What is SamzaSQL, and what might I use it for? Does this mean that Samza is turning into a database? What is a query optimizer, and what can it do for my streaming queries?
How does Apache Calcite parse, validate and optimize streaming SQL queries? How is relational algebra extended to handle streaming?
Apache Hive is a rapidly evolving project which continues to enjoy great adoption in big data ecosystem. Although, Hive started primarily as batch ingestion and reporting tool, community is hard at work in improving it along many different dimensions and use cases. This talk will provide an overview of latest and greatest features and optimizations which have landed in project over last year. Materialized view, micro managed tables and workload management are some noteworthy features.
I will deep dive into some optimizations which promise to provide major performance gains. Support for ACID tables has also improved considerably. Although some of these features and enhancements are not novel but have existed for years in other DB systems, implementing them on Hive poses some unique challenges and results in lessons which are generally applicable in many other contexts. I will also provide a glimpse of what is expected to come in near future.
Speaker: Ashutosh Chauhan, Engineering Manager, Hortonworks
Advanced Stream Processing with Flink and Pulsar - Pulsar Summit NA 2021 KeynoteStreamNative
In this talk, Till Rohrmann and Addison Higham discuss how Flink allows for ambitious stream processing workflows and how Pulsar and Flink enable new capabilities that push forward the state-of-the-art in streaming. They will also share upcoming features and new capabilities in the integrations between Flink and Pulsar and how these two communities are working together to truly advance the power of stream processing.
This presentation about HBase will help you understand what is HBase, what are the applications of HBase, how is HBase is different from RDBMS, what is HBase Storage, what are the architectural components of HBase and at the end, we will also look at some of the HBase commands using a demo. HBase is an essential part of the Hadoop ecosystem. It is a column-oriented database management system derived from Google’s NoSQL database Bigtable that runs on top of HDFS. After watching this video, you will know how to store and process large datasets using HBase. Now, let us get started and understand HBase and what it is used for.
Below topics are explained in this HBase presentation:
1. What is HBase?
2. HBase Use Case
3. Applications of HBase
4. HBase vs RDBMS
5. HBase Storage
6. HBase Architectural Components
What is this Big Data Hadoop training course about?
Simplilearn’s Big Data Hadoop training course lets you master the concepts of the Hadoop framework and prepares you for Cloudera’s CCA175 Big data certification. The Big Data Hadoop and Spark developer course have been designed to impart in-depth knowledge of Big Data processing using Hadoop and Spark. The course is packed with real-life projects and case studies to be executed in the CloudLab.
What are the course objectives?
This course will enable you to:
1. Understand the different components of the Hadoop ecosystem such as Hadoop 2.7, Yarn, MapReduce, Pig, Hive, Impala, HBase, Sqoop, Flume, and Apache Spark
2. Understand Hadoop Distributed File System (HDFS) and YARN as well as their architecture, and learn how to work with them for storage and resource management
3. Understand MapReduce and its characteristics, and assimilate some advanced MapReduce concepts
4. Get an overview of Sqoop and Flume and describe how to ingest data using them
5. Create database and tables in Hive and Impala, understand HBase, and use Hive and Impala for partitioning
6. Understand different types of file formats, Avro Schema, using Arvo with Hive, and Sqoop and Schema evolution
7. Understand Flume, Flume architecture, sources, flume sinks, channels, and flume configurations
8. Understand HBase, its architecture, data storage, and working with HBase. You will also understand the difference between HBase and RDBMS
9. Gain a working knowledge of Pig and its components
10. Do functional programming in Spark
11. Understand resilient distribution datasets (RDD) in detail
12. Implement and build Spark applications
13. Gain an in-depth understanding of parallel processing in Spark and Spark RDD optimization techniques
14. Understand the common use-cases of Spark and the various interactive algorithms
15. Learn Spark SQL, creating, transforming, and querying Data frames
Learn more at https://www.simplilearn.com/big-data-and-analytics/big-data-and-hadoop-training
ΚΤΣ 2016 - Μέθοδοι δοκιμών και ελέγχου ποιότητας υλικώνAlexander Tsiolis
Ποιες δοκιμές χρειάζεται να γνωρίζει όποιος ασχολείται με σκυρόδεμα, από την σκοπιά του νέου ΚΤΣ. Δοκιμές και πρότυπα για αδρανή, νωπό και σκληρυμένο σκυρόδεμα καθώς και σύντομη αναφορά στην δειγματοληψία υφιστάμενων κατασκευών και τις χρησιμοποιούμενες μεθόδους μη καταστροφικών μεθόδων.
Speaker: Jean-Daniel Cryans (Cloudera)
HBase Replication has come a long way since its inception in HBase 0.89 almost four years ago. Today, master-master and cyclic replication setups are supported; many bug fixes and new features like log compression, per-family peers configuration, and throttling have been added; and a major refactoring has been done. This presentation will recap the work done during the past four years, present a few use cases that are currently in production, and take a look at the roadmap.
Orchestration tool roundup kubernetes vs. docker vs. heat vs. terra form vs...Nati Shalom
Video recording: https://www.youtube.com/watch?v=tGlIgUeoGz8
It’s no news that containers represent a portable unit of deployment, and OpenStack has proven an ideal environment for running container workloads. However, where it usually becomes more complex is that many times an application is often built out of multiple containers. What’s more, setting up a cluster of container images can be fairly cumbersome because you need to make one container aware of another and expose intimate details that are required for them to communicate which is not trivial especially if they’re not on the same host.
These scenarios have instigated the demand for some kind of orchestrator. The list of container orchestrators is growing fairly fast. This session will compare the different orchestation projects out there - from Heat to Kubernetes to TOSCA - and help you choose the right tool for the job.
Session link from teh summit: https://openstacksummitmay2015vancouver.sched.org/event/abd484e0dedcb9774edda1548ad47518#.VV5eh5NViko
Parallel builds in Eclipse IDE workspace - EclispeCon Europe 2018Mickael Istria
https://www.eclipsecon.org/europe2018/sessions/parallel-builds-eclipse-ide-workspace-value-making-and-optimal-usages
Eclipse Platform 4.8 introduced support for parallel builds in the workspace. This feature is placed at the lowest level of the workspace so it can easily be profitable to any adopter without specific adoption effort, besides respecting usual good practices. This can in theory turn the Eclipse workspace into one of the fastest polyglot build engines in developer world.
In this presentation, we’ll explain what is the value to expect from this parallel builds (mostly performance gain and faster availability of project output).
Then we’ll also get into the details of how all that is implemented, reminding ourselves some basics of dependency resolution, graph processing and scheduling rules. We’ll remind the important best practices of builders and scheduling in the Eclipse resource model that are necessary to reach best results.
And we’ll go through a few case studies of project builders to highlight how those may be improved to take advantage of parallel builds, and get into deep details of how this can be enabled in JDT and other important Eclipse projects.
After this presentation, plugin developers will be empowered to make their project builders take advantage of parallel builds in Eclipse IDE, and may get ideas about really using this feature as an alternative build engine to speed up their headless builds as well.
What is SamzaSQL, and what might I use it for? Does this mean that Samza is turning into a database? What is a query optimizer, and what can it do for my streaming queries?
How does Apache Calcite parse, validate and optimize streaming SQL queries? How is relational algebra extended to handle streaming?
Apache Hive is a rapidly evolving project which continues to enjoy great adoption in big data ecosystem. Although, Hive started primarily as batch ingestion and reporting tool, community is hard at work in improving it along many different dimensions and use cases. This talk will provide an overview of latest and greatest features and optimizations which have landed in project over last year. Materialized view, micro managed tables and workload management are some noteworthy features.
I will deep dive into some optimizations which promise to provide major performance gains. Support for ACID tables has also improved considerably. Although some of these features and enhancements are not novel but have existed for years in other DB systems, implementing them on Hive poses some unique challenges and results in lessons which are generally applicable in many other contexts. I will also provide a glimpse of what is expected to come in near future.
Speaker: Ashutosh Chauhan, Engineering Manager, Hortonworks
Advanced Stream Processing with Flink and Pulsar - Pulsar Summit NA 2021 KeynoteStreamNative
In this talk, Till Rohrmann and Addison Higham discuss how Flink allows for ambitious stream processing workflows and how Pulsar and Flink enable new capabilities that push forward the state-of-the-art in streaming. They will also share upcoming features and new capabilities in the integrations between Flink and Pulsar and how these two communities are working together to truly advance the power of stream processing.
This presentation about HBase will help you understand what is HBase, what are the applications of HBase, how is HBase is different from RDBMS, what is HBase Storage, what are the architectural components of HBase and at the end, we will also look at some of the HBase commands using a demo. HBase is an essential part of the Hadoop ecosystem. It is a column-oriented database management system derived from Google’s NoSQL database Bigtable that runs on top of HDFS. After watching this video, you will know how to store and process large datasets using HBase. Now, let us get started and understand HBase and what it is used for.
Below topics are explained in this HBase presentation:
1. What is HBase?
2. HBase Use Case
3. Applications of HBase
4. HBase vs RDBMS
5. HBase Storage
6. HBase Architectural Components
What is this Big Data Hadoop training course about?
Simplilearn’s Big Data Hadoop training course lets you master the concepts of the Hadoop framework and prepares you for Cloudera’s CCA175 Big data certification. The Big Data Hadoop and Spark developer course have been designed to impart in-depth knowledge of Big Data processing using Hadoop and Spark. The course is packed with real-life projects and case studies to be executed in the CloudLab.
What are the course objectives?
This course will enable you to:
1. Understand the different components of the Hadoop ecosystem such as Hadoop 2.7, Yarn, MapReduce, Pig, Hive, Impala, HBase, Sqoop, Flume, and Apache Spark
2. Understand Hadoop Distributed File System (HDFS) and YARN as well as their architecture, and learn how to work with them for storage and resource management
3. Understand MapReduce and its characteristics, and assimilate some advanced MapReduce concepts
4. Get an overview of Sqoop and Flume and describe how to ingest data using them
5. Create database and tables in Hive and Impala, understand HBase, and use Hive and Impala for partitioning
6. Understand different types of file formats, Avro Schema, using Arvo with Hive, and Sqoop and Schema evolution
7. Understand Flume, Flume architecture, sources, flume sinks, channels, and flume configurations
8. Understand HBase, its architecture, data storage, and working with HBase. You will also understand the difference between HBase and RDBMS
9. Gain a working knowledge of Pig and its components
10. Do functional programming in Spark
11. Understand resilient distribution datasets (RDD) in detail
12. Implement and build Spark applications
13. Gain an in-depth understanding of parallel processing in Spark and Spark RDD optimization techniques
14. Understand the common use-cases of Spark and the various interactive algorithms
15. Learn Spark SQL, creating, transforming, and querying Data frames
Learn more at https://www.simplilearn.com/big-data-and-analytics/big-data-and-hadoop-training
ΚΤΣ 2016 - Μέθοδοι δοκιμών και ελέγχου ποιότητας υλικώνAlexander Tsiolis
Ποιες δοκιμές χρειάζεται να γνωρίζει όποιος ασχολείται με σκυρόδεμα, από την σκοπιά του νέου ΚΤΣ. Δοκιμές και πρότυπα για αδρανή, νωπό και σκληρυμένο σκυρόδεμα καθώς και σύντομη αναφορά στην δειγματοληψία υφιστάμενων κατασκευών και τις χρησιμοποιούμενες μεθόδους μη καταστροφικών μεθόδων.
Speaker: Jean-Daniel Cryans (Cloudera)
HBase Replication has come a long way since its inception in HBase 0.89 almost four years ago. Today, master-master and cyclic replication setups are supported; many bug fixes and new features like log compression, per-family peers configuration, and throttling have been added; and a major refactoring has been done. This presentation will recap the work done during the past four years, present a few use cases that are currently in production, and take a look at the roadmap.
Orchestration tool roundup kubernetes vs. docker vs. heat vs. terra form vs...Nati Shalom
Video recording: https://www.youtube.com/watch?v=tGlIgUeoGz8
It’s no news that containers represent a portable unit of deployment, and OpenStack has proven an ideal environment for running container workloads. However, where it usually becomes more complex is that many times an application is often built out of multiple containers. What’s more, setting up a cluster of container images can be fairly cumbersome because you need to make one container aware of another and expose intimate details that are required for them to communicate which is not trivial especially if they’re not on the same host.
These scenarios have instigated the demand for some kind of orchestrator. The list of container orchestrators is growing fairly fast. This session will compare the different orchestation projects out there - from Heat to Kubernetes to TOSCA - and help you choose the right tool for the job.
Session link from teh summit: https://openstacksummitmay2015vancouver.sched.org/event/abd484e0dedcb9774edda1548ad47518#.VV5eh5NViko
Parallel builds in Eclipse IDE workspace - EclispeCon Europe 2018Mickael Istria
https://www.eclipsecon.org/europe2018/sessions/parallel-builds-eclipse-ide-workspace-value-making-and-optimal-usages
Eclipse Platform 4.8 introduced support for parallel builds in the workspace. This feature is placed at the lowest level of the workspace so it can easily be profitable to any adopter without specific adoption effort, besides respecting usual good practices. This can in theory turn the Eclipse workspace into one of the fastest polyglot build engines in developer world.
In this presentation, we’ll explain what is the value to expect from this parallel builds (mostly performance gain and faster availability of project output).
Then we’ll also get into the details of how all that is implemented, reminding ourselves some basics of dependency resolution, graph processing and scheduling rules. We’ll remind the important best practices of builders and scheduling in the Eclipse resource model that are necessary to reach best results.
And we’ll go through a few case studies of project builders to highlight how those may be improved to take advantage of parallel builds, and get into deep details of how this can be enabled in JDT and other important Eclipse projects.
After this presentation, plugin developers will be empowered to make their project builders take advantage of parallel builds in Eclipse IDE, and may get ideas about really using this feature as an alternative build engine to speed up their headless builds as well.
This document summarizes a presentation about HBase storage internals and future developments. It discusses how HBase provides random read/write access on HDFS using tables, regions, and region servers. It describes the write path involving the client, master, and region servers as well as the read path. It also covers topics like snapshots, compactions, and future plans to improve encryption, security, write-ahead logs, and compaction policies.
This document summarizes research on isomorphism in Nahua dialectal morphology. It makes three main points:
1. Modern Nahua dialects are becoming more regular and less irregular compared to older dialects like Classical Nahuatl.
2. Some features identified as characteristic of "Urban Nahuatl" by Canger may actually be older features that predate the Aztec period.
3. Irregularities in Classical Nahuatl could date back to when the Mexica people were nomadic before founding Tenochtitlan.
The document examines specific morphological examples like reflexive prefixes and the verb 'to go' to support these arguments. It aims to re-evaluate the idea that Urban
The document provides an introduction to NoSQL and HBase. It discusses what NoSQL is, the different types of NoSQL databases, and compares NoSQL to SQL databases. It then focuses on HBase, describing its architecture and components like HMaster, regionservers, Zookeeper. It explains how HBase stores and retrieves data, the write process involving memstores and compaction. It also covers HBase shell commands for creating, inserting, querying and deleting data.
Accelerating Shuffle: A Tailor-Made RDMA Solution for Apache Spark with Yuval...Spark Summit
The opportunity in accelerating Spark by improving its network data transfer facilities has been under much debate in the last few years. RDMA (remote direct memory access) is a network acceleration technology that is very prominent in the HPC (high-performance computing) world, but has not yet made its way to mainstream Apache Spark. Proper implementation of RDMA in network-oriented applications can improve scalability, throughput, latency and CPU utilization. In this talk we are going to present a new RDMA solution for Apache Spark that shows amazing improvements in multiple Spark use cases. The solution is under development in our labs, and is going to be released to the public as an open-source plug-in.
This document provides an overview of optimizing and accelerating Ceph storage on Arm64 architecture. It discusses optimizations made to Ceph common libraries like UTF-8 handling and CRC calculation. It also covers integrating Ceph with technologies like SPDK and Seastar to improve performance. Benchmarks show performance gains from using 64K kernel pages and offloading storage functions to userspace frameworks. Overall, the document aims to improve the Ceph storage ecosystem on Arm servers through software and hardware optimizations.
Triplewave: a step towards RDF Stream Processing on the WebDaniele Dell'Aglio
The slides of my talk at INSIGHT Centre for Data Analytics (in NUI Galway) where I presented TripleWave (http://streamreasoning.github.io/TripleWave/), an open-source framework to create and publish streams of RDF data.
Continuous Integration and Deployment Patterns for MagentoAOE
Fabrizio Branca (@fbrnc) speaking about "Continuous Integration and Deployment Patterns for Magento" at the Meet Magento Conference 2015 in New York City (#mm15ny)
Jay Kreps on Project Voldemort Scaling Simple Storage At LinkedInLinkedIn
Jay Kreps on Project Voldemort Scaling Simple Storage At LinkedIn. This was a presentation made at QCon 2009 and is embedded on LinkedIn's blog - http://blog.linkedin.com/
This document provides an overview of big data. It begins with definitions of big data and its key characteristics, including volume, velocity, and variety. It then discusses how big data is stored, selected, and processed. Examples of big data sources and tools are provided. The document outlines several applications of big data across different industries like healthcare, manufacturing, and retail. It also discusses risks of big data like privacy issues and costs. The future of big data is presented, with projections that the big data market will grow significantly in coming years. In closing, references are provided for additional information on big data.
This document is a Labour Market Impact Assessment (LMIA) approval letter for Company Inc. to hire 1 foreign worker as a Chief Executive Officer & Director in Surrey, British Columbia. The LMIA is valid until May 12, 2021 and allows the named foreign worker(s) to apply for a Canadian work permit. The letter outlines the employer's responsibilities to comply with immigration regulations and notes the LMIA can be revoked if inaccurate information was provided or if the hiring has a negative impact on the Canadian labour market.
This document provides an overview of HazelCast IMDG (In-Memory Data Grid), which is middleware software that manages objects across distributed servers in RAM, enabling scaling and fault tolerance. It discusses cache access patterns, cache types, use cases for HazelCast including scaling applications and sharing data across clusters, features like dynamic clustering and distributed data structures, data partitioning, and configurations. It also covers advanced techniques, alternatives to HazelCast like Redis, and performance comparisons.
Percorso assistenziale del bambino con bisogni complessi. Dalla dimissione ospedaliera alla presa in carico da parte dei servizi riabilitativi
IL BAMBINO CON PATOLOGIA NEURODEGENERATIVA: LA DIMENSIONE DEL PROBLEMA
Dott.Giovanni Ciana, Centro Coordinamento Regionale Malattie Rare FVG
Udine,2 dicembre 2013
This document summarizes a presentation about HBase storage internals and future developments. It discusses how HBase provides random read/write access on HDFS using tables, regions, and region servers. It describes the write path involving the client, master, and region servers as well as the read path. It also covers topics like snapshots, compactions, and future plans to improve encryption, security, write-ahead logs, and compaction policies.
This document summarizes research on isomorphism in Nahua dialectal morphology. It makes three main points:
1. Modern Nahua dialects are becoming more regular and less irregular compared to older dialects like Classical Nahuatl.
2. Some features identified as characteristic of "Urban Nahuatl" by Canger may actually be older features that predate the Aztec period.
3. Irregularities in Classical Nahuatl could date back to when the Mexica people were nomadic before founding Tenochtitlan.
The document examines specific morphological examples like reflexive prefixes and the verb 'to go' to support these arguments. It aims to re-evaluate the idea that Urban
The document provides an introduction to NoSQL and HBase. It discusses what NoSQL is, the different types of NoSQL databases, and compares NoSQL to SQL databases. It then focuses on HBase, describing its architecture and components like HMaster, regionservers, Zookeeper. It explains how HBase stores and retrieves data, the write process involving memstores and compaction. It also covers HBase shell commands for creating, inserting, querying and deleting data.
Accelerating Shuffle: A Tailor-Made RDMA Solution for Apache Spark with Yuval...Spark Summit
The opportunity in accelerating Spark by improving its network data transfer facilities has been under much debate in the last few years. RDMA (remote direct memory access) is a network acceleration technology that is very prominent in the HPC (high-performance computing) world, but has not yet made its way to mainstream Apache Spark. Proper implementation of RDMA in network-oriented applications can improve scalability, throughput, latency and CPU utilization. In this talk we are going to present a new RDMA solution for Apache Spark that shows amazing improvements in multiple Spark use cases. The solution is under development in our labs, and is going to be released to the public as an open-source plug-in.
This document provides an overview of optimizing and accelerating Ceph storage on Arm64 architecture. It discusses optimizations made to Ceph common libraries like UTF-8 handling and CRC calculation. It also covers integrating Ceph with technologies like SPDK and Seastar to improve performance. Benchmarks show performance gains from using 64K kernel pages and offloading storage functions to userspace frameworks. Overall, the document aims to improve the Ceph storage ecosystem on Arm servers through software and hardware optimizations.
Triplewave: a step towards RDF Stream Processing on the WebDaniele Dell'Aglio
The slides of my talk at INSIGHT Centre for Data Analytics (in NUI Galway) where I presented TripleWave (http://streamreasoning.github.io/TripleWave/), an open-source framework to create and publish streams of RDF data.
Continuous Integration and Deployment Patterns for MagentoAOE
Fabrizio Branca (@fbrnc) speaking about "Continuous Integration and Deployment Patterns for Magento" at the Meet Magento Conference 2015 in New York City (#mm15ny)
Jay Kreps on Project Voldemort Scaling Simple Storage At LinkedInLinkedIn
Jay Kreps on Project Voldemort Scaling Simple Storage At LinkedIn. This was a presentation made at QCon 2009 and is embedded on LinkedIn's blog - http://blog.linkedin.com/
This document provides an overview of big data. It begins with definitions of big data and its key characteristics, including volume, velocity, and variety. It then discusses how big data is stored, selected, and processed. Examples of big data sources and tools are provided. The document outlines several applications of big data across different industries like healthcare, manufacturing, and retail. It also discusses risks of big data like privacy issues and costs. The future of big data is presented, with projections that the big data market will grow significantly in coming years. In closing, references are provided for additional information on big data.
This document is a Labour Market Impact Assessment (LMIA) approval letter for Company Inc. to hire 1 foreign worker as a Chief Executive Officer & Director in Surrey, British Columbia. The LMIA is valid until May 12, 2021 and allows the named foreign worker(s) to apply for a Canadian work permit. The letter outlines the employer's responsibilities to comply with immigration regulations and notes the LMIA can be revoked if inaccurate information was provided or if the hiring has a negative impact on the Canadian labour market.
This document provides an overview of HazelCast IMDG (In-Memory Data Grid), which is middleware software that manages objects across distributed servers in RAM, enabling scaling and fault tolerance. It discusses cache access patterns, cache types, use cases for HazelCast including scaling applications and sharing data across clusters, features like dynamic clustering and distributed data structures, data partitioning, and configurations. It also covers advanced techniques, alternatives to HazelCast like Redis, and performance comparisons.
Percorso assistenziale del bambino con bisogni complessi. Dalla dimissione ospedaliera alla presa in carico da parte dei servizi riabilitativi
IL BAMBINO CON PATOLOGIA NEURODEGENERATIVA: LA DIMENSIONE DEL PROBLEMA
Dott.Giovanni Ciana, Centro Coordinamento Regionale Malattie Rare FVG
Udine,2 dicembre 2013
Euro Huntington Disease Network: dagli studi osservazionali alla terapia geni...Aich Roma Onlus
ROMA DICEMBRE 2015
DAL SILENZIAMENTO GENICO ALLE SPERIMENTAZIONI FARMACOLOGICHE: IL FRONTE DELLA RICERCA SULLA MALATTIA DI HUNTIGTON, IN ITALIA E ALL’ESTERO
Cattaneo le urgenze in ematologia 21 maggio 2011cmid
This document discusses antiplatelet drugs used to prevent blood clots, including aspirin, clopidogrel, prasugrel, ticagrelor, cangrelor, and elinogrel. It summarizes the mechanisms of action, metabolism, efficacy, and safety profiles of these drugs. Resistance or variable response to certain antiplatelet drugs is also addressed. The ideal properties of an antithrombotic agent are outlined. Later generation drugs like prasugrel, ticagrelor, and cangrelor demonstrate more consistent platelet inhibition compared to clopidogrel. However, some such as ticagrelor and elinogrel are associated with higher rates of dyspnea.
