Constantin A. Dasanu, MD, PhD; Doru T. Alexandrescu, MD
Abstract
Major advances have occurred in understanding the biology, immunology, and modalities of treatment of chronic lymphocytic leukemia (CLL) in the last decade. B-cell CLL is the most common type of leukemia occurring in the US and Western nations. B-cell CLL is characterized by progressive defects in both cell-mediated and humoral-mediated immunity. B-lymphocyte defects, low gammaglobulin levels, and quantitative and functional T-cell defects have been documented in the setting of CLL. In concert with each other, they account for the increased susceptibility of the CLL patients to infectious agents. Moreover, several recent surveys have pointed out that CLL patients are at high risk of developing a large variety of second malignant neoplasms. Different therapeutic modalities used for CLL may further exacerbate immunosuppression by depleting both T- and B-immune effectors, thus favoring various infectious diseases and perhaps altering the immune surveillance. The occurrence of 2 or more second cancers is increasingly reported in the context of CLL. Increased awareness of this association is warranted. Future development of surveillance strategies may be needed for a growing population of surviving patients who are at risk for second nonlymphoid neoplasms.
Historical Perspectives and Early Data
Long before the purine analogs became available for the treatment of chronic lymphocytic leukemia (CLL), a few reports indicated that patients with CLL are at increased risk for lymphoid malignancies, and may be at increased risk for subsequent nonlymphoid malignant neoplasms as well ( Table 1 ).
In the United States, an early retrospective analysis in 1975 by Manusow and Weinerman[1] on 102 patients with CLL admitted between 1955 and 1974 at the University of Manitoba, Winnipeg, Manitoba, Canada, suggested that the incidence of second cancers was elevated when compared with that in the general population of Manitoba of the same age and sex distribution. The risk for all cancers developing in patients with CLL was found to be 3-fold that for the age- and sex-matched population, 8-fold for skin cancers, and 2-fold for all cancers — excluding skin cancer.
At about the same time, Santoro and colleagues,[2] at Instituto Nazionale Tumori of Milan, Italy, reported on the incidence of a second primary neoplasm in 82 consecutive cases of CLL. In 19.5% of patients, an associated neoplasm was diagnosed either subsequently or concurrently to CLL. Head and neck carcinomas and breast cancer had the highest incidence. The results of this study further supported the hypothesis that patients with CLL are prone to develop subsequent cancers.
A large retrospective study performed in 1978 by Greene and colleagues[3] at the National Cancer Institute in Bethesda, Maryland, looked at the outcome of 4869 patients with CLL. Second primary cancers developed in 234 patients compared with 204.9 expected. The risk was significantly elevated for malignant melanoma, soft-tissue sarcomas, and lung cancer. The study authors concluded that immunologic defects in CLL may be involved in the etiology of the excess risk for these sites, in part because a similar array of nonlymphoid tumors was seen following therapeutic immunosuppression among renal transplant recipients.
Subsequently, a large European study by Mellemgaard and colleagues[4] (n = 7391 CLL patients) found an increased overall risk for cancer in CLL patients. The standardized incidence ratios (ratio between the observed and the expected numbers) were 2.0 for men and 1.2 for women. Thus, increased risks were found for cancer of the lung and prostate in men (relative risk [RR], 2.0 and 1.5, respectively), along with even higher risks for nonmelanoma skin cancer (RR = 4.7 for men, RR = 2.4 for women), renal cell carcinoma (RR = 2.8 for men, RR = 3.6 for women), and sarcomas (RR = 3.3 for men, RR = 2.8 for women) in both sexes. The study authors similarly concluded that the risk was significantly increased for a number of cancer sites in persons with CLL.
Recent Population Surveys
A few recent epidemiologic studies, conducted in different geographic locations on the globe, confirmed the conclusions of previous studies, pointing toward an increased incidence of second cancers in patients with CLL.
A survey from the Division of Cancer Epidemiology and Genetics at the National Cancer Institute, performed by Hisada and colleagues,[5] quantified the risk for second cancers among 16,367 patients with CLL. Overall, the observed/expected ratio (O/E) was 1.20 (95% confidence interval [CI], 1.15-1.26). Significant excesses were found for Kaposi's sarcoma (O/E = 5.09), malignant melanoma (O/E = 3.18), cancers of the larynx (O/E = 1.72), and cancers of the lung (O/E = 1.66). Increased risks were also found for brain cancer among men (O/E =1.91) and for cancers of the stomach (O/E = 1.76) and bladder (O/E = 1.52) among women.
