American Journal of Hematology 41 :1-4 (1992)
Myelodysplastic Syndromes and Malignant Solid Tumors: Analysis of 21 Cases Jordi Sans-Sabrafen, Josep Buxo-Costa, Soledad Woessner, Lourdes Florensa, Carles Besses, Nuria Malats, and Miquel Porta Unit of Hematology and Oncology, Hospital de la Creu Roja, Universitat de Barcelona (J.S.-S.,J.B.-C., S.W., L.F.), Unit of Hematology and Oncology, Hospital Central La Alianca (C.B.), and lnstitut Municipal d'lnvestigacio Medica, Universitat Autonoma de Barcelona (N.M., M.P.), Barcelona, Spain.
We studied the association between myelodysplastic syndromes (MDS) and malignancies in a cohort of 155 patients with MDS, 21 of whom presented malignant solid tumors. Myelodysplasia was present after the diagnosis of cancer in eight patients (interval between the diagnosis of both conditions 18 months, median survival 49.5 months), simultaneously with diagnosis in 11 (median survival 8 months), and before malignancy in two patients (interval between the diagnosis of both conditions 47 and 7 months). One patient was given chemotherapy for lung cancer, and three patients received radiotherapy for adenocarcinomaof the kidney and cancer of the prostate. At the time of diagnosis of MDS, nine patients already presented metastatic spread. Fourteen patients died, ten as a result of tumor-related complications and four because of transformation to acute nonlymphocytic leukemia. The analysis of the incidence of malignancy in patients with MDS was statistically significant for males, and the relative risk was significant in both sexes. The results of this study show that MDS patients present a higher incidence of malignant tumors than the general population, that MDS may be of real paraneoplasticsignificance, and that the occurrence of MDS in cancer patients may be considered to be related to the malignancy rather than an independent phenomenon. o 1992 Wiley-Liss, Inc. Key words: FAB classification, second malignancies, paraneoplastic syndrome
INTRODUCTION
MATERIALS AND METHODS
Malignant tumors as second neoplasms are seen with increased frequency in patients with chronic lymphocytic leukemia (CLL) and in those treated for Hodgkin's disease [ 1-31. A relationship between myelodysplastic syndromes (MDS) and malignant solid tumors has also been documented, but the apparently increased incidence of malignant tumors diagnosed subsequent to the hematological disease is not easily explained [4-131. We studied the association between MDS and malignant solid tumors in a cohort of 155 patients with myelodysplasia, 21 of whom had carcinomas. The hypotheses established were as follows: 1) patients suffering MDS have a higher incidence of malignant tumors than the general population and 2) MDS could be a paraneoplastic syndrome and could, therefore, be present before, simultaneously with, or after the diagnosis of the malignant tumor.
Myelodysplastic syndromes were diagnosed in 155 patients attending our Department of Clinical Hematology and Oncology between 1975 and 1988. Follow-up data were obtained in 149 (96%) cases. Classification of the MDS was made according to the FAB criteria [ 141, which include refractory anemia (RA), refractory anemia with ring sideroblasts (RAS), refractory anemia with excess of blasts (RAEB), chronic myelomonocytic leukemia
0 1992 Wiley-Liss, Inc.
Received for publication April 5 , 1991; accepted January 30, 1992. J. Sans-Sabrafen, J. Buxo-Costa, S . Woessner, and L. Florensa are now at the Unit of Hematology and Oncology, Hospital Central La Al ianc;a, Barcelona, Spain. Address reprint requests to: Dr. Jordi Sans-Sabrafen, Bailtn 190, 08037 Barcelona, Spain.
2
Sans-Sabrafen et al.
(CMML), and RAEB “in transformation” (RAEB-t). Histologic features of tumors were assessed after biopsy, except in one case with disseminated osteolytic metastases when the primary tumor was unknown. The incidence of malignant solid tumors observed in the cohort of patients with MDS was compared with that expected in the general population. The population of the province of Tarragona was selected to represent the latter, since Tarragona is a geographic area located close to our cohort catchment area and data are available on the incidence of cancer from the Tarragona Tumor Registry for the years between 1980 and 1985 [IS]. The expected incidence of malignant tumors was obtained by multiplying the incidence rate of cancer in the tumor registry (grouped by age and sex) by the number of person-years of the MDS cohort. The eight patients (seven men and one woman) in whom MDS was diagnosed after the malignant solid tumor were excluded from this analysis. The x2 test with 1 degree of freedom was applied to the O/E ratio, where 0 is the number of malignancies observed and E the number of expected malignancies under the null hypothesis. The relative risk of patients suffering MDS developing solid tumors was calculated, taking the population of Tarragona as the control group [ 151 for age groups over age 20 years (1981 census). No subject was excluded from this analysis. The comparison of two qualitative variables was carried out using the x2 test. With qualitative variables of two or more categories, Student’s t test and analysis of variance were used for the comparison of means.
