Treatment of chronic myeloid leukaemia

9 Treatment of chronic myeloid leukaemia N O R M A N C. A L L A N P A T R I C I A C. A. S H E P H E R D

The treatment of chronic myeloid leukaemia (CML) remains a major problem. With the exception of allogeneic bone marrow transplantation which is available to only a minority of patients, no treatment significantly alters the natural history of the disease, although substantial if temporary improvement in a patient's well-being can usually be obtained. CML is an acquired clonal disorder of the haemopoietic stem cell which leads to an excessive, uncontrolled and inappropriate production ofmaturing myeloid elements. In 90% of patients a characteristic chromosome abnormality known as the Philadelphia (Ph) chromosome is found. If it is assumed correctly that this chromosome abnormality is a definitive marker of the disease, almost all the cells in the haemopoietic system including lymphocytes are involved. In almost all Ph positive cases 100% of the cells analysed show the Ph chromosome. Moreover some apparently Ph negative patients have the same molecular abnormality as the overtly Ph positive patients--the translocation of the A B L gene to chromosome 22 adjacent to the B C R region (Dreazen et al, 1987). However, even in Ph positive CML some residual haemopoietic elements not carrying the Ph chromosome and therefore presumably normal survive since they can reappear when the Ph positive clone is suppressed by aggressive chemotherapy (Singer et al, 1980) (see Chapter 4). The problem that confronts the therapist lies in the nature of the disease. No treatment has yet been devised which will eradicate the malignant clone while allowing the normal cell lines to survive, proliferate and repopulate the marrow. Aggressive treatment designed to eliminate the disease also destroys the residual normal stem cells and can be employed only in the context of bone marrow transplantation. The role of bone marrow transplantation and autografting is dealt with in the following chapter. This chapter deals with those forms of therapy that are essentially palliative. NATURAL HISTORY Treated or untreated the disease runs a chronic course for a randomly variable period of time. The median survival is about 33 months. In the majority of patients the end stage is characterized by a phase in which leukaemic blast cells Bailliere's Clinical Haematologk~Vol. I, No. 4, December 1987

1031

1032

N. C. ALLAN AND P. C. A. SHEPHERD

overtake the proliferating elements and establish a 'maturation arrest' type of disease typical of acute leukaemia. In many, but not all, patients this 'acute' or 'blast transformation' (BT) phase is preceded by a less readily defined stage in which the disease becomes less easily controlled by conventional treatment, the so-called 'accelerated' phase (AP). The haematological picture in AP still resembles the chronic phase but there may be an increasing percentage of immature forms. Very high platelet counts are common during the evolution of CML but thrombocytopenia can also be a problem. A steadily rising basophil count is also frequently found. Anaemia often reappears. Patients may be already in accelerated or even in blast transformation at diagnosis. A few patients never enter a BT phase but progress to an indolent refractory phase and some have features of myelofibrosis. Not all patients with CML require treatment at diagnosis. In some the diagnosis is an incidental finding in an asymptomatic patient. In such patients immediate treatment may be unnecessary although the patient should be kept under review. No prolongation of survival has been demonstrated from early treatment. However, the majority of patients have symptoms at diagnosis and treatment is then necessary to reduce splenomegaly and relieve hypermetabolic and other symptoms. RADIOTHERAPY The first successful treatment of this disease was achieved with radiotherapy (Pusey, 1902). Its value was reviewed 22 years later by Minot et al (1924). Until the 1950s radiotherapy, eithertotal body or splenic, was the mainstay of therapy. It caused a reduction in white cell and platelet counts, allowed a rise in haemoglobin, reduced the size of the spleen and improved the patient's wellbeing but did not improve survival (Minot et al, 1924). The benefit typically lasted about 6 months after which further therapy was required. This was continued until the disease became refractory, after which there was little that could be offered other than supportive care. Although radiotherapy is now seldom used as a primary treatment there has recently been renewed interest in the value of total body irradiation (Advani et al, 1983) because it may still be a reasonable alternative to chemotherapy. Radiotherapy is also valuable for treating certain complications such as extramedullary disease and massive splenomegaly when splenectomy is not possible. SINGLE AGENT CHEMOTHERAPY

Busulphan Galton reported the value of a new drug, busulphan, in 1953. Fifteen years later the results of a Medical Research Council (MRC, 1968) randomized controlled trial comparing radiotherapy with busulphan established that busulphan therapy was superior, since it produced better disease control for longer periods. Busulphan thereafter became the standard by which all other

TREATMENT OF CML

1033

single agents have been judged. It may be given at a low dose continuously, at a slightly higher dose intermittently or at a much higher dose with intervals of at least 4 weeks between doses. These various methods of administration are discussed in greater detail by Goldman and Baughan (1984). Busulphan therapy is not without problems. If not carefully monitored severe myelosuppression may occur which can prove fatal, although survival after aplasia.ls often associated with very long remissions (Greig, 1956; Finney et al, 19.2). Counts may continue to fall after the busulphan is stopped and the period of aplasia may continue for many months. In a small minority of patients a severe idiosyncratic pulmonary reaction with interstitial fibrosis associated with constitutional symptoms occurs, known as busulphan htng, which also can prove lethal (Kirschner and Esterley, 1971; Podoll and Winkler, 1974). Primary resistance to busulphan, although uncommon, is observed (Haut et al, 1961). Increasing skin pigmentation is found with prolonged use. In women disorders of menstruation occur and infertility is usual in both sexes. Cataract (Ravindranathan et al, 1972), sicca syndrome (Sidi et al, 1977), myasthenia gravis (Djaldetti et al, 1968) and endocardial fibrosis (Weinberger et al, 1975) have been reported. As other drugs were discovered each was tested and compared with busulphan. These included chlorambucil (Rundles et al, 1959), dibromomannitol (Eckhardt et al, 1963; Ramanan and Israels, 1969; Silver et al, 1987a), 6mercaptopurine (Huguley et al, 1963), 6-thioguanine (Spiers et al, 1975a), cyclophosphamide (Kaung et al, 1971) and melphalan (Hauch et al, 1978). None has proved better as primary therapy although some, particularly dibromomannitol, have their devotees.

Hydroxyurea The one drug that may take over the mantle as the best single agent for this diseaseis hydroxyurea, a cycle specific inhibitor of DNA synthesis. First manufactured 120 years ago in Germany (Dresler and Stein, 1869), its usefulness in CML was not reported until about 100 years later (Kennedy and Yarbro, 1966). Although it produces less smooth control of the leukocyte count than busulphan (Kennedy, 1969), one group has suggested that it may be the first drug to extend the duration of the chronic phase (Bolin et al, 1982). This has not however been confirmed in any prospectively randomized controlled trial and the survival difference between patients treated with busulphan and hydroxyurea was not statistically significant. Goldman and Baughan (1984) suggested that hydroxyurea was useful in three circumstances-for rapid reduction of very high leukocyte counts, in busulphan intolerance and for busulphan resistance. Hitherto it has tended to be held in reserve as second-line therapy since it is useful also in the accelerated and possibly blast crisis phases (Schwartz and Canellos, 1975). However, its use as first-line therapy is increasing. Although it is myelotoxic, prolonged or serious marrow suppression is rare (Kennedy, 1972) and it appears to be more platelet sparing than busulphan. The drug is well tolerated by most patients with very few side effects. There may be nausea and loss of appetite, usually transient.

1034

N. C. ALLAN AND P. C. A. SHEPHERD Table 1. Standard drug dosages for initial control of CML.

Drug

Dose

Reference

Busulphan

0.I mg/kg daily 4-10 mg daily Galton (1953) Dibromomannitol 125-250 mg daily Eckhardt et al (1963) Ramanan & I sraels (1969) Cyclophosphamide 2.0 mg/kg daily Kaung et al (1971) Melphalan 4-12 mg daily Hauch et al (1978) Hydroxyurea 0.5-2.0 g daily Bolin (1982)

Skin changes, such as atrophy and scaling, and partial alopecia may occur (Kennedy ct al, 1975). It causes considerable macrocytosis of the red cells and megaloblastic changes in the marrow.

