Leukemia Research Vo|. 14, No. 3, pp. 255-262, 1990. Printed in Great Britain.

0145-2126fl)0 $3.00 + .00 Pergamon Press pie

PROGNOSTIC FACTORS OF MYELODYSPLASTIC S Y N D R O M E S - A SIMPLIFIED 3-D SCORING SYSTEM* JEAN-ETIENNE GOASGUEN,t RICHARD GARAND,:~MARIE BIZET,§ JEAN=LOUISBREMOND,H JACQUES GARDAIS,¶ MARIE PIERRE CALLAT,§FRAN~OISEACCARD~ and JACQUES CHAPERONt Members of the "Groupe Fran~ais de Morphologie H6matologique Professeur Georges Flandrin", from the French University and Hospital Centers of t Rennes; ¢ Nantes; § Rouen; IITours; ¶ Angers

(Received 6 September 1989. Accepted 30 September 1989) Abstract--From 1 January 1982 to 31 December 1986 in five haematological centers of the west of France (Rennes, Rouen, Nantes, Tours and Angers), we have collected 503 cases of myelodysplastic syndrome (MDS). These cases were classified by FAB recommendation as followed: 85 refractory anemia with ring sideroblasts (RARS); 273 refractory anemia in which 86 were without blasts (RA), 153 were with excess of blasts (RAEB) and 34 were with excess of blasts and in transformation (RAEB-t); 111 chronic myelomonocytic leukaemia (CMML); and 34 cases with borderline features. The point date for statistical study was 31 December 1988, and the scoring method of Bournemouth was applied to compare with our findings (62% resulted in death, and 18% in leukemic transformation). It was demonstrated that haemoglobin, platelets, and bone marrow-blasts are the best factors to predict survival or leukaemic transformation (LT). But peripheral neutrophils don't affect the survival time excepted when lower than 500 ~tl (13 months vs 19.6 months). A scoring system based on haemoglobin (Hb), platelets (PI), and bone marrow blasts (BMB) may be represented in a three-dimensional space and is a good tool to know the own value of each parameter. This 3-D system shows that BMB and PI are the most important factors and are correlated with survival, per cent of death, and LT (p < 0.0001). The LT is observed in 18% of the whole population. RAEB and RAEB-t progress in AML2 (14.6%) or AML4 (1.4%), and CMML progress in AML2 (8.1%) or AML4 (11.7%). We observed that monocytes are not good parameters to predict the type of leukemic transformation. Furthermore, survival of RA treated with Ara-C(ld) or not treated was similar.

Key words: Myelodysplastic syndrome, refractory anaemia, sideroblastic anaemia, chronic myelomonocytic leukaemia, prognostic factors, survival.

INTRODUCTION

[6] in 1984 then 1988 [7]; the latest by Sanz [8] in 1988 established a severity coefficient which was calculated or a multiple regression analysis by Riccardi [9] in 1988. In order to have a simple and reliable prognostic system, we have recorded 503 running observations of MDS. That data-file allowed us to test each parameter's value. We only held the most significant p a r a m e t e r s in order to obtain a simple test.

SINCE the publication in 1982 by the F A B - G r o u p of the r e c o m m e n d a t i o n s for the classification of the myelodysplastic s y n d r o m e [1] and the Revisal for A M L in 1985 [2], several studies have tried to find prognostic criteria permitting a knowledge of the evolution and possible leukaemic transformation of these syndromes. These studies were different within themselves: some of t h e m were multiparametric like those of Coiffier [3] in 1987 and Garcia [4] in 1988; others applied the scoring system according to the publication of Varella and Bennett [5] in 1985 or Mufti

MATERIALS AND METHODS

Patients In five haematological centers in the west of France we recorded 526 new and consecutive cases of MDS from 1 January 1982 until 31 December 1988 inclusive. For each patient we collected information on: age; sex; clinical data (adenopathy, hepatomegaly, splenomegaly and cutaneous syndrome); biological data (peripheral blood, bone marrow aspiration and trephine biopsy); chromosome analysis

* This work was supported by a grant from "Association pour la Recherche contre le Cancer: ARC" and from "Fondation Langlois". Correspondence to: Dr Jean Goasguen, Laboratoire d'H6matologie, Hopital Sud, 16 Bd de Bulgarie, 35056 Rennes C6dex, France.

