Doctor-Patient-Artificial Intelligence Relations in Smart Healthcare

Introduction

Smart Healthcare leverages the latest mobile and digital advances in E-Health and mHealth, driving the development of smart and connected medical devices. The approach to medicine is also changing with smart trackers and other similar devices, doctors have much more opportunities to constantly monitor patient indicators outside of medical institutions and, accordingly, prevent diseases. In recent years, the term “Smart Medicine” has emerged and is becoming more widely used. By “Smart medicine” we mean intelligent healthcare, which uses the latest mobile and digital achievements in the field of eHealth and mHealth, which encourages the development of smart and connected medical devices that ensure constant monitoring of patient indicators outside of medical institutions and, accordingly, the prevention of diseases. In some cases, this type of monitoring can recognize or predict critical health condition of patients and it can warn health institutions if immediate first aid is needed. The Smart medicine will allow the doctor to quickly communicate with the patient, conduct a remote course of treatment. Through special sensors and chips installed on the human body, the doctor, regardless of the location, will be able to get acquainted with important information about the patient’s health status. For example, the doctor will be able to track body temperature, pulse, respiration rate, blood sugar, and blood pressure.

Together with the concept of “Smart Medicine”, the concept of “Smart Hospital” began to be actively applied. Smart Hospital can be defined as interactive intelligent digital environment that represents the meta-system to manage clinical pathways based on on-line monitoring of vital functions in combination with the operational personnel access and patient’s information (including virtual councils) with the wide use of mobile applications and robotics. Current development of technologies paves a way to the situations when most medical institutions will implement computer science and other contemporary ways of provision of medical support in their activities. This is not only a way to provide better healthcare, but also to minimize costs to uniform digital solutions. Doctor-Patient relations have already changed due the developments brought by Smart medicine. Further developments come to place with the active processes that bring Artificial Intelligence (AI) technologies into the medical field. It should be noted here that AI is not one technology, but rather a collection of them; and thus, specific processes and tasks that they support can vary widely [1]. The numerous uses of AI are already changing and will change even more the relationship between the patient and the doctor. AI systems can be beneficial for doctors and patients, as they can provide them with new tools and opportunities. However, many ethical and legal challenges need to be resolved, such as obtaining informed consent to the use of AI, security, reliability and reliability of data sets, transparency of algorithms, algorithmic fairness and impartiality, data ownership, data confidentiality. We will look at these issues in our paper.

Artificial Intelligence and how it Can Be Useful for Doctors and Patients

Today, artificial intelligence is widely used in various fields of medicine, appropriate systems are developed by many scientists and companies. The usage of AI in medicine has a number of potential benefits to both doctors and patients. Patients can benefit from the application of AI systems from the point they get access to medical services. Today, in the emergency medical centers of hospitals, the order of admission of patients depends on how urgently the patient needs help. New technologies, e.g., special apps with AI system in it, can analyse patient’s symptoms in order to determine the degree of urgency. AI can accelerate the diagnosis process and medical research; thus, patients can have benefits from AI during patient’s care. And even the patients are not in the medical institutions, they can anyway get personalized information and advice, for example, in case AI systems are used in chat-bots and provide answers to patient’s request. Thus, even when the patients are at home, they can use digital tools and modern telemedicine to get necessary advice, answers and prescriptions.

On the other hand, it helps doctors enhance their abilities in provision of healthcare services and solve a variety of problems: • Assistance in making personalised diagnoses and prescribing treatment. For example, AI systems can analyse data from Electronic Health Record (EHR) systems, biosensors, watches, smartphones, conversational interfaces and other instrumentation, software can tailor recommendations by comparing patient data to other effective treatment pathways for similar cohorts [2].

• Real-time data analysis of critically ill patients. This analysis may predict the risks of developing medical complications or conditions such as sepsis and ARDS, or use of clinical and physiological data to aid in the monitoring of patients undergoing ventilatory support [3].

• Assessing the likelihood of complications of diseases. Using AI and machine learning technologies, medical researchers can identify the relationship between the patient’s diseases, the conditions in which he lives, and his habits. Even the state of the environment can help to establish which patients in a given region are at the highest risk. It is also possible to find the most vulnerable regions or segments of the population, which can help to give them recommendations in advance, before serious medical care is required [4].

• Remote first aid. It may include real-time analysis, so as to prevention or notification of possible negative tendencies in their health conditions.

Application of AI can be valuable also in the process of overcoming patient’s non-compliance and absent of proper involvement into the process. As some studies show, less than 25 % of the patient are highly engaged in the healthcare process [5]. This means that patients do not take prescriptions, do not follow instructions from doctors and do not comply with the requirements of doctors. This can significantly influence the quality of the medical service. AI systems are viewed as a way to change patient’s behaviour, as they can analyse and address patient’s needs, by alerting patients at proper times, by providing them with targeted information and content to provoke actions and compliance with recommendations [6].

