- Email: [email protected]
Other secondary sequelae of treatments for myeloproliferative disorders
Other Secondary Sequelae of Treatments for Myeloproliferative Disorders Harriet S. Gilbert Polycythemia vera (PV) and essential thrombocythemia (ET) are chronic disorders for which there are no medical cures. Clinical sequelae of PV and ET fall into three categories: primary, such as thrombosis and hemorrhage; secondary, resulting from disease progression or treatment. The decision whether to treat the patient is based on the sequelae of no treatment versus short- and long-term toxicities of the three classes of drugs available for treatment: hydroxyurea, interferon-alpha, and anagrelide. Thrombosis is the most common short-term sequelae of untreated disease; the risk increases with age and after the first thrombotic complication. Hydroxyurea, a nonalkylating myelosuppressive agent, is mutagenic and probably leukemogenic over 5 to 15 years, which makes it unsuitable for treating most younger patients. Interferonalpha, a cytokine that is myelosuppressive and immunomodulatory, has been shown to have a therapeutic effect in both PV and ET. Tolerance to the initial flulike symptoms of interferon-alpha is usually developed, but dose-limiting symptoms of anorexia, asthenia, and neuropsychiatric disease can occur, along with exacerbation or development of autoimmune diseases. Anagrelide, a quinazoline that inhibits cyclic nucleotide phosphodiesterase, inhibits platelet aggregation and has an idiosyncratic effect of inhibiting megakaryocyte maturation and platelet budding at doses below those that affect platelet function. This agent is a vasodilator with positive inotropic activity and a side-effect profile that may include palpitations, forceful heartbeat, tachycardia, and headache. One in four patients develop fluid retention and/or edema that are controllable with diuretic therapy. Dizziness is frequent, but mild. Because these side effects usually abate in 2 to 4 weeks, successful management of patients taking anagrelide depends on encouraging them to maintain therapy. The availability of these three classes of drugs with differing modes of action suggests that combination therapy may offer the opportunity to achieve better control of proliferation while reducing short-term side effects as well as the risks of dose-related cumulative sequelae. Semin Oncol 29 (suppl 10):22-27. Copyright 2002, Elsevier Science (USA). All rights reserved.
From the Albert Einstein College of Medicine, Bronx, NY; and the Department of Medicine, Mount Sinai Medical Center, New York, NY. Address reprint requests to Harriet S. Gilbert, MD, 115 E 72nd St, New York, NY 10021. Copyright 2002, Elsevier Science (USA). All rights reserved. 0093-7754/02/2903-1005$35.00/0 doi:10.1053/sonc.2002.33757 22
T
HE MYELOPROLIFERATIVE disorders (MPDs) result from pluripotential hematopoietic stem cell monoclonal expansion and are differentiated by their respective dominant hematic components. Erythroid dominance characterizes polycythemia vera (PV), increased megakaryocytes and platelets dominate in essential thrombocythemia (ET), and myeloid metaplasia is a result of extramedullary hematopoiesis manifested as splenomegaly. Although ET is principally a disorder of individuals 60 years or older, a significant number of cases, including symptomatic cases, occur earlier. The median age of onset of PV, in contrast, is 58 or 60 years. The risk of complications, especially thrombosis, is age-related and influenced by the length of a patient’s survival. The patient’s risk stratum is the primary consideration in deciding whether or not to treat, and the dominant hematologic pathology and patient age at the time of treatment initiation are critical to the decision of how to treat. The sequelae of ET and PV fall into three categories: (1) primary sequelae, such as thrombosis and hemorrhage (not discussed here); (2) secondary sequelae determined by the natural progression of the disorder; and (3) secondary treatment-related sequelae. Natural progression is well understood in untreated MPDs, but the preventive benefits and the potential increased risk of these sequelae, if any, from therapy remain to be elucidated. Available therapies for MPDs are not benign. Hydroxyurea (HU), for example, is mutagenic and probably leukemogenic over a period of 5 to 15 years, making it unsuitable for most younger patients. The long-term pathogenic effects of interferon-alpha (IFN-␣) and anagrelide are less well known, but their short-term effects may not be tolerable to all patients. Hence, while age and hematologic pathology are the first considerations in a treatment decision for preventing or managing primary sequelae, potential secondary sequelae cannot be overlooked. At present there are three classes of drugs available for the treatment of MPDs, each of which has a different mechanism of action and acts at a Seminars in Oncology, Vol 29, No 3, Suppl 10 (June), 2002: pp 22-27
SECONDARY SEQUELAE OF TREATMENTS FOR MPDS
different point in the pathophysiology of the disorders. Of the three, HU has been used longest. It became the treatment of choice following publication of evidence that alkylating chemotherapeutic agents, such as chlorambucil, are leukemogenic in MPD patients.1 Hydroxyurea is a nonalkylating antimetabolite and a nonspecific myelosuppressive agent that affects ribonucleotide diphosphate reductase inhibition and, therefore, inhibits DNA synthesis. Hydroxyurea has been shown to reduce hematopoietic progenitor growth and CD34⫹ cells in patients with ET and PV.2 IFN-␣ is a biologic response-modifying cytokine that is both myelosuppressive and immunomodulatory. The mechanism of action of anagrelide, a prostaglandin synthetase inhibitor, is not entirely understood, but it is a platelet-specific agent that appears to affect platelet proliferation in postmitotic megakaryocytes by interfering with megakaryocyte maturation.3,4 Because of these different attributes, the three drugs vary in both their effects on MPDs and their treatment-induced secondary sequelae. PROGRESSION-RELATED SEQUELAE