A recent study from the University of Arkansas for Medical Sciences, Little Rock, Arkansas, by Kyasa and colleagues,[6] looked at the rate of second malignancy in 132 CLL/small lymphocytic lymphoma (SLL) patients and compared it with the rate of malignancy (excluding nonmelanoma skin cancer) in 38,000 veterans over a period of 11.5 years. The rate of second malignancy, diagnosed concomitantly or after CLL/SLL, and the age-adjusted rate of malignancy calculated from tumor registry reports and demographic data were used to calculate a standardized morbidity ratio (SMR) with a 95% CI. Twenty-one (16%) of the CLL/SLL patients had second malignancies (19 nonlymphoid, 1 Richter's transformation, and 1 Hodgkin's disease), which represented the cause of death for over two thirds of these patients. The SMR for the CLL/SLL population was 2.97 (95% CI, 1.84-4.55) for second malignancy and 2.69 (95% CI, 1.62-4.21) for nonlymphoid second malignancy. This study showed a significantly increased risk for second cancers, which were the primary causes of death for 9% of all CLL/SLL patients (34% of all patient deaths).
Another retrospective study from Italy, by Mauro and colleagues,[7] observed 1011 CLL patients (204 patients [20%] ≤ 55 years of age and 807 patients [80%] older than 55 years of age) over a period of 10 years. Both groups showed an elevated rate of second primary cancers (8.3% vs 10.7%).
A Japanese retrospective study by Suzuki and colleagues[8] evaluated the clinical prognosis and performed the analysis of causes of death of 75 CLL cases. Second malignancies represented 16% of all major causes of death in those patients. In terms of deaths from CLL complicated by cancer, there were 4 deaths from stomach cancer; 3 deaths from lung cancer; and 1 death from each liver, pancreas, and prostate cancers.
Isolated reports showed the later development of various gastrointestinal malignancies, including gastric cancer, colon cancer, or esophageal cancer, in patients with CLL.[9-11] Hsu and colleagues[11] reported a case of rapid progression of Barrett's esophagus to metastatic esophageal carcinoma in a patient with CLL.
Lung Cancer in CLL Population
An increased incidence of lung cancer of all major histologies has been identified in CLL in several observational studies. Thus, a large US retrospective registry-based cohort study by Parekh and colleagues[12] has shown that approximately 2% of patients with CLL develop lung carcinomas. However, that study showed that 85% of the lung cancer patients were smokers. The lung cancer was at least twice as common in men compared with its incidence in women, and the diagnosis was made approximately 8 years after the diagnosis of CLL. Along with a poor performance status, the presence of CLL was found to be a limiting factor for delivery of treatment for the lung cancer. Thirty-eight percent of patients also had a third malignancy that included melanoma, basal cell carcinoma, laryngeal carcinoma, and colon carcinoma. Patients who develop several second malignancies eventually die of lung carcinoma and not CLL or other solid tumors.
Another recent large retrospective study performed by Schollkopf and colleagues[13] in Denmark, involving 9541 consecutive patients with CLL, has shown a statistically significant increase in the relative risk for lung cancer. These findings applied to all major lung cancer histologies. No data on smoking history were available.
Multiple isolated case reports and case series have suggested an increased aggressiveness of lung cancer of both the non-small-cell and small-cell variety in patients with a concomitant diagnosis of CLL.[14,15] Potti and colleagues[16] concluded that HER-2/neu overexpression may unfavorably influence the prognosis of lung cancer in patients with CLL. Robak and colleagues[17] have reported an increased incidence of lung cancer in a cohort of patients treated with the purine analog cladribine.
Skin Cancers in Subjects With CLL
One of the frequently encountered second malignancies is skin cancer. The incidence of malignant melanoma, Merkel cell tumor, basal cell carcinoma (BCC), and squamous cell carcinoma (SCC) of the skin has been reported to be in excess in the setting of CLL.