TABLE 1. Clinical Data From 155 Patients With Myelodysplastic Syndromes
Data
RA
RAS RAEB
CMML RAEB-t
Patient (n)
19
22
17
14
50 50 70.0
58.8 51.8 41.2 48.2 71.0 66.1
57.9 42.1 76.3
85.7 14.3 68.9
0.0033 0.0356
11 52.4
6 28.6
2 9.5
0.0368
83
P
Sex
Males (%) Females (%) Age, mean (years) Second malignancy No. Percent
2 9.5
cinoma and one patient with renal tumor received local X-ray therapy. Myelodysplasia was present before malignancy in two patients, simultaneously with diagnosis in 1 1 patients, and after diagnosis of the malignant tumor in eight patients. A nonhomogeneous distribution of tumors in the subgroups of MDS was observed (RAS 2, RAEB 11, CMML 6, and RAB-t 2); n o solid tumor was found among patients with RA ( P = 0.036) (Table I). In the 1 1 patients in whom cancers were diagnosed at the same time as MDS, median survival was 8 months. In the eight patients in whom MDS developed after the diagnosis of malignancy, the median period of time between the diagnosis of both conditions was 18 months, and the median survival was 49.5 months after the initial detection of the malignant tumor. Only in two patients with RAS did the second tumor appear after the diagnosis of MDS, after intervals of 47 and 7 months. At the time of diagnosis of MDS, nine (47.4%) paRESULTS tients already presented metastatic spread (seven of l l The 155 patients with MDS consisted of 87 men and 68 patients in whom malignancy was diagnosed at the same women who ranged in age from 22 to 96 years (mean ? time as MDS and two of eight patients in whom MDS SD 68.7 k 12.9 years). There were 19 (12.3%) patients developed after the diagnosis of the neoplasms). In one of with RA, 22 (14.2%) patients with RAS, 83 (53.5%) these two patients, the diagnosis of MDS coincided with patients with RAEB, 17 ( 1 1 %) patients with CMML, and the apparent distant spread of the tumor, while the other 14 (9%) patients with RAEB-t. With the exception of was the only case to have previously been treated with RA, myelodysplasias were more frequently diagnosed in chemotherapy. males ( P = 0.0033). The higher predominance in males Of the 21 patients with malignant solid tumors, seven was particularly evident in patients with RAEB-t. Mean (33.3%) are still alive. Of them, six are apparently in a ages of patients with different subgroups of myelodyspla- clinical condition of complete remission. The remaining sia were also statistically significant ( P = 0.035) 14 patients have died, ten as a result of tumor-related (Table I). complications and four because of transformation to Malignant solid tumors developed in 2 1 ( 13.5%) pa- acute nonlymphocytic leukemia (ANLL). tients, 18 men and three women (Table 11). Cancer of the The number of expected cases of malignancy in the prostate occurred in six patients, lung cancer in five, cohort of 155 patients with MDS was 2.963 for males and carcinoma of the stomach in three, cancer of the large 1.30for females. The O/E ratio was 11/2.963 = 3.7 (X2 bowel in three, adenocarcinoma of the kidney in two, and = 21799, P < 0.01) for males and 2/1.3 = 1.54 (X2 = urothelial tumor of the bladder in one. In the remaining 0.377, P > 0.05) for females. Given that both prostate patient with disseminated osteolytic metastases, the pri- and lung cancer are common in males in the general mary tumor was unknown. None of the patients had two population, the O/E ratio for these tumor sites was also cancers identified. One patient with lung cancer was calculated. The O/E ratio for lung cancer was 3/0.4 = 7.5 given chemotherapy, and two patients with prostatic car- (X2 = 16.9, P < 0.01), and for prostate cancer the O/E
Myelodysplastic Syndromes and Maligancies
3
TABLE II. Clinical Characteristicsand Outcome in 21 Patients With MyelodysplasticSyndromes (MDS) and Solid Tumors
Age (years)
Sex
MDS
Tumor location
70 77 72
M M M
RAEB RAEB RAEB
Stomach Stomach Lung
81 61 80 14 67
M M
RAEB RAEB RAEB RAEB RAEB-t
Kidney Colon Kidney Stomach Prostate
F M M
Histopathology
Stage
Adenocarcinoma Adenocarcinoma Squamous cell carcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma
MI MI M1 T3NxMx Dukes A MI MI MI
72 80
RAEB-t CMML
Prostate Unknown
Adenocarcinoma
F
II MI
76
M
CMML
Lung
T3NxMx
61" 73" 83a 80a
M M M M
RAEB RAEB RAEB RAEB
U. bladder Prostate Prostate Prostate
Squamous cell carcinoma Urothelial cancer Adenocarcinoma Adenocarcinoma Adenocarcinoma
79" 75 " 12a
M
F M
CMML CMML CMML
Colon Colon Prostate
Adenocarcinoma Adenocarcinoma Adenocarcinoma
Dukes C Dukes A I11
75 a
M
CMML
Lung
M1
67c
M
RAS
Lung
Squamous cell carcinoma Adenocarcinoma
75c
M
RAS
Lung
Anaolastic
M
Y
TI I1
II III
Treatment None None None Radiation Surgery Progesterone None LH-RH agonists antiandrogens Surgery None None Surgery Radiation Surgery Orchiectom y antiandrogens Surgery Surgery idemkadiation
Survival (months)
6 8 6
Outcome Died Died Died
6 8 29 13 29
Died Remission Died Died Metastases
12 8
Remission Died, acute leukemia Died
8 71 105 28 28
Remission, 12b Remission, 24b Remission, 12b Remission, 12b
43 56 91
Died, 36b Remission, 36b Died, acute leukemia Died, S b
Chemotherapy
33
T2NoMo
Surgery
29
T3NxMx
None
3
Died, acute leukemia Died. 4Ib
aEight patients in whom solid tumors were diagnosed before myelodysplasia. bInterval (months) between the diagnosis of both conditions. CTwopatients in whom myelodysplasia was diagnosed before solid tumors.
