Early Human Development, 29 (1992) 143-147 Elsevier Scientific Publishers Ireland Ltd.
143
EHD 01264
Autoimmune thrombocytopenic purpura: maternal and fetal disease A. Pachi, E. Carapellaa,M.G. Mazzucconic,G.M. Gandolfob, R. Paesano,V. Petrelli, V. De SanctisCand F. MandelliC ‘Institute of Puericulture. ‘Department of Experimental Medicine cDepartment of Human Biopathology, Hematological Section, IV Obstetrical and Gynaecological Clinic, University o/ Rome, ‘La Sapienza’ (Italy)
Summary
We have followed up 63 pregnancies in women with autoimmune thrombocytopenic purpura (ATP). Of these, 15 were previously splenectomized. The characteristics of the sample can be summed up as follows: average age 27 years (17-41); platelets at the beginning of pregnancy, TI 129.5 x 109/1 (range 16-488); platelets at delivery, Z 133 x 109/1 (range 8-477); PA-IgG at delivery, TI 320 ng IgG/lO’ platlets (range 10-1000); SPB-IgG at delivery, Y262 ng IgG/lO’ platlets (range 10-1000). There were 30 spontaneous deliveries and 33 cesarean sections. Forty-two newborns had a platelet count within the normal range while nine had a platelet count I 150 x 109/1, while six had I 100 x 109/1 and a further six 5 50 x 109/l. The aim of this study is the evaluation of maternal risk and of possible feto-neonatal thrombocytopenia at birth. In this regard, the following parameters were considered: previous maternal splenectomy; the platelet count at the beginning of pregnancy; the platelet count and the titres of PA-IgG and SPB-IgG at delivery. Preliminary statistical evaluation of these parameters enabled us to identify a risk score. From this it was possible to obtain an optimum management of the final stage of pregnancy regarding the therapeutic approach and the timing of delivery. Key words: autoimmune thrombocytopenic purpura; feto-natal thrombocytopenia birth; previous maternal splenectomy; platelet count
at
Correspondence to: A. Pa&i, Institute of Puericulture, IV Obstetrical and Gynaecological Clinic, University of Rome, ‘La Sapienza’, Italy. 0378-3782/92605.00 0 1992 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland
144
Introduction Autoimmune thrombocytopenic purpura (ATP) is an immunologic hemostatic disorder characterized by a decrease in the number of circulating platelets caused by specific autoantibodies. It affects women in childbearing age and often manifests itself for the first time during pregnancy [l]. Maternal hemorrhages are now irrelevant because of the remarkable progress made in their treatment, but the possibility of the transmission of the disease to the fetus remains [2]. This happens in between 41 and 52% of cases [3]. The cause of fetal thrombocytopenia is supposed to be the maternal antiplatelet antibodies, for the most part, composed of IgG: platelet associated IgG (PA-&G) or platelet bindable IgG (SPB-IgG) circulating in the serum. These latter can cross the placenta and attack the fetal platelets. The resulting fetal thrombocytopenia exposes the fetus to a risk of bleeding especially intracranial hemorrhage during delivery [2,3]. Until now, no correlation has been found between neonatal pathology and maternal platelet levels just as the evaluation of the titres of the antiplatelet antibodies (PA-IgG and SPB-IgG) did not reach any definite conclusions [2-41. Subsequently, direct diagnosis on the fetus was attempted by scalp blood sampling [5] and, more recently, by obtaining blood from the umbilical cord (cordocentesis) [6]. This, in the case of a low platelet count, would leave open the option for a cesarean section so as to avoid the trauma of a vaginal delivery. The improvement of the technique of cordocentesis and of blood transfusion in utero in Rh immunization has recently suggested the transfusion of platelet concentrates to the thrombocytopenic fetus on the basis of an extemporary platelet count. This makes vaginal delivery possible even in the case of fetuses with severe thrombocytopenia [7,8]. Materials ad Methods Sixty-three pregnancies associated with ATP were followed up from 1982 until the present. The type and the timing of delivery, in the absence of obstetrical indications, depended on the evaluation of these parameters: (1) low maternal platelet count; (2) antibodies titres (especially SPB-IgG); (3) previous maternal splenectomy; (4) presence of hemorrhage; (5) necessity or not of therapy during pregnancy. The serum determination of antiplatelet autoantibodies was carried out for the SPB-IgG by using the indirect test of Dixon (normal values in pregnancy 5 200 ng IgG/107 platelets) and for the PA-&G with the direct test of Dixon (normal values in pregnancy 5 100 ng IgG/107 platelets [9]. The average age of the patients was 27 years (range 17-41 (mm-max)). In 54 cases (85.7%) ATP was diagnosed before pregnancy between 2 and 240 months (average 80.3 months) from the onset of the disease to the beginning of pregnancy. In 15 cases, splenectomy had already been performed. In 9 cases (14.3%) ATP was diagnosed at a gestational age from 8 to 33 weeks (average 16.7 weeks). An a posteriori preliminary statistical evaluation of the factors already considered clinically was carried out. We applied the x2 test to our data to verify the existence of a dependence between the variables under examination. In addition, an objective
145
description of ATP in pregnancy was given using the Multiple Correspondences Analysis (MCA). This is the non inferential method most indicated to demonstrate the structures of dependence which exist between the multiple characters studied.