The application of AI system is Smart medicine is a reality which needs to be assessed from various points of view. However, there’s a very important question which arises when we discuss the creation of any E-health system in the context of Smart medicine: who is the true owner of medical data of a concrete patient? Who is owner of Electronic Health Record of a patient? Who can dispose of them (patient, doctor, clinic, insurance company, employer, or computing service) and to what extent? These are the questions that are important not only from theoretical point of view; they need to be addressed and solved in practice, and necessary legislative amendment might be necessary to regulate in detail all the relations of the parties concerned so as to ensure security and personal health data protection. There are two main types of formation and storage of data about patient health in Electronic Health Record:

- Hospital-oriented system when EHR is owned by a hospital (polyclinic).

- Patient-oriented system when a patient is the owner of his EHR. It is he who decides what to store, where to store and to whom to give access to his EHR.

We considered the first case in our work [7]. In this paper, we pay more attention to the patient-oriented system. Thus, we will discuss the issues that are connected with the said approach, e.g., right of patient to obtain information about his health, to give informed consent to the application of AI, diagnosis and liability.

Patient and his Communication with AI and Doctor in Smart Healthcare

The patient oriented EHR system allows a patient to generate, administer and manage medical data from one central location using online technologies, which makes resource storage, retrieval and sharing extremely efficient. Each patient has absolute control over their medical records and can share medical data with a set of consumers, such as medical report providers, family members and friends. Although it is simple to provide access to EHR to anyone and everyone, there seems to be a number of security and privacy issues. The main cause of concern is whether patients have absolute control over their EHRs [8]. The ideal EHR includes personal medical information from various sources and provides complete and accurate personal medical information via the Internet or portable media, while maintaining security and confidentiality [9]. Cloud servers’ merit specific attention. Many EHR system are transferred to data storage on cloud servers as a result of the advent of cloud computing, which allows for more flexible use of resources and lower operating costs. However, when placing EHR data in the cloud, patients will face privacy issues. External cloud storage systems are often vulnerable to various attacks. It is extremely important to have precise “data access control” that works with untrusted servers to ensure that users (patients) manage their own EHRs. Thus, before storing data in cloud, it is advisable to encrypt it [10]. The EHR owner must choose how to encrypt the data and who has access to it. Only users who have been provided with the decryption key can access the EHR, while the rest of the clients must remain confidential. In addition, the patient should always be able not only to log in, but also to get authorization permission when he believes that it is really required. However, with such extensibility of the EHR system, patient-centered privacy is often at risk. Thus, it may be difficult to ensure proper access to medical information while maintaining flexibility and responsiveness in the encryption process [11].

Secure sharing of personal health records in the cloud is an area of specific concern, as patients are sometimes allowed to upload encrypted EHRs to the cloud, giving users access to certain parts of the EHR [12]. The owners grant each user in a group of users of a later type of access to the EHR to a certain extent, depending on the role of the user. Another important requirement of “patientoriented” EHR is that each patient has a specification about who has access to their personal EHR information. The EHR may be banned for some users [13].

Privacy and Security Concerns in Smart Healthcare

Medical information is a private area, which is even considered intimate by many people, so patient confidentiality is the most important issue. Thus, it is not only highly desirable, but strictly important to ensure that there are appropriate security measures in place, as digital data can be easily transferred anywhere in the world, as this is the mechanism of functioning of global networks. Storage and transmission of medical results, medical analysis and tests requires specific attention. Doctors need to ensure that they will not disclose private and sensitive information about the patient to any third party, as it is very simple when new devices are used, and there are many perpetrators that try to obtain this information [14]. Security is one of the most serious problems for artificial intelligence in healthcare. To realize the potential of AI, developers need to make sure of several key things [15,16]:

a) Reliability and reliability of data sets: the data sets used must be reliable and valid, because the better the training labeled data, the better the AI will work.

b) Data sharing: the need for huge amounts of data for analysis requires extensive data sharing.

c) Ensuring the transparency of algorithms: in the interests of the safety and trust of clinicians and patients, it is necessary to ensure some degree of transparency of algorithms, although in the real world there are problems related to the protection of investments and intellectual property, as well as cybersecurity.

Many AI systems that operate in healthcare rely on the existence of the big amounts of sensitive data, which sometimes conflicts with the data protection legislation. Of course, the data can be depersonalized, especially for the reasons of scientific research and big data analysis, which is often performed by AI. By performing depersonalization of data (anonymization), we will get both confidentiality and data integrity. This data will be useful for introducing innovations and strengthening cooperation between suppliers and partners, which will also benefit smart city medicine, including through the exchange of knowledge between doctors from around the world.

Informed Consent for AI Application in Smart Healthcare

A right to seek, receive and impart information and ideas through any media and regardless of frontiers is one the major human rights enshrined in Article 19 of the Universal Declaration of Human Rights. The right to access one’s personal information is not only part of respect for basic human dignity, but it is also central to effective personal decision-making; for example, access to medical records, for example, can help individuals make decisions about treatment, financial planning and so on [17]. Does the patient need or has the rights to give access to the information about the application of AI in diagnosing his or her condition? This is a question to be answered.

Issues related to obtaining informed consent for the use of AI: • Under what circumstances should the doctor notify the patient that AI is generally used for diagnosis, diagnosis and choice of treatment method?

• Under what circumstances should the principles of informed consent be applied in the field of AI?