was 6/0.3 = 20.0 (X2 = 108.3, P < 0.01). The relative risk of patients suffering MDS developing malignant solid tumors was 56.13 (confidence interval 95%, 37.68 and 83.63). DISCUSSION The association of malignant solid tumors as distinct from acute leukemia was first reported by Weisdorf et al. [4] in 17% of their patients with MDS. This association has been also recognized by others [S-lo], but whether the incidence is actually increased remains to be confirmed. Although it has been stated that the coexistence of CMML and malignant tumors is not purely fortuitous [9,12], the fact that this question has generally been omitted in important studies on MDS is striking [16-19]. Copplestone et al. [20] have, moreover, pointed to the high incidence of lymphoproliferative disorders in 10.5% of a series of 20 patients with MDS, which was also observed in 12 (7.5%) of the 155 patients included in our study. Juneja et al. [ 161 reported the occurrence of myeloproliferative disorders in 10.2% of 118 patients with MDS, which we observed in six (4%) of our patients.
In our study, the majority of cancers occurred in men (85.7%). When the incidence of solid tumors in patients with MDS was compared with that in the general population, it was found to be significantly higher in men than in women. Some authors have analyzed the association of SMD [ l l ] , CLL [ l ] , and CMML [12] with malignant tumors using the subject-years method, with controversial results. We observed that most cancers occurred simultaneously with MDS; in fact, eight patients developed MDS after the diagnosis of the malignant tumor. Hence, a link between both conditions was hypothesized. The present study is not strictly a cohort study [21-231 but a follow-up study of a group of patients without an intrinsic control group. Therefore, it was decided to use the methods previously described with the following assumptions: 1) that the total population of the province of Tarragona over the age of 20 years remained constant during the study period; 2) that the incidence rate of cancer remained also relatively stable during the same period of time; 3) that a surveillance bias may have occurred, since patients with MDS were subjected to closer supervision than the general population; and 4) that no case of MDS would be found in the population of Tarra-
4
Sans-Sabrafen et al.
gona. The presence of a surveillance bias would imply on our part an overestimation of the relative risk, that is, of the strength of the association between MDS and malignant solid tumor. However, such strength makes it unlikely that bias, if it did actually exist, would completely account for the observed association. With regard to the assumption that no case of MDS was present in the population of Tarragona, should there be an error in classification, this would produce an increase in the relative risk; that is to say, in our study, we would be underestimating the strength of the association. From a clinical point of view, it is interesting to note that MDS and cancer occurred simultaneously in 11 patients (RAEB 7, RAEB-t 2, CMML 2). It is also worthy of note that, in seven of the 11 patients, metastases were present at the time of diagnosis of MDS. In one patient, symptoms of MDS were noticed coincident with systemic spread of a localized tumor. In all seven patients with MDS and metastases, cancers were directly responsible for death. In three of the four patients who died due to ANLL, progression of the malignancy was not observed, which would suggest that MDS evolved independently of tumor development. Survival of patients, however, was closely related to tumor stage. Patients in whom tumors were localized and in complete remission were those who survived longest, even though three ANLL patients were among these. Only in two cases of ARS did myelodysplasia precede the development of second malignancy. The coincidence of ARS and malignant tumor has been known for many years, but it is not clear whether it is of real significance ~41. In conclusion, the results of this study suggest that MDS may be of real paraneoplastic significance, making it necessary to rule out the presence of a concurrent malignancy. On the other hand, the occurrence of MDS in cancer patients may be considered to be related to the malignancy rather than as an independent phenomenon.
ACKNOWLEDGMENT
The authors are grateful to Marta Pulido, MD, for editorial assistance and copyediting.
REFERENCES 1 . Lopez-Guillermo A, Reverter JC, Cervantes F, Viiiolas N, Rovira M,
Urbano-Ispizue A, Montserrat E, Rozman C: Neoplasias asociadas a la leucemia linfiica cronica. lncidencia y caracteristicas en una serie de 232 pacientes. Med Clin (Barc) 93:681-683, 1989. 2. Davis JC, Weiss NS, Armstrong BK: Second cancers in patients with chronic lymphocytic leukemia. JNCI 7 8 9 - 9 4 , 1987. 3 . Bookman MA, Longo DL, Young RC: Late complications of curative treatment in Hodgkin’s disease. JAMA 260:680-683, 1988.
