February 1992, Volume 42, Issue 2

Original Article

ACUTE LEUKAEMIA IN CHILDREN - FRENCH-AMERICAN-BRITISH (FAB) CLASSIFICATION AND ITS RELATION TO CLINICAL FEATURES

Khalid Hassan  ( Department of Pathology, Rawalpindi Medical College, Rawalpindi. )
Khalid P. Bukhari  ( Department of Pathology, Rawalpindi Medical College, Rawalpindi. )
Asif Zafar  ( Departments of Surgery, Rawalpindi Medical College, Rawalpindi. )
M. Zafarullah Kundi Malik  ( Departments of Paediatrics, Rawalpindi Medical College, Rawalpindi. )
M.Jamil Akhtar  ( Department of Pathology, Rawalpindi Medical College, Rawalpindi. )

ABSTRACT

A retrospective study of 45 patients of acute leukaemia in children (15 years age), in relation to FAB classification, is presented. Acute lymphoblastic leukaemias (ALL)were commoner than acute myeloid leukaemias (AML), with an ALL:AML ratio of 4.63:1. Amongst ALL, Li was the most frequent type (59%), followed by L2 (35%) and L3 (5.4%), respectively. FAB types M2, M3, M4 and MS were diagnosed in 2 patients each. An analysis of clinical features in relation to FAB classification is also presented (JPMA 42: 29, 1992).

INTRODUCTION

Leukaemias are amongst the commonest cancers under fifteen years age, constituting 13-30% of all the cancers diagnosed in this age group1,2. In children, acute lymphoblasticleukaemia (ALL) is the most frequent form, accounting for approximately 80% of leukaemias. Acute myeloid leukaemia is the next commonest (17%), with chronic myeloid leukaemia (CML) and juvenile CML constituting the remainder3. Acute leukaemias are usually classified according to French- American-British (FAB) classification. ALL is classified into three types, designated L1, L2 and L34. In order to differentiate between L1 and L2, ascoringsystem based on nuclear: cytoplasmic ratio, cell size, character of nuclear membrane and number of nucleoli, has been devised5. In L3, lymphoblasts are large and contain moderate to abundant basophilic, PAS negative cytoplasm which shows distinct vacuolation. FAB Li is the most frequent and is associated with the most livourable prognosis. L2 is less common and has a relatively poor prognosis6. L3 is rare and is considered by some as a haematogenous phase of Burkitt\\\'s lym­phoma7,8. AML can be classified according to FAB system into seven subgroups, designated as Mi, M2, M3, M4, M5, M6 and M7. FAB Mi to M3 are peroxidase and sudan black positive. M4 and M5 can be diagnosed by strongly positive NSE with partial (M4) or complete (M5) inhibi­tion with sodium fluoride9. The FAB classification system appears to have a prognostic value6. Patients with Li morphology have both a higher induction rate as well as more prolonged remission duration and survival. Patients with L2 mor­phology have a poor prognosis and this feature has been reported by some to function as an independent prognostic variable whereas patients with L3 disease have the worst prognosis10. Although the FAB classifica­tion may have prognostic usefulness, no biologic basis for the morphologic differences among these cell types have been elucidated. In this study, we have performed a retrospective analysis of 45 patients of acute leukaemias in children (<15 years) with reference to FAB classification and the presenting clinico-haematological features.

PATIENTS AND METHODS

A retrospective study of 45 leukaemic children diagnosed at Rawalpindi Medical College, during 1987­i990, was undertaken. A detailed account of clinical features was recorded in a proforma with special emphasis on fever, pallor, bleeding from various sites, lymphadenopathy, hepatomegaly and splenomegaly. Laboratory investigations included Hb estimation by cyanmethaemo- globin method, white cell and platelet leucocyte count after staining the smear by May-Grun­wald-Giemsa Stain. Bone marrow aspiration was performed at posterior superior iliac spine in most of the patients; however, in children less than twoyears age, marrowwas aspirated from tibia, below the tibial tuberosity. The smears were stained by May-Grunwald-Giemsa stain, periodic acid schiff stain (Sigma) and myelo-peroxidase stain (Sigma). When PAS and peroxidase were negative, a smear was also stained with naphthyl acetate esterase (NAE) stain. For diagnosis and classification of acute leukaemias, FAB classification was followed. In case of ALL, scoring was performed on the basis of nuclear: cytoplasmic ratio, cell size, outline of nuclear membrane and prominence and number of nucleoli. A score of 0 to + 2 led to the diagnosis of Li whereas a score of-i to -4 established L2. L3 was diagnosed when blast cells were large and contained moderate to abundant vacuolated PAS negative cytoplasm.