It is also needed to consider the limits of the provision of information on AI to public. AI might take important decisions as to one’s health. Many algorithms rely on very complex and difficult to deconvolute mathematics, sometimes called the “black box”. In the medical area there are situation where it can be extremely important to know the reasons for decisions because they can affect not only patient’s health, but his life in general. But can a patient ask to have the algorithm disclosed? On one hand, a patient who is diagnosed with a severe condition using AI system, or received a specific prescription (presumable, on the basis on big data analysis by AI) might be interested in knowing the reasons and algorithm that formed the basis for the decision.

On the other hand, the developer of the appropriate AI system might also have ground to object to such disclosure, so as to protect his investments and effort, as he might be afraid that in such a way his competitors will discover his know-how and violate other IP rights, including reverse engineering of the software. Let us look at one example from another sphere. In a Wisconsin v. Loomis case (USA), a criminal defendant challenged a state trial court’s use of a (non-machine learning) risk assessment algorithm (developed by a private company) to determine his sentence. He argued that his due process rights were violated, as the company refused to disclose how the risk score there determined, claiming that information was a “trade secret”, and due to the “proprietary nature” of the algorithm he could not assess the information that was used for sentencing. The Wisconsin Supreme Court rejected the defendant’s arguments, stating that the company had the right to protect its proprietary information; and it release sufficient information that satisfied due process requirements [18].

However, sometimes knowing of the algorithm is important, as it can be a way to overcome a so called “algorithmic bias”, which is cases by the decisions of AI which are based on factors that should not be in fact relevant to the case. For example, training data can be biased because they are based on discriminatory human decisions. Such situation occurred at a medical school in the UK in the 1980s, where a computer program was introduced to sort the applications. The training data were the admission files from earlier years, when selection of the applications was done by persons. And it turned out that computer program discriminated against women and against people with immigrant background [19]. Another possible algorithmic bias can be caused by the under-representation of poor people in a data set; poor people are less likely to have smartphones and other smart devices, and lower possibilities of access to paid medicine services in general, thus they might not be fully taken into account in the medical studies. Attentions should also be given to the ways of overcoming the so-called “automation bias”. Automation bias is a tendency to believe computers without additional consideration of the results; as human decision-makers tend to follow computer advice, either because they try to minimize their responsibility, or because they do not have enough time, context or skills to make an adequate decision in the individual case [20]. Thus, it is important to properly train doctors and other medical workers so as to ensure that they do not trust the AI algorithms blindly and take due care so as to ensure the accuracy of the results, taking into consideration other possible options.

Diagnosis-Making Using AI Systems and Liability

The most AI systems are used to help doctor to make a diagnosis. Most of diagnosis in medicine are made based on analysis of medical images. The use of AI in the analysis of medical images is under continuous evolution. There are already very good results that are shown by AI systems in detection of skin cancer: in 2017, the case was reported were researchers have trained a neural network (a dataset of 129450 clinical images consisting of 2032 different diseases was used); and the neural network achieved performance on par with all tested experts, demonstrating that AI was capable of classifying skin cancer at a level of accuracy comparable to that of dermatologists [21]. As some reports show, the impact of AI is especially relevant in neuroscience (neurosurgery, neurology). This area is based on the combination of AI-mediated technologies with advances in photonics (merging of applied optics and electronics) and engineering, together with other clinical disciplines (pharmacology, psychology) and related sciences (biology and genetics, biochemistry) [22]. AI systems are able to analyse complex data, moreover, they need data to learn and to operate properly. The quality of data affects the quality of the outcome. One part of the problem is the time and expenses that are needed to collect and insert this data into appropriate AI system in healthcare [23]. It is especially problematic in situations when patient’s data are stored in different institutions in a random way and incompatible formats, thus requiring additional resources for their collection and standardization.

As we discuss diagnosis, we can briefly note one more issue. AI systems can also create a new method of remuneration of medical workers. At the moment, doctors are encouraged to have many visits from patients, and take multiple tests (which are often viewed as unnecessary and burdensome), and thus basically their work is assessed on the volume of treatment, which is then reflected in the remuneration. AI systems might be able to assess the value of the treatment for the patient, that is, whether the treatment was successful or not, and how the strategy which was proposed by the doctor influenced patient’s condition in general. It is expected that a value-based remuneration will provide additional incentive for doctors to improve their skills and knowledge [24]. Another question: can a doctor rely entirely on AI? One of the common mechanisms of application of machine learning (which is a form of AI) is healthcare is precision medicine, that is, predicting what treatment protocols are likely to success on a patient based on various patient attributes and the treatment context; this requires a training dataset for which the outcome variable (e.g. onset of disease) is known (so called “supervised learning” [25]. However, the cognitive systems have problems with the quality and volume of medical information. The data accumulated in patients’ medical records may be incomplete, contain errors, inaccuracies, and nonstandard terms. There are not enough records of the patient’s life, habits, and behaviour. Effective mechanisms for collecting this information do not yet exist. In addition, many of the AI algorithms are considered as black box in which the decision-making process is hidden in network layers. This can be problematic especially in situations that are not present in data set used to train AI algorithms, which will likely result in inaccurate AI decisions.