4. Weisdorf DJ, Oken MM, Johnson GJ, Rydell RE: Chronic myelodysplastic syndrome: Short survival with or without evolution to acute leukaemia. Br J Haematol55:691-700, 1983. 5 . Raz I, Shinar E, Polliak A: Pancytopenia with hypercellular bone marrow-A possible paraneoplastic syndrome in carcinoma of the lung: A report of three cases. Am J Hematol 16:403408, 1984. 6. Sans-Sabrafen J, Buxo J, Woessner S, Florensa L, Lafuente R, PardoPeret P, Rodriguez-Ferrera JC, Pedro C: Anemia refractaria adquirida y neoplasia. Estudio de 1 I casos. Med Clin (Barc) 83:489491, 1984. 7. Hazdenar R: Pancytopenia with hypercellular bone marrow as a possible paraneoplastic syndrome. Am J Hematol 19:205-206, 1985. 8 . Foucar K, Langdon RM, Armitage JO, Olson DB, Carroll TJ: Myelodysplastic syndromes. A clinical and pathologic analysis of 109 cases. Cancer 56:553-561, 1985. 9. Sans-Sabrafen J, Woessner S , Besses C, Lafuente R, Florensa L, Buxo J: Association of chronic myelomonocytic leukemia and carcinoma: A possible paraneoplastic myelodysplasia. Am J Hematol 22: 109-1 10, 1986. 10. Perdiguer A, Girdldo MP, Cortes MT, Rubio Felix D, Giralt M: Sindromes mielodisplasicos asociados a neoplasia distinta de la leucosis aguda: Analisis de veintidos casos. Sangre 32: 194-201, 1987. I I . Stark AN, Thorogood J, Head C, Roberts BE, Scott CS: Prognostic factors and survival in chronic myelomonocytic leukaemia (CMML). Br J Cancer 56:59-63, 1987. 2. Rovira M, Cervantes F, Lozano M, Ribera JM, Reverter JC, Rozman C: Leucemia mielomonocitica cronica y neoplasias solidas: LAsociacion casual o fortuita? Sangre (Barc) 34:207-209, 1989. 3. Sans-Sabrafen J . Buxo J , Woessner S, Florensa L, Lafuente R, PardoPeret P, Pedro C, Rodriguez-Ferrera JC, Martin E, Salinas R, Zaragoza J , Anton I:Sindrome mielodisplasico. Estudio clinico de 124 cases y revision de la literatura. Med Clin (Barc) 91:481-487, 1988. 4. Bennet JM, Catovsky D, Daniel MT, Flandrin G, Galton DAG, Gralnick HR, Sultan C , The French-American-British (FAB) Cooperative Group: Proposals for the classification of the myelodysplastic syndromes. Br J Haematol51:189-199, 1982. 15. Borras J, Garcerin J, Anglada L, et al. “El cancer a Tarragona 19801985.Estudi Epidemiologic Descriptiu.” Tarragona: Associacio Espanyola Contra el Cancer, 1988. 16. Juneja SK, Imbert M, Jouault H, Scoazec JY, Sigaux F, Sultan C: Haematological features of primary myelodysplastic syndromes (PMDS) at initial presentation: a study of I18 cases. J Clin Pathol 36:1129-1135, 1983. 17. Kerkhofs H, Hermans J, Haak HL, Leeksma HW: Utility of the FAB classification for myelodysplastic syndromes: Investigation of prognostic factors in 237 cases. Br J Haematol 65:73-81, 1987. 18. VallespiT, TorrdbadellaM, Julia A, Irriguible D, Jaen A, AcebedoG, Triginer J: Myelodysplastic syndromes: A study of 101 cases according to the FAB classification. Br J Haematol 61:83-92, 1985. 19. Coiffier B, Adeleine P, Viala JJ, Bryon PA, F f r e D, Gentilhomme 0, Vuvan H: Dysmyelopoietic syndromes. A search for prognostic factors in 193 patients. Cancer 52:83-90, 1983. 20. Copplestone JA, Mufti GJ, Hamblin TJ, Oscier DG: Immunological abnormalities in myelodysplastic syndromes. 11. Coexistent lymphoid or plasma cell neoplasms: A report of 20 cases unrelated to chemotherapy. BrJ Haematol63:149-159, 1986. 21. Checkoway H, Pearce NE, Crawford-Brown DJ: “Research Methods in Occupational Epidemiology.” New York: Oxford University Press, 1989, pp 103-122. 22. Breslow NE, Day NE: “Statistical Methods for Cancer Research. Vol 2. The Design and Analysis of Cohort Studies.” IARC Scientific Publication No. 82. Lyon: IARC, 1987. 23. Alvarez-Dardet C, Bolumar F, Porta M: Tipos de estudios. Med Clin (Barc) 89:296301, 1987. 24. Farreras-Valenti P, Rozman C, Woessner S: AnCmies refractaires sideroblastiques prenioplasiques. Nouv Rev Fr HCmatol 4:s19-522, 1964.