Dosage in single agent therapy Most of these single agents are used in a similar manner. A standard dose of drug is given daily and the leukocyte and platelet counts are monitored regularly at one to two week intervals until the leukocyte count drops to between 25 and 20 x 109/litre. In the case of busulphan the drug is then stopped in order to avoid the risk ofmarrow aplasia. Once the leukocyte count rises above this figure again, busulphan is reintroduced at a smaller dose and the dose regulated to maintain the count around 10 x 109/litre. It is probably unwise to allow the count to drop much lower as there is then a risk of severe suppression of haemopoiesis. After a period of stability it is often possible to withdraw busulphan for varying periods of time with reintroduction of the drug when necessary. Occasionally no further therapy is required for long periods, even years. With hydroxyurea however the leukocyte count tends to rise fairly soon after stopping the drug. Resistance eventually appears, often with the onset of accelerated phase, and alternative therapy is required. Standard doses for this type of therapy for the more commonly used drugs are given in Table 1. There appears to be no advantage in switching from one drug such as hydroxyurea to another such as busulphan at a predetermined time before resistance to the first appears, since this approach does not delay blast transformation (Baccarani et al, 1981). INTENSIVE THERAPY IN CHRONIC PHASE Frustration with the purely palliative nature of single agent drug therapy led some workers to evaluate more aggressive regimens similar to those employed in acute leukaemia. The hope was that genuine remission could be established by eliminating the malignant clone and allowing normal cell lines to repopulate the marrow (Dowling et al, 1974; Sharp et al, 1979). In practice, considerable suppression of the Ph positive clone was achieved in some cases and temporary disappearance in a few. Cunningham et al (1979) treated 37

TREATMENT OF CML

1035

patients with the L-5 protocol and achieved temporary rcduction in Ph positivity in 12. In 7 of these no Ph chromosome was found for variable periods. Median survival was 50 months for all cases but those showing reduction in Ph positivity had a longer median survival that exceeded 108 months. As this was not a randomized trial it is impossible to know whether there was genuine prolongation of survival or whether the treatment was merely more 'successful' in those patients destined to have a long chronic phase. Unless survival can be extended the morbidity associated with aggressive regimens is unacceptable, not least because premature fatalities have occurred. NON-INTENSIVE COMBINATION THERAPY Another approach has been to combine two drugs active as single agents. Allan et al (1978) reported the combination of busulphan 2 mg and 6mercaptopurine 50 mg with allopurino1300 mg daily. Ifallopurinol could not be given the 6-mercaptopurine dose was doubled to 100 mg daily. In initial studies in untreated patients the effect on the leukocyte count was dramatic. It fell precipitously until it reached normal values, then plateaued and thereafter fell only gradually. Withdrawal of the drugs resulted in a rapid rise in the leukocyte count. Subsequent investigation confirmed that this rapid response occurred in almost all patients with little risk of severe marrow depression. Some dose reduction was usually required at the plateau and this was best achieved by reducing the number ofdays per week on which the standard daily dose was given. This drug combination was tried in a patient with eosinophilic CML with a double Ph chromosome (Stockdill et al, 1980). There was a rapid reduction in the eosinophils to normal levels with disappearance of the double Ph chromosome. The neutrophil counts however did not change significantly and it was thought that they represented a normal cell line without the Ph chromosome. If this interpretation were correct this combination might be significantly more toxic to leukaemic cells than to the normal lineage. In 1979 a randomized controlled trial comparing a similar combination With single agent busulphan was begun by the MRC. Instead of 6mercaptopurine, 6-thioguanine was used because of its lack of interaction with allopurinol. The use of this combination showed no survival advantage compared with busulphan alone but it did bring the disease under control more rapidly. Unfortunately the combination which included 6-thioguanine, which is known to be hepatotoxic, has been associated with the development of portal hypertension and oesophageal varices when used over long periods (Key et al, 1987). It cannot therefore be recommended for the long-term palliative management of CML. It remains however a very effective means of reducing very high leukocyte counts which may be associated with hyperviscosity. CYCLICAL THERAPY The search for a way to delay transformation or the emergence of drug resistance led some therapists to employ various drugs active in CML in

1036

N. C. ALLAN AND P. C. A. SHEPHERD

sequence. One such study was reported by Spiers (1976). It was abandoned because no survival advantage was obtained and there were too many disadvantages associated with the complicated use of a number of different drugs. INTERFERONS (IFNs) Interferons are cellular glycoproteins or cytokines with antiproliferative, antiviral and immunoregulatory properties. They are produced by various cell types in response to viral infection. There are three species, named ~, fl and ~. Alpha-interferons (IFN-~) derived from leukocytes (IFN-c~-Le) or lymphoblastoid cell lines (IFN-alfa-Nl) and beta-interferons (IFN-fl) derived from fibroblasts are classed together as Type I. Gamma-interferon (IFN-7) or immune interferon forms a separate class (Type II). IFN-~ differs from ~ and fl in some of its biological activities and there is some evidence that the combination of~ or fl with ~ interferon may be synergistic. IFN-~ is prepared from human leukocytes or lymphoblastoid cell lines following treatment with virus, or by recombinant techniques. The former method produces a mixture of 8 or more interferons whereas the latter yields only one pure molecular type. IFN-~ is produced by stimulating T cells with either mitogen or antigen or by recombinant technology. The use of IFN-c~ in CML has been evaluated by the Houston group over the past six years (Talpaz et al, 1986; 1987). In the most recent report 51 patients in chronic phase were treated with doses of 3-9 mega units daily (Talpaz et al, 1987). The majority had not received any prior treatment. IFN-~ resulted in satisfactory control of the disease in 36 patients (71%) with normalization of peripheral blood counts and gradual decrease in bone marrow hypercellularity. A partial response was seen in 5 (10%) patients and no response in 10 (19%). Of most interest was the finding that serial cytogenetic studies showed reduction in the percentage of Ph positive cells in 20 of the 36 patients in whom satisfactory control was established; in 18 patients these changes persisted for 6 months or longer and in eight the percentage fell to less than 35% of pretreatment levels. Yoffe et al (1987) from the same group reported two patients in whom the Ph positive clone disappeared, in one after 6 months of treatment and the other after 12 months, being sustained up to 18 months from start of therapy. Molecular analysis showed that this was accompanied by loss of B C R rearrangement providing further evidence that the leukaemia clone had been suppressed. However the Ph positive clone may not have been completely eradicated and it is uncertain whether survival for these patients will be prolonged. More recently Kantarjian reported (personal communication, 1988) that one patient who had received IFN-~ for four years and had been Ph negative for the last three had remained Ph negative for 15 months after stopping therapy. One other similar patient had shown reappearance of 5-10% Ph positive cells within a year of stopping treatment. Other workers have reported similar results. Niederle et al (1987) obtained good haematological control in 20 out of 29 patients (69%) in chronic phase

TREATMENT OF CML

1037

but in none was the Ph positive clone completely suppressed. Alimena et al (1987) described 63 patients in chronic phase given recombinant IFN~-2B for 2-12 months; 28 were untreated and 37 had been treated previously by conventional means. Complete haematologicai response was obtained in 44 (70%) and partial response in 16 (25%). Patients treated previously with other drugs were less responsive and tended to require higher doses of IFN-c~. One responder developed an abrupt blast transformation while on treatment with IFN-c~ after 6 months and two partial responders became IFN-c~resistant after 8 and 9 months respectively. Cytogenetic response was observed with reduction of Ph positive cells to a median of 78% (range 16-97). In none was there total suppression of the Ph positive clone. The consensus appears to be that IFN-cq whether lymphoblastoid or recombinant in origin, will produce good haematological control in about 70% of patients in chronic phase whether they have had previous therapy or not, although previous therapy may make the disease slightly more resistant. Control generally takes three or more months to achieve. It is also clear that it does not prevent progression to accelerated phase and blast crisis and resistance to IFN-~ can appear. Variable degrees of reduction ofPh positivity occur and may persist while on treatment but apparent elimination of the Ph positive clone is rare and cannot be proved. It is not known whether IFN-~ therapy extends the chronic phase. It may be of value in the accelerated phase but has little effect in blast crisis. Little is known about the mechanism of action of IFN. Studies in vitro suggest that the main effect of IFN-~ is to suppress proliferation (Opalka et al, 1987). Bergsagel et al (1987) studied the effect o f l F N on the growth in vitro of colonies from myeloid leukaemic blast cells and normal haemopoietic precursor cells. They found that the growth of malignant and normal precursors was inhibited equally but noted that the self-renewal capacity of acute myeloid leukaemic blasts was reduced by IFN. They argued that IFN should be tested for its ability to prolong remission and applied this concept to seven patients. The leukocyte doubling time was extended and length of disease control was prolonged after initial remission induction by busulphan. 'Most of the information on the value of IFN-~ in CML has come from studies of relatively small numbers of patients. The MRC have recently started a randomized multi-centre trial using lymphoblastoid IFN-~ (Wello feron) in which control of the disease is established by conventional chemotherapy and patients are then randomized to receive or not to receive IFN-~. Patients in both arms of the trial may receive further chemotherapy if and when the total leukocyte count exceeds 50• 109/litre. Thus the only 9variable is the use of IFN-cc in one arm of the study. Karyotype analysis is undertaken every 3 months in the first year and 6-monthly thereafter. This trial should provide information on the influence of IFN-a on the natural history of the chronic phase and cytogenetic changes. IFN-y also suppresses the proliferation of haemopoietic progenitor cells9 Kurzrock et al (1987) reported its use in 30 patients in chronic phase CML. Of the 26 evaluable patients, six (23%) achieved good control of the disease and four partial control. Twenty had previously had other chemotherapy. These results were less satisfactory than with IFN-~ but five patients had cytogenetic