Application of AI in healthcare thus causes concerns when we think about possible liability issues. We agree with the research that say that while it may be fairly easy to identify a wrongful act or effect resulting from the use of an AI system, it will often be less straightforward to identify the blameworthy actors [26]. Specifically, criminal liability generally requires showing knowledge or intention of the relevant actors, and it is clear that AI systems have no such mental state [27]. In case AI itself bears no legal liability, who is going to be liable? Which criteria do we need to use to choose the guilty one? Can be a doctor that relied on AI assistance? Or a technical worker that inserted, maybe unintentionally, wrong data, which lead to the wrong results? Or shall we put blame on the software developer, who did not think about possible options and did not teach AI system properly? This all-causes concerns and leads to the certain degree of unpredictability. However, it clear that legal solutions to these issues need to be further discussed. One the possible ways which is proposed by some scholars it to rethink the liability principles, so as maybe to split liability between manufacturers of AI systems (they can be held liable for their product causing harm under the genera product liability regime), physicians and patients [28]. There are already some developments in the legal sphere in considering liability for the actions of robots; e.g. under German tort law the following principles apply:

- There can be no fault-based liability if the malfunction of the robots is not foreseeable to the person using it;

- Liability is excluded if the patient has consented to the use of the medical robot; however, the patient is to be informed beforehead about all the circumstances and risks of the medical intervention, and also about all the alternative treatment measures which are equal at the achieving of the same treatment goal [29].

Of course, as robots “do not make independent movements and do not make their own decisions, but are completely controlled by the operator”, current law does not have any gaps in liability in this respect [30]. AI systems are of course different due to their nature. However, the approaches described above might be taken into consideration in developing legal rules on liability for the decisions and actions of AI systems. One of the main questions is: “What data processing can be considered an interpretation that has a real risk of harm to the patient’s health?” It is proposed to consider the processing of clinical data about a patient as such an interpretation, as a result of which new, clinically significant information missing from the initial data is produced (formed), which is necessary and used when making a clinical decision and/or performing a medical intervention [31]. It should be noted here that there’s no doubt that mistakes are inevitable. As some studies show, even not, according to data collected from several EU nations, medical errors and healthcare related adverse events occur in eight to twelve percent of hospitalizations; preventing such mistakes could help to prevent more than 3.2 million days of hospitalization each year within the EU [32,33]. Thus, application of AI by doctors needs to be done in such a way that minimizes the possibilities of mistake.

Having said that, it is important to note that legal regulations should also aim at taking due account of situations when AI technologies are used for evil purposes. For example, there’s a risk that medical worker would like to introduce changes into human genome or make other illegal activities. Thus, it is necessary to introduce appropriate safeguards, so as to ensure that patients are not at risk when there’s application of AI technologies by doctors. This may be achieved by introduction of specific forms of notification of authorities and control mechanism. And of course, it is necessary to raise patient’s awareness so that they know the signs that can show that there can be danger in dealing with the specific medical institution.

Conclusion

In this article, our intention was to consider AI use in Smart Healthcare from two different aspects: information technologies and legislation. There’s no doubt that information technologies are actively developing and will make medicine much more effective and will help people to have healthier life. However, at the moment legislative framework has not yet been fully adapted to the new reality where medical professionals increasingly rely on AI systems. In our mind, any further improvements to the legal framework of Smart Healthcare, need to be based on the study and development of the following aspects of the legislation:

- Defining responsibility rules for medical doctor, AI and patient in making diagnosis and choosing the treatment.

- Increasing the role of the patients in Hospital Information Systems, as a condition for the development of personalized medicine, with the possibility of limiting access to their Electronic Health Records.

- The right to information about their health and free access to information affecting the freedoms, rights, duties, interests of the patient, including the use of mobile applications.

- The use of web-services, remote interaction between the doctor and the patient through a variety of means: social networks, smartphone, tablet, etc.

- The protection of personal data and legally defined secrets.

- So, the triad: medical achievements, information technology and advanced legislation will change the medicine of the future.

For more Articles on: https://biomedres01.blogspot.com/

 

An Unusual Presentation of Rheumatoid Arthritis with Two Years Bilateral Knee Swelling in a Young Sero- Negative Patient; A Case Report

Introduction

Rheumatoid arthritis has widely been known as a highly heterogeneous disorder of Immune dys-regulation [1,2], Rheumatoid factor (RF) is an autoantibody, and is been used in the diagnosis of rheumatoid arthritis for years, recently, Anti-cyclic citrullinated peptide antibody (ACPA) has gained more attention, as a valuable and significant marker in predicting and diagnosing of RA [3]. Both RF and ACPA have the same sensitivity, however, ACPA much more specific in the diagnosis of RA [4]. Nowadays, seronegative RA has become a moving target, reflecting both new modification in RA categorization [5], and an evolving understanding of diseasesantibodies association [5]. The proportion of the RA patients which is reportedly sero-negative for both ACPA and RF ranges between 0.1-0.48 [6]. ACPA and RF have been recognized as poor prognostic markers of RA, and both are used evidence to justify the use of intensive therapy in seropositive RA patients (SPRA) [7]. The report aims to draw attention to unusual presentation of RA in a young sero-negative patient.

Case Presentation

A 24 years old male from South Sudan has been referred to our clinic from the orthopedics department complaining of bilateral knees swelling for two years (Figure 1). His condition started with progressive painless right knee swelling over several months then the left knee was affected similarly. There was no associated pain, redness or hotness, the swelling mildly affects his movement, other joints were not affected, there was no back pain nor morning stiffness, no fever, loss of weight or other constitutional symptoms. Other systems were unremarkable. Musculoskeletal examination shows both knees were swollen with effusion with slight hotness with no tenderness, deformity or restriction of movements. Other joints were normal including the back without any muscles weakness nor tenderness. There are no scalp, skin or nail changes. On general examination the patient looks well, afebrile, average weight, not pale. Normal pulse and blood pressure. The patient has no history of diabetic, hypertensive or any bleeding tendency. he has no history of trauma and or any family history of rheumatic disease or similar conditions. General works up is done and shows normal complete blood count. [Blood hemoglobin (13.8g/dl)( normal value 13.5 to 17.5g/dl), MCV 75 fl, MCH (28.2pg), MCHC (37.5g/dl), WCC (6.900)( Normal value: 4,000 and 11,000 per microliter of blood), platelet: 399,000 (Normal value 50,000 to 450,000 platelets per microliter of blood). ESR is 100 (Normal value: between 0 and 15 mm/hour), RFT and LFT were both normal, Urine analysis was also clear, viral screening is negative, rheumatoid factor (RF) and ACCP were also unremarkable, ANA profile is negative. Chest X-ray done and also normal. Right knew X-ray (Figure 2) and MRI along with synovial biopsy indicate the diagnosis of rheumatoid arthritis. Microscopic evaluation of synovial tissue inflammation shows marked cellular hyperplasia in the lining layer, T-cells, plasma cells, Macrophages, B-cells, neutrophils, mast cells, NK cells and dendritic cells accumulate in the synovial sub lining layer. The patient received prednisolone 20mg, hydroxychloroquine (HCQ) 200mg twice daily, methotrexate 15mg weekly, folic acid 5mg twice weekly and osteocare once daily. One month later the patient frankly improved and swelling subsided, ESR was 5mm/hr.

Figure 1: Shows bilateral knee swelling.

Figure 2.

Discussion

RA is usually diagnosed based on the classification of the American college of rheumatology [5]. According to the classification, A score of more or six is usually indicating RA, the classification includes features like Involved joint type, duration and serology markers such as CRP, ESR, RF and anti-CCP. When the patient still exhibit symptoms and signs of RA but his serological markers are negative this is an indication of seronegative rheumatoid arthritis. RA has long been recognized as an extremely heterogeneous disorder of immune dys-regulation. Despite an ever-growing appreciation of the role of circulating autoantibodies in the progress of ‘seropositive’ disease, the pathogenesis of sero-negative RA remains poorly understood [1]. The clinical presentation and scenario of this case can fit with many diagnoses such as undifferentiated arthritis which is an overlap syndrome with RA and not fit well known clinical disease categories ( e.g, reactive arthritis and sero-negative RA) [8,9]. Although, synovial knee biopsy confirmed the diagnosis. other possible diagnosis overlaps with this patient condition include reactive arthritis, but the absence of infection exclude the diagnosis as the swelling and joint pain in reactive arthritis mainly triggered by infection [10]. Psoriatic arthritis includes similar clinical features such as joints pain and stiffness and the diagnosis is primarily based on the clinical phenotype due to the diversity of the associated features which include nail and skin disease, uveitis, dactylitis and osteitis [11]. And these skin manifestations were not presented in this case.

The synovial biopsy is not usually used for routine diagnostic or therapeutic purpose in RA patients [12]. However, synovial tissue examination can assist in the diagnosis of some joint infection. Although there is no diagnostic role in acute RA, although, synovial tissue analysis that can provide significant prognostic information, recent studies have shown that examined mediators of joint damage and synovial tissue inflammation were found to be linked with unfavorable radiological and clinical outcomes [13]. Although genetic variation has long been supposed to account for around 60% of RA risk [14], however recent reports indicates that this figure for heritability may be significantly lower in sero-negative RA [15]. It follows that environmental factors such as smoking should play a more essential role in seronegative RA, but much has yet to be understood in this area [16,17].

Practice Points

1) Sero-negative RA signifies an important disease burden, for which prompt therapeutic intervention remains to be vulnerable to diagnostic challenges.

2) Currently, there is inadequate evidence to support a modified treatment strategy for the seronegative RA, even though the value of initial combination DMARD use, and that of B-cell depletion therapy, may be less than in seropositive disease.

Conclusion

Seronegative RA is a chronic, painful condition, where the patient exhibit features of RA with unremarkable serological value, their triggers and the extent to which they overlap or differ according to autoantibody status will define our ability to manage RA more effectively. This is a case of a young Sudanese patient who is been suffers from knees swelling for two years.

Clinical Message

Rheumatoid arthritis is an autoimmune disease affecting the joints and lead to their damage. Seronegative RA is a chronic, painful condition, where the patient exhibit features of RA however have negative diagnostic markers of RA.