Linfedema torino 4 e 5 marzo gaal palma [modalità compatibilità]
Matarazzo patrizia trattamento con gh nella sindrome di noonan pro e contro-torino gennaio 2011-14à conv
1. 14° CONVEGNO
PATOLOGIA IMMUNE E MALATTIE ORFANE 2010
Torino, 20-22 gennaio 2011
Trattamento con GH nella sindrome di Noonan: pro e contro
Patrizia Matarazzo (Torino)
SCDU Endocrinologia-Diabetologia
Ospedale Infantile Regina Margherita
2. La Sindrome di Noonan
ETEROGENEITA’ CLINICA
Bassa statura
Aspetto tipico del volto
Sordità
Problemi visivi
Cardiopatia congenita: stenosi valvolare polmonare, cardiomiopatia ipertrofica
Deformità toracica. Linfedema. Ipertermia maligna.
Diatesi emorragica; splenomegalia. da Romano A: Pediatrics 2010
Criptorchidismo
Disabilità nell’apprendimento; ipotonia poi iperattività; difficoltà nell’alimentarsi da
lattanti
Incidenza: 1:1000-2500 nati vivi
Eredità autosomica dominante, ma sporadica nel 60% dei casi
3. La Sindrome di Noonan
ETEROGENEITA’ GENETICA
PTPN11 gene mutations sul braccio lungo del cromosoma 12 (circa 50% dei casi)
K-ras/MAPK mutations (1-3%)
SOS1 mutations (10%)
RAF1 mutations (3-17%) 20-25% dei casi
MEK1 mutations
La diagnosi di NS è solo clinica in molti casi!
Fenotipo faciale tipico + almeno 2 degli altri segni
MAPK = mitogen-activated protein kinase
4. La Sindrome di Noonan
ETEROGENEITA’ NEGLI ASPETTI INTELLETTIVI,
PSICOLOGICI E SOCIALI
Ritardo mentale lieve: 1/3 dei soggetti
Disturbi del comportamento
Disturbi dell’attenzione/iperattività
Problemi di competenza sociale e di comunicazione
Problemi di percezione emotiva di sé e degli altri e di verbalizzazione dei
sentimenti “no words for feelings” (alexitimia)
Ridotta vita sociale ma qualità di vita “soddisfacente”
Disturbi psichiatrici rari (case reports) in genere in soggetti con ritardo mentale
Verhoeven W. Am J Med Genet Part A 2008; 146A: 191-196
Wingbermuhle E. Horm Res 2009; 72 (Suppl 2): 15-23
5. La Sindrome di Noonan
LA CRESCITA SPONTANEA
Alla nascita, peso e altezza normali
Durante l’infanzia, ritardo di età ossea di circa 2 anni
Possibile ritardo puberale (più frequente nei maschi)
Statura definitiva : 145-162.5 cm nei maschi e 135-151 cm nelle femmine, i più alti si collocano a -2 sds
rispetto alla popolazione generale
FEMMINE MASCHI
6. La Sindrome di Noonan
PROBLEMATICA DEL TRATTAMENTO CON
ORMONE DELLA CRESCITA
•Patogenesi della bassa statura
in termini di velocità di crescita
•Risultati del trattamento in termini di statura definitiva
sugli aspetti metabolici
•Rischi del trattamento GH e cuore
GH e tumori
FDA
• La normativa attuale: EMEA
AIFA
problematiche di salute del bambino/a
• Chi trattare con GH e perché aspetto psicologico del bambino/a
motivazione della famiglia
7. La Sindrome di Noonan
PATOGENESI DELLA BASSA STATURA
Disfunzione dell’asse GH-IGF: GHD classico (raro); neurosecretory dysfunction;
IGF ridotte (specie nelle NS PTPN11 pos.)
Insensibilità al GH : la mutazione PTPN11/SHP-2 causa modesta resistenza al
GH attraverso un difetto di segnale post-recettoriale (in vitro)
Problemi associati: cardiopatia, ipotiroidismo in tiroidite autoimmune, alterazioni
del timing puberale (pubertà tarda)
Parziale displasia scheletrica: tronco sproporzionato rispetto agli arti, scoliosi,
cifosi, sinostosi radio-ulnare
8. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
Limiti degli studi:
Studi osservazionali o case reports
Non studi randomizzati caso-controllo
I controlli sono i vecchi pazienti non trattati con GH
Eterogeneità delle casistiche (medico che fa diagnosi, quadro clinico più o meno
evidente, età alla diagnosi, età all’inizio del trattamento)
Dose di GH impiegata e durata del trattamento
Interpretazione dei risultati non uniforme (SDS HV, SDS H, ∆SDS H, curve di
crescita della popolazione generale o curve specifiche per la NS).
9. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
Romano A. Noonan Syndrome: clinical features, diagnosis, and management guidelines.
Pediatrics 2010; Vol. 126 N° 746-759; Oct 2010
4,
Ross J. Factors influencing the one- and two-year growth response in children treated with growth hormone: analysis
from an observational study. Int J Pediatr Endocrinol. 2010;2010:494656. Epub 2010 Oct 12.
Binder G Response to growth hormone in short children with Noonan syndrome: correlation to genotype..
Horm Res. 2009 Dec;72 Suppl 2:52-6. Epub 2009 Dec 22.
Dahlgren J. GH therapy in Noonan syndrome: Review of final height data.
Horm Res. 2009 Dec;72 Suppl 2:46-8. Epub 2009 Dec 22
Ranke MB. Noonan syndrome: growth to growth hormone - the experience of observational studies.
Horm Res. 2009 Dec;72 Suppl 2:36-40. Epub 2009 Dec 22
Savage MO Abnormal growth in noonan syndrome: the challenge of optimal therapy.
Pediatr Endocrinol Rev. 2009 Jun;6 Suppl 4:523-8
Romano AA Growth response, near-adult height, and patterns of growth and puberty in patients with noonan
syndrome treated with growth hormone. J Clin Endocrinol Metab. 2009 Jul;94(7):2338-44. Epub 2009 Apr 28.
Castinetti F Noonan's syndrome and growth hormone treatment
Ann Endocrinol (Paris). 2008 Sep;69 Suppl 1:S2-5
10. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
Padidela R Abnormal growth in noonan syndrome: genetic and endocrine features and optimal treatment.
Horm Res. 2008;70(3):129-36. Epub 2008 Jul 29
Noordam C Long-term GH treatment improves adult height in children with Noonan syndrome with and without
mutations in protein tyrosine phosphatase, non-receptor-type 11.
Eur J Endocrinol. 2008 Sep;159(3):203-8. Epub 2008 Jun 18.
Raaijmakers R Response to growth hormone treatment and final height in Noonan syndrome in a large cohort of
patients in the KIGS database.
J Pediatr Endocrinol Metab. 2008 Mar;21(3):267-73
Walton-Betancourth S Excellent growth response to growth hormone therapy in a child with PTPN11-negative Noonan
syndrome and features of growth hormone resistance. J Endocrinol Invest. 2007 May;30(5):439-41
Osio D Improved final height with long-term growth hormone treatment in Noonan syndrome.
Acta Paediatr. 2005 Sep;94(9):1232-7.
Binder G PTPN11 mutations are associated with mild growth hormone resistance in individuals with Noonan
syndrome. J Clin Endocrinol Metab. 2005 Sep;90(9):5377-81. Epub 2005 Jun 28.
Ferreira LV PTPN11 (protein tyrosine phosphatase, nonreceptor type 11) mutations and response to growth hormone
therapy in children with Noonan syndrome.
J Clin Endocrinol Metab. 2005 Sep;90(9):5156-60. Epub 2005 Jun 14.
Brown DC Growth hormone therapy in Noonan's syndrome: non-cardiomyopathic congenital heart disease does not
adversely affect growth improvement.
J Pediatr Endocrinol Metab. 2002 Jun;15(6):851-2
11. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
Noordam K The relationship between clinical severity of Noonan's syndrome and growth, growth hormone (GH)
secretion and response to GH treatment. J Pediatr Endocrinol Metab. 2002 Feb;15(2):175-80
Noordam C Bone mineral density and body composition in Noonan's syndrome: effects of growth hormone
treatment. J Pediatr Endocrinol Metab. 2002 Jan;15(1):81-7
.Noordam C Growth hormone treatment in children with Noonan's syndrome: four year results of a partly controlled
trial. Acta Paediatr. 2001 Aug;90(8):889-94
MacFarlane CE Growth hormone therapy and growth in children with Noonan's syndrome: results of 3 years' follow-up.
J Clin Endocrinol Metab. 2001 May;86(5):1953-6.
Kirk JM Short stature in Noonan syndrome: response to growth hormone therapy.
Arch Dis Child. 2001 May;84(5):440-3.
12. STUDIES ON GH THERAPY IN NS RISULTATI A BREVE TERMINE
13. La Sindrome di Noonan
RISULTATI A BREVE TERMINE
Prepubertal gain in height SDS in
all patients with Noonan syndrome
according to Noonan reference
values. Results are expressed as
mean+SEM for the number of
patients in parentheses. * p50.001
compared with height at start of
GH treatment
Osio D. Acta Paed 2005; 94: 1232-37
14. La Sindrome di Noonan
RISULTATI A MEDIO E LUNGO TERMINE
Short stature in Noonan Syndrome: Results of growth hormone treatment in KIGS. Barto J Otten, Kees Noordam.
In Growth Hormone therapy in Pediatrics-20 years of KIGS, Ranke MB, Price DA, Reiter EO eds; pp 347-355
429 NS
85 trattati con GH 38 stature adulte definitive
per 3 anni (19 M, 19 F)
HV 4.8 (t 0’) 7.3 (t 1) 5.9 (t 2)
HSDS t0 t1 t2 t3 HSDS t0 NAH
PRADER -3.6 -3.1 -2.8 -2.4 PRADER -3.5 -2.4
NOONAN -1.2 -0.6 -0.3 -0.2 NOONAN -1.1 0.0
KIGS (Pfizer Endocrine Care)
15. La Sindrome di Noonan
RISULTATI A BREVE TERMINE
Factors Influencing the One- and Two-Year Growth Response in Children Treated with Growth Hormone: Analysis
from an Observational Study. Judith Ross,1* Peter A. Lee,2 Robert Gut,3 and John Germak3
Int J Pediatr Endocrinol. 2010; 2010: 494656
Children with short stature associated with Noonan syndrome (NS): 79 pazienti (58 M, 21 F)
Age, yr 9.6 ± 3.76
HSDS −2.7 ± 0.70
IGF-I SDS −2.8 ± 1.27
Bone age, yrs 8.0 ± 3.75
Baseline GH dose, 0.33
mg/kg/wks
Peak GH, ng/mL 10.8 ± 9.13
Baseline HSDS (males) −2.6 ± 0.69
Baseline HSDS (female) −2.9 ± 0.69
ANSWER Program Registry (Novo)
16. La Sindrome di Noonan
RISULTATI A BREVE TERMINE
Factors Influencing the One- and Two-Year Growth Response in Children Treated with Growth Hormone: Analysis
from an Observational Study. Judith Ross,1* Peter A. Lee,2 Robert Gut,3 and John Germak3
Int J Pediatr Endocrinol. 2010; 2010: 494656
Children with short stature associated with Noonan syndrome (NS): 79 pazienti (58 M, 21 F)
Risultati di 2 anni di trattamento con GH nella NS:
-non differenze tra M e F
-∆HSDS significativamente maggiore nei pazienti più giovani a inizio terapia
-∆HSDS a 1 anno: 0.4 sds
-∆HSDS alla fine dei 2 anni: 0.6 sds
ANSWER Program Registry (Novo)
17. La Sindrome di Noonan
RISULTATI A LUNGO TERMINE: STATURE FINALI
Pediatrics 2010; Vol. 126 N° 746-759; Oct 2010
4,
Noonan Syndrome: clinical features, diagnosis, and management guidelines.
Romano A, Allanson J, Dahlgren J, Gelb B, Hall B, Pierpont , Roberts A, Robinson W, Takemoto C,
Noonan J.
Reference Patients, n Baseline Baseline GH dose Duration Delta Height Height
(n, Females) Age, y SDS mg/kg/wk Therapy, y SDS Gain
Near Final Hight
Municchi 4 (4) 13.5 (–0.6) 0.19 for 1 y, 3.5 (0.48) NA
1995 then 0.31
Kirk 2001 10 (4) 12 –3.1 (–0.7) 0.30 5.3 3.1 cm
Osio 2005 18 (11) 8.6 (boys), –2.9 (–0.3) 0.23 (n = 7.5 1.7 13 cm
7.7 (girls) 10), 0.46 (n (boys), 9.8
= 15) for 2 y, cm(girls)
then dose
titrationb
Raaijmakers 24 (NA) 10.2 –3.24 0.24 7.59 0.61 (0.97) 4-5 cm.
KIGS 2008 (median) (median) (median (median)
[range:
0.17–0.77])
Noordam 29 (8) 11 –2.8 (0.0) 0.35 6.4 (median) 1.3 (1.3) 9.5 cm
2008 (boys), 9.0
cm (girls)
Romano 65 (30) 11.6 –3.5 0.33 5.6 1.4 10.9 cm
2009 (boys) 9.2
cm (girls)
18. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
.Risultati:
-guadagno in H varia da 0.6 a 2 sds ( = 4-13 cm) e dipende soprattutto dall’età
di inizio del trattamento (più precoce il trattamento, migliori i risultati) e
dall’età di inizio della pubertà (più tardiva la pubertà, migliori i risultati)
-nessuno studio documenta una correlazione tra dose impiegata e guadagno
in H definitiva
-non correlazione tra guadagno in H definitiva e presenza della mutazione
PTPN11
-i maschi guadagnano più delle femmine, verosimilmente per una più lenta
progressione dei caratteri puberali (che in alcuni è dovuta alla disfunzione
testicolare)
Dahlgren J. Horm Res. 2009 Dec;72 Suppl 2:46-8. Epub 2009 Dec 22.
19. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
Risultati:
-discrepanza nei risultati dipende dalla diversa selezione dei pazienti e metodi
di valutazione applicati: nel KIGS diagnosi effettuata da Clinici Pediatri, negli
altri 2 studi da Genetisti o Endocrinologi esperti nella NS; nel KIGS H inizio
terapia era più compromessa che negli altri 2 studi; compliance al trattamento
migliore nello studio svedese; età all’inizio del trattamento sembra il fattore
prognostico più importante
-non segnalato peggioramento della cardiopatia
-non segnalato peggioramento della scoliosi
-nello studio svedese, 1 paziente ha sviluppato un linfoma dopo 3 anni di
trattamento con GH (peraltro reintrodotto dopo la remissione di malattia)
Westphal O. Horm Res. 2009 Dec;72 Suppl 2:41-5. Epub 2009 Dec 22.
20. La Sindrome di Noonan
RISULTATI DEL TRATTAMENTO CON GH
ASPETTO METABOLICO
Hb1, trigliceridi, colesterolo non si modificano in 3 anni di trattamento (McFarlane 2001)
BMD a livello corticale migliora nei 2 anni di trattamento rispetto al valore basale.
Massa magra e contenuto di acqua libera corporea aumentano nei primi 3 mesi di
trattamento con GH e poi poco successivamente. La % di massa grassa decresce
nei primi 6 mesi (effetto lipolitico) poi ri-aumenta (Noordam 2002)
21. La Sindrome di Noonan
RISCHI DEL TRATTAMENTO CON GH
GH e cuore: a breve termine, il GH ha effetto positivo sul cuore attraverso il rilascio di ossido
nitrico (NO), che mantiene normale reattività vascolare e normali resistenze vascolari periferiche.
Sul lungo termine, o per dosi elevate, il GH può indurre ispessimento della parete ventricolare o
aumentata incidenza di eventi cardiaci acuti
2 studi prospettici caso-controllo (NS trattate con GH e non) NON hanno documentato differente
progressione della cardiopatia tra i 2 gruppi né maggiore incidenza di eventi acuti
Negli studi di sorveglianza post-marketing, il KIGS negli 85 NS trattati con GH ha segnalato 7 eventi
cardiaci avversi (1 cianosi, 3 aritmie, 1 angina pectoris, 1 severa ipertrofia VS, 1 severa
cardiomiopatia necessitante il trapianto cardiaco), ritenendo peraltro “poco probabile” la
correlazione tra questi eventi e la terapia con GH.
Nei soggetti con NS affetti da cardiomiopatia severa (spessore parete ventricolare > 1
cm) è SCONSIGLIATO l’impiego del GH.
Nei soggetti con mutazione RAF1 è preferibile non impiegarlo per rischio di
cardiopatia severa evolutiva
I soggetti con NS e cardiopatia sottoposti a trattamento con GH vanno attentamente
e regolarmente monitorizzati sotto il profilo clinico, ECGrafico ed ecocardiografico
(ogni 6 mesi durante il trattamento, 1 volta/anno nei 3 anni successivi)
Noordam C. Horm Res 2009; 7 (Suppl 2): 49-51
22. La Sindrome di Noonan
RISCHI DEL TRATTAMENTO CON GH
Le persone con NS sembrano esposte a maggior rischio di neoplasie
Mutazioni germinali dei geni RAS-MAP costituiscono un pathway mitogeno
iperattivo, che si traduce in un rischio aumentato di sviluppare tumori.
I soggetti affetti da NS hanno una maggiore incidenza di leucemia
mielomonocitica giovanile (JMML) che peraltro decorre con
andamento più benigno (regressione spontanea) che nei soggetti
non-Noonan (30 neonati descritti con NS e PTPN11 pos), LLA,
LMA.
Altri tumori segnalati sporadicamente nella NS: Neuroblastoma,
rabdomiosarcoma, tumori testicolari (1 carcinoma, 1 seminoma, 1
t. cellule del Sertoli), carcinoma mammario
Mortalità nella NS è dovuta alle cardiopatie complesse (ipertrofia
miocardica), alla linfopatia severa e sue complicanze, alle neoplasie
Hasle H. Horm Res 2009; 7 (Suppl 2): 8-14
23. Fattori di rischio per sviluppo di tumori in
soggetti trattati con GH
Trattamento Precedenti Chemio e/o Sindromi genetiche
con GH Neoplasie radioterapia o alterazioni
cromosomiche
S. Fanconi, S. Down,
Retinoblastoma,T.Wilm’s, Carney complex, NF1,
LLA, tumori SNC Noonan
4-45 Gy LH Ca. mammella
Ciclofosfamide Clorambucil
Non sussiste un rischio aggiuntivo di sviluppare una leucemia de novo in
soggetti trattati con GH che non abbiano fattori di rischio predisponenti
Rischio aumentato di Ca. colo-rettale (Swerdlow and Coll. 2002), correlazione tra
livelli di IGF e Ca. mammella, prostata, polmone; LH
24. La Sindrome di Noonan
LA NORMATIVA ATTUALE SUL TRATTAMENTO CON GH
The pediatric indications for recombinant human GH approved by:
•FDA (the United States Food and Drug Administration) include treatment of children with
growth failure due to GH deficiency; children with short stature associated with Noonan
syndrome (NS), Turner syndrome (TS), and Prader-Willi syndrome (PWS); children with short stature born
small for gestational age (SGA) who have not reached normal growth range by age 2–4 years; short stature with
homeobox-containing gene deficiency; children with chronic renal insufficiency; children with idiopathic short
stature (ISS) who are >2.25 standard deviations (SD) below the mean in height and who are unlikely to catch up in
height .
