor TP53 [33]. Deletion of the TP53 gene is associated with poor overall survival (OS) and chemoresistance [33, 34]. In CLL cases with monoallelic loss of TP53, the other allele is mutated in the vast majority [35]. TP53 mutation even in the absence of 17p13 abnormality is associated with poor OS in CLL [36]. The frequency of TP53 mutations is higher among patients with relapsed/refractory disease.
Somatic Hypermutation of Immunoglobulin Heavy Chain Variable Gene
Recombination of variable (V), diversity (D), and joining (J) genes and insertion of nontemplated nucleotides at the V–D and D–J junction occur in the pregerminal phase of B-cell development. In addition to the diversity brought about by such VDJ recombination, somatic hypermutations are introduced in the V(D)J rearrangement in normal B cells in the germinal center to increase the B-cell repertoire. Assessment for somatic hypermutation in the immunoglobulin heavy chain variable gene (IgVH) defines two “subsets” of CLL. Approximately 50 % of CLL cases have somatic hypermutation of the IgVH gene and thus appear to arise from postgerminal B cells while the subset of CLL lacking IgVH gene hypermutation appears to arise from naive B cells [37]. The mutation status of CLL cells appears fixed and mutational status is not gained or lost during the course of disease. The outcome of CLL patients with unmutated IgVH gene is significantly worse than that of patients with IgVH mutations; the unmutated population has advanced stage and progressive disease [37, 38].
Stereotyped B-Cell Receptors
Study of VDJ rearrangement of immunoglobulin gene in CLL indicates that certain gene segments are overrepresented across different patients, indicating a stereotyped use of these gene segments (an event occurring more frequently than by chance). A higher proportion of unmutated CLL cases carry stereotyped VDJ rearrangements resulting in similar complementarity-determining regions (CDRs) [37]. More than 20 % of CLL cases can carry stereotyped BCRs which is statistically very unlikely and may suggest exposure to similar antigens which are related to the pathogenesis [39–43]. Use of stereotyped IgVH genes also affects prognosis. The use of IgVH 3-21 gene is associated with an aggressive clinical course independent of the IgVH mutation status [44].
ZAP-70 and CD38 Expression
As IgVH mutation analysis may be difficult to perform in routine diagnostic laboratories; an attempt was made to identify surrogate markers for IgVH mutation status. A gene expression analysis identified ZAP-70, a tyrosine kinase protein normally expressed in T and NK cells, to be differentially expressed between mutated and unmutated CLL cases [45]. Immunophenotypic analysis of CLL cases with known mutation status identified higher CD38 expression in cases with unmutated IgVH [38]. Thus CLL cell expression of ZAP-70 [46–49] and CD38 [38] tends to correlate with unmutated IgVH and predict poor prognosis. However, their correlation with unmutated IgVH is not absolute [50, 51], and CD38 and IgVH mutation status can be independent prognostic factors. Moreover, unlike IgVH mutation status, CD38 expression can change with time.
Factors That Identify Patients More Likely to Need Treatment
The CLL Research Consortium evaluated the relative value of ZAP-70, CD38 expression, and IgVH mutation status for predicting time to treatment in patients with newly diagnosed CLL [52]. Based on analysis of these three parameters, patients can be divided into three risk groups: low, ZAP-70 negative and IgVH mutated; intermediate, ZAP-70 negative and IgVH unmutated; and high, ZAP-70 positive irrespective of mutation status. Though ZAP-70 expression can predict for need for treatment, there are concerns about standardization of the procedure to detect ZAP-70 including whether flow cytometry or immunohistochemistry should be used and appropriate gating methods to detect ZAP-70 by flow cytometry. Information about CD38 expression did not appear to add any further prognostic information.
Combining FISH and IgVH mutation status, the group in Ulm divided patients with early-stage CLL into three risk groups [53]: high risk, del (17p13) irrespective of IgVH mutation status; intermediate risk, del (11q22) and/or unmutated IgVH; and low risk, IgVH mutated in absence of del (17p13) or del (11q22).
Factors That Impact Response to Therapy
Deletion 17p13 and TP53 mutations are associated with resistance to treatment with nucleoside analogs or alkylating agents and their combinations. These patients respond to high-dose steroids, rituximab [54] and alemtuzumab [55, 56], though responses are short-lived. Lenalidomide may be another agent with activity in patients with del (17p) [57]. IgVH mutation status does not predict for response, but responses in patients with IgVH mutations last longer [58, 59].
Factors That Impact Overall Survival
Based on clinical and laboratory characteristics of 1,674 patients with previously untreated CLL presenting to MDACC, Wierda et al. [60] developed a nomogram comprising of easily available parameters that include age, β2-microglobulin, absolute lymphocyte count, sex, Rai stage, and number of involved lymph node groups. This nomogram predicts for survival probability at 5 and 10 years.
Indications for Treatment
Outside the auspices of a clinical trial, patients with asymptomatic early-stage CLL (Rai stage = 0, Binet stage = A) should be observed until there is evidence of disease progression [2]. Patients with advanced disease (modified Rai stage intermediate or high or Binet B and C) may benefit from therapy. In addition to advanced stage, evidence of active disease should be present to initiate therapy. Such indicators of active disease include (1) marrow failure indicated by cytopenia, (2) splenomegaly (>6 cm below the costal margin or symptomatic splenomegaly), (3) massive (>10 cm) or symptomatic lymphadenopathy, (4) lymphocyte doubling time of <6 months, (5) autoimmune hemolytic anemia or thrombocytopenia poorly responsive to corticosteroids, and (6) a minimum of one disease-related symptom: (a) unintentional weight loss of ≥10 % within the previous 6 months, (b) significant fatigue, (c) fevers for ≥2 weeks without any evidence of infection, and (d) night sweats for more than 1 month.
Treatment of CLL
Use of single alkylating agents such as chlorambucil for frontline therapy of CLL is now of historical interest though such an approach may have a role in the treatment of elderly patients with CLL and/or patients with significant co-morbidities. See Table 9.5 for commonly used regimens in CLL. Purine analogs have shown single-agent activity and combinations built around the use of purine analogs have become the standard of care for patients with CLL.
Table 9.5
Commonly used regimens in CLL
Regimen | Schedule | ORR/CR | MRD negative | Remission duration |
|---|---|---|---|---|
Flurarabine + rituximab (FR>R) [61] | F 25 mg/m2 days 1–5 R 375 mg/m2 day 1 For 6 cycles ± followed by R 375 mg/m2 weekly × 4 | 90/47 | N/A | 70 % at 2 years |
F 25 mg/m2 days 1–3 C 250 mg/m2 days 1–3 R 500 mg/m2 day 1 For 6 cycles | 95/70 | 78 % of CRs | 68 % at 5 years | |
FCR3 | FC as above R 500 mg/m2 days 1–3 For 6 cycles | 94/65 | N/A | N/A |
FCR + mitoxantrone (FCRM) [64] | F 25 mg/m2 days 1–3 C 250 mg/m2 days 1–3 R 500 mg/m2 day 1 M 6 mg/m2 day 1 Followed by R 375 mg/m2 every 3 months for 2 years | 93/82 | 46 % of all patients | N/A |
Pentostatin + rituximab (PR) [65] | P 4 mg/m2 days 8, 5, 12 R 375 mg/m2 days 1, 8, 15, 22 For 2–4 cycles | 76/27 | N/A | N/A |
Pentostatin + cyclophosphamide + rituximab (PCR) [66] | P 2 mg/m2 day 1 C 600 mg/m2 day 1 R 375 mg/m2 day 1 For 6 cycles | 91/41 | 73 % of CRs | 34 Months |
Purine Analogs
As CLL cells were thought to be quiescent and nonproliferative, cell cycle active agents such as purine analogs were not expected to be effective in the treatment of CLL. Evidence of a significant proportion of CLL cells being in cycle [67] and demonstration of clinical activity of fludarabine in frontline and salvage settings has changed this perception. Moreover, there is ongoing DNA strand break and repair in both dividing and nondividing cells of CLL, and nonhomologous end-joining repair (NHEJ) process repairs these DNA strand breaks and prevents CLL cells from undergoing apoptosis following DNA strand breaks [68, 69]. Purine analogs are converted intracellularly into their triphosphates by deoxycytidine kinase, are incorporated at sites of DNA repair, and cause double-stranded DNA breaks in dividing and nondividing cells.
Fludarabine
Fludarabine is the most extensively tested nucleoside analog in CLL. Cellular pharmacology suggests that intracellular accumulation of fludarabine triphosphate is dependent on both drug concentration and duration of exposure [70]. Thus a low-dose repeated dosing schedule was adopted for treatment of CLL. Fludarabine alone (25 or 30 mg/m2 daily for 5 days) or in combination with prednisone resulted in overall response rate (ORR) of 78 % and complete remission (CR) plus nodular PR rate of 61 % in patients with previously untreated CLL [9, 71]. Median OS was 63 months (74 months in responders) with older patients (>70 years) and patients with advanced stage disease (Rai stage II and III) having shorter survival. Randomized comparisons of single-agent fludarabine in the frontline setting, to chlorambucil, or combination of chemotherapies (cyclophosphamide, adriamycin, and prednisone: CAP or CAP with vincristine: CHOP) have shown higher complete remission rates and longer remission duration with fludarabine [72, 73].
Pentostatin or Deoxycoformycin
Deoxycoformycin is an inhibitor of adenosine deaminase (ADA) and based on lymphopenias observed in patients with ADA deficiency, pentostatin has been tested as single agent in CLL, both in frontline and salvage settings, yielding modest activity [74, 75]. Pentostatin is perceived to be less myelosuppressive than fludarabine and that led to it being investigated as part of chemoimmunotherapy regimens that will be discussed later.
Cladribine
Cladribine (2-CDA) is a nucleoside analog resistant to inactivation by adenosine deaminase and accumulates inside the cell. Exposure to 2-CDA can cause depletion of NAD in lymphocytes; as NAD is necessary for energy use in DNA repair, 2-CDA can induce massive double-stranded DNA breaks [76]. A continuous intravenous administration schedule (0.1-mg/kg/day 7-day continuous intravenous infusion every 28–35 days until maximum response or prohibitive toxicity) [77] yielded a CR rate of 25 % and an ORR of 85 %. Alternate schedules of 0.12 mg/kg/day as 2-h i.v. infusions for 5 days, until maximum response or excessive toxicity, or oral administration (10 mg/m2/day orally for 5 consecutive days in monthly courses) have been explored with comparable responses [78].
Therapy with purine analogs in CLL has been complicated by cytopenias and immune dysfunction [8, 79, 80]. While cytopenias and associated risk of infections last for relatively short periods of time, immune dysfunction particularly involving T cells can persist for long periods. This lymphopenia mandates prophylaxis for opportunistic infections such as pneumocystis carinii or herpes zoster reactivation.
Combination of Purine Analogs with Other Chemotherapeutic Agents
Alkylating agents have been tested in combination with purine analogs. The rationale for such combinations is based on the fact that alkylating agents induce base excision, nucleotide excision, and mismatch repair. This involves removal of damaged nucleotides followed by resynthesis. Exposure to alkylating agents results in more CLL cells requiring DNA resynthesis. At this resynthesis step purine analogs are incorporated in DNA strand repair patch, stop elongation of DNA strands, and induce apoptosis [81].
Fludarabine/Pentostatin and Cyclophosphamide
In a cohort of 128 patients with CLL that included untreated and previously treated (including fludarabine refractory) patients, the combination of fludarabine (30 mg/m2 intravenously daily for 3 days) and cyclophosphamide (FC) showed an ORR of ≥80 % [82]. The cyclophosphamide dose was decreased from 500 mg/m2/day for 3 days to 300 mg/m2/day for 3 days because of myelosuppression in the early part of the study. The response to FC was higher compared to historical responses to single-agent fludarabine among patients undergoing salvage therapy, with a 38 % response rate among patients refractory to fludarabine. While the CR rate (35 %) was comparable to fludarabine alone among previously untreated patients, minimal residual disease elimination at the end of therapy was achieved at a higher rate compared to fludarabine alone. The German CLL Study Group (GCLLSG) reported similar activity with FC [83]. Myelosuppression leading to infections was the most common side-effect of therapy in both studies.
The South-west Oncology Group (SWOG) conducted a study in previously untreated patients with CLL using a higher dose of cyclophosphamide (1.5 g/m2 every 3 weeks for a total of four cycles) sequentially with fludarabine (days 1–5 at 25 mg/m2 every 4 weeks or later, depending on the recovery of blood counts, for a total of six cycles or until disease progression) [84]. GM-CSF was started 48 h after cyclophosphamide and continued for 10 days or until count recovery. The overall response rate was 55 %.
Two randomized studies (Table 9.6), one conducted by the US Intergroup and the other by GCLLSG, compared the combination of FC to fludarabine [85, 86]. The FC arm in both studies showed higher CR and ORR rates and was associated with longer progression-free survival (PFS). The LRF CLL4 study randomized previously untreated patients with CLL among F, FC, or chlorambucil [87]. Although none of the arms showed overall survival (OS) benefit, the primary aims of the study, i.e. response rate and relapse-free survival (RFS), were consistently better in the FC arm. Myelosuppression was more frequent in the FC arm but did not result in increased infectious complications. Post hoc analysis showed PFS was longer after FC for the subgroups with unmutated IgVH (p = 0.005), no cytogenetic aberration (p = 0.010), 11q− (p = 0.017), unmutated TP53 (p = 0.001), CD38 > 7 % (p = 0.011), and β2M < 5 mg/L (p = 0.009).
Table 9.6
Randomized trials in previously untreated patients with CLL
Treatment | N | Comparator arm | N | Overall response | Complete response | Outcome |
|---|---|---|---|---|---|---|
Treatment arm/comparator (%) | ||||||
Trial A [15] Chlorambucil | 303 | Observation | 309 | 76/0 | 45/0 | OS at 10 years 47/54 % |
Trial B [16] Chlorambucil + prednisone | 460 | Observation | 460 | 69/0 | 28/0 | OS at 7 years 69/69 % |
Fludarabine [73] | 170 | Chlorambucil | 181 | 63/33 | 20/4 | PFS 25 vs. 15 months |
Fludarabine + cyclophosphamide A: GCLLSG [85] | 164 | Fludarabine | 164 | 94.5/82.9 | 23.8/6.7 | PFS 48 vs. 20 months |
B: US Intergroup [86] | 147 | 141 | 74.3/59.5 | 23.4/4.6 | PFS 31.6 vs. 19.2 months | |
Fludarabine + cyclophosphamide [87] | 196 | Fludarabine, chlorambucil | 194, 387 | 94/80/72 | 38/15/7 | PFS at 5 years 36/10/10 % |
Fludarabine + cyclophosphamide + rituximab [88] | 409 | Fludarabine + cyclophosphamide | 408 | 92/85 | 44/23 | PFS 43 vs. 32 months |
Bendamustine [89] | 162 | Chlorambucil | 152 | 31/2 | 68/31 | PFS 22 vs. 8 months |
Alemtuzumab [90] | 149 | Chlorambucil | 148 | 83/55 | 24/2 | PFS 14.6 vs. 11.7 months |
The combination of pentostatin with cyclophosphamide (both administered on day 1 of a 3 week cycle) was reported to have a comparable response rate to FC in a smaller cohort of patients with previously treated CLL [91]. As myelosuppression is frequently encountered in combination treatments with nucleoside analogs and alkylating agents, prophylactic use of trimethoprim/sulfamethoxazole and acyclovir/valacyclovir with or without myeloid growth factor is common practice.
Purine Analogs and Anthracycline
In indolent lymphomas, the combination of fludarabine (F), mitoxantrone (N), and dexamethasone (D) showed a high rate of activity. However, combinations of FN [92] or F with adriamycin [93] have not been proven to be superior to FC. A three-drug combination of FC with mitoxantrone (FCM) also showed good activity in previously untreated patients [94] with CLL (ORR = 90 %, MRD negative CR = 28 %) or at salvage (ORR ≈ 78 %) [95, 96]. While these combinations remain options for fludarabine refractory disease, superiority over FC has not been established.
Chemoimmunotherapy
Fludarabine and Rituximab
Initial reports of use of single-agent rituximab, a chimeric CD20 antibody, showed low level of activity in patients with CLL [97]. This was attributed to low expression of target on CLL cells and circulating CD20 antigen acting as “antibody sink” to reduce effective plasma half-life of rituximab [98]. The combination of fludarabine and rituximab has been shown to have a synergistic effect against an indolent lymphoma cell line [99]. One potential mechanism of such synergy is that rituximab acts partly through complement-dependent cytotoxicity [100] and treatment of indolent lymphoma cells with fludarabine can reduce the surface expression of CD55 and CD59, two proteins with complement inhibitory activity. This led to clinical investigations involving the combination of rituximab with chemotherapeutic agents (chemoimmunotherapy) in CLL. CALGB 9712 study compared concomitant fludarabine and rituximab with a sequential regimen of fludarabine followed by rituximab in patients with previously untreated CLL [61]. The ORR (90 vs. 77 %) and CR rate (47 vs. 28 %) were better in the concomitant arm. A retrospective comparison of the 104 patients treated in the CALGB 9712 study with 171 patients treated with fludarabine alone in CALGB 9011 study indicated better OS and PFS in patients enrolled in the CALGB 9712 study [101], suggesting meaningful clinical benefit from the addition of rituximab to fludarabine.
Fludarabine, Cyclophosphamide, and Rituximab
Based on single-agent activity of rituximab in other studies and its synergism with fludarabine as discussed previously, the group at MD Anderson Cancer Center pioneered the combination of rituximab with the most effective chemotherapy combination of fludarabine and cyclophosphamide (FCR) [62]. In their initial report, the group from MD Anderson Cancer Center (MDACC) reported an ORR of 95 % (CR = 70 %, nodular PR = 10 %, PR = 15 %) with the combination regimen of fludarabine, cyclophosphamide, and rituximab (FCR). The CR rate with this regimen was significantly higher than that reported with FC. Moreover, 78 % of the patients achieving CR also achieved MRD negative status as assessed by flow cytometry [defined as CD5- and CD19-coexpressing cells of less than 1 %, with normalization of the kappa:lambda ratio (<3:1 in patients with monotypic kappa and >1:3 in patients with monotypic lambda)]. Cytopenias precluded completion of the planned six cycles of treatment in 13 % of patients. Neutropenia (≥ grade 3) was encountered in 52 % of courses administered to all patients, but only 2.6 % of these courses were associated with serious infectious episodes.
Long-term follow-up (median follow-up 6 years) results of this chemoimmunotherapy regimen administered to 300 patients showed a 6-year overall survival of 77 % and progression-free survival of 51 % with a median time to progression of 80 months [63].
Though all response parameters are superior with the FCR regimen compared to historical data with the FC regimen, a demonstrated survival benefit in a randomized comparison was lacking till recently. The German CLL Study Group performed a multicenter randomized phase III trial [88] involving 817 patients; they reported a better overall survival in patients with previously untreated CLL with the FCR regimen compared to the FC regimen [84.1 % in the FCR arm versus 79.0 % in the FC arm (p = 0.01)]. This improvement in survival was seen in patients with Binet stage A and B CLL. Though cytopenias were more common in the FCR arm, no increased serious infectious episodes were seen (compared to the rate with FC). A multivariate analysis confirmed the beneficial effect of FCR regimen on OS and PFS.
In patients with previously treated CLL, the MD Anderson group reported that the FCR regimen produced a 73 % ORR and 25 % CR rate [102].
Pentostatin, Cyclophosphamide, and Rituximab
Clinical activity of the combination of pentostatin and cyclophosphamide also encouraged the chemoimmunotherapy regimen of pentostatin, cyclophosphamide, and rituximab (PCR) (see discussion to come). In an initial report of 64 patients, the ORR was 91 % and the CR rate was 41 % with 23 % of patients in CR achieving a MRD negative status (≤1 % positive CD5+/CD19+ cells) [103]. A total of five patients required transfusions; grade 3/4 cytopenias were encountered in 14.5 % of cycles with grade 3/4 infectious complication in only 2 % of the cycles.
The initial expectation with the PCR regimen was that the infectious complications would be less than with FCR regimen. However, a randomized community-based trial in previously untreated or minimally pretreated patients comparing PCR regimen to FCR reported a better CR rate with FCR with a comparable overall response rate, as well as rate of cytopenias and infectious complications [104].
Antibodies
CD20 Antibodies
Rituximab and ofatumumab are two anti-CD20 antibodies approved for the treatment of CLL. CD20 is an antigen expressed on the surface of CLL cells (dim expression by flow cytometry) and is tightly bound to the cell surface. The mechanisms of action of anti-CD20 antibodies against CLL cells include antibody-dependent cellular cytotoxicity [105] (ADCC) and complement-mediated cytotoxicity [100]. In addition, exposure of CLL cells to anti-CD20 antibody has been shown to reduce levels of antiapoptotic proteins including XIAP and Mcl-1 and to induce caspase activation and PARP cleavage [106].
Rituximab
Rituximab is a chimeric monoclonal antibody approved for the treatment of low-grade B-cell lymphomas. The pivotal trial in patients with relapsed low-grade B-cell lymphomas using 375 mg/m2 weekly for 4 weeks showed responses in 48 % of patients, but the response rate among patients with SLL (tissue equivalent of CLL) was 12 % (Table 9.7) [97]. This was attributed to the fact that expression of CD20 was low in SLL/CLL cells compared to cells of follicular lymphomas. O’Brien et al. [110] conducted a dose escalation study of rituximab (375 mg/m2 dose 1 and dose 2–4 at an escalated rituximab dose). For each patient, the dose of rituximab was kept constant and escalation range was 500–2,250 mg/m2. The ORR among patients with CLL was 36 %, all responses being PR. Byrd et al. [109] treated 33 (18 % previously untreated) patients with SLL/CLL on a dose dense three times a week rituximab schedule for 4 weeks and reported an ORR of 45 %. These studies established the principle that despite low expression of CD20 on CLL/SLL cells, rituximab had activity against CLL and paved the way for the combination regimens discussed previously.
Table 9.7
Results with rituximab and alemtuzumab single agent studies
N | Alemtuzumab | N | Rituximab | |||
|---|---|---|---|---|---|---|
Schedule | ORR/CR | Schedule | ORR/CR | |||
Salvage | 93 [107] 24 [108] | 30 mg IV TIW 12 weeks 30 mg IV TIW 16 weeks [108] | 33/2 29/0 | 30 [97] 33 [109] 50 [110] | 375 mg/m2/week × 4 weeks 375 mg/m2 TIW × 4 weeks 500–2,250 mg/m2/week × 4 weeks | 13/0 52/4 22–75/0 |
Frontline | 41 [111] | 30 mg sc TIW 18 weeks | 87/19 | 44 [112] | 375 mg/m2/week × 4 weeks Repeat treatment every 6 months | 51/4 |
Frontline therapy with rituximab (weekly four times) followed by repeated therapy every 6 months for a total of four cycles resulted in an ORR of 51 % (CR 4 %) [112]. The single-agent activity of rituximab is modest. However, the combination of rituximab (weekly for 4 weeks) and methylprednisolone (1 g/m2 daily for 5 days) demonstrated an ORR of 93 % and CR rate of 36 % among patients with relapsed CLL [113]. Rituximab also has (single agent or in combination) activity in treating autoimmune complications of CLL [114–116].
Ofatumumab
Ofatumumab is a fully humanized anti-CD20 monoclonal antibody that binds to an epitope that is distinct and closer to the cell surface than the epitope bound by rituximab; ofatumumab also has a slower off rate after binding to CD20 [120]. This slow off rate could potentially make CLL cells more susceptible to complement-mediated lysis [121]. In the initial study in patients with relapsed CLL, the ORR was 50 % in the highest dose cohort (first infusion of 500 mg and three subsequent infusions of 2,000 mg) without any DLT [122]. In an international study [123] involving patients with fludarabine and alemtuzumab refractory (FA-ref) (N = 59 patients) or fludarabine refractory, bulky CLL (BF-ref) (N = 79 patients), ORR was 58 % and 47 %, respectively. Patients in this study received eight weekly infusions of ofatumumab followed by four monthly infusions during a 24-week period (dose 1 = 300 mg; doses 2–12 = 2,000 mg). Though there was no comparator arm in this study, the responses reported are clearly superior to historical expectations with any antibody therapy in the same setting. Chemoimmunotherapy studies with ofatumumab are in progress.
Lumiliximab (CD23 Antibody)
Lumiliximab is a partially humanized monoclonal antibody targeting CD23, a transmembrane antigen highly expressed on the surface of CLL cells (minimal expression on other cells), that induces apoptosis of CD23 bearing CLL cells and has synergistic activity with fludarabine or rituximab [124]. A phase 1/2 trial combining lumiliximab with FCR in previously treated patients with CLL demonstrated an ORR of 62 % (52 % CR) that compares favorably with results with FCR in a similar population of patients [125]. A randomized study comparing FCR with FCR+ lumiliximab in patients with relapsed CLL is in progress.
CD52 Antibody
Alemtuzumab, originally known as Campath 1G, is a human immunoglobulin G1 (IgG1) anti-CD52 monoclonal antibody (MAb) that binds to nearly all B- and T-cell lymphomas and leukemias. Early phase 2 study with an administration schedule of a 30-mg 2-h intravenous (IV) infusion thrice weekly for a maximum period of 12 weeks reported an ORR of 42 % [126]. Most disease elimination was seen in blood, bone marrow, and spleen while lymph node response was less. Keating et al. [107] reported on an international study involving 93 patients with fludarabine-refractory CLL using alemtuzumab 30 mg IV three times a week for 12 weeks. In the first week, the initial dose was 3 mg, which was increased to 10 mg, and then to 30 mg as soon as infusion-related reactions were tolerated. Infection prophylaxis with trimethoprim/sulfamethoxazole and famciclovir was mandatory. The intent-to-treat analysis showed an ORR of 33 % (CR = 2 %, PR = 31 %). Though there was reduction in lymphadenopathy and other organomegalies, response in lymph nodes >2 cm was modest. In this pivotal trial, 25 patients had grade 3/4 infectious complications. Viral reactivation (cytomegalovirus = seven patients and herpes simplex virus = six patients) was seen in 13 patients. Infectious complications were more frequent in nonresponders to alemtuzumab than in responders.
Lundin et al. [111] reported on the frontline use experience with alemtuzumab using a subcutaneous administration schedule with a target dose of 30 mg three times a week for 18 weeks. Patients also could self-administer alemtuzumab. Infection prophylaxis was with famciclovir, fluconazole, and trimethoprim/sulfamethoxazole. The ORR was 87 % (CR = 19 %, PR = 68 %) in 38 evaluable patients (81 % by ITT). CLL cells were cleared from peripheral blood in 95 % of patients, and as in the salvage study, though reduction in lymph node size was seen in 87 %, the response in bulky lymph nodes was modest. None of the patients with lymph nodes >5 cm achieved CR. Low-grade fever and local injection site reaction in the early part of treatment were common, but grade 3 skin reaction was seen in only 2 % of patients. Most skin reactions disappeared after 2 weeks of treatment. Grade 4 neutropenia was encountered in 21 % of patients without any episodes of neutropenic fevers; CMV reactivation was seen in four patients (10 %) and promptly responded to ganciclovir treatment.
Related posts:
Diagnosis and Treatment of Essential Thrombocythemia and Primary Myelofibrosis
Diagnosis and Treatment of Childhood Acute Myeloid Leukemia
The Myelodysplastic Syndromes
Non-Hodgkin Lymphoma of Childhood
Psychological Aspects of Hematological Neoplasms
Histopathology, Morphology and Immunophenotyping of CLL
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