A few large population-based studies documented a significant association between CLL and malignant melanoma. In a study by Hisada and colleagues,[5] the O/E ratio of melanoma in patients with CLL was 3.18, exceeded only by that of Kaposi's sarcoma. Analyzing data for 9456 patients diagnosed with CLL, Travis and coworkers[18] reported O/E ratios of 2.79 for cutaneous melanoma and 3.97 for intraocular melanoma. Conversely, Swerdlow and colleagues[19] found an increased risk of developing CLL in a retrospective analysis of a large cohort of patients with cutaneous and ocular melanoma.
The association between Merkel cell tumor, an aggressive skin neoplasm, and CLL has been extensively published.[20-23] Although encountered infrequently, Merkel cell carcinoma (MCC) is most commonly found on sun-exposed areas of the body. Ultraviolet radiation together with drug-induced and/or CLL-induced immunosuppression may be the underlying mechanisms in the observed relationship between CLL and skin cancers, including MCC.[22] However, the association of MCC with a multitude of other primary cancers has also been documented, which points toward either common etiologic factors or a shared predisposition to develop these cancers.[23] A heightened awareness of the associations of lymphohematopoietic malignancies with MCC may also facilitate early clinical recognition of these cancers. Cohen and colleagues[24] have described the development and quick dissemination of MCC soon after receiving chemoimmunotherapy with fludarabine and rituximab for relapsed small lymphocytic lymphoma.
In a small case series, Hartley and colleagues[25] documented the high tendency of cutaneous SCC toward local recurrence and lymph node metastasis in patients with CLL. In that study, 60% of patients had multiple primary carcinomas. Larsen and colleagues[26] reported a case of SCC of the skin in a patient with CLL who relapsed locally after excision and subsequently metastasized to multiple distant sites while on treatment with fludarabine.
In another case series by Weimar and colleagues,[27] the behavior of 4 SCC and 3 BCC in 7 patients with CLL or small lymphocytic lymphoma (SLL) was monitored. The skin tumors recurred repeatedly after conventional treatment and grew to large sizes. The SCC metastasized in all 4 of the CLL patients. The increased clinicohistologic atypia of SCC and BCC in CLL[28] and significantly increased recurrence rates of this malignancy after Mohs surgery have also been documented in a few other studies.[29]
Of interest, a cutaneous lymphocytic infiltrate is frequently observed in patients with CLL in association with skin malignancies, with one third of BCC and SCC lesions containing dense infiltrates. At least 20 patients with this condition have been reported in the analysis of Mehrany and colleagues,[29] and 8 other patients were described by Smoller and Warnke.[30] The cardinal features of this phenomenon, such as a predominance of dense leukemic infiltrates as opposed to a benign immune response, the ability of such infiltrates to herald the diagnosis of CLL, the uncertainty as to the appropriate amount of resection during Mohs surgery, and the worse prognosis of patients with nonmelanoma skin cancers, have all been recognized.
Two or More Second Cancers
Two or even 3 second malignancies in CLL patients have been reported with increased frequency in more recent publications. In a large retrospective study of 1329 patients, Parekh and colleagues[12] have shown that more than a third of patients with CLL and lung cancer also developed at least one of the following malignancies: melanoma, BCC, laryngeal carcinoma, or colon carcinoma.
Pigeaud-Klessens and van der Valk[31] described a female CLL patient who first developed a malignant melanoma of her left lower eyelid and subsequently an SCC of the same eyelid. Prosvic and colleagues[32] reported a case of a male patient with CLL who subsequently developed a clear cell kidney cancer, and ultimately prostate cancer.
Immune Deficiencies in CLL
Over the last decades, it became clear that B-CLL is characterized by progressive defects in both cell-mediated and humoral-mediated immunity. B-lymphocyte defects, low gammaglobulin levels, and quantitative and functional defects of various subsets of T cells have all been documented in the setting of CLL. Further immune impairments result from both disease progression and the management of CLL. Such immune dysregulations may play a role in the development of second malignancies.