From our data it emerges that 19 cases (30.2%) maintained normal platelet values during the whole of pregnancy. In 37 cases (58.7%) the values were between 50 and 150 x log/l; and in 7 (11.1%) there was always a platelet count below 50 x log/l. The patterns of maternal parameters are shown in the table.
Total cases 63 Splenectomy patients 15 Non-splenectomy patients 48
Begining of prewcy Platlets iY ( x 109/l)
Pre Partum Platlets X ( x 10%)
PA-IgG (ng IgG/lO’ platelets)
SPBIgG (ng IgG/lO’ platlets)
129.5 (16-488) 219.4 (32-488) 98.6 (11-484)
133 (8-477) 219.3 (22-477) 106.2 (8-257)
320 (10-1000) 414 (10-900) 290.8 (10-700)
262 (10-1000) 449 (10-1000) 203.4 (10-710)
Slight hemorrhage symptoms were evident in 10 cases (15.8%) with blood loss from the external genitals and in 3 cases (4.7%) there was gum bleeding, nose bleeding and cutaneous petechiae. On the basis of a clinical evaluation, treatment was necessary in 14 pregnancies (22.2%): 10 patients were treated with prednisone from 5 to 25 mg/day associated in two cases with antifibrinolytics; one patient with only antitibrinolytics; three patients with prednisone associated with intravenous high dose gamma-globulin (400 mgikg per day for a 5 day course Endobulin, manufactured by the pharmaceutical company IMMUNO). The pregnancies had the following outcome: cesarean section (CS) in 33 cases (52.4%) carried out at an average gestational age (GA) of 37.5 weeks (range 33-41 (mm-max)); spontaneous delivery (SD) in 30 cases (47.6%) at an average gestational age of 39.2 (range 37-41 (mm-max)). There were no major hemorrhage complications in any of the cases. However, in 15 cases (23.8%) therapy was necessary: the use of platelet concentrates in eight cases, packed red cell blood in four cases and both in three cases. Cesarean section was carried out: in 12 cases (36.4%) for solely obstetrical reasons (abruptio placentae, fetal distress and previous CS ); in 12 other cases (36.4%) for hematological reasons (high titres of PA-IgG and SPB-IgG, low platelet values, and previous splenectomy); and in the remaining nine cases (27.2%) because of the combined evaluation of obstetrical and hematological factors. Of the newborns (63 in number), all alive, 42 (66.6%) had platelet count values
146
> 150 x log/l; 9 (14.2%) had I 150 x 109/1; 6 (9.6%) had I 100 x 109/1 and in six cases (9.6%) the platelet count was 150 x 109/l. Of the latter six newborns, all of whom were delivered by cesarean section, four had skin hemorrhages (petechiae and widespread bruises) and were treated with prednisone, platelet concentrates associated in two cases with blood transfusions and in one case with high dose intravenous gamma-globulin. The mothers of the six newborns with platelet counts I 50 x log/1 had: an average number of platelets at the beginning of pregnancy 60.5 x log/l (range 16-135 (min-max)) and, at the time of delivery 44.8 x IO911 (range 18-108 (min-max)). The average pre-partum values of PA-IgG and SPB-IgG were 463.3 ng IgG/107 platelets (range 80-700 (mm-max)) and 425.8 ng IgG/107 platelets (range 10-1000 @in-max)), respectively. Three of these had been splenectomized. Diiion Using MCA, we distinguished four groups of variables. In the first group, ‘maternal thrombocytopenia in the third trimester < 50 x 109/1’, appears to be correlated with: (1) maternal hemorrhage; (2) neonatal platelet count < 50 x log/l; (3) neonatal hemorrhage; (4) neonatal therapy. The dependence between maternal thrombocytopenia in the third trimester and neonatal hemorrhage was also confirmed by x2. Instead, maternal thrombocytopenia in the first trimester of pregnancy appears to be much less interesting for a prognostic evaluation of the newborn. In the second group the correlated variables were: (1) splenectomy; (2) high SPBIgG (>500 ng IgG/107 platelets); (3) high PA-IgG (>500 ng IgG/107 platelets). This confirms that splenectomy does not cure the disease itself but induces an eventual remission of it. The variables ‘splenectomy’ and ‘neonatal thrombocytopenia’ are not statistically correlateable, even though in our data there are four neonatal thrombocytopenics (platelet count < 100 x log/l) born from already splenectomized mothers (15 in number). In the third group the variables correlated are the following: (1) nonsplenectomized patients; (2) PA-IgG and SPB-IgG < 500 ng &G/lo7 platelets; (3) platelets of 1st and 3rd trimester > 50 and < 150 x 109/l; (4) neonatal platelets > 150 x log/l; (5) absence of neonatal or maternal hemorrhages. It would seem that the variable ‘non-splenectomized patients’ is associated with a more favourable materno-fetal situation even if this dependence is not confirmed by x2. Instead, the dependence between PA-IgG and neonatal thrombocytopenia has been confirmed by x2. Finally, in the fourth group the variables are: (1) maternal platelet count in the 1st and 3rd trimester > 150 x 109/1; (2) neonatal platelet count > 50 and < 100 x 109/l; (3) diagnosis of ATP > 50 and < 100 months before pregnancy. So, there is a possible relationship between greater distance of diagnosis of pregnancy and less serious maternal consequences. Conclusions The results of our study show the possibility of bringing to a favourable term an
147
ATP associated pregnancy. Looking to the future, we can hypothesize a lesser need of substitute therapies during delivery. In fact, the use of intravenous gammaglobulin increases the values of the platelets and brings the patient to optimum conditions for delivery. Neonatal morbidity (9.3%) is greatly reduced and the mortality and morbidity linked to intracranial hemorrhage is reduced to zero. The cesarean section prevents it even in newborns with a platelet count < 50 x 109/l. Finally, the statistical elaboration in our prospective study showed that the variables to take into greatest consideration to evaluate the feto-neonatal risk are: (1) maternal platelets in the 3rd trimester; (2) PA-IgG in the 3rd trimester. x2 showed that SPB-IgG does not correlate with neonatal thrombocytopenia. Using the MCA with regard to splenectomy, there is a tendency to associate the variable ‘non splenectomized patients’ and the variable ‘newborns with a platelet count > 150 x 109/1’ but not the opposite. These latter data, in any case, need further statistical evaluation since the dependence trend observed was not confirmed by x2. The results obtained confirmed our tendency not to perform direct diagnosis by cordocentesis. Finally, we disagree with performing a platelet transfusion in urero before delivery in the case of fetal thrombocytopenia from maternal ATP. This should be reserved only for thrombocytopenic alloimmune fetuses. They, in fact, can present bleeding symptoms in the uterus, at a very early stage and could benefit from the transfusion of the mother’s platelets (81. Acknowledgements We would like to thank Professor Anna Grassi, Department of Statistics, Faculty of Economics and Commerce, University of Rome, ‘La Sapienza’ for the statistical elaboration of the data. References Martin, J.N., Morrison, J.C. and Files, J.C. (1984): Autoimmune thrombocytopenic purpura: current concepts and recommended practices. Am. J. Obstet. Gynaecol., 150, 86-96. Mazzucconi, M.G., Francesconi, M., Fidani, P., Conti, L. Martelli, M.C., Paesano, R., Pachi, A., Purpura, M. and Gandolfo, G.M. (1985): Pregnancy, delivery and detection of antiplatelet antibodies in women with idiopathic thrombocytopenic purpura. Hematologia, 70, 506-509. Kelton, J.G., (1983): Management of the pregnant patient with idiopathic thrombocytopenic purpura. Ann. Intern. Med., 99, 796. Samuels, P., Bussel, J.B. and Braitman, L.E. (1990): Estimation of the risk of thrombocytopenia in the offspring of pregnant women with presumed immune thrombocytopenic purpura. New Engl. J. Med., 323, 229-235 Christianens Godelieve, C.M.L. and Helmerhorst, F.M. (1987): Validity of intrapartum diagnosis of fetal thrombocytopenia. Am. J. Gbstet. Gynaecol., 157, 865-865. Kaplan, C., Daffos, F., Forestier, F., Tertian, G., Catherine, N. and Pons, J.C. (1990): Tchernia Fetal platelet counts in thrombocytopenic pregnancy. Lancet, 336, 979-982. Paniel, B.J., True, J.B. and Poittout P.H. (1987): Thrombccytopenie autoimmune et grossesse. J. Gynaecol. Obstet. Biol. Repord, 16, 17-26. Kaplan, C., Daffos, F., Forestier, F., Cox, W.L., Lyon-Caen, D., Dupuy-Montbrun, MC. and Salmo, Ch. (1988): Management of alloimmune thrombocytopenia: antenatal diagnosis and in utero transfusion of maternal platelets. Blood, 72, 340-343. Dixon, R.H., Rosse, W. and Ebbert, L. (1975): Quantitative determination of antibody in idiopathic thrombocytopenic purpura. N. Engl. J. Med., 292, 239.
143
EHD 01264
Autoimmune thrombocytopenic purpura: maternal and fetal disease A. Pachi, E. Carapellaa,M.G. Mazzucconic,G.M. Gandolfob, R. Paesano,V. Petrelli, V. De SanctisCand F. MandelliC ‘Institute of Puericulture. ‘Department of Experimental Medicine cDepartment of Human Biopathology, Hematological Section, IV Obstetrical and Gynaecological Clinic, University o/ Rome, ‘La Sapienza’ (Italy)
Summary
We have followed up 63 pregnancies in women with autoimmune thrombocytopenic purpura (ATP). Of these, 15 were previously splenectomized. The characteristics of the sample can be summed up as follows: average age 27 years (17-41); platelets at the beginning of pregnancy, TI 129.5 x 109/1 (range 16-488); platelets at delivery, Z 133 x 109/1 (range 8-477); PA-IgG at delivery, TI 320 ng IgG/lO’ platlets (range 10-1000); SPB-IgG at delivery, Y262 ng IgG/lO’ platlets (range 10-1000). There were 30 spontaneous deliveries and 33 cesarean sections. Forty-two newborns had a platelet count within the normal range while nine had a platelet count I 150 x 109/1, while six had I 100 x 109/1 and a further six 5 50 x 109/l. The aim of this study is the evaluation of maternal risk and of possible feto-neonatal thrombocytopenia at birth. In this regard, the following parameters were considered: previous maternal splenectomy; the platelet count at the beginning of pregnancy; the platelet count and the titres of PA-IgG and SPB-IgG at delivery. Preliminary statistical evaluation of these parameters enabled us to identify a risk score. From this it was possible to obtain an optimum management of the final stage of pregnancy regarding the therapeutic approach and the timing of delivery. Key words: autoimmune thrombocytopenic purpura; feto-natal thrombocytopenia birth; previous maternal splenectomy; platelet count
at
Correspondence to: A. Pa&i, Institute of Puericulture, IV Obstetrical and Gynaecological Clinic, University of Rome, ‘La Sapienza’, Italy. 0378-3782/92605.00 0 1992 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland
144
Introduction Autoimmune thrombocytopenic purpura (ATP) is an immunologic hemostatic disorder characterized by a decrease in the number of circulating platelets caused by specific autoantibodies. It affects women in childbearing age and often manifests itself for the first time during pregnancy [l]. Maternal hemorrhages are now irrelevant because of the remarkable progress made in their treatment, but the possibility of the transmission of the disease to the fetus remains [2]. This happens in between 41 and 52% of cases [3]. The cause of fetal thrombocytopenia is supposed to be the maternal antiplatelet antibodies, for the most part, composed of IgG: platelet associated IgG (PA-&G) or platelet bindable IgG (SPB-IgG) circulating in the serum. These latter can cross the placenta and attack the fetal platelets. The resulting fetal thrombocytopenia exposes the fetus to a risk of bleeding especially intracranial hemorrhage during delivery [2,3]. Until now, no correlation has been found between neonatal pathology and maternal platelet levels just as the evaluation of the titres of the antiplatelet antibodies (PA-IgG and SPB-IgG) did not reach any definite conclusions [2-41. Subsequently, direct diagnosis on the fetus was attempted by scalp blood sampling [5] and, more recently, by obtaining blood from the umbilical cord (cordocentesis) [6]. This, in the case of a low platelet count, would leave open the option for a cesarean section so as to avoid the trauma of a vaginal delivery. The improvement of the technique of cordocentesis and of blood transfusion in utero in Rh immunization has recently suggested the transfusion of platelet concentrates to the thrombocytopenic fetus on the basis of an extemporary platelet count. This makes vaginal delivery possible even in the case of fetuses with severe thrombocytopenia [7,8]. Materials ad Methods Sixty-three pregnancies associated with ATP were followed up from 1982 until the present. The type and the timing of delivery, in the absence of obstetrical indications, depended on the evaluation of these parameters: (1) low maternal platelet count; (2) antibodies titres (especially SPB-IgG); (3) previous maternal splenectomy; (4) presence of hemorrhage; (5) necessity or not of therapy during pregnancy. The serum determination of antiplatelet autoantibodies was carried out for the SPB-IgG by using the indirect test of Dixon (normal values in pregnancy 5 200 ng IgG/107 platelets) and for the PA-&G with the direct test of Dixon (normal values in pregnancy 5 100 ng IgG/107 platelets [9]. The average age of the patients was 27 years (range 17-41 (mm-max)). In 54 cases (85.7%) ATP was diagnosed before pregnancy between 2 and 240 months (average 80.3 months) from the onset of the disease to the beginning of pregnancy. In 15 cases, splenectomy had already been performed. In 9 cases (14.3%) ATP was diagnosed at a gestational age from 8 to 33 weeks (average 16.7 weeks). An a posteriori preliminary statistical evaluation of the factors already considered clinically was carried out. We applied the x2 test to our data to verify the existence of a dependence between the variables under examination. In addition, an objective
145
description of ATP in pregnancy was given using the Multiple Correspondences Analysis (MCA). This is the non inferential method most indicated to demonstrate the structures of dependence which exist between the multiple characters studied.
From our data it emerges that 19 cases (30.2%) maintained normal platelet values during the whole of pregnancy. In 37 cases (58.7%) the values were between 50 and 150 x log/l; and in 7 (11.1%) there was always a platelet count below 50 x log/l. The patterns of maternal parameters are shown in the table.