Many people with ET have a chronic, indolent form of the disorder that goes on for many years with minimal complications. Similarly, many PV patients are controlled by phlebotomy for long periods of time and need no pharmacologic intervention. However, untreated patients face the risk of secondary sequelae that occur with the natural progression of ET or PV over time. The most common of these are splenomegaly caused by myeloid metaplasia, myelofibrosis (an epiphenomenon of proliferation), acute myelogenous leukemia, and myelodysplastic syndrome. In addition, ET may progress to PV. At least one specific genetic mutation has been identified in leukemic transformation.5 The primary therapeutic objective in the treatment of PV is to minimize the risk of thromboembolic complications by maintaining a hematocrit of 45% or less, following only modest reductions in thromboembolic complications seen in the Polycythemia Vera Study Group trials that used 52% as the maintenance goal.6 In ET, the traditional primary treatment objective has been to maintain a platelet count of less than 600 ⫻ 109/L. This was dictated in part by the capabilities of the treatment alternatives then available. As newer agents offer the prospect of lower platelet counts without
23
significant increases in toxicity, clinicians are looking more toward maintenance levels of 400 ⫻ 109/L. Three recent studies indicate that thrombotic complications can occur at low platelet counts, suggesting that the traditional point of cytoreductive intervention may be dangerously high and that the treatment goal should be the high end of the normal range.7-9 In most patients, long-term treatment with HU will maintain a count of approximately 600 ⫻ 109/L, but not much lower (except at doses that may adversely affect hematopoiesis in general because of the drug’s lack of specificity). Anagrelide produces a dose-dependent response with little accentuation of toxicities and is usually capable of maintaining a platelet count in the vicinity of 400 ⫻ 109/L following dose titration. One small prospective study of 17 newly diagnosed ET patients concluded that anagrelide-maintained platelet counts of less than 400 ⫻ 109/L in symptomatic patients are associated with the “disappearance” of spontaneous platelet aggregation and the “complete relief” of thrombotic and hemorrhagic manifestations.10 This requires continuous treatment. With IFN-␣, however, treatment can be suspended following titration and attainment of the platelet count goal. Many patients achieve stable remission for a few months up to 2 years or more before requiring the resumption of IFN-␣ therapy.11 Secondary treatment goals deal with the secondary sequelae associated with disease progression. These include preventing splenomegaly or reducing the splenic mass, eliminating monoclonal stem cells, promoting polyclonal hematopoiesis, and preventing both myelofibrosis and leukemic conversion. Together these are considered optimal treatment goals until either new medications or hematopoietic stem cell transplantation, now in its infancy for MPDs,12 offers a cure. Among the currently available therapeutic alternatives, only IFN-␣ effectively reduces splenomegaly. After approximately 2 months of treatment, the spleen begins to shrink and the decrease in size continues for a period of approximately 6 to 12 months. The reduction often persists after withdrawal of the agent and patients respond quickly to subsequent treatment courses following detection of renewed growth. A meta-analysis of 279 patients with PV enrolled in 16 prospective, nonrandomized trials and three case reports reported
24
reduction of splenomegaly in 77% of patients following IFN-␣ therapy.13 This is an effective treatment for splenomegaly in PV, active myeloid metaplasia, and post-PV myeloid metaplasia. Hydroxyurea may shrink the spleen on a short-term basis, but subsequent growth renewal is the norm during chronic HU therapy. Anagrelide is not effective for treating splenomegaly because it is a platelet-specific agent that acts only on megakaryocytes. Anagrelide also has no effect on clonal versus polyclonal pluripotential stem-cell proliferation because its action is on the differentiated megakaryocyte. As a nonselective myelosuppressive agent, HU will interfere with the MPD clone as well as the polyclones. IFN-␣ reduces or changes the MPD clone and minimizes the number of megakaryocytes produced. Theoretically, IFN-␣ and, to a lesser extent, anagrelide should be beneficial for preventing or reversing myelofibrosis in MPDs. If fibrosis is promoted by the release of platelet-derived growth factor or other secretions of megakaryocytes, then changing the MPD clone, altering the marrow milieu, and preventing the maturation of megakaryocytes (IFN-␣) or suppressing platelet budding (anagrelide) might have antifibrogenic effects. Although there is scattered anecdotal evidence that the reversal of fibrosis has occurred with IFN-␣, this concept has not been held to the higher standard of controlled and randomized clinical trials. The concept does not extend to HU, because PV patients treated for long periods with HU eventually evolve to a spent phase characterized in part by myeloid metaplasia. Because acute leukemic conversion was observed in the phlebotomy-only study arm of the landmark Polycythemia Vera Study Group trial, it is generally accepted that leukemia is part of the natural history of disease progression in MPDs.6 The incidence was 1.5%. However, clinical data vary widely on whether or not the risk for acute leukemic conversion is increased by HU therapy. Because it is a nonalkylating agent, it was initially expected to be nonleukemogenic, thus explaining its replacement of alkylating agents as a preferred agent. More recently, however, some studies have reported the risk of leukemic conversion as 10% or more over 20 years,14 but others have reported the added risk as negligible.15,16 Nonetheless, until the issue is resolved, most clinicians assume an added risk and try to avoid using HU in patients under 60
HARRIET S. GILBERT