1038

N. C. ALLAN AND P. C. A. SHEPHERD

improvement with reduction in the Ph positive clone to as low as 55%. Sideeffects appeared to be worse than with IFN-c~. Silver et al (1987b) reported haematological response to IFN
TREATMENT OF CML

1039

in the terminal phase. There is some evidence that patients with major karyotypic abnormalities, especially double Ph and hyperdiploidy, benefit least from splenectomy. Nevertheless splenectomy is not totally valueless in CML. It may be appropriate for symptomatic splenomegaly. When acceleration or blast crisis has supervened it may produce benefit in about 15% of cases (Gomez et al, 1976), but there is no way of predicting this response before splenectomy. It may also help patients with thrombocytopenia or high transfusion requirements (Gomez et al, 1976). The value of splenectomy at this stage has to be balanced against the increased risk of the operation. In addition, splenectomy may be indicated for the patient in accelerated phase with splenomegaly that is refractory to therapy. It may save the patient the discomfort associated with massive enlargement. Splenectomy is not without hazard and should be undertaken by experienced surgeons. Donor platelets may be valuable irrespective of the patient's platelet count (McBride and Hester, 1977). If the splenic bed is large, antifibrinolytic agents such as tranexamic acid may be administered while the abdomen is still open and the splenic bed can be viewed until dry. Topical antibiotics may reduce the incidence of postoperative infection, particularly of subphrenic abscess.

MANAGEMENT OF ACCELERATED PHASE

Definition The phase of acceleration, intermediate between chronic phase and frank blast transformation, is not easily defined. Suggested criteria are listed by Goldman and Baughan (1984) and include amongst others increasing leukocyte count, splenomegaly, thrombocytosis (in the non-splenectomized patient), blast cells in the blood or marrow in excess of 10%, blasts and promyelocytes more than 20%, anaemia below 100 g/litre unresponsive to treatment, acquisition of additional chromosomal abnormalities, basophilia or eosinophilia in excess of 20% in the blood and increasing marrow fibrosis. Any one of these criteria alone may not be sufficient grounds to establish accelerated phase.

Cytotoxic drugs If the patient has been treated initially with busulphan a switch to hydroxyurea may be useful. The value ofhydroxyurea in the accelerated phase has been cited as a reason for reserving this drug for this phase of the disease. While hydroxyurea is undoubtedly the best drug, other single agents may be tried if they have not been previously used. There is little enthusiasm for the use of more aggressive therapeutic regimens in acceleration. This probably reflects the lack of evidence that they provide any real benefit and the belief that they should be reserved for the phase of BT.

1040

N. C. ALLAN AND P. C. A. SHEPHERD

SPECIFIC PROBLEMS IN CHRONIC AND ACCELERATED PHASES

Splenic infarction Infarction of the spleen occurs most often in patients who have substantial splenic enlargement. The onset is often rapid and may be associated with systemic upset and fever. The patient complains of severe pleuritic-type pain in the left abdomen or left lower chest, exacerbated by breathing deeply, coughing or sneezing. Left shoulder tip pain may occur if the infarcted area is in contact with the diaphragm. The spleen is often exquisitely tender to palpation and a friction rub may be heard over the affected part. It may mimic and be mistaken for pleurisy. Treatment involves rest, strong analgesics while the pain is severe and adequate hydration. Concomitant infections should be vigorously treated. The pain usually subsides in 7-10 days. Recurrent infarctions may require splenectomy.

Thrombocytosis Marked thrombocytosis with a platelet count in excess of 1000 x 109/litre is found at diagnosis in 8% of patients. Thrombocytosis is significantly more frequent in female patients for reasons that are unknown (males 5.2% versus females 11.7%, P= <0.001; observations from the MRC CML II Trial). Patients with platelet counts > 1000 x 109/litre at presentation have an 81% chance of having counts above this level later in the course of their disease, whereas patients who do not present with such high counts have only a 29% likelihood of having such high counts subsequently. If absent at diagnosis, thrombocytosis may develop early in treatment and persist for a variable period. It responds eventually to treatment in most cases. Failure to control the thrombocytosis at this stage suggests that the patient may be already in accelerated phase. Thrombocytosis appearing later during chronic phase usually indicates acceleration of the disease and can be very resistant to drug therapy. Attempts to reduce the thrombocytosis may be effective but often result in severe leukopenia. Some forms of therapy are less likely to control thrombocytosis than others. In the MRC CML II Trial in which single agent busulphan therapy was compared with a combination of busulphan at a lower dose and 6thioguanine, thrombocytosis (> 1000 • 109/litre) occurred in 90 out of 338 patients (27%) on busulphan only but in 129 of 333 patients (36%) on the combination (P= <0.02). The total number of weeks during which the platelet count was in excess of 1000 • 109/litre in the busulphan only arm was 1332 versus 2430 for the combination arm. Thus busulphan appears to be more effective in preventing the development of marked thrombOcytosis. Thrombocytosis in myeloproliferatiVe disorders is generally thought to increase the risk of haemorrhagic and thromboembolic complications but in CML this increased risk is relatively small. Mason et al (1974) reviewed 111 Ph positive patients and found no evidence of problems arising when the platelet count was between 500 and 1000 x 109/iitre. Twenty-nine of the patients had counts in excess of 1000 • 109/iitre. Only four thrombohaemorrhagic compli-

TREATMENT OF CML

1041

cations were found, three of haemorrhage, one of which followed dental extraction. There was one thrombotic complication (priapism). None was fatal. However, severe haemorrhagic complications of CML may occur without significant thrombocytosis following major surgery. One of the authors (NCA) observed a patient not previously known to have CML who had intractable bleeding for 5 days following partial gastrectomy for pyloric stenosis reqtiiring numerous transfusions. Further surgical intervention failed to stop the bleeding and no definite site ofsurgical bleeding was found. At this point a diagnosis of CML was made. The platelet count was 452 x 109/litre. A transfusion of normal platelets stopped the bleeding. Marked and sustained thrombocytosis occurs frequently following splenectomy. The Italian Cooperative Study Group (1984) on CML reported 83 splenectomized patients. A platelet count in excess of 500 x 109/1 was observed in 52%. In eight cases thromboembolic or vascular complications occurred but in only three of these was a thrombocytosis probably contributory. Leucocytosis was the most likely cause of priapism in three cases and in two others the platelet count was less than 500 x 109/1. The rarity of serious complications from marked thrombocytosis in CML means that the clinician can avoid major therapeutic efforts to reduce the high platelet count. None of the drugs commonly used will achieve this without the risk of excessive depression of the white cell count. However, Kurzrock et al (1987) reported reduction of high platelet counts to normal within 5 weeks in 4 patients using IFN-~, without significant depression of the leukocyte count. Similar effects have been reported with IFN-ct by Talpaz et al (1983) although there is greater likelihood that IFN-c~ will also reduce the leukocyte count. IFN therapy may therefore be the best way of controlling excessive thrombocytosis in CML. Administration of 32p is not recommended because of its unpredictable effect (Mason et al, 1974). Aspirin is probably oflittle help and may exacerbate the bleeding tendency. Rarely, central nervous system symptoms such as headaches, transient ischaemic attacks and transient amaurosis may require urgent reduction of platelet count. Plateletpheresis may be employed in conjunction with measures mentioned above.