RESULTS

In a total of 160 patients of various types of leukaemia diagnosed at pathology department of Rawal­pindi Medical College, from 1987 to 1990, 47 (29.4%) were children (<15 years of age). Among these children 45 manifested acute leukaemias, whereas 2 had juvenile CML The breakup of patients of acute leukaemias is given in Table I, which also shows sex distribution of various subtypes. Acute lymphoblastic leukaemias were more com­mon (82%) as compared to acute myeloid leukaemias (i8%). Amongst ML, Li was the commonest subtype (59%) followed by L2 (35%) and L3 (5.4%) respectively. FAB subtypes M2, M3, M4 and MS were diagnosed in 2 patients each.
Age distribution
In ALL-Li, 4i% of patients were between i and 4 years of age and 50% between 6 and 14 years. In L2 patients, 30% were between i and 3 years and 70% between 5 and i4 years. The age representation of the remaining FAB subtypes is given in Table I.


Sex distribution (Table II)


L1 was commoner in females (M:F ratio 1:1.44), whereas L2 was more common in males (M:F ratio 1.6:1).

Clinical features (Table III) Fever and pallor were invariably observed as presenting features in all the types of leukaemia. Bleeding manifestations, especially bleeding from gums and ecchymoses, were more frequent in Li as compared to L2 subtype. Both the patients of M3 manifested severe forms of bleeding; one of them died of cerebral haemorrhage before the treatment could be started.
Laboratory Investigations
Hb concentration: Mean values of fib level were 5.6 G/dl(+1.43 S.D.) in Li, 5.5 G/dl (±132 S.D.) in L2 and 6.0 G/dl (± 1.42 S.D) in AML patients. White cell count: Amongst the 22 Li cases, 9 showed white cell count below4.Ox 109/1, 5 between 4.1 x 109/1 and 11.0x 109/1 and8 between i2.5x 109/1 and 25.5 x 10~/1. in a total of 13 L2 cases, 4 had a white cell count below 4.0 x 109/1, 6 between 4.1 and 8.0 x i09/I and3 between 11.5x109/1 and 18.8x109/1. In AML (8 patients), only one showed a TLC less than 4.0x 109/1, two between 4.1x 109/1 and 8.Ox 109/1 and the remaining 5 between 12.0x 109/1 and 40x 109/1. Platelet count (Table IV): Variable degree of throm­bocytopenia was invariably present. Itwas more severe in AML patients. The degree of thrombocytopenia corre­lated with bleeding manifestations.

DISCUSSION

FAB classification of acute leukaemias is now widely accepted. In children, acute lymphoblastic leukaemia (80%) is commoner than acute myeloid leukaemia3. In ALL patients, Li is the commonest, followed by L2 and L3, respectively. Zafar reported a series of 80 patients of ALL in children in Karachi; 55% of them had Li and 40% L2 subtype11. Alvi et al a study of 36 children having ALL, have observed Li in 92% and L2 in 8%9. In both these series L3 was not observed. In other childhood series, the incidence of L1 has ranged between 71% and 85%, whereas L2 was from 14% to 25%12-14. In the present series, ALL: AML ratio was 4.63:1. In ALL, Li was the commonest subtype (59.5%), followed by L2 (35.i%) and L3 (5.4%) respectively. These results are in agreement with those presented by Zafar11. L1 was more common in females, whereas L2 was commoner in males. In a total of 8 patients of AML, two had M2, two M3, two M4 and two MS. Both the patients of MS were under 1 year of age. Whereas FAB classification system itself has prog­nostic implications6, there are certain additional features which have been shown to act as prognostic indicators in various types of acute leukaemias. The age at diagnosis and initial leucocyte count have been the two most reliable indicators of prognosis, both for remission rate and survival14. The relationship between the initial count and prognosis is linear; the patients with a high count have a worse prognosis14,15. An age less than two years and more than ten years, at diagnosis, is an indicator of a relatively poor prognosis10. In Miller’s series, good prognosis or low risk patients were between 3 and 7 years of age14. In the present series, 24% of Li and L2 patients were below three years, 30% between 3 and 7 years and the remaining 46% more than 10 years of age. Therefore, only 30% of our patients corresponded to the good prognosis group of Miller’s series. Further, in our patients, the initial leucocyte count was less than 10.0 x 109/1 in 64% of L1 and 45% of L2 patients. In this series, patients of ALL invariably presented with fever, pallor and lymphadenopathy. Bleeding manifestations were more common in AML than in ALL. Amongst the ALL patients, bleeding from gums and ecchymoses were more frequent in Li as compared to L2. Splenomegaly was equally frequent in Li (90%) and L2 (91%) subtypes, whereas hepatomegaly was slightly more frequent in L2 (75%) than in Li (68%). In AML, fever, pallor and splenomegaly were invariable; hepatomegaly was observed in 75% and lymphadenopathy in 50% of patients. Bleeding manifes­tations were slightly more frequent as compared to ALL patients. Both the patients of AML-M3 showed severe bleeding manifestations; one of them presented with intracerebral haemorrhage and died before the treat­ment could be started.