years of age. An increased incidence of acute leukemia has not been observed in patients treated for 15 years with either anagrelide or IFN-␣, leading to increasing confidence (without proof) that neither is leukemogenic. The results of a long-term safety study of anagrelide in over 4,000 MPD patients are eagerly awaited. Long-term stable control with these agents may, on the other hand, interrupt the natural progressive history of the disease and postpone indefinitely the development of such secondary sequelae as acute leukemia and myelodysplastic syndrome. Data from one recent prospective study indicate, however, that treatment with anagrelide offers no benefit to patients with myelofibrosis with myeloid metaplasia.4 TREATMENT-RELATED SEQUELAE
A patient’s risk stratum is the primary determinant of treatment.17 In ET, patients who are less than 40 years of age, have platelet counts in the range of 600 to 1,000 ⫻ 109/L, are asymptomatic, and have no comorbidities are considered low-risk patients. These patients are not candidates for myelosuppressive or cytoreductive therapy, but may benefit from low-dose aspirin as prophylaxis against thrombosis. The benefit of aspirin for its antithrombotic properties must be balanced against the risk of hemorrhage for each patient. Patients who are 60 years of age or above, or patients of any age who are symptomatic and have at least one comorbidity, are considered high-risk patients. A history of thrombosis is the leading indicator of risk for subsequent thrombosis.18 Recent evidence, based on a study of 65 patients, suggests that subnormal plasma erythropoietin levels in newly diagnosed ET patients may be a risk factor for vascular complications.19 These patients are candidates for myelosuppressive (HU, IFN-␣) or megakaryocyte-specific cytoreductive (anagrelide) therapy to control platelet counts in the range of 400 to 600 ⫻ 109/L. All other patients, particularly asymptomatic individuals who are between 40 and 60 years of age and have no comorbidities, are at indeterminate risk for thrombotic complications. A decision to treat other than with low-dose aspirin must include a careful balancing of the potential long-term adverse effects of therapy against the potential benefits. Phlebotomy, the standard first-line treatment for PV, should be undertaken with all diagnosed patients with a goal of maintaining a maximum
SECONDARY SEQUELAE OF TREATMENTS FOR MPDS
hematocrit of 45%. Although some investigators distinguish between 45% for men and 42% for women, there are no controlled and randomized trial data to substantiate the clinical significance of the difference. The most common side effect of phlebotomy is iron deficiency. However, supplemental iron is not recommended because it accelerates erythropoiesis. The frequency of phlebotomy usually decreases over time, however, reducing the risk of chronic iron deficiency. Phlebotomy and iron deficiency may increase thrombocythemia. Low-dose aspirin is frequently used to reduce the risk of symptomatic platelet aggregation that may cause thrombosis or microvascular complications of PV, such as neurologic changes,20 visual disturbance,21 and erythromelalgia.22,23 Occasionally low-dose aspirin induces hemorrhage, but it is usually minor. It is contraindicated in patients with marked thrombocythemia because of the risk of cerebral hemorrhage. Patients whose hematocrits are not controlled by phlebotomy require myelosuppressive therapy. Platelet-specific cytoreduction treatment is of no value to PV patients except in the event that thrombocythemia accompanying erythrocythemia is not controlled by myelosuppression. The drug selected determines the treatmentrelated sequelae. Hydroxyurea is an antimetabolite and a dose-dependent, nonspecific myelosuppressive agent. The use of high doses to control platelet counts in ET increases the risks of neutropenia and anemia. In general, the dose required for platelet control at 600 ⫻ 109/L does not induce these complications, explaining in part why control at this level, now regarded as high, was considered adequate for so long. Adjustments in the dose of HU may give rise to marked oscillation of platelet counts,24 including a rebound to even higher platelet counts than existed before the initiation of therapy. It is possible, however, that dose-limiting factors and the relatively low longterm cumulative ingestion may result in less treatment-induced leukemic conversion than might have occurred otherwise. In PV patients, HU doses required for effective control of hematocrits may induce thrombocytopenia and/or neutropenia. Dose reduction designed to prevent these complications may diminish efficacy with respect to erythrocytes, thus requiring subsequent phlebotomies in patients treated with HU. Other side effects of HU include cutaneous ulceration,25,26
25
mucositis, and hyperpigmentation. Several cases have been reported of painful oral ulcers27 and leg ulcers, usually in the malleolar region, that are unresponsive to secondary treatment and require discontinuation of HU.28,29 Hydroxyurea treatment has also been associated with drug fever30 and life-threatening hypersensitivity pneumonitis.31 Gastrointestinal sequelae include nausea, vomiting, bloating, and diarrhea. Headache and asthenia are the most common systemic side effects. Hydroxyurea is known to be mutagenic in humans, and there is inconclusive evidence that it is leukemogenic. IFN-␣ may also have many side effects, the most common of which are flu-like symptoms including chills and fever. Patients may experience myalgia, fatigue, anorexia, headache, depression, and/or neuropathy. Potential hematologic side effects are neutropenia and anemia. Mild alopecia is common. Its action on the intestinal mucosa may result in diarrhea. As an immunomodulator, IFN-␣ may cause symptomatic episodes of autoimmune disease including hypothyroidism. This high profile of potential side effects makes IFN-␣ intolerable to approximately 30% of patients, and those who tolerate it may incur moderate expense to relieve symptoms of the sequelae. The side effects of anagrelide are because it is a prostaglandin inhibitor and a vasodilator. Hence, it may cause headache and dizziness. As