Extramedullary disease The term extramedullary disease (EMD) describes the appearance of deposits or infiltrates of primitive myeloid cells at sites other than spleen, liver or bone marrow. Its reported frequency varies. Terjanian et ai (1987) described 24 cases out of 303 patients (7.9%). In contrast Rosenthal et al (1977) found EMD in 37% of 73 cases. In almost halfofthese the EMD presented at about the same time as blast crisis. Similarly Marks et al (1978) noted that EMD was present in 16 of 50 patients (32%) and usually heralded blast crisis. However, in some of these EMD was only demonstrated at post mortem. From these studies it is clear that EMD is associated with the accelerated or blast crisis phase of the disease and carries a poor prognosis. In the study by Terjanian the interval to blast crisis from diagnosis of EMD was 4 months and median survival 5 months. Lymph nodes were most commonly affected (54%), followed by bones (37%) and skin and subcutaneous tissues (29%).

1042

N. C. ALLAN AND P. C. A. SHEPHERD

Seven patients had multiple sites of involvement. The development of EMD was not associated with any particular form of preceding treatment; both single agent and combination therapy had been used. The appearance of EMD should alert the clinician to the possibility that the disease is accelerating and that a change of therapy may be required. Chromosomal analysis to confirm the clinical suspicion of evolution is worthwhile. Where the EMD is localized, local therapy is probably the most effective. Skin lesions can be treated with topical nitrogen mustard (Murphy et al, 1985). Other tumorous lesions respond well to local radiotherapy. If the underlying disease is changing to an accelerated or blast phase or if the EMD is widespread, treatment appropriate to the acute stage of the disease may produce improvement in the local lesions. Hyperviscosity Hyperviscosity is a relatively rare complication of CML and is related to high leukocyte counts (reviewed by Goldman and Baughan, 1984). In the MRC CML II Trial it occurred in about 1% of patients. The total leukocyte count was > 3 0 0 x 109/litre in all but one case (l15x 109/litre); in this patient however the platelet count was 1250 x 109/litre and typical fundal changes were seen. Hyperviscosity cannot be attributed solely to high leukocyte counts since some patients with very high counts have no evidence of the syndrome. Very high platelet counts alone do not usually cause hyperviscosity. Nevertheless treatment is usually designed to reduce the leukocyte count as rapidly as possible by urgent leukapheresis and chemotherapy. The combination of busulphan 2 mg and 6-mercaptopurine 50 mg or 6-thioguanine 80 mg daily orally is very effective if given with 300 mg of aliopurinol daily. The counts may be reduced and the hyperviscosity relieved within a few days. Alternatively hydroxyurea in doses up to 6 g daily may achieve similar effects. Anaemia is usually present but blood transfusion should be delayed until the hyperviscosity has been treated.

Priapism Priapism occurs in 1-2% of men with CML and may be a presenting feature (Suri et al, 1980; Goldman and Baughan 1984). It is often associated with high leukocyte and platelet counts. The patient may give a history of previous transient episodes of sustained erection. Treatment is difficult and often the patient presents late. Direct aspirat!on of the stale blood may relieve congestion (Schreibman et al, 1974) but sexual potency may still be lost. Rapid reduction ofwhite cell and platelet counts by apheresis is seldom effective and thrombolytic therapy has been disappointing. Relief for the patient often involves surgical intervention which is crude and usually destroys sexual function.

Hyperuricaemia Moderately raised levels of uric acid arc common in untreated CML. With treatment uric acid levels can rise unless the production of uric acid is curbed

TREATMENT OF CML

1043

by allopurinol therapy. If the uric acid level in the blood has been high the administration of allopurinol may precipitate an attack of acute gout by lowering the blood uric acid level causing mobilization of uric acid from deposits in the joints. The production of a large uric acid load by treatment also carries the risk of inducing urate nephropathy by obstruction of the ureters or renal tubules with urate crystals. Treatment involves urinary alkalinization with bicarbonate and gentle rehydration. Blockage of the ureters may require retrograde catheterization and lavage.

Hypercalcaemia Hypercalcaemia is a well known complication ofmetastatic malignant disease and myeloma but is relatively rare in leukaemia, most cases occurring in association with acute leukaemia. Only very few cases have been recorded in CML. Ballard and Marcus (1970) detail two cases; both had evidence of bone erosion and died with hypercalcaemia. Joyner et al (1977) described three patients in whom hypercalcaemia occurred during the accelerated phase ofthe disease. The second may have been a case of Ph positive acute lymphoblastie "leukaemia but the first and third cases were genuinely CML. The third had severe bone pain and the first lyric lesions and blood vessel calcification. Both patients were young, 19 and 23 years respectively. Despite the use of steroids, intravenous phosphate, calcitonin and mithramycin, the hypercalcaemia proved difficult to control and both patients died.

Hypokalaemia Hypokalaemia in the chronic phase of CML is rare. Evans and Bozdech (1981) reported two cases and attributed the hypokalaemia to hyperkaluresis secondary to a large tumour burden and lysozymuria. Treatment was by adequate potassium supplementation, chemotherapy and splenic irradiation.

Tumour lysis syndrome Tumour lysis syndrome secondary to massive cell destruction and release of metabolic products induced by chemotherapy has only rarely been reported in the accelerated or blastie phase of CML (Cervantes et al, 1982; Schiffer et al, 1982; Thomas et al, 1984). The usual clinical findings are hypocalcaemia, hyperphosphataemia, hyperkalaemia, hyperuricaemia and renal insufficiency. Acidosis and shock are often present. It has occurred despite prophylactic measures including hydration and allopurinol. Management is symptomatic with careful observation and monitoring of the patient and reversal of any metabolic abnormalities. Renal dialysis may be necessary. Careful watch for this syndrome is necessary particularly when intensive chemotherapy is given.

1044

N. C. ALLAN AND P. C. A. SHEPHERD

MANAGEMENT OF BLAST TRANSFORMATION The management of blast transformation is uniformly difficult. The patient is usually aware that the disease has become much more aggressive. Symptoms may be due to hypermetabolism, bleeding and infection. Increasing splenomegaly may cause pressure symptoms and painful infarcts may occur. Survival is a matter of a few weeks or months. Psychological support for the patients and their families is most important. This stage is also difficult for the physician who knows that available therapeutic options have little chance of improving survival. A decision has to be made whether to administer chemotherapy or to provide supportive care only. The latter may well be the more humane course particularly for older patients and a reasonable quality of life can sometimes be sustained for six months or more. However, chemotherapy does have a part to play in helping to alleviate symptoms, reducing organomegaly and producing some prolongation of life in responders.

Type of transformation Generally the initial choice of therapy is dictated by the type of transformation, which should be characterized by assessment of morphological, cytochemical and cell surface antigenic features. The majority are myeloid (60%), including rarely predominant involvement of the megakaryoblastic or erythroid cell lineages. Lymphoblastic characteristics are seen in about 30%; 4% are biphenotypic with both myeloid and lymphoid characteristics and the remainder are undifferentiated (Griffin et al, 1983; Bettelheim et al, 1985).

Lymphoblastic transformation Vincristhle and steroid combhlations.

Terminal deoxynucleotidyl transferase (Tdt) positive lymphoblastic transformations respond in 60% of cases to regimens which include vincristine (or other vinca alkaloids) and prednisolone (Marks et al, 1978). Response rates may be higher for those whose blasts are both Tdt and CD 10 positive (Janossy et al, 1979; Griffin et al, 1983). The median duration ofsurvival is still however short, around 4-6 months. Recently Waiters et al (1987) reported the use of continuous infusions of vincristine and doxorubicin (adriamycin) with dexamethasone (VAD) in patients with lymphoid transformation. Although the complete remission rates are similar, 55%, the median duration of remission was longer at 10 months. Muehleck et al (1984) however noted that patients whose blast cells had lymphoid characteristics but no cytogenetic abnormalities other than the Ph chromosome did better with a median survival of over one year.