Myelodysplastic Syndromes and Malignant Solid Tumors: Analysis of 21 Cases Jordi Sans-Sabrafen, Josep Buxo-Costa, Soledad Woessner, Lourdes Florensa, Carles Besses, Nuria Malats, and Miquel Porta Unit of Hematology and Oncology, Hospital de la Creu Roja, Universitat de Barcelona (J.S.-S.,J.B.-C., S.W., L.F.), Unit of Hematology and Oncology, Hospital Central La Alianca (C.B.), and lnstitut Municipal d'lnvestigacio Medica, Universitat Autonoma de Barcelona (N.M., M.P.), Barcelona, Spain.
We studied the association between myelodysplastic syndromes (MDS) and malignancies in a cohort of 155 patients with MDS, 21 of whom presented malignant solid tumors. Myelodysplasia was present after the diagnosis of cancer in eight patients (interval between the diagnosis of both conditions 18 months, median survival 49.5 months), simultaneously with diagnosis in 11 (median survival 8 months), and before malignancy in two patients (interval between the diagnosis of both conditions 47 and 7 months). One patient was given chemotherapy for lung cancer, and three patients received radiotherapy for adenocarcinomaof the kidney and cancer of the prostate. At the time of diagnosis of MDS, nine patients already presented metastatic spread. Fourteen patients died, ten as a result of tumor-related complications and four because of transformation to acute nonlymphocytic leukemia. The analysis of the incidence of malignancy in patients with MDS was statistically significant for males, and the relative risk was significant in both sexes. The results of this study show that MDS patients present a higher incidence of malignant tumors than the general population, that MDS may be of real paraneoplasticsignificance, and that the occurrence of MDS in cancer patients may be considered to be related to the malignancy rather than an independent phenomenon. o 1992 Wiley-Liss, Inc. Key words: FAB classification, second malignancies, paraneoplastic syndrome
INTRODUCTION
MATERIALS AND METHODS
Malignant tumors as second neoplasms are seen with increased frequency in patients with chronic lymphocytic leukemia (CLL) and in those treated for Hodgkin's disease [ 1-31. A relationship between myelodysplastic syndromes (MDS) and malignant solid tumors has also been documented, but the apparently increased incidence of malignant tumors diagnosed subsequent to the hematological disease is not easily explained [4-131. We studied the association between MDS and malignant solid tumors in a cohort of 155 patients with myelodysplasia, 21 of whom had carcinomas. The hypotheses established were as follows: 1) patients suffering MDS have a higher incidence of malignant tumors than the general population and 2) MDS could be a paraneoplastic syndrome and could, therefore, be present before, simultaneously with, or after the diagnosis of the malignant tumor.
Myelodysplastic syndromes were diagnosed in 155 patients attending our Department of Clinical Hematology and Oncology between 1975 and 1988. Follow-up data were obtained in 149 (96%) cases. Classification of the MDS was made according to the FAB criteria [ 141, which include refractory anemia (RA), refractory anemia with ring sideroblasts (RAS), refractory anemia with excess of blasts (RAEB), chronic myelomonocytic leukemia
0 1992 Wiley-Liss, Inc.
Received for publication April 5 , 1991; accepted January 30, 1992. J. Sans-Sabrafen, J. Buxo-Costa, S . Woessner, and L. Florensa are now at the Unit of Hematology and Oncology, Hospital Central La Al ianc;a, Barcelona, Spain. Address reprint requests to: Dr. Jordi Sans-Sabrafen, Bailtn 190, 08037 Barcelona, Spain.
2
Sans-Sabrafen et al.
(CMML), and RAEB “in transformation” (RAEB-t). Histologic features of tumors were assessed after biopsy, except in one case with disseminated osteolytic metastases when the primary tumor was unknown. The incidence of malignant solid tumors observed in the cohort of patients with MDS was compared with that expected in the general population. The population of the province of Tarragona was selected to represent the latter, since Tarragona is a geographic area located close to our cohort catchment area and data are available on the incidence of cancer from the Tarragona Tumor Registry for the years between 1980 and 1985 [IS]. The expected incidence of malignant tumors was obtained by multiplying the incidence rate of cancer in the tumor registry (grouped by age and sex) by the number of person-years of the MDS cohort. The eight patients (seven men and one woman) in whom MDS was diagnosed after the malignant solid tumor were excluded from this analysis. The x2 test with 1 degree of freedom was applied to the O/E ratio, where 0 is the number of malignancies observed and E the number of expected malignancies under the null hypothesis. The relative risk of patients suffering MDS developing solid tumors was calculated, taking the population of Tarragona as the control group [ 151 for age groups over age 20 years (1981 census). No subject was excluded from this analysis. The comparison of two qualitative variables was carried out using the x2 test. With qualitative variables of two or more categories, Student’s t test and analysis of variance were used for the comparison of means.