an inotrope it may induce palpitations, a sense of an unusually strong heartbeat, and even cardioarrhythmias. Anagrelide-induced cardiomyopathy has been reported, and is thought to be caused by phosphodiesterase inhibition leading to cardiomyopathy and vasodilation.32 Anagrelide also changes the renal blood flow and induces fluid retention. At present the vehicle of the drug is lactose, so the bloating, nausea, diarrhea, and vomiting that some patients experience may be due to some extent by lactose sensitivity as well as the fact that prostaglandins do not affect gastrointestinal function. Patients who are in borderline cardiac function may experience pulmonary edema following ingestion of anagrelide, and patients with poor cardiac function may lapse into congestive heart failure. It appears, however, that elderly patients with these risk factors may tolerate anagrelide when treated concurrently with good cardiotonic regimens. In a recent review of 35 young patients (age range, 17 to 48 years) treated
26
HARRIET S. GILBERT
with anagrelide for ET before 1992, 24% experienced mild-to-moderate anemia.33 CONCLUSION
It is unique that there are now three agents with different mechanisms of action and with effects on the progeny of the MPD clone that differ markedly. Because these permit good control of platelet counts and hematocrits, the next steps in treating ET and PV should be to reduce toxicity and to capture synergies by using combined-agent therapy. One example in which that has been performed successfully is the addition of anagrelide to HU for incremental improvement in platelet counts without chronic overexposure to HU and without increasing the risk of neutropenia. Very low-dose IFN-␣ has also been added to anagrelide therapy to improve platelet counts. By exploiting the different mechanisms of these three classes of drugs, ET and PV can be treated successfully with little risk of cumulative treatment-related sequelae and minimal exposure to potential unforeseen consequences of long-term therapy. REFERENCES 1. Berk PD, Goldberg JD, Fruchtman SM, et al: Therapeutic recommendations in polycythemia vera based on Polycythemia Vera Study Group protocols. Semin Hematol 23:132-143, 1986 2. Andraesson B, Swolin B, Kutti J: Hydroxyurea treatment reduces haematopoietic progenitor growth and CD34 positive cells in polycythemia vera and essential thrombocythemia. Eur J Haematol 64:188-193, 2000 3. Spencer CM, Brogden RN: Anagrelide: A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in the treatment of thrombocythemia. Drugs 7:809-822, 1994 4. Yoon SY, Li CY, Mesa RA, et al: Bone marrow effects of anagrelide therapy in patients with myelofibrosis with myeloid metaplasia. Br J Haematol 106:682-688, 1999 5. Tabata M, Imagawa S, Tarumoto T, et al: Essential thrombocythemia transformed to acute myelogenous leukemia with t(3;17)(p24;q12), del(5)(q13q34) after treatment with carboquone and hydroxyurea. Jpn J Clin Oncol 30:310-312, 2000 6. Berk PD, Wasserman LR, Fruchtman SM, et al: Treatment of polycythemia vera: A summary of clinical trials conducted by the Polycythemia Vera Study Group, in Wasserman LR, Berk PD, Berlin NI (eds.): Polycythemia Vera and the Myeloproliferative Disorders. Philadelphia, PA, Saunders, 1995, pp 102-113 7. Regev A, Stark P, Blickstein D, et al: Thrombotic complications in essential thrombocythemia with relatively low platelet counts. Am J Hematol 56:168-172, 1997 8. Lengfelder E, Hochhaus A, Kronawitter U, et al: Should a platelet count of 600 ⫻ 109/L be used as a diagnostic criterion
in essential thrombocythaemia? An analysis of the natural course including early stages. Br J Haematol. 199:15-23, 1998 9. Sacchi S, Vinci G, Gugliotta L, et al: Diagnosis of essential thrombocythemia at platelet counts between 400 and 600 ⫻ 109/L. Gruppo Italiano Malattie Mieloproliferative Croniche (GIMMC). Haematologica 85:492-495, 2000 10. Laguna MS, Kornblihtt LI, Marta RF, et al: Effectiveness of anagrelide in the treatment of symptomatic patients with essential thrombocythemia. Clin Appl Thromb Hemost 6:157161, 2000 11. Gilbert HS: Long term treatment of myeloproliferative disease with interferon-alpha-2b: Feasibility and efficacy. Cancer 83:1205-1213, 1998 12. Jurado M, Deeg HJ, Gooley T, et al: Hemopoietic stem cell transplantation for advanced polycythemia vera or essential thrombocythemia. Br J Haematol 112:392-396, 2001 13. Lengfelder E, Berger U, Hehlmann R: Interferon alpha in the treatment of polycythemia vera. Ann Hematol 79:103109, 2000 14. Fruchtman SM, Mack K, Kaplan M, et al: From efficacy to safety: A Polycythemia Vera Study Group report on hydroxyurea in patients with polycythemia vera. Semin Hematol 34:17-23, 1997 15. Nand S, Stock W, Godwin J, et al: Leukemogenic risk of hydroxyurea therapy in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Am J Hematol 52:42-46, 1996 16. Finazzi G, Ruggeri M, Rodeghiero F, et al: Second malignancies in patients with essential thrombocythaemia treated with busulphan and hydroxyurea: Long-term follow-up of a randomized clinical trial. Br J Hematol 110:577-583, 2000 17. Gilbert HS: Diagnosis and treatment of thrombocythemia in myeloproliferative disorders. Oncology (Huntingt) 15: 989-998, 2001 18. Cortelazzo S, Viero P, Finazzi G, et al: Incidence and risk factors for thrombotic complications in a historical cohort of 100 patients with essential thrombocythemia. J Clin Oncol 8:556-562, 1990 19. Andraesson B, Lindstedt G, Kutti J: Plasma erythropoietin in essential thrombocythaemia: At diagnosis and in response to myelosuppressive treatment. Leuk Lymphoma 38: 113-120, 2000 20. Jabaily J, Iland HJ, Laszlo J, et al: Neurologic manifestations of essential thrombocythemia. Ann Intern Med 99:513518, 1983 21. Koudstall PJ, Koudstall A: Neurologic and visual symptoms in essential thrombocythemia: Efficacy of low-dose aspirin. Semin Thromb Hemost 23:365-370, 1997 22. VanGenderen PJJ, Michiels JJ: Erythromelalgia: A pathognomonic microvascular thrombotic complication in essential thrombocythemia and polycythemia vera. Semin Thromb Hemost 23:357-363, 1997 23. Michiels JJ: Erythromelalgia and vascular complications in polycythemia vera. Semin Thromb Hemost 23:441-454, 1997 24. Tefferi A, Elliott MA, Kao PC, et al: Hydroxyureainduced marked oscillations of platelet counts in patients with polycythemia vera. Blood 96:1582-1584, 2000 25. Young HS, Khan AS, Kendra JR, et al: The cutaneous side-effects of hydroxyurea. Clin Lab Hematol 22:229-232, 2000
SECONDARY SEQUELAE OF TREATMENTS FOR MPDS
26. McLintock C, Ockelford PA: Hydroxyurea-induced skin ulceration in myeloproliferative disorders. Aust N Z J Med 30:373-376, 2000 27. Paleri V, Lindsey L: Oral ulcers caused by hydroxyurea. J Laryngol Otol 114:976-977, 2000 28. Poros A, Nadasdy K: Leg ulcer in hydroxyurea-treated patients. Haematologia (Budap) 30:313-318, 2000 29. Murphy J, Morley SM: Hydroxyurea-induced leg ulcers. Hosp Med 61:510a, 2000 (abstr)
27
30. Braester A, Quitt M: Hydroxyurea as a cause of drug fever. Acta Haematol 104:50-51, 2000 31. Sandhu HS, Barnes PJ, Hernandez P: Hydroxyurea-induced hypersensitive pneumonitis. Can Resp J 7:491-495, 2000 32. James CW: Anagrelide-induced cardiomyopathy. Pharmacotherapy 20:1224-1227, 2000 33. Storen EC, Tefferi A: Long-term use of anagrelide in young patients with essential thrombocythemia. Blood 97:863866, 2001
From the Albert Einstein College of Medicine, Bronx, NY; and the Department of Medicine, Mount Sinai Medical Center, New York, NY. Address reprint requests to Harriet S. Gilbert, MD, 115 E 72nd St, New York, NY 10021. Copyright 2002, Elsevier Science (USA). All rights reserved. 0093-7754/02/2903-1005$35.00/0 doi:10.1053/sonc.2002.33757 22
T
HE MYELOPROLIFERATIVE disorders (MPDs) result from pluripotential hematopoietic stem cell monoclonal expansion and are differentiated by their respective dominant hematic components. Erythroid dominance characterizes polycythemia vera (PV), increased megakaryocytes and platelets dominate in essential thrombocythemia (ET), and myeloid metaplasia is a result of extramedullary hematopoiesis manifested as splenomegaly. Although ET is principally a disorder of individuals 60 years or older, a significant number of cases, including symptomatic cases, occur earlier. The median age of onset of PV, in contrast, is 58 or 60 years. The risk of complications, especially thrombosis, is age-related and influenced by the length of a patient’s survival. The patient’s risk stratum is the primary consideration in deciding whether or not to treat, and the dominant hematologic pathology and patient age at the time of treatment initiation are critical to the decision of how to treat. The sequelae of ET and PV fall into three categories: (1) primary sequelae, such as thrombosis and hemorrhage (not discussed here); (2) secondary sequelae determined by the natural progression of the disorder; and (3) secondary treatment-related sequelae. Natural progression is well understood in untreated MPDs, but the preventive benefits and the potential increased risk of these sequelae, if any, from therapy remain to be elucidated. Available therapies for MPDs are not benign. Hydroxyurea (HU), for example, is mutagenic and probably leukemogenic over a period of 5 to 15 years, making it unsuitable for most younger patients. The long-term pathogenic effects of interferon-alpha (IFN-␣) and anagrelide are less well known, but their short-term effects may not be tolerable to all patients. Hence, while age and hematologic pathology are the first considerations in a treatment decision for preventing or managing primary sequelae, potential secondary sequelae cannot be overlooked. At present there are three classes of drugs available for the treatment of MPDs, each of which has a different mechanism of action and acts at a Seminars in Oncology, Vol 29, No 3, Suppl 10 (June), 2002: pp 22-27
SECONDARY SEQUELAE OF TREATMENTS FOR MPDS
different point in the pathophysiology of the disorders. Of the three, HU has been used longest. It became the treatment of choice following publication of evidence that alkylating chemotherapeutic agents, such as chlorambucil, are leukemogenic in MPD patients.1 Hydroxyurea is a nonalkylating antimetabolite and a nonspecific myelosuppressive agent that affects ribonucleotide diphosphate reductase inhibition and, therefore, inhibits DNA synthesis. Hydroxyurea has been shown to reduce hematopoietic progenitor growth and CD34⫹ cells in patients with ET and PV.2 IFN-␣ is a biologic response-modifying cytokine that is both myelosuppressive and immunomodulatory. The mechanism of action of anagrelide, a prostaglandin synthetase inhibitor, is not entirely understood, but it is a platelet-specific agent that appears to affect platelet proliferation in postmitotic megakaryocytes by interfering with megakaryocyte maturation.3,4 Because of these different attributes, the three drugs vary in both their effects on MPDs and their treatment-induced secondary sequelae. PROGRESSION-RELATED SEQUELAE
Many people with ET have a chronic, indolent form of the disorder that goes on for many years with minimal complications. Similarly, many PV patients are controlled by phlebotomy for long periods of time and need no pharmacologic intervention. However, untreated patients face the risk of secondary sequelae that occur with the natural progression of ET or PV over time. The most common of these are splenomegaly caused by myeloid metaplasia, myelofibrosis (an epiphenomenon of proliferation), acute myelogenous leukemia, and myelodysplastic syndrome. In addition, ET may progress to PV. At least one specific genetic mutation has been identified in leukemic transformation.5 The primary therapeutic objective in the treatment of PV is to minimize the risk of thromboembolic complications by maintaining a hematocrit of 45% or less, following only modest reductions in thromboembolic complications seen in the Polycythemia Vera Study Group trials that used 52% as the maintenance goal.6 In ET, the traditional primary treatment objective has been to maintain a platelet count of less than 600 ⫻ 109/L. This was dictated in part by the capabilities of the treatment alternatives then available. As newer agents offer the prospect of lower platelet counts without