TREATMENT OF CML

1045

Myeloid transformation Combhlation chemotherapy schedules. Myeloid transformations do not generally respond to vincristine and steroid combinations. The results ofchemotherapy have generally been poor, despite the use of a wide variety of regimens. This contrasts unfavourably with results obtained in de novo AML. Haematological remissions defined by reduction of the blast count in the marrow to less than 5% have occurred in only a small minority of patients, usually less than 20%, although some response with decrease in peripheral blood and marrow blasts is noted in many. True complete remission implies disappearance of the Ph chromosome. In most cases this does not occur and reversion to chronic phase is all that is possible, and this is usually short lived (Janossy et al, 1979). A number of authors have reported results with a variety ofdrugs in combination including doxorubicin, daunorubicin, vincristine, cytarabine, etoposide, cyclophosphamide, 5-azacytidine, bleomycin, hydroxyurea and prednisolone (Vallejos et al, 1974; Canellos et al, 1976b; Beard et al, 1976; Spiers et al, 1977; Marks et al, 1978; Coleman et al, 1980; Winton et al, 1981; Schiffer et al, 1982; Kantarjian et al, 1987; reviewed by Allen & Coleman, 1985). The number of reported regimens is a reflection of the poor results and the continued search for improvement. Comparisons between regimens are difficult due to the variable criteria used to judge response. In some reports patients were not classified into lymphoblastic and myeloid sub-types. Kantarjian et al (1987) in the largest reported series on therapy in blast crisis identified prognostic factors for induction of remission and survival--the most important being blast cell type and platelet count. The presence ofclonal evolution also had prognostic significance. Regardless of the regimen selected survival figures were very similar. This suggests that the major problem is the inherent biological resistance of the blasts. Even in patients who respond, median duration of survival is short, usually about 6 months. Patients who show no significant response have a median survival of about 2 months (Coleman et al, 1980).

High dose cytarabine (Ara-C). More recent studies have attempted to improve response rates using high dose cytarabine. Preisler et al (1984) treated 30 patients in myeloid blast crisis with high dose cytarabine alone or in combination with other agents. Autografting with cryopreserved peripheral blood stem cells was used in addition in 17 of these patients. Marrow aplasia was induced in over 90% of cases. Of the 13 patients receiving chemotherapy alone 7 patients died before regeneration of their marrow and drug resistance was present in the remaining 6. None of these patients achieved a second chronic phase. For the 17 who had the same chemotherapy plus autologous peripheral blood cells, 5 entered a second chronic phase, 5 had selective drug resistance after initial hypoplasia with early re-emergence of blasts and 6 died shortly after induction or with failure to engraft. Remarkably 1 patient failed to engraft but recovered cytogenetically normal haemopoiesis after a two month period of aplasia. He had

1046

N. C. ALLAN AND P. C. A. SHEPHERD

survived for over a year at the time of the report. Of the 5 patients restored to chronic phase, 3 maintained this for over a year. The conclusion was that autotransplantation and consolidation chemotherapy were necessary to reestablish and maintain the chronic phase following such aggressive chemotherapy. Iacoboni et al (1986) treated 21 previously treated patients with high dose cytarabine alone and again showed that marrow aplasia occurred in the majority of patients. However haematological remission was only achieved in 5 out of 17 and its median duration was only 3 months despite consolidation therapy. Median survival was 6 months. Response rates were similar for both myeloid and lymphoblastic transformations.

Mitoxantrone and etoposide. Mitoxantrone is helpful in relapsed or refractory leukaemia but has yet to be fully evaluated in the blast crisis ofCML. Kantarjian (1987) treated 5 patients with mitoxantrone and high dose cytarabine; 4 achieved complete remission. Neither the duration of complete remission nor the Ph status after therapy were reported but this drug combination may warrant further consideration. Paciucci et al (1987) treated 5 patients with Tdt positive blastic phase CML with mitoxantrone, vincristine and prednisolone and obtained complete remissions in 2. Mailliard et al (1986) treated 11 patients in blast crisis with etoposide with only very limited response.

Low dose cytarabhle. An alternative approach is to attempt to induce differentiation of the leukaemic blasts. Low dose cytarabine (L-DAC) has been used for the purpose in the treatment of blast phase CML. Schey et al (1985) reported its use combined with daunorubicin, standard dose cytarabine and 6-thioguanine (DAT) in one patient and obtained a beneficial response for 9 months but the role of the low dose cytarabine was not evaluable. Gallo et al (1985) described a case with reversion to chronic phase after 6 weeks continuous therapy; the patient relapsed in blast crisis after cessation oftherapy. Waage et al (1985) reported one patient in accelerated phase who was treated with 6mercaptopurine and low dose cytarabine and obtained a response for 5 months. A second response with low dose cytarabine alone was obtained for 2 months after which the patient was unresponsive. However, Di Raimondo et al (1985) treated 7 patients in blast crisis with L-DAC alone but obtained a response lasting 4 months in only one patient.

Mithramychl. Koller and Miller (1986) used mithramycin (plicamycin) and hydroxyurea in 9 patients with blast crisis; 6 had myeloid blast cells, 2 lymphoid blast cells and 1 undifferentiated cells. The 6 patients with myeloid blast crisis all responded but no haematological remissions (less than 5% blasts in the bone marrow) were obtained. Continuing therapy was required to maintain the response and in 4 cases, control was maintained from 5 to more than 19 months. This was achieved without significant toxicity and with improvement in clinical

TREATMENT OF CML

1047

symptoms. None of those with lymphoid blasts responded. Antimi et al (1988) have reported their experience with the same regimen in 11 patients, seven of whom were in blast crisis and four in accelerated phase. A return to chronic phase was seen in all the patients in accelerated phase but in none of those in blast crisis. This emphasises the need for further studies with these agents to define their role in accelerated or blast phase. The main problem with therapy that aims to promote maturation of the leukaemia blast cells is that it must be given continuously. Relapse occurs as soon as treatment is stopped.

Less aggressive therapy For patients for whom allogeneic bone marrow transplantation is not an option, the probability of obtaining a substantial increase in survival with intensive chemotherapy is low, even with autografiing. Results that stand comparison with those obtained by more intensive regimens can be obtained using less intensive chemotherapy (Silver and Gale, 1986). The quality of life for the patient may be substantially better. Coleman et al (1980) using a combination of hydroxyurea, 6-mercaptopurine and corticosteroids for all patients showed that 34% of 202 patients responded, 12% completely and -22% partially with median remission duration of 30 weeks. The addition of vincristine did not improve response rates. These results are as good as those obtained with more intensive regimens which induce marked myelosuppression. When the blast crisis has lymphoid features a vincristine and steroid combination may be tried first, although the median duration of remission does not appear to be longer than that achieved in cases with myeloid transformation which respond to treatment.

Symptomatic therapy Symptomatic therapy such as blood product support reduces symptoms of anaemia and thrombocytopenia and may be the only appropriate therapy in elderly patients or in those for whom chemotherapy is not suitable for other reasons. Splenectomy may be indicated to reduce splenic pooling and relieve pain from recurrent splenic infarcts and discomfort due to massive enlargement. It may also improve the patient's tolerance and response to chemotherapeutic agents and ameliorate the problems of thrombocytopenia (Sadamori and Sandberg, 1984).

INVOLVEMENT OF THE CENTRAL NERVOUS SYSTEM Involvement of the central nervous system (CNS) is generally not seen during the stable chronic phase. CNS lesions may precede the onset of blast crisis or occur at the time of transformation. They include extradural deposits, infiltration of peripheral nerves by myeloblasts and meningeal leukaemia (Rosenthal et al, 1977; Marks et al, 1978). More commonly however meningeal leukaemia develops in patients with relatively long survival after 9 the onset of blast crisis--usually those who have had a favourable response to