Hypogammaglobulinemia represents a widespread and clinically important immune defect. Its frequency and severity progress as the duration of CLL advances.[33] Some studies have suggested that it is more likely to be related to a higher tumor burden than to either certain stages of B-cell differentiation or a particular cellular phenotype.[33,34] The biological mechanisms of hypogammaglobulinemia in CLL are not entirely understood. It occurs in more than 50% of patients with CLL. Although at the time of diagnosis it is noted in about 8% of patients, its prevalence increases considerably (up to 70% or more) paralleling the disease progression.[34,35] Commonly, all 3 immunoglobulin classes (IgG, IgA, and IgM) are decreased, but in some patients only 1 or 2 may be low. Often, the IgM and IgA levels are decreased to a greater extent than the IgG level.[34] Some evidence has suggested that the decrease in immunoglobulins is not an independent process and may be related to the alterations in the proportion of peripheral blood lymphocytes.[36] A possible explanation for hypogammaglobulinemia may be that a suppressive effect is seen on the malignant cells rather than just a replacement process by the neoplastic clone.[37]
Occurrence of malignant tumors was demonstrated in patients having a background of humoral immunodeficiency, such as common variable immunodeficiency, which was associated with lymphomas at an increased relative risk of 12-30, and with gastric carcinoma at an incidence ratio of 10-47). The increased incidence of neoplasias was mainly attributed to the immunodeficiency itself because the relatives of patients, who would otherwise share an abnormal procarcinogenic genetic background, did not manifest an increased risk for cancer.[38-40]
A complex array of abnormal immunoregulatory T-cell functions was demonstrated by Perri and Kay[41] in early-stage CLL. Those include a prominent T-helper dysfunction and a more variable excessive T-suppressor activity. Important cytokine production abnormalities were found in patients with CLL.[40,41] In the context of T-cell cross talk with antigen-presenting cells, B-CLL B cells were found to be poor antigen presenters.[40] These findings led to the conclusion that the T cells in B-CLL may be unable to initiate, maintain, and complete an immune response to the antigens presented by malignant B cells and other antigens.
Development of neoplasias has been linked to abnormalities in T-cell function. One example is ataxia-telangiectasia, which is characterized by mutations in ATM (ataxia-telangiectasia mutated), a kinase that monitors double-strand DNA breaks and the response to damaged DNA, resulting in deficient T-cell development and a predisposition to T-cell leukemia and other lymphoid neoplasms.[42,43] Furthermore, similar T-cell abnormalities to those seen in CLL are encountered in HIV disease and are associated with the development of lymphoid and nonlymphoid malignancies: Kaposi's sarcoma, non-Hodgkin's lymphoma, and cervical cancer. A direct relationship was demonstrated between the risk for occurrence of cancer (Kaposi's sarcoma and non-Hodgkin's lymphoma) and CD4 cell count.[44]
Novel CLL Therapies: Are They Increasing the Risk for Cancer?
Over the past decade, the introduction of nucleoside analogs and monoclonal antibodies into the treatment of patients with CLL has resulted in higher rates and longer duration of response.[45] Beyond any doubt, this is a significant step toward achieving the ultimate goal of disease eradication and improved survival. However, the question has persisted whether the existing immune dysfunctions in CLL would further be exacerbated by the actions of these potent therapeutic agents
Therapy With Nucleoside Analogs and Alkylating Agents
Both the indications to initiate treatment in patients with CLL and the treatment of the disease itself have changed radically over the last years. The purine analogs fludarabine and cladribine have been demonstrated to be the most active drugs in CLL. Among the side effects of purine analogs, immunosuppression is widely recognized and results in prolonged decreases in CD4 lymphocytes, which mandates prophylaxis for opportunistic infections for long periods of time ( Table 2 ).[46] In one study, low levels of CD4 and CD8 T lymphocytes down to 150-200/microliters (mcL) were detected long after completion of the first 3 courses of treatment with fludarabine.[47]
Although the short-term mortality in treated CLL is currently very low, long-term disease-related and treatment-related complications are increasingly observed. Mutagenic effects and possibly iatrogenic immunosuppression raise the possibility of an increase in the rate of second malignancies in patients whose disease already places them at a greater risk. The purpose of a large study by Cheson and colleagues,[48] published in 1999, was to assess the frequency of second tumors in over 2000 patients with relapsed and refractory CLL who had been treated with fludarabine. The rate of second malignancies was found to be increased in those patients compared with the rate in the general population; nonetheless, it was consistent with the increase already associated with CLL itself. The final conclusion of the study was that fludarabine can be safely administered to patients with CLL without a significantly increased risk for second malignancies.
A retrospective study done by Robak and colleagues[17] in 2004 analyzed whether treatment with cladribine during the course of CLL had an impact on the subsequent occurrence of either second solid tumors or Richter's syndrome. The conclusion was that cladribine does not seem to increase the risk for second malignancies in CLL patients except for lung cancers ( Table 2 ). However, further studies are necessary to establish the real risk for lung cancer in CLL patients treated with cladribine.