23
significant increases in toxicity, clinicians are looking more toward maintenance levels of 400 ⫻ 109/L. Three recent studies indicate that thrombotic complications can occur at low platelet counts, suggesting that the traditional point of cytoreductive intervention may be dangerously high and that the treatment goal should be the high end of the normal range.7-9 In most patients, long-term treatment with HU will maintain a count of approximately 600 ⫻ 109/L, but not much lower (except at doses that may adversely affect hematopoiesis in general because of the drug’s lack of specificity). Anagrelide produces a dose-dependent response with little accentuation of toxicities and is usually capable of maintaining a platelet count in the vicinity of 400 ⫻ 109/L following dose titration. One small prospective study of 17 newly diagnosed ET patients concluded that anagrelide-maintained platelet counts of less than 400 ⫻ 109/L in symptomatic patients are associated with the “disappearance” of spontaneous platelet aggregation and the “complete relief” of thrombotic and hemorrhagic manifestations.10 This requires continuous treatment. With IFN-␣, however, treatment can be suspended following titration and attainment of the platelet count goal. Many patients achieve stable remission for a few months up to 2 years or more before requiring the resumption of IFN-␣ therapy.11 Secondary treatment goals deal with the secondary sequelae associated with disease progression. These include preventing splenomegaly or reducing the splenic mass, eliminating monoclonal stem cells, promoting polyclonal hematopoiesis, and preventing both myelofibrosis and leukemic conversion. Together these are considered optimal treatment goals until either new medications or hematopoietic stem cell transplantation, now in its infancy for MPDs,12 offers a cure. Among the currently available therapeutic alternatives, only IFN-␣ effectively reduces splenomegaly. After approximately 2 months of treatment, the spleen begins to shrink and the decrease in size continues for a period of approximately 6 to 12 months. The reduction often persists after withdrawal of the agent and patients respond quickly to subsequent treatment courses following detection of renewed growth. A meta-analysis of 279 patients with PV enrolled in 16 prospective, nonrandomized trials and three case reports reported
24
reduction of splenomegaly in 77% of patients following IFN-␣ therapy.13 This is an effective treatment for splenomegaly in PV, active myeloid metaplasia, and post-PV myeloid metaplasia. Hydroxyurea may shrink the spleen on a short-term basis, but subsequent growth renewal is the norm during chronic HU therapy. Anagrelide is not effective for treating splenomegaly because it is a platelet-specific agent that acts only on megakaryocytes. Anagrelide also has no effect on clonal versus polyclonal pluripotential stem-cell proliferation because its action is on the differentiated megakaryocyte. As a nonselective myelosuppressive agent, HU will interfere with the MPD clone as well as the polyclones. IFN-␣ reduces or changes the MPD clone and minimizes the number of megakaryocytes produced. Theoretically, IFN-␣ and, to a lesser extent, anagrelide should be beneficial for preventing or reversing myelofibrosis in MPDs. If fibrosis is promoted by the release of platelet-derived growth factor or other secretions of megakaryocytes, then changing the MPD clone, altering the marrow milieu, and preventing the maturation of megakaryocytes (IFN-␣) or suppressing platelet budding (anagrelide) might have antifibrogenic effects. Although there is scattered anecdotal evidence that the reversal of fibrosis has occurred with IFN-␣, this concept has not been held to the higher standard of controlled and randomized clinical trials. The concept does not extend to HU, because PV patients treated for long periods with HU eventually evolve to a spent phase characterized in part by myeloid metaplasia. Because acute leukemic conversion was observed in the phlebotomy-only study arm of the landmark Polycythemia Vera Study Group trial, it is generally accepted that leukemia is part of the natural history of disease progression in MPDs.6 The incidence was 1.5%. However, clinical data vary widely on whether or not the risk for acute leukemic conversion is increased by HU therapy. Because it is a nonalkylating agent, it was initially expected to be nonleukemogenic, thus explaining its replacement of alkylating agents as a preferred agent. More recently, however, some studies have reported the risk of leukemic conversion as 10% or more over 20 years,14 but others have reported the added risk as negligible.15,16 Nonetheless, until the issue is resolved, most clinicians assume an added risk and try to avoid using HU in patients under 60
HARRIET S. GILBERT
years of age. An increased incidence of acute leukemia has not been observed in patients treated for 15 years with either anagrelide or IFN-␣, leading to increasing confidence (without proof) that neither is leukemogenic. The results of a long-term safety study of anagrelide in over 4,000 MPD patients are eagerly awaited. Long-term stable control with these agents may, on the other hand, interrupt the natural progressive history of the disease and postpone indefinitely the development of such secondary sequelae as acute leukemia and myelodysplastic syndrome. Data from one recent prospective study indicate, however, that treatment with anagrelide offers no benefit to patients with myelofibrosis with myeloid metaplasia.4 TREATMENT-RELATED SEQUELAE