TABLE 1. Clinical Data From 155 Patients With Myelodysplastic Syndromes
Data
RA
RAS RAEB
CMML RAEB-t
Patient (n)
19
22
17
14
50 50 70.0
58.8 51.8 41.2 48.2 71.0 66.1
57.9 42.1 76.3
85.7 14.3 68.9
0.0033 0.0356
11 52.4
6 28.6
2 9.5
0.0368
83
P
Sex
Males (%) Females (%) Age, mean (years) Second malignancy No. Percent
2 9.5
cinoma and one patient with renal tumor received local X-ray therapy. Myelodysplasia was present before malignancy in two patients, simultaneously with diagnosis in 1 1 patients, and after diagnosis of the malignant tumor in eight patients. A nonhomogeneous distribution of tumors in the subgroups of MDS was observed (RAS 2, RAEB 11, CMML 6, and RAB-t 2); n o solid tumor was found among patients with RA ( P = 0.036) (Table I). In the 1 1 patients in whom cancers were diagnosed at the same time as MDS, median survival was 8 months. In the eight patients in whom MDS developed after the diagnosis of malignancy, the median period of time between the diagnosis of both conditions was 18 months, and the median survival was 49.5 months after the initial detection of the malignant tumor. Only in two patients with RAS did the second tumor appear after the diagnosis of MDS, after intervals of 47 and 7 months. At the time of diagnosis of MDS, nine (47.4%) paRESULTS tients already presented metastatic spread (seven of l l The 155 patients with MDS consisted of 87 men and 68 patients in whom malignancy was diagnosed at the same women who ranged in age from 22 to 96 years (mean ? time as MDS and two of eight patients in whom MDS SD 68.7 k 12.9 years). There were 19 (12.3%) patients developed after the diagnosis of the neoplasms). In one of with RA, 22 (14.2%) patients with RAS, 83 (53.5%) these two patients, the diagnosis of MDS coincided with patients with RAEB, 17 ( 1 1 %) patients with CMML, and the apparent distant spread of the tumor, while the other 14 (9%) patients with RAEB-t. With the exception of was the only case to have previously been treated with RA, myelodysplasias were more frequently diagnosed in chemotherapy. males ( P = 0.0033). The higher predominance in males Of the 21 patients with malignant solid tumors, seven was particularly evident in patients with RAEB-t. Mean (33.3%) are still alive. Of them, six are apparently in a ages of patients with different subgroups of myelodyspla- clinical condition of complete remission. The remaining sia were also statistically significant ( P = 0.035) 14 patients have died, ten as a result of tumor-related (Table I). complications and four because of transformation to Malignant solid tumors developed in 2 1 ( 13.5%) pa- acute nonlymphocytic leukemia (ANLL). tients, 18 men and three women (Table 11). Cancer of the The number of expected cases of malignancy in the prostate occurred in six patients, lung cancer in five, cohort of 155 patients with MDS was 2.963 for males and carcinoma of the stomach in three, cancer of the large 1.30for females. The O/E ratio was 11/2.963 = 3.7 (X2 bowel in three, adenocarcinoma of the kidney in two, and = 21799, P < 0.01) for males and 2/1.3 = 1.54 (X2 = urothelial tumor of the bladder in one. In the remaining 0.377, P > 0.05) for females. Given that both prostate patient with disseminated osteolytic metastases, the pri- and lung cancer are common in males in the general mary tumor was unknown. None of the patients had two population, the O/E ratio for these tumor sites was also cancers identified. One patient with lung cancer was calculated. The O/E ratio for lung cancer was 3/0.4 = 7.5 given chemotherapy, and two patients with prostatic car- (X2 = 16.9, P < 0.01), and for prostate cancer the O/E
Myelodysplastic Syndromes and Maligancies
3
TABLE II. Clinical Characteristicsand Outcome in 21 Patients With MyelodysplasticSyndromes (MDS) and Solid Tumors
Age (years)
Sex
MDS
Tumor location
70 77 72
M M M
RAEB RAEB RAEB
Stomach Stomach Lung
81 61 80 14 67
M M
RAEB RAEB RAEB RAEB RAEB-t
Kidney Colon Kidney Stomach Prostate
F M M
Histopathology
Stage
Adenocarcinoma Adenocarcinoma Squamous cell carcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma
MI MI M1 T3NxMx Dukes A MI MI MI
72 80
RAEB-t CMML
Prostate Unknown
Adenocarcinoma
F
II MI
76
M
CMML
Lung
T3NxMx
61" 73" 83a 80a
M M M M
RAEB RAEB RAEB RAEB
U. bladder Prostate Prostate Prostate
Squamous cell carcinoma Urothelial cancer Adenocarcinoma Adenocarcinoma Adenocarcinoma
79" 75 " 12a
M
F M
CMML CMML CMML
Colon Colon Prostate
Adenocarcinoma Adenocarcinoma Adenocarcinoma
Dukes C Dukes A I11
75 a
M
CMML
Lung
M1
67c
M
RAS
Lung
Squamous cell carcinoma Adenocarcinoma
75c
M
RAS
Lung
Anaolastic
M
Y
TI I1
II III
Treatment None None None Radiation Surgery Progesterone None LH-RH agonists antiandrogens Surgery None None Surgery Radiation Surgery Orchiectom y antiandrogens Surgery Surgery idemkadiation
Survival (months)
6 8 6
Outcome Died Died Died
6 8 29 13 29
Died Remission Died Died Metastases
12 8
Remission Died, acute leukemia Died
8 71 105 28 28
Remission, 12b Remission, 24b Remission, 12b Remission, 12b
43 56 91
Died, 36b Remission, 36b Died, acute leukemia Died, S b
Chemotherapy
33
T2NoMo
Surgery
29
T3NxMx
None
3
Died, acute leukemia Died. 4Ib
aEight patients in whom solid tumors were diagnosed before myelodysplasia. bInterval (months) between the diagnosis of both conditions. CTwopatients in whom myelodysplasia was diagnosed before solid tumors.