REFERENCES

1. Silverberge, B. and Lubera, 3. cancerStatistica 1986. cancer, 1986; 36:9-25.
2. Hanif, SM. and Ashraf, M. Childhood malignant diaeaaea at Lahore. Pakistan Paediatr. 3., 1980; 4:170-178.
3. Neglia, if’, and Robison, LL. Epidemiology of the childhood acute leukemias. Pediatr. Clin. North Am., 1988; 35:675-692.
4. Bennett, J.M., Catovsky, D., Daniel, M.T., Flandrin, G.,Galton, D.A., Gralnick, HR., Sultan, C. Proposala for the clasaification of acute leukemias. Br. J. Haematol., 1976; 33:451.
5. Bennett, J.M., Catovsky, D., Daniel, MT., Flandrin, C., Galton, D.A., Gralnick, H.R., Sultan, C. and (F.A.B.). The morphological classification of acute lymphoblaatic leukemias; concordance among obaervera and clinical correlation. Br, .1. Haematol., 1981; 47:553.
6. Miller, D.R., Krailo, M., Bleyer, W.A. et al Prognostic implication of blast cell morphology in childhood acute lymphoblastic leukemia: a report from the children cancer group. Cancer Treat. Rep., 1985; 69:1211-1221.
7. Aur, A., Simone, 3., Huster, 0. et al. Multiple combination therapy for childhood lymphoblastic leukemia (ALL). Blood, 1978; 52:238.
8. Riccardi, It, Vigersky, LA., Barnea, S., Bleyer, W.A. and Poplack, 0. Methotrexate levels in the interstitial apace in the aeminiferoua tubule of rat teatia. cancer Res., 1982; 42:1617. 1619.
9. Alvi, E.A., Saleem, M., Ahmad, P. et al. A atudy of 100 cases of acute leukemia in northern Pakiatan, with reference to French- American-British (FAR) classification. PJP., 1990; 1(2):87-92.
10. Poplack, D.C. and Reaman, 0. Acute lymphoblastic leukemia in childhood. Pediatr. Gin. North Am., 1988; 35:903-931
11. Zafar, MN. FAR classification ofacute lymphoblastic leukemia (ALL) and itarelevance to ALL in Karachi children. JPMA, 1985;35:233-236.
12. Hann, LM., Evans, D.I., Palmer, M.D., Morris-Jones, P.J> and Haworth, C. The prognostic significance of morphological features in childhood acute lymphoblastic leukemia. Clin. Lab. Haematol., 1979; 1:215-216.
13. Viana, M.B., Maurer, H.S. and Ferene, C. Sub-classification of acute lymphoblastic leukemia in children; analysis of the reproducibility of morphological criteria and prognostic implications. Br.J. Haematol., 1980; 44:383.
14. Miller, DR., Leikin, S., Albo, V., Sather, H. and Hammond, D. Prognostic importance of morphology (FAB classification) in childhood acute lymphoblsstic leukemia (ALL). Br. 3. HaematoL, 1981; 48:199-206.
15. Robison, L, Sathan, H., Coccia, P. et sI. Assessment of the inter-relationship of prognostic factors in childhood acutelymphoblastic leukemia.Am.J. Pediatr. Haematol. Oncol., 1980; 2(1):3-5.

Journal of the Pakistan Medical Association has agreed to receive and publish manuscripts in accordance with the principles of the following committees: