Predicting in-Hospital Mortality of Patients with COVID-19 Using Machine Learning Techniques.

The present work aims to identify the predictors of COVID-19 in-hospital mortality testing a set of Machine Learning Techniques (MLTs), comparing their ability to predict the outcome of interest. The model with the best performance will be used to identify in-hospital mortality predictors and to build an in-hospital mortality prediction tool. The study involved patients with COVID-19, proved by PCR test, admitted to the "Ospedali Riuniti Padova Sud" COVID-19 referral center in the Veneto region, Italy. The algorithms considered were the Recursive Partition Tree (RPART), the Support Vector Machine (SVM), the Gradient Boosting Machine (GBM), and Random Forest. The resampled performances were reported for each MLT, considering the sensitivity, specificity, and the Receiving Operative Characteristic (ROC) curve measures. The study enrolled 341 patients. The median age was 74 years, and the male gender was the most prevalent. The Random Forest algorithm outperformed the other MLTs in predicting in-hospital mortality, with a ROC of 0.84 (95% C.I. 0.78-0.9). Age, together with vital signs (oxygen saturation and the quick SOFA) and lab parameters (creatinine, AST, lymphocytes, platelets, and hemoglobin), were found to be the strongest predictors of in-hospital mortality. The present work provides insights for the prediction of in-hospital mortality of COVID-19 patients using a machine-learning algorithm.

Serum Thrombopoietin and cMpl Expression in Thrombocytopenia of Different Etiologies.

The relationship between thrombopoietin (TPO) and its receptor cMpl in thrombocytopenic conditions has not been entirely clarified. To elucidate this interplay may expand the spectrum of indications of TPO mimetics. In this study we have explored the relationship between TPO and cMpl in platelets and megakaryocytes of 43 patients with thrombocytopenia due to idiopathic thrombocytopenic purpura (ITP), bone marrow hypoplasia, myelodysplastic syndromes (MDS), and familial thrombocytopenia. Data were compared to cMpl and TPO in patients with a normal platelet count and in patients with thrombocytosis due to essential thrombocythemia (ET). All but familial patients showed higher TPO compared to controls. All thrombocytopenic states were invariably associated with increased expression of platelet cMPL compared to healthy controls. ET patients showed normal TPO and a trend toward a reduced cMpl expression. Immunofluorescence of bone marrow sections from patients with ITP and MDS failed to show a peculiar pattern compared to controls. Multiple mechanisms regulate TPO and cMpl in thrombocytopenic conditions.

Epidemiology of primary and secondary thrombocytopenia: first analysis of an administrative database in a major Italian institution.

Administrative databases can be a reliable source for estimating the epidemiology of blood disorders. No data are available estimating the epidemiology of thrombocytopenia from administrative data in Italian institutions. We analyzed the administrative database of the Padua University Hospital with the aim to study the epidemiology of thrombocytopenia in patients discharged with an International Classification of Disease, 9th Revision, Clinical Modification (ICD9-CM) code of thrombocytopenia. The database from year 2004 to 2008 was evaluated and all cases of thrombocytopenia (Code 287.1, 3, 4, 5) were identified and analyzed with regard to age, sex, associated diseases, therapeutics procedures and bleeding complications. The accuracy of electronic records was validated in all available medical records of patients discharged in 2009, by applying the ICD9-CM update 2007 version (Code 287.1, 4, 5; 287.30, 31, 32, 33, 39). We found 368 patients discharged from 2004 to 2008 with an ICD9-CM code of thrombocytopenia, correspondent to 0.1% of discharge rate and to a rate of 73.6 patients/year. The incidence of thrombocytopenia for this period was 14.8 cases per 100,000 per year. When considering patients with an ICD9-CM diagnosis of immune thrombocytopenia (ITP: Code 287.3), the incidence was of 6.8 cases per 100,000 per year. The clinical records of 40 patients with a discharge diagnosis of thrombocytopenia during year 2009 were reviewed for clinical consistency with ICD9-CM codes. A concordant diagnosis between clinical records and discharge code was found in 82.5% of cases. Following validation of ICD9-CM code, the incidence of ITP (Code 287.31) was 2.6 cases per 100,000 per year. When evaluated for sensitivity and specificity, we found the ICD-9-CM to be useful in studying thrombocytopenia using administrative data.

A A386G biallelic GPIbα gene mutation with anomalous behavior: a new mechanism suggested for Bernard-Soulier syndrome pathogenesis.

Platelet glycoprotein GPIbα mutations are the basic defect behind Bernard-Soulier syndrome, a rare inherited macrothrombocytopenia characterized by anomalies of the GPIbα, GPIbß and GPIX subunits of von Willebrand factor receptor. A 32-year old man was investigated for suspected Bernard-Soulier syndrome. Ristocetin induced agglutination was absent. Flow cytometry and Western blot analysis showed a severe reduction in GPIbα, but sequencing revealed only a biallelic c.386A>G substitution, theoretically leading to a p.Asn110Glu variation. To further clarify the data, megakaryocyte cultures were set. Though the maturation of megakaryocytes was normal, proplatelet formation was defective and GPIbα mRNA was not detectable. GPIX protein was slightly reduced and GPIbß polypeptide almost absent. Computational analysis showed that the c.386A>G mutation disrupted an exon splicing enhancer motif involved in the proper maturation of the GPIbα transcript. The c.386A>G mutation suggests a unique mutational mechanism causing the virtual absence of GPIbα without creating a stop codon.

JAK2V617F mutation is common in old patients with polycythemia vera and essential thrombocythemia.

BACKGROUND: JAK2V617F mutation occurs in 90% of polycythemia vera (PV) and in 50% of essential thrombocythemia (ET) patients. MATERIALS AND METHODS: 253 consecutive patients affected by myeloproliferative disorders (MPD, 121 PV, 132 ET) were evaluated and stratified in 4 age groups: 18-39, 40-59, 60-75 and over 75 years (>75). The JAK2V617F mutation was searched and its allele burden was evaluated. RESULTS: The percentage of mutated patients increased progressively with age mainly in patients >75 (p=0.0015 vs 18-39, p=0.0021 vs 40-59 and p=0.012 vs 60-75). We also found a progressive increase in allele burden with age (R2=0.042). Thrombotic events were more common in patients carrying the mutation in comparison with wild type (WT) (p=0.006, coefficient risk 1.94). No differences in the percentage of patients carrying the JAK2V617F mutation were found, in spite of different follow-up durations (<5 yrs, 5-10 yrs, 10-15 yrs, >15 yrs). The JAK2V617F allele burden was similar in patients with (57 ± 31%) and without (45 ± 26%) long-term hydroxyurea treatment. CONCLUSIONS: JAK2V617F mutation is more common in old than in young patients with MPD. Older patients have an higher allele burden.

Heparin-induced thrombocytopenia in patients with Philadelphia-negative myeloproliferative disorders and unusual splanchnic or cerebral vein thrombosis.

BACKGROUND: Philadelphia-negative myeloproliferative disorders (Ph-MPD) are common causes of unusual splanchnic or cerebral vein thrombosis, which is treated with unfractionated heparin (UFH) or low-molecular-weight heparin (LMWH). Heparin-induced thrombocytopenia (HIT) is a dangerous potential complication of this therapy, but it has rarely been reported in Ph-MPD. PATIENTS AND METHODS: We retrospectively reviewed clinical records of 29 patients with Ph-MPD who have been treated with UFH or LMWH for unusual splanchnic or cerebral vein thrombosis (3 cerebral sinus, 6 portal and 20 hepatic vein). The goal of the study was to determine the occurrence of new thrombotic events during heparin therapy secondary to HIT (HITT). RESULTS: During heparin therapy, 5 out of the 29 patients (17%) developed a new thrombotic episode (pulmonary embolism) with a high clinical probability of HIT based on the 4 T's score even though not all the patients developed 'true' thrombocytopenia. A diagnosis of HIT was established in 2 patients (6.8%) through the presence of heparin-related antibodies. CONCLUSIONS: Ph-MPD patients on heparin warrant careful monitoring and HIT has to be suspected whenever platelet counts drop or a new thrombosis is detectable.

Hydroxyurea in old patients with essential thrombocythemia.

BACKGROUND AND AIMS: A previous thrombotic event and advanced age are well-known risk factors for thrombosis in essential thrombocythemia (ET). In these patients, therefore, cytotoxic drugs are needed to reduce platelet count. In spite of this convincing idea, in clinical practice, some old patients do not use platelet-reducing drugs, for a variety of causes, and few specific studies in old patients with ET are available. Our retrospective study reports single-center experience in 54 old ET patients with long follow-ups. METHODS: We compared the clinical outcome of 27 ET old patients not taking cytotoxic drugs (group A) with 27 cases treated with hydroxyurea (HU) (group B), evaluating the incidence of thrombosis and thrombosis-free survival. In 16 patients in group A and in 18 in group B, V617FJak2 mutation was sought. About 20% of HU-treated patients developed major side-effects. RESULTS: No significant difference was found in the occurred thrombosis between the 2 groups in either clinical or laboratory features. V617FJak2 was equally common in groups A and B, and in patients with or without thrombosis. CONCLUSIONS: This study is not randomised and includes a small number of patients. However, it shows that it is necessary to identify better patients who really need treatment, as the side-effects of HU are relatively common in old people and their treatment should be discontinued. V617FJak2 does not define the thrombotic risk in old ET patients.

Increased risk of pregnancy complications in patients with essential thrombocythemia carrying the JAK2 (617V>F) mutation.