was 6/0.3 = 20.0 (X2 = 108.3, P < 0.01). The relative risk of patients suffering MDS developing malignant solid tumors was 56.13 (confidence interval 95%, 37.68 and 83.63). DISCUSSION The association of malignant solid tumors as distinct from acute leukemia was first reported by Weisdorf et al. [4] in 17% of their patients with MDS. This association has been also recognized by others [S-lo], but whether the incidence is actually increased remains to be confirmed. Although it has been stated that the coexistence of CMML and malignant tumors is not purely fortuitous [9,12], the fact that this question has generally been omitted in important studies on MDS is striking [16-19]. Copplestone et al. [20] have, moreover, pointed to the high incidence of lymphoproliferative disorders in 10.5% of a series of 20 patients with MDS, which was also observed in 12 (7.5%) of the 155 patients included in our study. Juneja et al. [ 161 reported the occurrence of myeloproliferative disorders in 10.2% of 118 patients with MDS, which we observed in six (4%) of our patients.
In our study, the majority of cancers occurred in men (85.7%). When the incidence of solid tumors in patients with MDS was compared with that in the general population, it was found to be significantly higher in men than in women. Some authors have analyzed the association of SMD [ l l ] , CLL [ l ] , and CMML [12] with malignant tumors using the subject-years method, with controversial results. We observed that most cancers occurred simultaneously with MDS; in fact, eight patients developed MDS after the diagnosis of the malignant tumor. Hence, a link between both conditions was hypothesized. The present study is not strictly a cohort study [21-231 but a follow-up study of a group of patients without an intrinsic control group. Therefore, it was decided to use the methods previously described with the following assumptions: 1) that the total population of the province of Tarragona over the age of 20 years remained constant during the study period; 2) that the incidence rate of cancer remained also relatively stable during the same period of time; 3) that a surveillance bias may have occurred, since patients with MDS were subjected to closer supervision than the general population; and 4) that no case of MDS would be found in the population of Tarra-
4
Sans-Sabrafen et al.
gona. The presence of a surveillance bias would imply on our part an overestimation of the relative risk, that is, of the strength of the association between MDS and malignant solid tumor. However, such strength makes it unlikely that bias, if it did actually exist, would completely account for the observed association. With regard to the assumption that no case of MDS was present in the population of Tarragona, should there be an error in classification, this would produce an increase in the relative risk; that is to say, in our study, we would be underestimating the strength of the association. From a clinical point of view, it is interesting to note that MDS and cancer occurred simultaneously in 11 patients (RAEB 7, RAEB-t 2, CMML 2). It is also worthy of note that, in seven of the 11 patients, metastases were present at the time of diagnosis of MDS. In one patient, symptoms of MDS were noticed coincident with systemic spread of a localized tumor. In all seven patients with MDS and metastases, cancers were directly responsible for death. In three of the four patients who died due to ANLL, progression of the malignancy was not observed, which would suggest that MDS evolved independently of tumor development. Survival of patients, however, was closely related to tumor stage. Patients in whom tumors were localized and in complete remission were those who survived longest, even though three ANLL patients were among these. Only in two cases of ARS did myelodysplasia precede the development of second malignancy. The coincidence of ARS and malignant tumor has been known for many years, but it is not clear whether it is of real significance ~41. In conclusion, the results of this study suggest that MDS may be of real paraneoplastic significance, making it necessary to rule out the presence of a concurrent malignancy. On the other hand, the occurrence of MDS in cancer patients may be considered to be related to the malignancy rather than as an independent phenomenon.
ACKNOWLEDGMENT
The authors are grateful to Marta Pulido, MD, for editorial assistance and copyediting.
REFERENCES 1 . Lopez-Guillermo A, Reverter JC, Cervantes F, Viiiolas N, Rovira M,
Urbano-Ispizue A, Montserrat E, Rozman C: Neoplasias asociadas a la leucemia linfiica cronica. lncidencia y caracteristicas en una serie de 232 pacientes. Med Clin (Barc) 93:681-683, 1989. 2. Davis JC, Weiss NS, Armstrong BK: Second cancers in patients with chronic lymphocytic leukemia. JNCI 7 8 9 - 9 4 , 1987. 3 . Bookman MA, Longo DL, Young RC: Late complications of curative treatment in Hodgkin’s disease. JAMA 260:680-683, 1988.