255

J.-E. GOASGUEN et al.

256

TABLE 1. DISTRIBUTION OF AGE, SEX RATIO, MEAN AGE BY DIAGNOSIS

Males

Females

Total

n

Mean Age

n

Mean Age

M:F

Mean Age

Cases

RARS

50

71

35

75

1.42

73

85

RA RAEB RAEB-t

42 88 14

70 72 72

44 64 20

69 71 66

0.95 1.37 0.70

69 72 68

86 152 34

145

71

128

70

1.13

70

273

CMML

63

72

48

76

1.31

74

111

Non classified

211

72

14

68

1.42

70

34

280

72

223

72

1.25

72

503

RA + R A E B + R A E B - t

Total Per cent fatality

65%

(when they were done); preceding features before diagnosis; evolution with or without therapy; type of therapy; and survival duration. The point date for statistical analysis was 31 December 1988 and finally we obtained all the data for 503 patients. We were not able to obtain survival duration for 23 cases. Patients researched in 1982 and 1983 were already included in a previous data about the French Registry [10[.

Cytological analysb Previous to the recording, the five biologists had two workshops for cytological evaluation. They tested the FAB classification against a serial of 20 slides prepared by one of them. Secondly during that study, a control test on ten slides was realized. That test had established a good correlation of up to 80% between the cytologists for FAB classification and cytological recognition.

Scoring system The Bournemouth scoring system [6] was applied to our cases with the same criteria: score 1 is attributed when haemoglobin is equal or lower than 10g/dl; score 1 if platelets are equal or lower than 100/1; score 1 if bone marrow blasts are equal or up to 5%. The sum of each one of the three criteria is the final score.

59%

1.10

62%

RESULTS In this s t u d y we r e c o r d e d 280 m a l e s a n d 223 f e m a l e s (503 p a t i e n t s , sex r a t i o = 1.25). T h e m e a n age was 72 y e a r s , a n d d i s t r i b u t i o n by a g e , sex a n d diagnosis is s h o w n in T a b l e 1.

Preceeding features A l l the p a t i e n t s w e r e r e c o r d e d at the first s t e p o f the d i s e a s e , w h e n d i a g n o s i s o f M D S was c a r r i e d out. A m o n g t h e m , 40 p a t i e n t s h a d p r e v i o u s l y h a d a c a n c e r , 17 h a d a l y m p h o p r o l i f e r a t i v e d i s e a s e (3 Hodgkin's lymphoma, 4 non-Hodgkin's lymphoma, 8 myeloma, 2 chronic lymphocytic leukaemia), and t h r e e o t h e r s h a d a l o c a l i z e d i r r a d i a t i o n for a Z o s t e r disease. A l l p a t i e n t s w e r e t a k e n off c h e m o t h e r a p y o r i r r a d i a t i o n m o r e t h a n 3 y e a r s ago. In 196 cases ( 3 9 % ) a p e r i p h e r a l c y t o p e n i a h a d b e e n f o u n d : 156 cases of a n a e m i a ( < 1 1 g / d l ) s e e n 24 m o n t h s ago; 24 cases of n e u t r o p e n i a (2500 P M N / ~ d ) s e e n 24 m o n t h s ago; 40 cases o f t b r o m b o p e n i a ( < 1 0 0 G / l ) s e e n 18 m o n t h s ago. T r i c y t o p e n i a o n l y e x i s t e d in 14 p a t i e n t s .

Survival duration Survival duration was obtained by the clinician of each hospital, by private consultants of each patient from the Town Hall of the town of birth.

Statistical analysis All the data were collected on a computer to exchange the complete database by floppy disk between the different haematological centers. Statistical analysis were done by the Public Health Service of Rennes Hospital. The survival curves were established by the method of Kaplan & Meier [11], and the median survival was obtained from these curves. The Log Rank Test method permitted a comparison of the groups of patients.

TABLE

2.

Age range (years) 45-60 1-70 71-80 >81

DISTRIBUTION OF AGE AND SURVIVAL (IN MONTHS) FOR 503 PATIENTS

Cases

Median survival

65 86 184 117

13 20 14 12

Establishment of a simplified 3-D scoring system in 503 cases of myelodysplastic syndrome 0

TABLE 3. DISTRIBUTION OF SURVIVAL AND DEATH WITH

09

HAEMOGLOBIN, PLATELETS, NEUTROPHILS AND BONEMARROW-MEDULLAR-BLASTS (BMB) PARAMETERS, SUMMARY OF FIGS 1, 2, 3 AND 4

Cases

Median survival

Death

H b ~ 10 Hb > 10

306 197

19 32

206 108

67% 55%

P N ~ 2500 PN > 2500

279 224

38 37

174 140

62% 62%

2.81

PI ~

100 PI > 100

181 322

12 32

139 175

77% 54%

49.02

BI~5 BI < 5

267 236

16 37

197 117

74% 50%

35.59

Total

503

19

314

62%

Parameters

chi 2

257

~'05

>i®

>,

14.9

02 0 I

~]_~ets

I

I

I0

20

I

I

< =I00

I

II

I

40 50 60 Survival in months

70

30

FIG. 2. Survival curves for 322 patients with platelets > 100.10/1 and 181 patients with platelets ~< 100.10/1. Median survival are respectively 32 and 12 months and chi 2 test is 49.02. I0OB

O6

Suroioal Sex and age were first tested for survival. N o correlation was f o u n d for sex distribution. Results for age distribution and survival are given in Table 2. N o statistical difference appears. H a e m o g l o b i n , platelets, m e d u l l a r blasts, neutrophil p o l y m o r p h o n u c l e a r ( P M N ) leucocytes were tested for survival duration. Survival curves are given in Figs 1, 2, 3 and 4. M e d i a n survival, per cent fatality and leukaemic t r a n s f o r m a t i o n rate are s u m m a r i z e d in Table 3. M e d i a n survival for P M N was tested for 500, 1000, 1500, 2000, 2500, 3000, 5000 and 8000 cells/ixl. N o difference was n o t e d for each one of these value except w h e n u n d e r 500/~d. M e d i a n survival is 19.6 m o n t h s w h e n neutrophils are up to 500/Ixl and only

~

0

4-

El< 5S/I

o3

I

o 2 0

~_BMB>=5

I

0

20

I

50

I

I

I

I

40

,50

60

70

S u r v i w l in months

FIG. 3. Survival curves for 236 patients with bone marrow blasts < 5% and 267 patients with bone marrow blasts >/5%. Median survival are respectively 37 and 16 months and chi 2 test is 35.59.

~0

~o ~o xo -60 .~o

I

5

O8

~'05 -~ ,oo4

HD>IO I

~o3 ~

=

1

I

IO

I

t5

I

L

I

I

I

I

20 25 30 55 40 45 Survival in months

I

50

I

55

FIG. 4. Survival curves of 503 MDS with the number of neutrophils in the peripheral blood delimited by 3000, 3000 to 8000, up to 8000/~tl. Median survival delimited by 2500 (as in Ref.) is respectively 38 or 37 months (not significant).

0 06

0

L

I0

0

tO_

0 2

Yo

L

I--

I__

0

I

0

I

I

to

2o

I

i

L

i

30 40 50 60 Survival in month

I 70

FIG. 1. Survival curves for 197 patients with haemoglobin > 10 g/dl and 306 patients with haemoglobin ~ 10 g/dl-median survival are respectively 32 and 19 months, and chi 2 is 14.9.

13.1 m o n t h s ber of P M N point in the Absolute for survival

w h e n lower t h a n 500/~tl. W h e n the n u m was indexed and listed, no b r e a k - d o w n curve a p p e a r e d . n u m b e r s of blast cells in PB was tested but no correlation was found.

258

J.-E. GOASGUEN et al. Thrombocytes

>,oo

/

>1oo ~ 1 1 i

.

I,'i"dl~

•

,,

5o/~II~ r

~

BLosts~.~

~

Hemoglobin

.~ogto~,.,

!i!m!l ~

-

Btasts

~,

Thrombocytes

21 21

< I00

II,

11

I

>1o