More recently, an Italian retrospective study by Callea and colleagues,[49] involving 389 CLL patients, failed to demonstrate that prolonged chlorambucil therapy had any impact on the development of second neoplasias.
Thus, there is no compelling evidence that either purine nucleoside analogs or alkylating agents are associated with an increased incidence of second malignancies in patients with CLL. However, new and appropriately powered studies are necessary to validate these conclusions.
Treatment With Monoclonal Antibodies
Used in the treatment of a large range of B-cell malignancies, the chimeric (mouse-human) monoclonal antibody rituximab is an active agent in CLL.[50,51] The drug targets the CD20 antigen present on the surface of neoplastic and normal B lymphocytes. The experience with this agent suggests that escalating doses of rituximab may be more effective in the treatment of nonbulky disease. Two studies have shown depletion of normal B cells by a single standard dose of rituximab, which persisted for 6-9 months.[52,53] Rituximab interfered with both the humoral response to recall antigens as well as with the memory response when administered prior to antigen exposure ( Table 2 ).[53] A potential concern is that higher doses of rituximab can also deplete the normal B-cell clones and impair the normal antibody response against the tumor antigens as well as the antigen-presenting function of normal B cells.
Alemtuzumab is a humanized rat IgG1 antibody that binds to the cell membrane of virtually all normal as well as malignant B and T lymphocytes via the CD52 antigen. In the vast majority of CLL patients, alemtuzumab causes a constant reduction of the leukemic lymphocytes, in usually less than 4 weeks.[54] Both complement and non-complement-mediated mechanisms are responsible for alemtuzumab-mediated killing of T-cell subtypes.[55] Concerns have been expressed that the profound and prolonged T and B lymphopenia would also impair the cell-mediated and antibody-mediated antitumoral immune defenses, further increasing the already augmented rate of second cancers in patients with CLL ( Table 2 ).
Combination therapies with nucleoside analogs and rituximab, with or without cyclophosphamide, are more active than monotherapy with nucleoside analogs in regard to response rate and possibly survival, and have emerged as new and effective combinations.[56] Moreover, the combination rituximab/alemtuzumab was tested in CLL patients resistant to fludarabine, and their synergistic interactions with cytotoxic drugs suggested that a combination of these agents may lead to further therapeutic progress. However, the rate of infectious complications as well as the occurrence of second cancers in this context are still unknown.
Long-term Follow-up
The situation created by treatment-induced immunosuppression as well as the pattern of second cancers seen in CLL is analogous to the ones seen in solid organ or autologous stem cell transplant recipients.[57] This resemblance, along with the prolonged survival seen in a vast majority of patients with CLL, makes the case for heightened awareness for development of cancer in patients with CLL.
Clinical evaluation, with a careful review of history and every organ system, traditionally represents an important tool for identification of high-risk patients. In the general population, internists often do routine screenings, but for the patients with CLL hematologists and/or medical oncologists may assume the role of primary care providers. Patient education and involvement are an important tool in maintaining the best possible long-term health in this group of patients.[58]
Guidelines for screening are not currently available for patients with CLL. Until the cost-efficacy and clinical utility of screening for second cancers in CLL are demonstrated, regular medical follow-up remains the standard of care for these patients.
Conclusions and Future Directions
CLL is currently viewed as a long-term disease, for which most available treatments (with the exception of allogeneic transplantation) do not result in a cure. Awareness of physicians following these patients needs to be raised to the fact that significant morbidity and mortality stem not only from CLL itself, but also from superimposed processes, most notably infections and second malignancies.
Careful consideration of existing risk factors may allow for the development and implementation of prophylactic strategies against second nonlymphoid neoplasms in CLL patients. Nonetheless, with respect to minimizing the mortality from second cancers, more studies are necessary to determine the value of screening and, if shown to be useful, an appropriate follow-up schedule for patients with CLL. An improved understanding of the risk factors associated with the development of second malignancies in patients with CLL could also translate into improved clinical outcomes.
Looking forward, improvements in selection of chemotherapeutic and/or immunotherapeutic regimens will hopefully reduce long-term complications and will allow their curative potential to be utilized to the fullest.