For more Articles on: https://biomedres01.blogspot.com/

 

Pulmonary Metastasis of Ewing’s Sarcoma in Pediatric Age Literature Review and Case Report

Introduction

Ewing’s Sarcoma (ES) is a malignant neoplasm of mesenchymal origin that affects children and adolescents, with a peak incidence in the second decade of life. The average age being between 13 and 16 years [1,2] It was described by the American histopathologist, oncologist, and hematologist James Stephen Ewing in 1921. [1,2] The annual incidence is [1-3] cases per million children under 15 years of age. Ewing’s sarcoma accounts for about 1.5% of all childhood cancers and is the second most common type of bone sarcoma. [1,2] It appears in patients in the stage of life of greater exercise and changes in the body typical of adolescence, so it can be confused at first with bone growth pains, with banal local traumas of exercise in daily activities. [1.3-5] Clinically, it presents as a picture characterized by intermittent bone pain at the site of the injury, which increases at night and increases in intensity, may be accompanied by increased local volume, in the form of a mass that may be painful on palpation, sometimes It is accompanied by fever and constitutional syndrome due to weight loss, asthenia and anorexia. [1,3,4] The most affected bone sites are: metaphysis of the long bones (around 56%, costal arches approximately 15 to 17%, flat bones 16% and skull between 3 to 4%) [1,2,6,7]. Ewing’s sarcoma produces metastases via the hematogenous route to the lung in more than half of the cases, to the bone, brain and bone marrow, less frequently via the lymphatic route to distant nodes.

It is good to point out that in most cases these metastases are present from the beginning of the disease. [2,7,8] Diagnosis is based on imaging studies: X-rays of the affected limb or site, Computerized Axial Tomography (CAT) and Nuclear Magnetic Resonance Imaging (MRI). [1,6,9] Primary tumor biopsy is mandatory. Definite diagnosis requires histological evaluation and confirmation with cytogenetic or molecular biology techniques. The defining characteristic of these tumors is the presence of a series of chromosomal translocations that culminate in the fusion of the EWSR1 gene, on chromosome 22, with one of several members of the ETS family of transcription factors. The most common of these translocations, t (11; 22) (q24; q12), which fuses the EWSR1 gene with the FLI1 gene on chromosome 11, is present in approximately 90% of cases immunohistochemistry demonstrates CD 99 positive [9]. Treatment is based on extensive surgery of the lesion with free section margins whenever the site or location of the tumor allows it, chemotherapy with high-dose drug association, and radiotherapy. [2,3,8,10] The prognosis of this variant of tumor depends on age, clinical stage at diagnosis, since the presence of metastases at the time of diagnosis overshadows the prognosis, especially in the lung, causing a low survival rate of around 21%, [2,11] while patients diagnosed early in the initial stages of the disease have a better survival of around 50%.[1,2] Mortality is high, especially in the first year after diagnosis in cases with lung metastases. Fiveyear disease-free survival is very low, around 10% to 31% of cases [2,6,12].

Case Presentation

We present the case of a mestizo female patient from the province of Camaguey, who at 2 years of age begins with an increase in volume in the anterolateral face of the left lower limb, of two months of evolution accompanied by intermittent fever, in her area of health interpret the symptoms as if it were acute osteomyelitis of the tibia, so they sent her to the Pediatric Hospital of her native province they admitted her with a diagnosis of possible osteomyelitis and started antibiotic treatment with phosphocin first intravenously and then orally, also with non-steroidal antiinflammatory drugs, evolutionarily they did not see clinical improvement and radiographic studies of the lower left extremity were performed, detecting osteolytic lesion and bone destruction. A CT scan is performed, verifying a destructive bone tumor lesion of the proximal third of the anterior face of the left tibia. Clinically, the visible and palpable tumor lesion in the anteromadial aspect of the upper third of the left lower extremity continued to increase in size, so the relatives decided to travel to the capital of the country where they were admitted to our hospital in December 2019, studies were completed and a CT scan of the lung is performed, a single metastatic lesion is confirmed in the lower lobe projection of the left lung, without associated respiratory symptoms. The diagnosis was made by surgical biopsy of the tibial bone lesion, yielding the histopathological study: Ewing’s sarcoma. The general physical examination maintained good coloration of the skin and mucous membranes.

Chest

symmetrical with adequate ventilatory movements, without alterations of the vesicular murmur in the lung fields, no rales. Rf: 28xmin

Cardiovascular

Rhythmic heart tones, not murmurs. HR: 88xmin AT: 90/50. Examination of the lower left extremity below the patella detected a painful tumor lesion of approximately 8 to 12 centimeters in diameter, somewhat painful on palpation, firm consistency, welldefined borders, non-movable. No peripheral lymph nodes. Free sensory. No meningeal signs. Sensory and motor reflexes preserved. It is discussed in a multidisciplinary team and it is decided to start treatment with high-dose chemotherapy and evaluate at the end of the first 4 cycles VAC/IE (vincristine, cyclophosphamide, doxorubicin alternating with ifosfamide/etoposide, it was evaluated and the response was very poor with only a clinical and imaging reduction of the tumor lesion from 1.5 to 2 cm, it is agreed in Within the tumor committee with the authorization of the parents to perform the surgery in the month of May 2020, the supracondylar amputation of the affected limb was performed, with very good postoperative evolution and healing of the stump.