Total cases 63 Splenectomy patients 15 Non-splenectomy patients 48
Begining of prewcy Platlets iY ( x 109/l)
Pre Partum Platlets X ( x 10%)
PA-IgG (ng IgG/lO’ platelets)
SPBIgG (ng IgG/lO’ platlets)
129.5 (16-488) 219.4 (32-488) 98.6 (11-484)
133 (8-477) 219.3 (22-477) 106.2 (8-257)
320 (10-1000) 414 (10-900) 290.8 (10-700)
262 (10-1000) 449 (10-1000) 203.4 (10-710)
Slight hemorrhage symptoms were evident in 10 cases (15.8%) with blood loss from the external genitals and in 3 cases (4.7%) there was gum bleeding, nose bleeding and cutaneous petechiae. On the basis of a clinical evaluation, treatment was necessary in 14 pregnancies (22.2%): 10 patients were treated with prednisone from 5 to 25 mg/day associated in two cases with antifibrinolytics; one patient with only antitibrinolytics; three patients with prednisone associated with intravenous high dose gamma-globulin (400 mgikg per day for a 5 day course Endobulin, manufactured by the pharmaceutical company IMMUNO). The pregnancies had the following outcome: cesarean section (CS) in 33 cases (52.4%) carried out at an average gestational age (GA) of 37.5 weeks (range 33-41 (mm-max)); spontaneous delivery (SD) in 30 cases (47.6%) at an average gestational age of 39.2 (range 37-41 (mm-max)). There were no major hemorrhage complications in any of the cases. However, in 15 cases (23.8%) therapy was necessary: the use of platelet concentrates in eight cases, packed red cell blood in four cases and both in three cases. Cesarean section was carried out: in 12 cases (36.4%) for solely obstetrical reasons (abruptio placentae, fetal distress and previous CS ); in 12 other cases (36.4%) for hematological reasons (high titres of PA-IgG and SPB-IgG, low platelet values, and previous splenectomy); and in the remaining nine cases (27.2%) because of the combined evaluation of obstetrical and hematological factors. Of the newborns (63 in number), all alive, 42 (66.6%) had platelet count values
146
> 150 x log/l; 9 (14.2%) had I 150 x 109/1; 6 (9.6%) had I 100 x 109/1 and in six cases (9.6%) the platelet count was 150 x 109/l. Of the latter six newborns, all of whom were delivered by cesarean section, four had skin hemorrhages (petechiae and widespread bruises) and were treated with prednisone, platelet concentrates associated in two cases with blood transfusions and in one case with high dose intravenous gamma-globulin. The mothers of the six newborns with platelet counts I 50 x log/1 had: an average number of platelets at the beginning of pregnancy 60.5 x log/l (range 16-135 (min-max)) and, at the time of delivery 44.8 x IO911 (range 18-108 (min-max)). The average pre-partum values of PA-IgG and SPB-IgG were 463.3 ng IgG/107 platelets (range 80-700 (mm-max)) and 425.8 ng IgG/107 platelets (range 10-1000 @in-max)), respectively. Three of these had been splenectomized. Diiion Using MCA, we distinguished four groups of variables. In the first group, ‘maternal thrombocytopenia in the third trimester < 50 x 109/1’, appears to be correlated with: (1) maternal hemorrhage; (2) neonatal platelet count < 50 x log/l; (3) neonatal hemorrhage; (4) neonatal therapy. The dependence between maternal thrombocytopenia in the third trimester and neonatal hemorrhage was also confirmed by x2. Instead, maternal thrombocytopenia in the first trimester of pregnancy appears to be much less interesting for a prognostic evaluation of the newborn. In the second group the correlated variables were: (1) splenectomy; (2) high SPBIgG (>500 ng IgG/107 platelets); (3) high PA-IgG (>500 ng IgG/107 platelets). This confirms that splenectomy does not cure the disease itself but induces an eventual remission of it. The variables ‘splenectomy’ and ‘neonatal thrombocytopenia’ are not statistically correlateable, even though in our data there are four neonatal thrombocytopenics (platelet count < 100 x log/l) born from already splenectomized mothers (15 in number). In the third group the variables correlated are the following: (1) nonsplenectomized patients; (2) PA-IgG and SPB-IgG < 500 ng &G/lo7 platelets; (3) platelets of 1st and 3rd trimester > 50 and < 150 x 109/l; (4) neonatal platelets > 150 x log/l; (5) absence of neonatal