ACUTE MYELOID LEUKAEMIA--A REVIEW OF SIXTY-SEVEN CASES Una O'Call~ghan, Donald McCarthy, John Hegarty, Frank Goodwin, James J. Fennelly and Liam G. O'Connell The Departments of Haemato/ogy and Onco/ogy, St. Vincent's Hospita/, E/m Park, Dub~in 4. Summary SIXTY-SEVEN consecutive patients with acute myeloid leukaemia (AML), who presented to the unit over a 5-year period, were reviewed. Remission rates, length of remission and duration of survival were studied. Twenty-seven patients (40.3%) achieved complete remission (CR). Eleven patients (16.4%) achieved a second CR. The median survival of patients who achieved a CR was 15 months. Median survival of patients who failed to go into remission was one month. Median duration of the first remission was 7 months. Three patients are still alive and in remission. These results are discussed in relation to those reported from other centres. Introduction Although in recent years the newer chemotherapy regimens for acute lymphoblastic leukaemia in children have resulted in significantly improved survival, there has been much less progress in the treatment of AML and it remains one of the most difficult and challenging problems in medicine. In the early 1960s, prednisone and 6-mercaptopurine were used for induction of patients with AML and the results obtained were poor (Medical Research Council Working Party, 1963). Since then, more effective chemotherapeutic agents have been introduced and results have improved considerably. In 1971 we started using the combination of cystosine arabinoside, vincristine and daunorubicin for induction of patients with AML and we report here the results obtained using these agents in all patients who presen-

ted with AML over a 5-year period. The results reported in this paper are as of May 1 1978.

Patients and Methods Patients. Sixty-seven previously untreated, unselected and consecutive patients with AML occurring de n ovo, who presented to St. Vincent's Hospital from February 1972 to January 1977 (inclusive), were included in the study. Patients presenting with AML superimposed on a pre-existing myeloproliferative disorder or any other disease recognised as being pre-leukaemic were omitted from the review. Of the 67 patients, 34 were female and 33 male. The mean age at the time of presentation was 41 years (range 9 years to 80 years). The mean duration of presenting symptoms was 7 weeks. The main presenting features were weakness (60%), bleeding (51% ), dyspnoea (33%), hepatomegaly (31%), splenomegaly (24%), lymphadenopathy (15%) and infection (20%). Mean blood values at the time of presentation were: haemoglobin 8.1g/ dl; white cell count 48.0 x 109/I; and platelet count 46.0 x 10'/I. All except 3 had haemoglobin values less than 12g/dl. Morphology. The diagnosis of AML was made on the basis of Wright-Giemsa staining and cytochemical reactions of the bone marrow and peripheral blood smears. Considerable care was taken to establish an accurate diagnosis of the cell type. While reviewing the cases, the bone marrow and peripheral blood smears were re-examined and re-classified according to the proposals for the classification of AML suggested by the

173

174

IRISH JOURNAL OF MEDICAL SCIENCE

FAB Co-operative Group, which described 6 main types: myeloblastic leukaemia without maturation (M1), myeloblastic leukaemia with maturation (M2), hypergranular promyelocytic leukaemia (M3), myelo-monocytic leukaemia (M4), monocytic leukaemia (M5) and erythroleukaemia (M6) (Bennett et al, 1976). The numbers in the various subgroups are shown in Table I. The 6 slides unavailable for re-classification were all originally diagnosed as myeloblastic by the same observers and would therefore be classified as either M1 or M2 according to the new classification. TABLE I Subdivision of 61 patients with AML according to the FAB classification (see text). Type of AML

No.

MI

18

Ms

28

M~

7

M,,

4

M5

2

M6

2

Chemotherapy. The basic induction regimen was a modification of that used by Rosenthal and Moloney (1972). It consisted of a series of 5-day courses of cytosine arabinoside, vincristine and daunorubicin, interrupted by a suitable treatment-free interval to allow time for recovery of the normal haematopoietic cells. The time required for recovery varied, but the average treatment free interval was 7 days. Cytosine arabinoside 2mg/kg was given intravenously on days one to 5. Vincristine l mg/m 2 was given intravenously on day 2. Daunorubicin was initially given as a single dose of l mg/kg intravenously on day one, but this was later changed to l mg/kg on days 2 to 4, and