1048

N . C. A L L A N A N D P. C. A. S H E P H E R D

remission induction. It occurs in both lymphoblastic and myeloblastie types as assessed by morphological criteria. The reported incidence of clinical CNS leukaemia varies from 3.5-12% (Wolk et al 1974; Schwartz et al, 1975b; Rosenthai et al, 1977). The median time to onset from blast crisis is around 4 months (Schwartz et al, 1975b). Autopsy studies have however shown CNS involvement in up to 40% of cases (Wolk et al, 1974; Reske-Nielsen et al, 1974; Rosenthal et al, 1977). The incidence of clinical meningeal involvement may well be higher today than in the reported series since systemic therapy, particularly in the lymphoblastic transformation, is now more effective. Therapy for CNS leukaemia using standard intrathecal chemotherapy regimens with or without cranial irradiation often clears the leukocytosis in the cerebrospinal fluid and relieves symptoms. However, morbidity from CNS leukaemia is still considerable. Death is usually associated with haematological relapse (Schwartz et al, 1975b). The place of prophylaxis for those with successful remission induction has not been defined (Smith et al, 1983). Although effective in reducing the incidence and morbidity of CNS leukaemia in adult acute lymphoblastic leukaemia, survival is not prolonged (Omura et al, 1979). If the incidence of clinical CNS leukaemia were now considerably higher than the reported 12%, prophylactic measures may well be justified. However, the use of agents which cross the blood-brain barrier such as high dose methotrexate and high dose cytarabine in systemic chemotherapy protocols may effectively reduce the incidence of CNS leukaemia in CML. Ph NEGATIVE CML Approximately 10% of patients have a myeloproliferative syndrome consistent with CML but lack the Ph chromosome. In general their median survival is shorter than that of patients with Ph positive chronic granulocytic leukaemia (CGL) (Whang-Peng et al, 1968; Ezdinli et al, 1970; Canellos et al, 1976c; Gomez et al, 1981; Kantarjian et al, 1986). The clinical features in these cases, including cytogenetic abnormalities and characteristics associated with poor prognosis such as anaemia, severe thrombocytopenia, blood peripheral blast count and age, have been described by Kantarjian et al (1986). Recently it has been appreciated that some of these patients have a disease morphologically indistinguishable from Ph positive CML and have similar survival (see Chapter 2). The remainder constitute a heterogeneous group morphologically distinct from CML with dyshaemopoiesis and often large numbers of monocytes. The treatment of Ph negative CML in general resembles that of Ph positive CML in regard to single agent therapy, combination chemotherapy or more intensive chemotherapy. Certain problems may arise when administering chemotherapy, particularly thrombocytopenia which may necessitate stopping treatment before control of the leukocyte count is achieved. There were 61 patients with Ph negative disease in the MRC CML II trial. Platelet counts were less than 200 x 109/litre at presentation in 50% compared to 10% in Ph positive CML. Thrombocytopenia (< I00 x 109/litre) progressed in 82% on treatment and chemotherapy was stopped for this reason in 31%. Thus for Ph

TREATMENTOF CML

1049

negative patients without significant leukocytosis, organomegaly or symptoms it may be safer to withhold chemotherapy. If required, hydroxyurea is probably preferable to busuiphan for controlling the leukocyte count as it is more platelet sparing. Blast crises occur in 30% of cases; 50% of patients die without entering a blastic phase, usually as a result o f refractory disease with infections or bleeding. Myeloid blast crises are the rule and only occasional cases o f lymphoid transformation are described (Hughes et ai, 1981; Bettelheim and Knapp, 1986). Ph negative patients treated in blast crisis fare as poorly as those with Ph positive disease. CONCLUSIONS All forms o f treatment used in chronic phase C M L are palliative with the exception o f allogeneic bone marrow transplantation. No form of treatment has extended patient survival significantly but patient well-being is greatly improved by disease control. Busulphan and hydroxyurea remain the most commonly used drugs. Interferon is the first non-myelotoxic drug to cause a marked reduction in Ph positivity in some patients. Early results are promising but it is not yet clear whether a significant increase in survival will be achieved. For the acute phase of disease no substantial progress has been made towards eradication of the neoplastic clone despite a great deal of effort. 10 to 20% o f patients with myeloid transformations and up to 60% of those with lymphoid blast transformations achieve a second chronic phase but this is generally short-lived. The combination of hydroxyurea and mithramycin in accelerated disease or myeloid transformation may represent an improvement. In general the use of regimens tailored to the patient's needs that suppress counts without marrow ablation appears to be as satisfactory as more intensive therapy. The basic problem is the inherent biological resistance o f the neoplastic stem cell in C M L and the lack o f selectivity o f available chemotherapeutic agents. Significant progress will be made only when more selectively active drugs are found or drugs with different modes o f action are discovered. In the future, suppression of the B C R - A B L chimeric gene product, thought to be related to the proliferative advantage o f the neoplastic clone, by 'biological response modifiers' may improve the long-term survival of patients with CML. At present the only prospect for long-term survival involves bone marrow transplantation. REFERENCES Advani SH, Dinshaw KA, Nair CN & Ramakrishnan G (1983)Total body irradiationin chronic myeloidleukemia. International Journal o f Radiation Oncology, Biology and Physics 9: 497500. Alimena G, Morra E, Lazzarino Met al (1987)Treatment of Ph positivechronic myelogenous leukaemia (CML) with recombinantinterferonalfa-2B. Abstracts o f the Thirdlnternational Conference on 211alignantL ymphoma ( ICML ) : current status and prospects, pp 6-7, June 1013, Lugano, Switzerland.

1050

N. C. ALLAN AND P. C. A. SHEPHERD

Allan NC, Duvall E & Stockdill G (1978) Combination chemotherapy for chronic granulocytic leukaemia. Lancet ii: 523. Allen SL & Coleman M (1985) Terminal-phase chronic myelogenous leukemia: approaches to treatment. Cancer Investigation 3: 491-503. Antimi M, Poeta GD, Cianciulli P e t al (1988) Plicamycin (Ply) and hydroxyurea (Hy) in the treatment of blast phase (BP) and accelerated phase (AP) of the Ph j chronic myelogenous leukemia (CML). Proceedings of the American Society of Clinical Oncology 7: 181. Baccarani M, Corbelli G, Tura S & the Italian Co-operative Study Group on Chronic Myeloid Leukaemia (1981) Early splenectomy and polychemotherapy versus polychemotherapy alone in chronic myeloid leukemia. Leukemia Research 5: 149-157. Ballard HS & Marcus AJ (1970) Hypercalcemia in chronic myelogenous leukemia. Medical Intelligence 282:12 663-665. Beard M, Gauci C, Sikora E et al (1976) Blast crisis ofchronic myeloid leukaemia: The effect of intensive chemotherapy. Scandinavian Journal of Haematology, 16: 258-262. Bergsagel DE, Haas RH & Messner HA (1987) Interferon alfa-2b in the treatment of chronic granulocytic leukemia. Inrestigational New Drugs 5: Supplement 9-17. Bettelheim P & Knapp W (1986) Philadelphia chromosome negative chronic myelogenous leukemia with lymphoid stem cell blastic transformation. American Journal of Hematology 21: 115-116. Bettelheim P, Lutz D, Majdic O et al (1985) Cell lineage heterogeneity in blast crisis of chronic myeloid leukaemia. British Journal ofttaematology 59: 395--409. Bolin RW, Robinson WA, Sutherland J & Hamman RF (1982) Busulfan versus hydroxyurea in long term therapy of chronic myelogenous leukemia. Cancer 50: 1683-1686. Canellos GP, DeVita VT, Whang-Peng J e t al (1976a) Chemotherapy of the blastic phase of chronic genulocytic leukemia: hypodiploidy and response to therapy. Blood 47:1003-1009. Canellos GP, Schwartz JH & DeVita VT (1976b) Splenectomy in chronic phase of chronic granulocytic leukemia: effects in 32 patients. Annals oflnternal Medicine 84: 17-21. Canellos GP, Whang-Peng J & DeVita VT (1976c) Chronic granulocytic leukemia without the Philadelphia chromosome. American Journal of Clinical Pathology 65: 467-470. Cervantes F, Ribera JM, Granena A et al (1982) Tumour lysis syndrome with hypocalcaemia in accelerated chronic granulocytic leukaemia. Acta Haematologica (Basel) 68: 157-159. Coleman M, Silver R & Pajak T et al (1980) Combination chemotherapy for terminal phase chronic myelocytic leukemia. Cancer and Leukemia Group B studies. Blood 55: 29-36. Cunningham I, Gee T, Dowling M et al (1979) Results of treatment o f P h ' + chronic myelogenous leukemia with an intensive treatment regimen (L-5 protocol). Blood 53: 375-395. Di Raimondo F, Milone F, Guglielmo G e t al (1985) Treatment ofCM L blast crisis with low dose Ara-C. British Journal of Haematology 60: 773-774. Djaldetti M, Pinkhas J, De Vries A e t al (1968) Myasthenia gravis in a patient with chronic myeloid leukemia treated by busulfan. Blood 32: 336-340. Dowling MD, ttopfan S, Knapper WH et al (1974) Attempt to induce true remission in chronic myel _ogenous leukemia (CM L) Proceedingsof American Society of Clinical Oncology 15: 822. Dreazen O, Klisak 1, Rassool F, Goldman JM, Gale RP (1987) Do oncogenes determine cl~nical features in chronic myeloid leukaemia? Lancet i: 1402-1405. Dresler WFC & Stein R (I 869) Uer den hydroxyharnstoff. Justus Liebig's Annals of Chemical Pharmacolog_y 150: 242-252. Eckhardt S, Sellei C, Horvath IP & Institorisz L (1963) Effect of 1,6-dibromo-l, 6-dideoxy-dmannitol on chronic granulocytic leukemia. Cancer Chemotherapy Reports 33: 57-61. Evans JJ & Bozdech MJ (1981) Hypokalemia in nonblastic chronic myelogenous leukemia. Archires of hzternal Medicbze 141: 786-787. Ezdinli EZ, Sokal JE, Crosswhite L & Sandberg AA (1970) Philadelphia chromosome positive and negative chronic myelocytic leukemia. Annals of Internal Medichze 72: 175-182. Finney R, McDonald GA, Baikie AG & Douglas AS (I 972) Chronic granulocytic leukaemia with Ph' negative cells in bone marrow and a 10 year remission after busulphan hypoplasia. British Journal of tlaematology 23: 283-288. Gallo JH, McKinley R, Grace CS & Rozenberg MC (1985) Control of acute transformation in chronic granulocytic leukaemia with low dose cytosine arabinoside. British Journal of Haematology 61: 586-588. Galton DAG (1953) Myleran in chronic myeloid leukaemia. Results of treatment. Lancet i: 208.