GHD, sindrome di Turner citogeneticamente dimostrata
•AIFA, nota 39: deficit staturale nell’insufficienza renale cronica
soggetti prepuberi affetti dalla sindrome di Prader Willi
Small for Gestational Age con età uguale o superiore a 4 anni
Età evolutiva
In soggetti con caratteristiche clinico-auxologiche in accordo con il punto I e con
normale secrezione di GH (punto II), la terapia può essere effettuata solo se
autorizzata dalla Commissione Regionale preposta alla sorveglianza epidemiologica
ed al monitoraggio dell’appropriatezza del trattamento con GH.
25. La Sindrome di Noonan
LA NORMATIVA ATTUALE SUL TRATTAMENTO CON GH
EMEA (European Medicines Agency)
Update on somatropin-containing medicines
Review of somatropin officially started (16.12.2010)
The European Medicines Agency’s Committee for Medicinal Products for Human Use (CHMP) has started a
review of the safety of somatropin-containing medicines authorised centrally or by national procedures in the
European Union (EU). The CHMP will look into all available data on somatropin to reassess the benefit-risk
balance of these medicines.
While this review is ongoing, the CHMP confirms that there is no immediate concern. However, prescribers are
reminded to strictly follow the indications and the approved doses. The maximum recommended dose of
50 µg/kg weight/day for somatropin-containing medicines should not be exceeded.
Somatropin is a human growth hormone, manufactured using recombinant DNA technology. It is used to treat a
number of conditions associated with a lack of growth hormone and short stature. This includes children who fail
to grow due to a lack of growth hormone, Turner syndrome or chronic renal insufficiency.
This review was initiated further to information received from the French medicines agency on a long-term
epidemiological study in patients treated during childhood for idiopathic lack of growth hormone and idiopathic or
gestational short stature with somatropin-containing medicines. The study results suggest an increased risk of
mortality with somatropin therapy compared with the general population. This study on the safety and
appropriateness of growth hormone treatments is funded by the European Commission and conducted by a
European consortium of paediatric endocrinologists, epidemiologists and biostatisticians, involving eight EU
countries. The study is still ongoing and further results are expected in the future.
Further updates on this review will be made as appropriate
26. La Sindrome di Noonan
CHI TRATTARE CON GH E PERCHE’
OCCORRE VALUTARE:
•Problemi medici associati: tipo di cardiopatia, scoliosi, tipo di mutazione,
predisposizione alle neoplasie:
•Problemi psicologici, comportamentali, intellettivi del bambino:
come vive la sua condizione di bassa statura, come questa interferisce
negativamente sul proprio self-esteem, sui rapporti col gruppo dei “pari”:
•Aspettative della famiglia sulla statura del figlio/a, motivazione al
trattamento, conoscenza dei limiti dei risultati, consapevolezza dei rischi:
27. La Sindrome di Noonan
CONCLUSIONI 1. I PRO
•Il trattamento con ormone della crescita migliora la velocità di crescita negli anni
della pre-pubertà dove il disagio psicologico è maggiore. La risposta è massimale
nei primi 2 anni di trattamento. Il fattore predittivo più importante di risposta al
trattamento con GH è l’età di inizio (comunque dopo i 5 anni).
•Il trattamento con ormone della crescita corregge la bassa statura adulta definitiva.
Il guadagno in termini di altezza definitiva è di 4-10 cm dopo 5-7 anni di terapia a
dosi farmacologiche
•Sotto il profilo metabolico il trattamento con GH non sembra aumentare il rischio di
diabete né di dislipidemia.
28. La Sindrome di Noonan
CONCLUSIONI 2. I CONTRO
•Limiti: studi su casistiche piccole, non caso-controllo, non randomizzazione,
diagnosi è spesso solo clinica (possibili errori), studi short-term
•Guadagno in termini di altezza definitiva è di 4-10 cm dopo 5-7 anni di
terapia a dosi farmacologiche
•Possibili rischi in termini di peggioramento della cardiopatia. E’
comunque controindicato nei soggetti con cardiomiopatia ipertrofica
severa. Occorre una sorveglianza cardiologica (visita, ECG, Ecocardio)
stretta e prolungata
•Possibili rischi in termini di sviluppo neoplasie (rischio intrinseco alla NS,
rischio specifico del GH).
29. La Sindrome di Noonan
CONCLUSIONI 3. COME COMPORTARSI
•In attesa di studi più controllati, nella pratica clinica occorre valutare caso per
caso l’indicazione/opportunità del trattamento con GH, tenendo conto degli aspetti
fisici (cardiopatia, scoliosi, problemi ematologici), genetici (tipo di mutazione, se
nota), emotivi, intellettivi, sociali del bambino e della sua famiglia
•Come spesso avviene nelle malattie rare, la scelta deve essere “condivisa”,
“negoziata” tra medico, famiglia, minore e deve scaturire dal bilancio dei pro e dei
contro e delle motivazioni, paure, aspettative di queste tre figure protagoniste.
30. GRAZIE A VOI PER L’ATTENZIONE E GRAZIE AI MIEI COLLEGHI!
31. La Sindrome di Noonan
NOSTRA CASISTICA
H adulta 158 cm
A.D.
GH
GH dopo GHRH+Arg:
picco 2.35 ng/ml
IGF: 87 ng/ml
32. La Sindrome di Noonan
NOSTRA CASISTICA
P.Y.
GH+LHRH-a
GH+LHRH-a
GH dopo Arginina: picco 9.9 ng/ml
GH dopo Clonidina: picco 22 ng/ml
IGF: 129 ng/ml
Gene PTPN11: mutazione Tyr63Cys
33. La Sindrome di Noonan
NOSTRA CASISTICA
Gene PTPN11: mutazione Tyr63Cys.
P.S.
GH
GH
GH dopo Arginina: picco 7.7 ng/ml
GH dopo Clonidina: picco 12 ng/ml
IGF: 55.2 ng/ml