4. Weisdorf DJ, Oken MM, Johnson GJ, Rydell RE: Chronic myelodysplastic syndrome: Short survival with or without evolution to acute leukaemia. Br J Haematol55:691-700, 1983. 5 . Raz I, Shinar E, Polliak A: Pancytopenia with hypercellular bone marrow-A possible paraneoplastic syndrome in carcinoma of the lung: A report of three cases. Am J Hematol 16:403408, 1984. 6. Sans-Sabrafen J, Buxo J, Woessner S, Florensa L, Lafuente R, PardoPeret P, Rodriguez-Ferrera JC, Pedro C: Anemia refractaria adquirida y neoplasia. Estudio de 1 I casos. Med Clin (Barc) 83:489491, 1984. 7. Hazdenar R: Pancytopenia with hypercellular bone marrow as a possible paraneoplastic syndrome. Am J Hematol 19:205-206, 1985. 8 . Foucar K, Langdon RM, Armitage JO, Olson DB, Carroll TJ: Myelodysplastic syndromes. A clinical and pathologic analysis of 109 cases. Cancer 56:553-561, 1985. 9. Sans-Sabrafen J, Woessner S , Besses C, Lafuente R, Florensa L, Buxo J: Association of chronic myelomonocytic leukemia and carcinoma: A possible paraneoplastic myelodysplasia. Am J Hematol 22: 109-1 10, 1986. 10. Perdiguer A, Girdldo MP, Cortes MT, Rubio Felix D, Giralt M: Sindromes mielodisplasicos asociados a neoplasia distinta de la leucosis aguda: Analisis de veintidos casos. Sangre 32: 194-201, 1987. I I . Stark AN, Thorogood J, Head C, Roberts BE, Scott CS: Prognostic factors and survival in chronic myelomonocytic leukaemia (CMML). Br J Cancer 56:59-63, 1987. 2. Rovira M, Cervantes F, Lozano M, Ribera JM, Reverter JC, Rozman C: Leucemia mielomonocitica cronica y neoplasias solidas: LAsociacion casual o fortuita? Sangre (Barc) 34:207-209, 1989. 3. Sans-Sabrafen J . Buxo J , Woessner S, Florensa L, Lafuente R, PardoPeret P, Pedro C, Rodriguez-Ferrera JC, Martin E, Salinas R, Zaragoza J , Anton I:Sindrome mielodisplasico. Estudio clinico de 124 cases y revision de la literatura. Med Clin (Barc) 91:481-487, 1988. 4. Bennet JM, Catovsky D, Daniel MT, Flandrin G, Galton DAG, Gralnick HR, Sultan C , The French-American-British (FAB) Cooperative Group: Proposals for the classification of the myelodysplastic syndromes. Br J Haematol51:189-199, 1982. 15. Borras J, Garcerin J, Anglada L, et al. “El cancer a Tarragona 19801985.Estudi Epidemiologic Descriptiu.” Tarragona: Associacio Espanyola Contra el Cancer, 1988. 16. Juneja SK, Imbert M, Jouault H, Scoazec JY, Sigaux F, Sultan C: Haematological features of primary myelodysplastic syndromes (PMDS) at initial presentation: a study of I18 cases. J Clin Pathol 36:1129-1135, 1983. 17. Kerkhofs H, Hermans J, Haak HL, Leeksma HW: Utility of the FAB classification for myelodysplastic syndromes: Investigation of prognostic factors in 237 cases. Br J Haematol 65:73-81, 1987. 18. VallespiT, TorrdbadellaM, Julia A, Irriguible D, Jaen A, AcebedoG, Triginer J: Myelodysplastic syndromes: A study of 101 cases according to the FAB classification. Br J Haematol 61:83-92, 1985. 19. Coiffier B, Adeleine P, Viala JJ, Bryon PA, F f r e D, Gentilhomme 0, Vuvan H: Dysmyelopoietic syndromes. A search for prognostic factors in 193 patients. Cancer 52:83-90, 1983. 20. Copplestone JA, Mufti GJ, Hamblin TJ, Oscier DG: Immunological abnormalities in myelodysplastic syndromes. 11. Coexistent lymphoid or plasma cell neoplasms: A report of 20 cases unrelated to chemotherapy. BrJ Haematol63:149-159, 1986. 21. Checkoway H, Pearce NE, Crawford-Brown DJ: “Research Methods in Occupational Epidemiology.” New York: Oxford University Press, 1989, pp 103-122. 22. Breslow NE, Day NE: “Statistical Methods for Cancer Research. Vol 2. The Design and Analysis of Cohort Studies.” IARC Scientific Publication No. 82. Lyon: IARC, 1987. 23. Alvarez-Dardet C, Bolumar F, Porta M: Tipos de estudios. Med Clin (Barc) 89:296301, 1987. 24. Farreras-Valenti P, Rozman C, Woessner S: AnCmies refractaires sideroblastiques prenioplasiques. Nouv Rev Fr HCmatol 4:s19-522, 1964.