TREATMENT OF CML

1051

Goldman JM & Baughan ASJ (1984) Chronic granulocytic leukemia: treatment. In Goldman JM & Preisler HD (eds) Leukemias, Butterworth hlternational Medical Ret'iews, Volume 8 pp 239-265. London: Butterworth and Co. Gomez G, Hossfeld DK & Sokal JE (1975) Removal of abnormal clone of leukaemia cells by splenectomy. British Medical Journal 2: 421-423. Gomez GA, Sokal JE, Mittelman A & Aungst CW (1976) Splenectomy for palliation of chronic myelocytic leukemia. American Journal of Medicine 61: 14-22. Gomez GA, Sokal JE & Walsh D (1981) Prognostic features at diagnosis of chronic myelocytic leukemia. ~ancer 47: 2470-2477. Greig HBW (1956) Myleran in the treatment of chronic granulocytic leukaemia. Acta Haematologica (Basel) 16: 171-180. Griffin JD, Todd RF, Ritz J e t al (1983) Differentiation patterns in the blastic phase of chronic myeloid leukemia. Blood 61: 85-91. Hauch T, Logue G, Laszlo J, Cox E & Rundles W (1978) Treatment of chronic granulocytic leukemia with melphalan. Blood51: 571-577. Haut DM, Abbott WS, Wintrobe hiM & Cartwright GE (1961) Busulfan in the treatment of chronic myelocytic leukemia. The effect of long term intermittent therapy. Blood 17: 1-19. Hughes A, McVerry BA, Walker H, Bradstock KF, Hoffbrand AV & Janossy G (1981) Heterogeneous blast cell crises in Philadelphia negative chronic granulocytic leukaemia. British Journal o f ltaematology 47: 563-569. Huguley CM Jr, Grizzle JR, Rundles RW et al (1963) Comparison of 6-mercaptopurine and busulfan in chronic granulocytic leukemia. Blood 21: 89-101. lacoboni SJ, Plunkett W, Kantarjian HM et al (1986) High-dose cytosine arabinoside: treatment and cellular pharmacology of chronic myelogenous leukemia blast crisis. Journal o f Clinical Oncology 4: 1079-1088. Italian Cooperative Study Group on Chronic Myeloid Leukemia (1984) Results of a prospective randomized trial of early splenectomy in chronic myeloid leukemia. Cancer 54: 333-338. Janossy G, Woodruff RK, Pippard MJ et al (1979) Relation of 'lymphoid' phenotype and response to chemotherapy incorporating vincristine-prednisolone in the acute phase of Ph' positive leukemia. Cancer 43: 426-434. Joyner MV, Dujardin P, Cassuto JP & Audoly P (1977) ttypercalcaemia as complication of accelerated chronic granulocytic leukaemia. British Medical Journal 2: 1060. Kantarjian HM, Keating M J, Walters RS et al (1986) Clinical and prognostic features of Philadelphia chromosome negative chronic myelogenous leukemia. Cancer 58: 2023-2030. Kantarjian HM, Keating M J, Talpaz M e t al (1987) Chronic myelogenous leukemia in blast crisis--analysis of 242 patients. American Journal o f Medichle 83: 445-454. Kaung DT, Close HP, Whittington RM & Patno ME (1971) Comparison of busulfan and cyclophosphamide in the treatment of chronic myelocytic leukemia. Cancer 27" 608-612. Kennedy BJ (1969) Hydroxyurea in chronic myelogenous leukemia. Annals ofhlternal Medichle 70: 1084-1085. Kennedy BJ (1972) Hydroxyurea therapy in chronic myelogenous leukemia. Cancer 29: 10521056. Kennedy BJ & Yarbro JW (i966) Metabolic and therapeutic effects of hydroxyurea in chronic myeloid leukemia. Journal o f the American Medical Association 195: 1038. Kennedy BJ, Smith LR & Goltz RW (1975) Skin changes secondary to hydroxyurea. Archices o f Dermatology 111: 183. Key NS, Emerson PM, Allan NC et al (1987) Oesophageal varices associated with busulphan/ thioguanine combination therapy for chronic myeloid leukaemia. Lancet ii: 1050-1052. "Kirschner RH & Esterley JR (1971) Pulmonary lesions associated with busulfan therapy of chronic myelogenous leukemia. Cancer 27:1074-1080. Koller CA & Miller DM (i 986) Preliminary observations in the therapy ofmyeloid blast phase of chronic granulocytic leukemia with plicamycin and hydroxyurea. New England Journal o f Medich~e 315: 1433-1438. Kurzrock R, Talpaz M, Kantarjian t Iet al (! 987) Therapy o fchronic myelogenous leukemia with recombinant interferon gamma. Blood 70: 943-947. Mailliard JA, Letendre L, Dalton RJ et al (1986) Phase I-I I trial ofVP- 16 in the treatment ofacute nonlymphocytic leukemia and blast crisis of chronic granulocytic leukemia. Medical and Pediatric Oncology 14: 306-309.