Pathological Anatomy Result

Ewing’s sarcoma of the upper third of the tibia, with extensive necrosis, bone section edges and soft tissue free of tumor. Immunohistochemical studies showed CD 99 positive. The postoperative chemotherapy scheme with cytostatic drugs is continued for 6 complete cycles in order to reduce or disappear the metastatic lesion of the left lung, which from the beginning measured 19 mm x 22 mm, thus maintaining that diameter throughout the systemic treatment. which ends in September 2021. She is studied again at the end of the chemotherapy treatment and the patient maintains a pulmonary metastatic lesion in projection of the lower lobe of the left lung, but with slight growth to 20x28mm in the chest X-ray, but in the CT the lesion was 28x34mm. (Figure 1) The multidisciplinary team discussed it again and it was agreed to perform the metastasis through the left anteroposterior thoracotomy. On November 4, 2021, she was operated on, and a wide resection of the lesion was performed at the level of the lower segment of the lingula of the left lung (Figures 2A & 2B).

Figure 1: Contrast-enhanced CT scan of the chest sagittal view, where a metastatic lesion of the lower lobe of the left lung is displayed.

Figure 2: A. Wide resection surgery of the left lung metastasis together with adjacent tissue of the lingula where it was inserted. Figure 2B.

Pathological Anatomy Result

Small blue cell tumor metastasis with mitotic index greater than 10 per field, tumor necrosis index 5%, presence of vascular permeation, histological picture compatible with Ewing Sarcoma metastasis. The postoperative evolution was good, the wound healed without complications. He presented fever 11 days after surgery and a mild inflammatory process was found at the base of the left lung without pleural effusion by ultrasound or chest X-rays. He completed antibiotic treatment first with Piperacillin with Tazobactan and Amikacin, but the fever persisted and on the fifth day the antibiotic treatment was changed due to Staphylococcus epidermitidis growth in blood culture performed to Vancomycin and Meronem, the fever disappearing, and he always maintained a very good general condition. Locoregional radiotherapy of the left lung was then performed for six sessions during the month of December 2021. The postoperative chemotherapy treatment was completed in the months of January and February 2022. To date, the patient is stable. Two months after finishing treatment, a positron emission tomography (PEC/CT) scan was performed from the apex of the skull to the middle third of the femurs, which did not reveal abnormal areas of glycolytic hypermetabolism. Conclusions of the study no tumor metabolic lesion was observed at the bone, lung or ganglionic chain level. This case is brought to collation due to the rarity of this tumor variant in early ages of life, and due to its presentation in an advanced stage of the disease from the beginning (Stage IV). Highlighting the value of surgery in disease control combined with other therapeutic weapons: chemotherapy and ionizing radiation.

Discussion

Ewing’s sarcoma is a primary malignant bone tumor that is common in children, preceded in frequency by osteosarcoma. [2,9] It is a very aggressive tumor, its etiology is unknown, it is more common in males, it is located more common in the lower half of the skeleton. [1,3,4] It tends to metastasize early, mainly to the lung. It should be noted that primary lung neoplasms are infrequent in children, they are usually secondary lesions or metastases of malignant solid tumors typical of childhood, and their clinical manifestations are initially confused with an infectious process. [2,5] There are few case reports of this tumor variant in early life. (Izaguirre, et al. [3]) reported a 13-year-old patient with Ewing’s sarcoma of the left fibula, but at such a young age as in our case it is extremely rare. The symptoms are common to other bone tumors: intermittent pain and inflammation located in the affected area. [2,4,9] Generally, these symptoms are confused with inflammatory processes. In order to establish an early diagnosis of the bone lesion, in addition to the symptoms, imaging studies of the chest X-ray, CT and MRI are required. It is also important to evaluate early for disease recurrence [3,9,10-14].

The main treatment weapon for this variant of malignant tumor is surgery with en bloc resection of the tumor and nearby soft tissues or amputation of the limb, in very specific cases limb salvage surgery is proposed [2,3, 9,15]. Sánchez Saba, et al. [9] in their study of 88 patients diagnosed with Ewing’s sarcoma of bone treated with preoperative chemotherapy and limb-sparing surgery, the overall survival rates were 79.5% at 2 years, 69% at 5 years and 64% at 10 years considered that limb-sparing surgery associated with pre- and postoperative chemotherapy should be the treatment for Ewing’s sarcoma of bone that meets certain requirements that allow its performance. In our case, it did not meet the necessary requirements and demands for its performance since the disease was metastatic since its diagnosis. [9,10] Most authors agree that chemotherapy is mandatory for disease control before (neoadjuvant) and after surgery (adjuvant), the most widely used regimens worldwide are VACA (vincristine, actinomycin, cyclophosphamide doxorubicin) and VAC/IE (vincristine, cyclophosphamide, doxorubicin alternating with ifosfamide/ etoposide) [3,9].

Other protocols are VAI (vincristine – actinomycin – ifosfamide) and VIDE between 6 and 8 cycles. For a definitive diagnosis, histopathological studies with immunohistochemical technique are required, the realization of the CD99 immunomarker is necessary to confirm this entity since it gives us a sensitivity of more than 90%. [2,3,7,9,11] Definite diagnosis requires histological evaluation and confirmation with cytogenetic or molecular biology techniques. Under the light microscope, Ewing’s sarcoma appears as a tumor of small, round, blue cells, with different architectural patterns and variable protein expression, detectable by immunohistochemical techniques [3,9,10]. Stéphanie Foulon et al. [16] in a study conducted stated that although the role of radiotherapy has been discussed, it demonstrated that the use of radiotherapy helps in local control of the disease and reduction of tumor volume [15-17].