Essential thrombocythemia (ET) may occur in women of childbearing age. To investigate the risk of pregnancy complications, we studied 103 pregnancies that occurred in 62 women with ET. The 2-tailed Fisher exact test showed that pregnancy outcome was independent from that of a previous pregnancy. The rate of live birth was 64%, and 51% of pregnancies were uneventful. Maternal complications occurred in 9%, while fetal complications occurred in 40% of pregnancies. The Mantel-Haenszel method showed that fetal loss in women with ET was 3.4-fold higher (95% confidence interval [CI]: 3-3.9; P < .001) than in the general population. Half of the women studied carried the JAK2 (617V>F) mutation, and a multivariate logistic regression model identified this mutation as an independent predictor of pregnancy complications (P = .01). Neither the platelet count nor the leukocyte count was a risk factor. JAK2 (617V>F)-positive patients had an odds ratio of 2.02 (95% CI: 1.1 - 3.8) of developing complications in comparison with JAK2 (617V>F)-negative patients. Aspirin did not prevent complication in JAK2 (617V>F)-positive patients and appeared to worsen outcome in JAK2 (617V>F)-negative patients. A relationship was found between JAK2 (617V>F) and fetal loss (P = .05). This study indicates that patients carrying the JAK2 (617V>F) mutation have higher risk of developing pregnancy complications.

Arterial thrombosis in young women after ovarian stimulation: case report and review of the literature.

Both venous and arterial thrombosis have been described in women after ovarian stimulation and/or hyperstimulation for infertility management. The ratio between venous and arterial thrombosis in this condition is about 2 or 3 to 1, contrary to what seen during pregnancy or oral contraception where it is 5 or 10 to 1. An accurate perusal of the literature and of personal files has yielded 34 cases of arterial thrombosis after assisted reproductive technologies (ART) which entailed ovarian stimulation. There were 15 cases of ischemic strokes; 7 cases of carotid and/or vertebral artery occlusion; 6 of aorta and peripheral vessel thrombosis, 2 of mesenteric artery occlusion, 3 with myocardial infarction and 2 with intracardiac thrombosis. Associated risk factors were as follow: smoking in 4 women; antiphospholipid antibodies in 2; decrease in protein S in 1. Furthermore, polycystic ovaries were present in two women. Ovarian hyperstimulation was obtained with several protocols which included human chorionic gonadotropin (HCG) in all but a few instances. Age of women varied between 22 and 41 with an average of 32. Nineteen women were pregnant at the time of thrombosis. Only seven of these 19 pregnancies were brought to term with fetal survival. Abortion was spontaneous in 5 cases and therapeutic in the additional 7. There were two maternal fatalities.

Leg ulcers in elderly on hydroxyurea: a single center experience in Ph- myeloproliferative disorders and review of literature.

Hydroxyurea (HU) is effective in controlling thrombocytosis while reducing the risk of thrombosis in essential thrombocythemia (ET), polycythemia vera (PV) and myelofibrosis (MF). However, HU may carry more or less severe side-effects. Rare cases of patients with painful leg ulcers have been published. We report our experience on such a side-effect in a large cohort of patients with ET and PV treated with HU and review the literature on the topic. Five (4%) out of our 124 patients (69 ET, 51 PV, 4 MF; 49 males, 75 females; mean age at diagnosis 59.1+/-11.8 years) treated with HU developed painful leg ulcers. Sixty-one other patients affected with Phmyeloproliferative disorders (Ph- MPD) developing HU-related painful leg ulcers are described in the English literature. All our five patients were women and developed leg ulcers over the age of 75. Sixty-five percent of all described cases are women; 59% were over 65 years of age and 45% over 70. Most cases received over 1 gr HU per day for at least 1 year. The pathogenesis of HU-induced skin ulcers remains elusive. Treatment is difficult and requires prompt cessation of HU therapy.

Toxicity and side effects of hydroxyurea used for primary thrombocythemia.

Over the last 20 years a vast array of data has been accumulated on the efficacy of hydroxyurea (HU) in patients with Philadelphia-negative myeloproliferative disorders (MPD). However, several side effects have been described as well. Besides many anecdotal reports, no evaluation of their prevalence and type exists in large series of treated patients. We report here the side effects of HU in a retrospective, single institution, cohort study of 152 patients suffering from MPD with thrombocytosis (median follow-up 8.13 years). In 6.5% of patients drug failure was registered. Unwanted side-effects (five symptomatic macrocytic anemia, two fever reactions, two allergic reactions, four cases each of leg painful ulcers, three acute leukemia or myelodysplasia) induced to withdraw therapy in 16 patients. Three cases of nail pigmentation were observed. In our experience, HU showed to be an effective and safe drug in most patients with MPD. Prompt recognition of side effects, which have been mostly minor and rapidly subsiding on drug withdrawal, is in any case crucial to avoid more severe complications.

Safety profile of hydroxyurea in the treatment of patients with Philadelphia-negative chronic myeloproliferative disorders.

The efficacy of hydroxyurea (HU) in myeloproliferative disorders is well documented. HU controls thrombocytosis both in polycythemia vera (PV) and in essential thrombocythemia (ET), while reducing the risk of thrombosis.1 Despite many anectodal reports, no evaluation of the prevalence and type of side effects of HU exists in large series of patients.