1052

N. C. ALLAN AND P. C. A. SHEPHERD

Marks SM, McCaffrey R, Rosenthal DS & Moloney WD (1978) Blastic transformation in chronic myelogenous leukemia: experience with 50 patients. Medical and Pediatric Oncology 4: 159-167. Mason JE, DeVita VT & Canellos GP (1974) Thrombocytosis in chronic granulocytic leukemia. Incidence and clinical significance. Blood 44: 483-487. McBride CM & Hester JP (1977) Chronic myelogenous leukemia: management of splenectomy in a high risk population. Cancer 39: 653-658. ,Medical Research Council (1968) Chronic granulocytic leukaemia: comparison of radiotherapy and busulphan therapy. British Medical Journal 1: 201-208. Medical Research Council (1983) Randomised trial of splenectomy in Ph positive chronic granulocytic leukaemia including an analysis of prognostic features. British Journal o f Haematology 54:415-430. Minot GR, Buckman TE & Isaacs R (1924) Chronic myelogenous leukemia. Age incidence, duration and benefit derived from irradiation. Journal o f the American Medical Association 82: 1489-1494. Muehleck SD, McKenna RW, Arthur DC, Parkin JL & Brunning RD (1984) Transformation of chronic myelogenous leukemia: clinical, morphologic and cytogenetic features. American Journal o f Clinical Pathology 82: 1-14. Murphy WG, Fotheringham GH, Busuttil A & Allan NC (1985) Skin lesions in chronic granulocytic leukemia: treatment of a patient with topical nitrogen mustard. Cancer 55: 2630-2633. Niederle N, Kloke O, May D et al (1987) Treatment of chronic myelogenous leukemia with recombinant interferon alpha-2B, hwestigational New Drugs 5; Supplement 19-25. Opalka B, Wandl U, Koppe J e t al (1987) Molecular and in vitro stem cell analysis on patients (PTS) with chronic myelogenous leukemia (CML) under therapy with recombinant interferon alfa (IFN alfa-2B). Proceedings o f the American Association o f Cancer Research 28: 206. Omura GA, Moffitt S, Vogler WR & Salter MM (1979) Combination chemotherapy of adult acute lymphoblastie leukemia with randomized central nervous system prophylaxis. Blood 55: 199-204. Paciucci PA, Keaveney C, Cuttner J & Holland JF (1987) Mitoxantrone, vincristine and prednisone in adults with relapsed or primarily refractory acute lymphocytic leukemia and terminal deoxynucleotidyl transferase positive blastic phase chronic myelocytic leukemia. Cancer Research 47: 5234-5237. Podoll LN & Winkler SS (1974) Busulfan lung: report of two cases and review of the literature. American Journal o f Roentgenology 120: 151-156. Preisler HD, Raza A, Higby D et al (1984) Treatment of myeloid blastic crisis of chronic myelogenous leukemia. Cancer Treatnwnt Reports 68: 1351-1355. Pusey WA (1902) Report on cases treated with roentgen rays. Journal o f the American Medical Association 38:91 I-919. Ramanan CV & Israels MCG (1969) Treatment of chronic myeloid leukaemia with dibromannitol. Lancet ii: 125-128. Ravindranathan MP, Paul VJ & Kuriakose ET (1972) Cataract after busulphan treatment. British Journal I: 218-219. Reske-Nielsen E, Petersen JH, Sogaard H & Jensen KB (1974) Leukaemia of the central nervous system. Lancet i: 211-212. Rosenthal S, Canellos GP, DeVita VT & Gralnick HR (1977) Characteristics of blast crisis in chronic granulocytic leukemia. Blood 49:705-714. Rundles RW, Grizzle J, Bell WN et al (1959) Comparison ofchlorambucil and busulfan in CML. American Journal o f Medichze 27: 424-432. Sadamori N & Sandberg AA (1984) Chromosome changes and splenectomy in Ph'-positive chronic myeloid leukemia. Cancer 54: 2456-2459. Schey SA, Greene MI, Ardeman S & Ruddell K (1985) The blast crisis of chronic granulocytic leukaemia: a new approach to treatment. British Journal ofltaematology 59: 190-192. Schiffer CA, De Bellis R, Kasdorf H, & Wiernik PH (1982) Treatment of the blast crisis ofchronic myelogenous leukemia with 5 aza-cytidine and VP-16-213. Cancer Treatnwnt Reports 66: 267-271.

TREATMENT OF CML

[ 053

Schreibman SM, Gee TS & Grabstald H (1974) Management ofpriapism in patients with chronic granulocytic leukemia. Journal of Urology ! I I: 786. Schwartz JH & Canellos GP (1975) ttydroxyurea in the management of hematologic complications of chronic granulocytic leukemia. Blood 46: I I-16. Schwartz JH, Canellos GP, Young RC & DeVita VT (1975) Meningeal leukemia in the blastic phase of chronic granulocytic leukemia. American Journal of Medicine 59: 819-828. Sharp JC, Joyner MV, Wayne AW et al (1979) Karyotypic conversion in Ph'-positive chronic myeloid leukaemia with combination chemotherapy. Lancet i: 1370-1372. Sidi Y, Douer D & Pinkhas J (1977) Sicca syndrome in a patient with toxic reaction to busulfan. Journal of the American Medical Association 238: 1951. Silver RT & Gale RP (I 986) Chronic myeloid leukemia. American Journal of Medicine 80:11371148. Silver RT, Mick R, Cooper R et al (1987a) A comparative study of dibromomannitol and busulfan in the treatment of chronic myeloid leukemia. Cancer 60: 1442-1448. Silver RT, Coleman M, Benn Pet al (1987b) Interferon (rlFN) alfa-2B has activity in treating chronic myeloid leukemia (CML) in rlFN gamma failures. Journal of Clinical Oncology 6: 149. Singer JW, Arlin ZA & Najfeld V (1980) Restoration of non-clonal hematopoiesis in chronic myelogenous leukemia (CML) following a chemotherapy-induced loss of the Ph' chromosome. Blood 56: 356-360. Smith AG, Prentice AG, Lucie NP & Dagg JH (1983) Meningeal relapse in Ph'-positive acute lymphoblastic and lymphoid blast crisis of chronic granulocytic leukemia. Is CNSprophylaxis indicated? Cancer 51: 2031-2034. Spiers ASD (1976) Annotation: the treatment of chronic granulocytic leukaemia. British Journal of Haematology 32:291-298. Spiers ASD, Baikie AG, Galton DAG et al (1975b) Chronic granulocytic leukaemia: effect of elective splenectomy on the course of disease. British Medical Journal 1:175-179. Spiers ASD, Galton DAG, Kaur J & Goldman JM (1975a) Thioguanine as primary treatment for chronic granulocytic leukaemia. Lancet i: 829-832. Spiers ASD, Goldman JM, Catovsky D et al (1977) Multiple-drug chemotherapy for acute leukemia. The TRAMPCOL regime: results in 86 patients. Cancer 40: 20-29. Stockdill G, Hartley SE & Allan NC (1980) Eosinophilic leukaemia in association with a double Philadelphia chromosome. Postgraduate Medical Journal 56: 268-270. Suri R, Goldman JM, Catovsky D et al (1980) Priapism complicating chronic granulocytic leukemia. American Journal of Hematology 9: 295-299. Talpaz M, Mavligit G, Keating M, Waiters RS & Gutterman JU (1983) Human leukocyte interferon to control thrombocytosis in chronic myelogenous leukemia. Annals oflnternal Medicine 99: 789. Talpaz M, Kantarjian HM, McCredie K, Trujillo JM, Keating MJ & Gutterman JU (1986) Hematologic remission and cytogenetic improvement induced by recombinant human interferon alphaA in chronic myelogenous leukemia. New England Journal of Medicine 314: 1065-1069. Talpaz M, Kantarjian HM, McCredie KB et al (1987) Clinical investigation of human alpha interferon in chronic myelogenous leukemia Blood 69: 1280-1288. Terjanian T, Kantarjian H, Keating Met al (1987) Clinical and prognostic features of patients with Philadelphia chromosome positive chronic myelogenous leukemia and extramedullary disease. Cancer 59: 297-300. Thomas MR, Robinson WA, Mughal TI & Glode M (1984) Tumor lysis syndrome following VP16-213 in chronic myeloid leukemia in blast crisis. American Journal of Hematology 16:185188. Tura S, Baccarani M, Gugliotta L et al (Italian Cooperative Study Group on Chronic Myeloid Leukaemia) (1975) A clinical trial of early splenectomy, hydroxyurea and cyclic arabinosyl cytosine, vincristine and prednisone in chronic myeloid leukaemia. Series Haematologica 8:4 121-142. Vallejos CS, Trujillo JMC, Bodey A e t al (1974) Blastic crisis in chronic granulocytic leukemia: experience in 39 patients. Cancer 34: 1806-1812. Waage A, Hammerstrom J & Lamvik J (1985) Chronic granulocytic leukaemia in accelerated stage treated with low dose cytosine arabinoside. British Journal of Haematology 61: 599589.

|054

N . C . ALLAN AND P. C. A. SHEPHERD

Walters RS, Kantarjian HM, Keating MJ et al (1987) Therapy of lymphoid and undifferentiated chronic myelogenous leukemia in blast crisis with continuous vincristine and adriamycin infusions plus high-dose decadron. Cancer 60: 1708-1712. Weinberger A, Pinkhas J, Sandbank U et al (1975) Endocardial fibrosis following busulfan treatment. Journal of the American Medical Association 231: 495. Whang-Peng J, Canellos GP, Carbone PP & Tjio JH (1968) Clinical implications of cytogenetie variants in chronic myelocytic leukemia (CML). Blood 32: 755-766. Winton EF, Miller D & Vogler WR (1981) Intensive chemotherapy with daunorubicin, 5azacytidine, 6-thioganine and cytarabine (DATA) for the blastic transformation ofchronie granulocytic leukemia. Cancer Treatment Reports 65: 389-392. Wolk RW, Masse SR, Conklin R & Freireich EJ (1974) The incidence of central nervous system leukemia in adults with acute leukemia. Cancer 33: 863-869. Yoffe G, Blick M, Kantarjian H et al (1987) Molecular analysis of interferon induced suppression of Philadelphia chromosome in patients with chronic myeloid leukemia. Blood 69: 961-963.