the vincristine was then given on the first day of the cycle. Prior to initiating chemotherapy, each patient was commenced on allopurinol 100mg orally 3 times daily in an effort to prevent complications due to hyperuricaemia. The basic induction course was repeated at intervals until remission occurred or until a total of 5 courses had been given. If remission occurred, the patient was given a further 5-day consolidation course similar to the initial induction course when the bone marrow had recovered from the induction regimen. When a patient survived the initial induction schedule but failed to go into remission, various other established regimens for induction in AML were tried. These involved the use of cyclophosphamide, vincristine, cytosine-arabinoside, prednisone, 6-mercaptopurine, methotrexate, I-asparaginase and thioguanine in various combinations. The basic maintenance therapy for patients who achieved remission was 6-mercaptopurine 1-2 mg/kg daily orally. and cyclophosphamide 100-150 mg/m ~ orally once weekly. This was commenced as soon as possible after the first consolidation course. Patients who were continued on allopurinol because of persisting hyperuricaemia had their dose of 6-mercaptopurine lowered accordingly. The regular maintenance therapy was interrupted at 3-month intervals for a further 5-day consolidation course which was similar in all respects to the initial inducation course. When possible the consolidation courses were given on an out-patient basis. Nine of the patients in the study were given maintenance immunotherapy in the form of BCG and unirradiated allogeneic blast cells, as previously reported by Fennelly et al (1976). This was commenced following various intervals of consolidation chemotherapy. When immunotherapy was commenced, maintenance chemotherapy was discontinued.

ACUTE MYELOID LEUKAEMIA

Apart from patients receiving immunotherapy, maintenance chemotherapy was continued indefinitely for patients remaining in remission, except for one patient who developed a renal carcinoma when it was felt that continuation of chemotherapy in this instance might promote rapid dissemination of the turnour. Maintenance chemotherapy was interrupted briefly for consolidation courses and occasionally where drug-induced marrow hypoplasia occurred. In the latter event, it was recommenced at a lower dose when the marrow recovered sufficiently. In patients who relapsed on maintenance therapy re-induction was attempted where possible with a regimen similar to the initial induction regimen. Some patients who relapsed died before reinduction was attempted. Supportive management. All patients required intensive supportive management during induction therapy as pancytopenia invariably occurred. Bleeding secondary to thrombocytopenia was managed with transfusions of fresh platelet concentrate and fresh blood where necessary. Platelet transfusions were given prophylactically during induction if the platelet count was less than 10.0 x 10'/I. Bleeding in patients with the particular coagulation disturbance seen in the M3 type was managed with epsilonaminocaproic acid (EACA) together with platelet transfusions as previously outlined by Keane et al (1976). Anaemia was treated with transfusions of packed cells when indicated. Prophylactic antibiotics were not used except in the case of the oral antifungal agent, nystatin, which was given during induction to limit the growth of candida in the mouth. Infections were treated vigorously when they occurred. Any patient thought to have an infection was treated with a broad spectrum antibiotic combination - - usually gentamicin together with cephalosporin--after appropriate cultures were taken. When antibiotic sensitivity became available the antibiotics were changed accordingly. If

175

infection with pseudomonas was suspected, or if the patient did not appear to be responding to the above antibiotic regimen, carbenicillin was given in addition. Patients with a history of tuberculosis were covered with antituberculous therapy throughout the illness. Patients with a respiratory infection unresponsive to broad spectrum antibiotics were also commenced on antituberculous therapy, without necessarily isolating the organism.

Definition of remission. The criteria for complete remission used in this study were those used by the Midwest Chemotherapy Study Group (Hewlett et al, 1964). Complete remission was judged to have occurred when the patient no longer had any symptoms or signs ascribable to leukaemia; the bone marrow cellularity was normal with less than 5% of apparently normal blasts and no abnormal blast cells; the peripheral blood showed a haemoglobin greater than 12g/dl not maintained by transfusion, a white cell count greater than 3.0 x 109/I with more than 1.5 x 10'/I neutrophils and a platelet count greater than 100 x 10~/I. Results

Eight patients died within the 5 days following diagnosis and did not live long enough to receive full induction course. However, no ient was excluded from analysis.

first thus one pat-

Twenty-seven patients achieved CR (40.3%). Eleven patients achieved a second CR (16.4%). The change in dosage and administration of daunorubicin had no effect on the percentage of remissions achieved. Twelve out of 30 achieved CR on the lower dose of daunorubicin (40%) while 15 (40.5%) of the 37 patients who received the higher dose achieved CR. None of the patients who survived the initial induction regimen but failed to achieve remission went into remission with the other induction schedules used.

176

IRISH JOURNAL OF MEDICAL SCIENCE TABLE II

Number and percentage of remissions achieved related to the type of AML ( N = 6 7 ) . Type of AML

No.

No. achieving CR

CR rate

M, "[ M= f

52

19

36.5%

M,

7

5

71.4%

M~

4

2

50%

M,

2

0

0%

M~

2

1

50%

TABLE III Survival related to the clinical response. Mean survival (months)

Median survival (months)

9.6

5.0

20.2

15.0

2.5

1.0

All patients ( N = 6 7 ) Patients who achieved CR (N--27) Patients who failed to achieve CR ( N = 4 0 )

Table II shows the remission rates of the various subgroups. The remission rates of M1 and M2 are not calculated separately as it was not possible to reallocate to either subgroup the 6 patients who were originally diagnosed as having myeloblastic leukaemia' but whose slides were subsequently unavailable. Remission rates of patients in the M3 subgroup were better than those in the other groups. Neither patient in the M5 group achieved remission. The mean survival of all patients in the series was 9.6 months (range 1 day to 62 months). Mean survival of patients achieving CR was 20.2 months (Table Ill) while that of patients who failed to achieve CR was 2.5 months. Median survival of patients who achieved CB was 15 months whereas that of noriresponders was one month. Survival of patients who went into remission is shown in Fig. 1 while Fig. 2 demonstrates the survival of non-responders. Three patients are still alive and in CR at 30+ months, 33+ months and 44+ months respectively following diagnosis. One patient died while in complete haematological remission from a disseminated renal carcinoma. All other pat-

1-0. 0 Continues in remission. 9 Died following relapse. 9 Died in remission.

0.90'8 ~

z O'?B

>~ 0-6-

88

u) 0.54 z 0.40 0'3-~ 0

0.2'

0.1|

I

4

6

I

!

I

i

!

I

I

i

|

I

i

I

I

I

I

I

I

i

I

I

i

8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48

DURATION OF SURVIVAL (months) Fig. 1--Survival of patients who achieved complete remission ( N = 2 7 ) .

62

ACUTE MYELOIO LEUKAEMIA

9 Died during attempted induction.

0"9

0'8

z

in the unit, using cytosine arabinoside, vincristine and daunorubicin for induction and 6-mercaptopurine and cyclophosphamide for maintenance of patients with AML, to compare these results with those recently published by other centres and to plan future management of patients with AML accordingly. it is difficult to compare the results reported from different centres for a variety of reasons. Varying degrees of patient selection occur. The age of the patients and the number of children included vary from series to series. Some reporters exclude patients who die soon after diagnosis and therefore have received little or no chemotherapy. Certain varieties of acute myeloid leukaemia are excluded from some series. All of these factors, together with varying criteria for assessing responses, can markedly influence success rates. The number of

0"6-

> = 0~ Z 0

F-

0 B. 0 t,,

177

Oi

TABLE IV Duration of remission in patients with AML.

1 2 3 MONTHS

4

5 6 7 8 9 10 FROM DIAGNOSIS

Fig. 2--Survival of patients who failed to achieve complete remission ( N = 4 0 ) .

ients who achieved remission died following relapse of the leukaemic process. Mean duration of the first remission was 12.8 months (Table IV). Median duration of the first remission was 7 months while that of the second remiss.ion was 5 months. Table V shows survival figures for the various subgroups of AML. Patients with the M3 type had the longest median survival, while patients with M2 had a considerably longer median survival than the less differentiated M1.

Discuss,ion The purpose of the present study was to analyse the results obtained

Remission

Mean duration (months)

First Second

Median duration (months)

12.8

7.0

6.3

5.0

TABLE V Survival of patients related to the type of AML (N=61). Type of AML

Median survival (months)

M,

1.0

M=

6.0

M~

18.0

M~

11.0

M.~

0.6

M6

6.0

178

IRISH JOURNAL OF MEDICAL SCIENCE

patients included in a series is also important. Six out of 9 patients (67%) treated with the same regimen in our unit prior to the present series achieved CR (Gorman et al, 1972) but we were unable to maintain the high remission rate when a larger number of patients was involved. There was no selection, to our knowledge, in the present series. All patients referred to us were accepted for treatment and all were included in the review, including those who died before chemotherapy was instituted (if the 8 patients who died before receiving one full course of chemotherapy were excluded the CR rate would rise to 46%). The series contained only 3 patients under 15. All 6 types of AML described in the FAB classification (Bennett et al, 1976) were included in the series. Nonetheless the results are comparable in many respects with the results recently published by most other centres. Rosenthal and Moloney (1972) using the same drug combination for induction achieved CR in 11 out of 23 patients. Crowther et al (1973) had a CR rate of 49% in 94 patients induced with cytosine arabinoside and daunorubicin. Boiron et al (1969) reported a CR rate of 55% in patients treated with cytosine arabinoside, daunorubicin and thioguanine. One of the largest series published involved 301 patients in ,the Fourth and Fifth MRC Therapeutic Trials (MRC Working Party, 1974) and of these 101 achieved remission (34%). In contrast Gale and Cline (1977), using an intensive regimen of thioguanine, daunorubicin and cytosine arabinoside for induction have achieved a CR rate of 7 9 % - - a rate superior to most others published. Few series refer to the number of second remissions achieved. Eleven patients (16.4%) in the present series achieved a second CR which compares favourably with the number achieved by Clarkson et al (1975).

As previously noted, the dosage and timing of daunorubicin administration was changed during the study. This was done because it was felt that daunorubicin was the most effective of the 3 induction drugs. However the altered regimen had no effect on the percentage of remissions achieved. The median survival of patients achieving CR (15 months) compares favourably with most other series. Wiernik et al (1976) reported a median survival of 13.8 months for patients achieving CR on daunorubicin alone while that for patients achieving CR on daunorubicin, cytosine arabinoside and thioguanine was 12.8 months. Median survival of patients achieving CR on the regimen used by Gale and Cline (1977) was 375 days. Clarkson et al (1975) and Spiers et al (1977) achieved a median survival of 2 years in their patients who achieved complete remission. However, it must be remembered that their maintenance schedules were very intensive by most standards. Patients failing to achieve remission in our series had a median survival of one month, which was remarkably similar to that reported by others (Wiernik et al, 1976; Clarkson et af, 1975). The median duration of the first remission (7 months) is comparable with that reported in many other series. Wiernik et al (1976) reported a median remission duration of 6.8 months and 5.6 months in patients achieving remission with daunorubicin alone and a combination of cytosine arabinoside, vincristine and daunorubicin respectively. Median remission duration in the Fourth and Fifth MRC. Therapeutic Trials (MRC Working Party, 1974) was 6 months. The more intensive maintenance regimes of Spiers et at (1977) and Clarkson et al (1975) produced median remission durations of 66 weeks and 10 months respectively. There are few published reports on the duration of second remissions. The med-

ACUTE MYELOID LEUKAEMIA

ian duration of second remission obtained in our series (5 months) was marginally better than that seen in patients treated by the L-6 protocol (Clarkson et al, 1975). Although the numbers in the various subgroups are small, analysis of the results showed certain trends. The results obtained with the M3 type were considerably better than the other subgroups in terms of remissions achieved and duration of survival. These results contrast with other published reports. Clarkson et al (1975) failed to achieve remission in any of 6 patients with this type of AML treated in his series, whereas Crowther et al (1973) had a CR rate of 33% (2 out of 6 patients) while their overall CR rate was 49%. We feel that patients with M3 have a relatively good prognosis which has hitherto been obscured by the high initial mortality due to haemorrhage and ascribe our good results to the early use of EACA in the management of the coagulopathy which is frequently encountered in this type of AML. The results obtained using immunetherapy were not analysed separately as the number of patients who received it was small and moreover they were a selected group. It should be mentioned however that the longest survivor in the series was maintained on immunotherapy from 10 months after diagnosis until her death 52 months later, except for a brief period following relapse of her leukaemia when she was reinduced with the induction regimen. Considering our results in the light of other published series, we feel that we must aim for a higher remission rate. Consequently, we have discontinued the induction regime used on patients in this series and are currently using .the combination which gave Gale and Cline (1977) a CR rate of 79%. If their results are reproducible the outlook for patients with AML may not be so grim in the future.