or maternal hemorrhages. It would seem that the variable ‘non-splenectomized patients’ is associated with a more favourable materno-fetal situation even if this dependence is not confirmed by x2. Instead, the dependence between PA-IgG and neonatal thrombocytopenia has been confirmed by x2. Finally, in the fourth group the variables are: (1) maternal platelet count in the 1st and 3rd trimester > 150 x 109/1; (2) neonatal platelet count > 50 and < 100 x 109/l; (3) diagnosis of ATP > 50 and < 100 months before pregnancy. So, there is a possible relationship between greater distance of diagnosis of pregnancy and less serious maternal consequences. Conclusions The results of our study show the possibility of bringing to a favourable term an
147
ATP associated pregnancy. Looking to the future, we can hypothesize a lesser need of substitute therapies during delivery. In fact, the use of intravenous gammaglobulin increases the values of the platelets and brings the patient to optimum conditions for delivery. Neonatal morbidity (9.3%) is greatly reduced and the mortality and morbidity linked to intracranial hemorrhage is reduced to zero. The cesarean section prevents it even in newborns with a platelet count < 50 x 109/l. Finally, the statistical elaboration in our prospective study showed that the variables to take into greatest consideration to evaluate the feto-neonatal risk are: (1) maternal platelets in the 3rd trimester; (2) PA-IgG in the 3rd trimester. x2 showed that SPB-IgG does not correlate with neonatal thrombocytopenia. Using the MCA with regard to splenectomy, there is a tendency to associate the variable ‘non splenectomized patients’ and the variable ‘newborns with a platelet count > 150 x 109/1’ but not the opposite. These latter data, in any case, need further statistical evaluation since the dependence trend observed was not confirmed by x2. The results obtained confirmed our tendency not to perform direct diagnosis by cordocentesis. Finally, we disagree with performing a platelet transfusion in urero before delivery in the case of fetal thrombocytopenia from maternal ATP. This should be reserved only for thrombocytopenic alloimmune fetuses. They, in fact, can present bleeding symptoms in the uterus, at a very early stage and could benefit from the transfusion of the mother’s platelets (81. Acknowledgements We would like to thank Professor Anna Grassi, Department of Statistics, Faculty of Economics and Commerce, University of Rome, ‘La Sapienza’ for the statistical elaboration of the data. References Martin, J.N., Morrison, J.C. and Files, J.C. (1984): Autoimmune thrombocytopenic purpura: current concepts and recommended practices. Am. J. Obstet. Gynaecol., 150, 86-96. Mazzucconi, M.G., Francesconi, M., Fidani, P., Conti, L. Martelli, M.C., Paesano, R., Pachi, A., Purpura, M. and Gandolfo, G.M. (1985): Pregnancy, delivery and detection of antiplatelet antibodies in women with idiopathic thrombocytopenic purpura. Hematologia, 70, 506-509. Kelton, J.G., (1983): Management of the pregnant patient with idiopathic thrombocytopenic purpura. Ann. Intern. Med., 99, 796. Samuels, P., Bussel, J.B. and Braitman, L.E. (1990): Estimation of the risk of thrombocytopenia in the offspring of pregnant women with presumed immune thrombocytopenic purpura. New Engl. J. Med., 323, 229-235 Christianens Godelieve, C.M.L. and Helmerhorst, F.M. (1987): Validity of intrapartum diagnosis of fetal thrombocytopenia. Am. J. Gbstet. Gynaecol., 157, 865-865. Kaplan, C., Daffos, F., Forestier, F., Tertian, G., Catherine, N. and Pons, J.C. (1990): Tchernia Fetal platelet counts in thrombocytopenic pregnancy. Lancet, 336, 979-982. Paniel, B.J., True, J.B. and Poittout P.H. (1987): Thrombccytopenie autoimmune et grossesse. J. Gynaecol. Obstet. Biol. Repord, 16, 17-26. Kaplan, C., Daffos, F., Forestier, F., Cox, W.L., Lyon-Caen, D., Dupuy-Montbrun, MC. and Salmo, Ch. (1988): Management of alloimmune thrombocytopenia: antenatal diagnosis and in utero transfusion of maternal platelets. Blood, 72, 340-343. Dixon, R.H., Rosse, W. and Ebbert, L. (1975): Quantitative determination of antibody in idiopathic thrombocytopenic purpura. N. Engl. J. Med., 292, 239.