A patient’s risk stratum is the primary determinant of treatment.17 In ET, patients who are less than 40 years of age, have platelet counts in the range of 600 to 1,000 ⫻ 109/L, are asymptomatic, and have no comorbidities are considered low-risk patients. These patients are not candidates for myelosuppressive or cytoreductive therapy, but may benefit from low-dose aspirin as prophylaxis against thrombosis. The benefit of aspirin for its antithrombotic properties must be balanced against the risk of hemorrhage for each patient. Patients who are 60 years of age or above, or patients of any age who are symptomatic and have at least one comorbidity, are considered high-risk patients. A history of thrombosis is the leading indicator of risk for subsequent thrombosis.18 Recent evidence, based on a study of 65 patients, suggests that subnormal plasma erythropoietin levels in newly diagnosed ET patients may be a risk factor for vascular complications.19 These patients are candidates for myelosuppressive (HU, IFN-␣) or megakaryocyte-specific cytoreductive (anagrelide) therapy to control platelet counts in the range of 400 to 600 ⫻ 109/L. All other patients, particularly asymptomatic individuals who are between 40 and 60 years of age and have no comorbidities, are at indeterminate risk for thrombotic complications. A decision to treat other than with low-dose aspirin must include a careful balancing of the potential long-term adverse effects of therapy against the potential benefits. Phlebotomy, the standard first-line treatment for PV, should be undertaken with all diagnosed patients with a goal of maintaining a maximum
SECONDARY SEQUELAE OF TREATMENTS FOR MPDS
hematocrit of 45%. Although some investigators distinguish between 45% for men and 42% for women, there are no controlled and randomized trial data to substantiate the clinical significance of the difference. The most common side effect of phlebotomy is iron deficiency. However, supplemental iron is not recommended because it accelerates erythropoiesis. The frequency of phlebotomy usually decreases over time, however, reducing the risk of chronic iron deficiency. Phlebotomy and iron deficiency may increase thrombocythemia. Low-dose aspirin is frequently used to reduce the risk of symptomatic platelet aggregation that may cause thrombosis or microvascular complications of PV, such as neurologic changes,20 visual disturbance,21 and erythromelalgia.22,23 Occasionally low-dose aspirin induces hemorrhage, but it is usually minor. It is contraindicated in patients with marked thrombocythemia because of the risk of cerebral hemorrhage. Patients whose hematocrits are not controlled by phlebotomy require myelosuppressive therapy. Platelet-specific cytoreduction treatment is of no value to PV patients except in the event that thrombocythemia accompanying erythrocythemia is not controlled by myelosuppression. The drug selected determines the treatmentrelated sequelae. Hydroxyurea is an antimetabolite and a dose-dependent, nonspecific myelosuppressive agent. The use of high doses to control platelet counts in ET increases the risks of neutropenia and anemia. In general, the dose required for platelet control at 600 ⫻ 109/L does not induce these complications, explaining in part why control at this level, now regarded as high, was considered adequate for so long. Adjustments in the dose of HU may give rise to marked oscillation of platelet counts,24 including a rebound to even higher platelet counts than existed before the initiation of therapy. It is possible, however, that dose-limiting factors and the relatively low longterm cumulative ingestion may result in less treatment-induced leukemic conversion than might have occurred otherwise. In PV patients, HU doses required for effective control of hematocrits may induce thrombocytopenia and/or neutropenia. Dose reduction designed to prevent these complications may diminish efficacy with respect to erythrocytes, thus requiring subsequent phlebotomies in patients treated with HU. Other side effects of HU include cutaneous ulceration,25,26
25
mucositis, and hyperpigmentation. Several cases have been reported of painful oral ulcers27 and leg ulcers, usually in the malleolar region, that are unresponsive to secondary treatment and require discontinuation of HU.28,29 Hydroxyurea treatment has also been associated with drug fever30 and life-threatening hypersensitivity pneumonitis.31 Gastrointestinal sequelae include nausea, vomiting, bloating, and diarrhea. Headache and asthenia are the most common systemic side effects. Hydroxyurea is known to be mutagenic in humans, and there is inconclusive evidence that it is leukemogenic. IFN-␣ may also have many side effects, the most common of which are flu-like symptoms including chills and fever. Patients may experience myalgia, fatigue, anorexia, headache, depression, and/or neuropathy. Potential hematologic side effects are neutropenia and anemia. Mild alopecia is common. Its action on the intestinal mucosa may result in diarrhea. As an immunomodulator, IFN-␣ may cause symptomatic episodes of autoimmune disease including hypothyroidism. This high profile of potential side effects makes IFN-␣ intolerable to approximately 30% of patients, and those who tolerate it may incur moderate expense to relieve symptoms of the sequelae. The side effects of anagrelide are because it is a prostaglandin inhibitor and a vasodilator. Hence, it may cause headache and dizziness. As an inotrope it may induce palpitations, a sense of an unusually strong heartbeat, and even cardioarrhythmias. Anagrelide-induced cardiomyopathy has been reported, and is thought to be caused by phosphodiesterase inhibition leading to cardiomyopathy and vasodilation.32 Anagrelide also changes the renal blood flow and induces fluid retention. At present the vehicle of the drug is lactose, so the bloating, nausea, diarrhea, and vomiting that some patients experience may be due to some extent by lactose sensitivity as well as the fact that prostaglandins do not affect gastrointestinal function. Patients who are in borderline cardiac function may experience pulmonary edema following ingestion of anagrelide, and patients with poor cardiac function may lapse into congestive heart failure. It appears, however, that elderly patients with these risk factors may tolerate anagrelide when treated concurrently with good cardiotonic regimens. In a recent review of 35 young patients (age range, 17 to 48 years) treated
26
HARRIET S. GILBERT
with anagrelide for ET before 1992, 24% experienced mild-to-moderate anemia.33 CONCLUSION