179

In comparing the results of different maintenance schedules one has to consider not only the duration of remission and survival achieved but also the quality of life during the patient's remission. Our maintenance regimen was a simple one which caused little inconvenience to the patient and required few hospital visits. More intensive maintenance schedules, while prolonging remissions and survival among responders by a matter of months are much more inconvenient for the patient and carry their own risks. Overall we feel that the most important consideration at this stage is the quality of life enjoyed by the patient, and thus we plan to continue our present maintenance regimen until a schedule is discovered which can prolong remissions without compromising the quality of life for the patient in remission, or alternatively until we feel that the advantages offered to the patient by the more intensive maintenance regimens outweigh the inconvenience they necessarily involve.

References

Bennett, J. M., Catovsky, D., Daniel MarieTherese, Flandrin, G., Galton, D. A. G., Gralnick, H. R. and Sultan, C. 1976. Proposals for the classification of the acute leukaemias. Brit. J. Haemat. 33, 451. Boiron, M., Jacquillat, C., Weil, M., Tanzer, J., Levy, D., Sultan, C. and Bernard, Jean. 1959. Daunorubicin in the treatment of acute myelocytic leukaemia. Lancet i, 330. Clarkson, B. D., Dowling, M. D., Gee, T. S., Cunningham, Isabel B. and Burchienal, J. H. 1975. Treatment of acute leukemia in adults. Cancer 36, 775. Crowther, D., Powles, R. L., Bateman, C. J. T., Beard, M. E. J., Gauci, C. L., Wrigley, P. F. M., Malpas, J. S., Hamilton Fairley, G. and Bodley Scott, R. 1973. Management of adult myelogenous leukaemia. Brit. Med. J. i, 131. Fennelly, J. J., McBride, A. and O'Connell, L. G. 1976. Maintenance immunotherapy of acute non-lymphocytic leukaemia using unirradiated blast cells and B.C.G. Biomedicine 25, 198. Gale, R. P. and Cline, M. J. 1977. High remission induction rate in acute myeloid leukaemia. Lancet i, 497.

180

IRISH JOURNAL OF MEDICAL SCIENCE

Gorman, A., Fennelly, J. J. and O'Connell, L. G. 1972. Intensive chemotherapy of acute myelogenous leukaemia. Irish Med. J. 65, 509. Hewlett, J. S., Battle, J. D., Bishop, R. C., Fowler, W. M., Schwartz, S. O., Hagen, P. S. and Louis, J. 1964. The Midwest Chemotherapy Study Group--Phase II of Study of A--8103 (NSC25154) in acute leukaemia in adults. Cancer chemotherapy Rep. 42, 25. Keane, T. J., Gorman, A. M. and O'Connell, L. G. 1976. Epsilon-Amino-Caproic acid in the management of acute proymelocytic leukaemia. Acta Haemat. 56, 202. Medical Research Council Working Party on the evaluation of different methods of therapy in Leukaemia. 1963. Treatment of acute leukaemia in adults: Comparison of steroid therapy at high and low dosage in conjunction with 6-mercaptopurine. Brit. Med. J. i, 7. Medical Research Council Working Party on leukaemia in adults. 1974: Treatment of acute

myeloid leukaemia with daunorubicin, cytosine, arabinoside, mercaptopurine, I-asparaginase, prednisone and thioguanine; Results of treatment with five multiple-drug schedules. Brit. J. Haemat. 27, 373. Rosenthal, D. S. and Moioney, W. C. 1972. The treatment of acute granulocytic leukaemia in adults. N. Eng. J. Med. 286, 1176. Spiers, A. S. D., Goldman, J. M., Catovsky, D., Costello, C., Galton, D. A. G. and Pitcher, C. S. 1977. Prolonged remission maintenance in acute myeloid leukaemia. Brit. Med. J. ii, 544. Wiernik, P. H., Schimpff, S. C., Schiffer, C. A., Leonard Lichtenfeld, J., Aisner, J., O'Connell, M. J. and Fortner, C. 1976. Randomized clinical comparison of daunorubicin alone with a combination of daunorubicin, cytosine arabinoside, 6-thioguanine and pyrimethamine for the treatment of acute non-lymphocytic leukaemia. Cancer Treat. Rep. 60, 41.