It is unique that there are now three agents with different mechanisms of action and with effects on the progeny of the MPD clone that differ markedly. Because these permit good control of platelet counts and hematocrits, the next steps in treating ET and PV should be to reduce toxicity and to capture synergies by using combined-agent therapy. One example in which that has been performed successfully is the addition of anagrelide to HU for incremental improvement in platelet counts without chronic overexposure to HU and without increasing the risk of neutropenia. Very low-dose IFN-␣ has also been added to anagrelide therapy to improve platelet counts. By exploiting the different mechanisms of these three classes of drugs, ET and PV can be treated successfully with little risk of cumulative treatment-related sequelae and minimal exposure to potential unforeseen consequences of long-term therapy. REFERENCES 1. Berk PD, Goldberg JD, Fruchtman SM, et al: Therapeutic recommendations in polycythemia vera based on Polycythemia Vera Study Group protocols. Semin Hematol 23:132-143, 1986 2. Andraesson B, Swolin B, Kutti J: Hydroxyurea treatment reduces haematopoietic progenitor growth and CD34 positive cells in polycythemia vera and essential thrombocythemia. Eur J Haematol 64:188-193, 2000 3. Spencer CM, Brogden RN: Anagrelide: A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in the treatment of thrombocythemia. Drugs 7:809-822, 1994 4. Yoon SY, Li CY, Mesa RA, et al: Bone marrow effects of anagrelide therapy in patients with myelofibrosis with myeloid metaplasia. Br J Haematol 106:682-688, 1999 5. Tabata M, Imagawa S, Tarumoto T, et al: Essential thrombocythemia transformed to acute myelogenous leukemia with t(3;17)(p24;q12), del(5)(q13q34) after treatment with carboquone and hydroxyurea. Jpn J Clin Oncol 30:310-312, 2000 6. Berk PD, Wasserman LR, Fruchtman SM, et al: Treatment of polycythemia vera: A summary of clinical trials conducted by the Polycythemia Vera Study Group, in Wasserman LR, Berk PD, Berlin NI (eds.): Polycythemia Vera and the Myeloproliferative Disorders. Philadelphia, PA, Saunders, 1995, pp 102-113 7. Regev A, Stark P, Blickstein D, et al: Thrombotic complications in essential thrombocythemia with relatively low platelet counts. Am J Hematol 56:168-172, 1997 8. Lengfelder E, Hochhaus A, Kronawitter U, et al: Should a platelet count of 600 ⫻ 109/L be used as a diagnostic criterion
in essential thrombocythaemia? An analysis of the natural course including early stages. Br J Haematol. 199:15-23, 1998 9. Sacchi S, Vinci G, Gugliotta L, et al: Diagnosis of essential thrombocythemia at platelet counts between 400 and 600 ⫻ 109/L. Gruppo Italiano Malattie Mieloproliferative Croniche (GIMMC). Haematologica 85:492-495, 2000 10. Laguna MS, Kornblihtt LI, Marta RF, et al: Effectiveness of anagrelide in the treatment of symptomatic patients with essential thrombocythemia. Clin Appl Thromb Hemost 6:157161, 2000 11. Gilbert HS: Long term treatment of myeloproliferative disease with interferon-alpha-2b: Feasibility and efficacy. Cancer 83:1205-1213, 1998 12. Jurado M, Deeg HJ, Gooley T, et al: Hemopoietic stem cell transplantation for advanced polycythemia vera or essential thrombocythemia. Br J Haematol 112:392-396, 2001 13. Lengfelder E, Berger U, Hehlmann R: Interferon alpha in the treatment of polycythemia vera. Ann Hematol 79:103109, 2000 14. Fruchtman SM, Mack K, Kaplan M, et al: From efficacy to safety: A Polycythemia Vera Study Group report on hydroxyurea in patients with polycythemia vera. Semin Hematol 34:17-23, 1997 15. Nand S, Stock W, Godwin J, et al: Leukemogenic risk of hydroxyurea therapy in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Am J Hematol 52:42-46, 1996 16. Finazzi G, Ruggeri M, Rodeghiero F, et al: Second malignancies in patients with essential thrombocythaemia treated with busulphan and hydroxyurea: Long-term follow-up of a randomized clinical trial. Br J Hematol 110:577-583, 2000 17. Gilbert HS: Diagnosis and treatment of thrombocythemia in myeloproliferative disorders. Oncology (Huntingt) 15: 989-998, 2001 18. Cortelazzo S, Viero P, Finazzi G, et al: Incidence and risk factors for thrombotic complications in a historical cohort of 100 patients with essential thrombocythemia. J Clin Oncol 8:556-562, 1990 19. Andraesson B, Lindstedt G, Kutti J: Plasma erythropoietin in essential thrombocythaemia: At diagnosis and in response to myelosuppressive treatment. Leuk Lymphoma 38: 113-120, 2000 20. Jabaily J, Iland HJ, Laszlo J, et al: Neurologic manifestations of essential thrombocythemia. Ann Intern Med 99:513518, 1983 21. Koudstall PJ, Koudstall A: Neurologic and visual symptoms in essential thrombocythemia: Efficacy of low-dose aspirin. Semin Thromb Hemost 23:365-370, 1997 22. VanGenderen PJJ, Michiels JJ: Erythromelalgia: A pathognomonic microvascular thrombotic complication in essential thrombocythemia and polycythemia vera. Semin Thromb Hemost 23:357-363, 1997 23. Michiels JJ: Erythromelalgia and vascular complications in polycythemia vera. Semin Thromb Hemost 23:441-454, 1997 24. Tefferi A, Elliott MA, Kao PC, et al: Hydroxyureainduced marked oscillations of platelet counts in patients with polycythemia vera. Blood 96:1582-1584, 2000 25. Young HS, Khan AS, Kendra JR, et al: The cutaneous side-effects of hydroxyurea. Clin Lab Hematol 22:229-232, 2000
SECONDARY SEQUELAE OF TREATMENTS FOR MPDS
26. McLintock C, Ockelford PA: Hydroxyurea-induced skin ulceration in myeloproliferative disorders. Aust N Z J Med 30:373-376, 2000 27. Paleri V, Lindsey L: Oral ulcers caused by hydroxyurea. J Laryngol Otol 114:976-977, 2000 28. Poros A, Nadasdy K: Leg ulcer in hydroxyurea-treated patients. Haematologia (Budap) 30:313-318, 2000 29. Murphy J, Morley SM: Hydroxyurea-induced leg ulcers. Hosp Med 61:510a, 2000 (abstr)
27
30. Braester A, Quitt M: Hydroxyurea as a cause of drug fever. Acta Haematol 104:50-51, 2000 31. Sandhu HS, Barnes PJ, Hernandez P: Hydroxyurea-induced hypersensitive pneumonitis. Can Resp J 7:491-495, 2000 32. James CW: Anagrelide-induced cardiomyopathy. Pharmacotherapy 20:1224-1227, 2000 33. Storen EC, Tefferi A: Long-term use of anagrelide in young patients with essential thrombocythemia. Blood 97:863866, 2001