For more Articles on: https://biomedres01.blogspot.com/

 

Dilemma in Management of Endometriotic Ovarian Cyst

Introduction

Ovarian endometrioma or chocolate cyst is ectopic endometrial tissue lining ovarian cyst [1]. Presence of endometrial glands and stroma outside the uterine cavity is known as endometriosis [2]. Endometriosis may present as peritoneal lesions, deep endometriosis and ovarian endometriomas. Due to relative ease of diagnosis endometriomas are the most diagnosed form of endometriosis. Endometriosis have been reported in 17-44% [3]. With 10-20% in women in reproductive age most commonly between 40 and 44 years and approximately 17% of women suffering from infertility [4]. This condition is prevalent among the East Asian race [5]. Endometriosis and endometriomas have the multifactorial ethology [6]. Endometriomas are mostly unilateral, commonly left-sided, they are either asymptomatic or may present with symptoms. It is not infrequent to have under diagnosis or misdiagnosis and this is quite common in adolescent women [7]. Treatment of endometrioma is a clinical dilemma that if found in imaging then whether to treat or not and if yes then how to treat. Symptoms of a patient will guide for available options either conservative, medical, surgical or combination of both or sclerotherapy [8].

Discussion

The presence of ovarian endometrioses has been found to be associated with deep endometriosis and multifocal deep lesions [9]. Endometriomas are mostly unilateral, commonly left-sided [10]. Left sided predisposition is explained by anatomic barriers like sigmoid colon that may delay in eliminations of endometriotic tissue from left side of pelvis and promotes left sided cysts, in support this is explained by theory of retrograde menstruation [11]. The pathogenesis of endometrioma is explained as implantation of endometrial cells on ovarian surface via tubular lumen that causes persistent inflammation, bleeding, cyst formation at implantation site resulting invagination of ovarian cortex, adhesions secondary to metaplasia which may result in progressive damage of healthy ovarian tissue [12]. Endometrioma pseudo capsule is ovarian epithelium containing oocytes and follicular structures. The reason of endometrioma-related infertility remains unclear. Possible theories may be damage to affected ovary or tubo-ovarian distortion anatomy and cellular damage resulting in follicular loss [13]. Other factors may be involved including immune factors, inflammatory factors, environmental toxins, and genetic factors [14]. Endometrioma may present with dyspareunia, dysmenorrhea, pelvic pain, bleeding, infertility, and dysuria. It is not infrequent to have under diagnosis or misdiagnosis and this is quite common in adolescent women. Ovarian endometriomas may predispose to ovarian malignancies, especially clear cell carcinoma and endometrioid adenocarcinoma.

For Endometrioma diagnosis, transvaginal ultrasonography is a very sensitive and specific. Unilocular cyst with a “ground glass” homogeneity, low levels of echogenicity, and poor vascularization, one to four compartments and no papillary structures with detectable blood flow [15], which had been adopted in the ESHRE guidelines [16] are typical ultrasound characteristics of endometriomas [17]. One useful diagnostic indicator is immobility as adherent to pelvic side wall. Diagnosis and treatment of endometrioma another useful tool is laparoscopy. A new promising biomarker is Human epidydimal secretory protein E4 used in the differential diagnosis of endometriosis cyst. The combination of HE4 and CA 125 assay could discriminate ovarian endometriosis cysts from malignant ovarian tumours effectively [18]. The advantage of HE4 over CA125 is mainly in the detection of borderline ovarian tumours and early-stage epithelial ovarian and tubal cancers. After diagnosis, possible options are either expectant management or treatment depending on symptoms, age, fertility concerns, ovarian reserve and previous history of treatment with specific reference to past surgical interventions; nature of the cyst; and the fertility wishes of the woman [19]. Treatment of incidental disease in otherwise asymptomatic women is currently not recommended, as still the natural progression and development history of endometriomas is not well understood.

Treatment of endometrioma is a clinical dilemma that if found in imaging then whether to treat or not and if yes then how to treat. Symptoms of a patient will guide for available options either medical treatment progestins , oestrogen suppression or surgical or combination of both. An incidental finding of an ovarian endometrioma in young women with regular menstrual cycles and without suspicion of malignancy who wish to conceive should be encouraged for natural conception before seeking fertility treatment. While the evidence of the impact of an endometrioma on spontaneous conception is limited. Aim of surgical treatment is removal of endometriotic tissue, to have sufficient sample for histopathology and to preserve maximum ovarian tissue in cases where fertility is desired and to avoid risk of menopause. With surgical treatment risk is unintentional removal of ovarian follicles which is later shown by reduced levels or antral follicle count on ultrasound or reduction in serum anti- Müllerian hormone (AMH) [20]. To reduce recurrence after surgery medical therapy may be used. Recurrence rate of endometriomas after surgical treatment are 30-40 % [21]. So to delay recurrence of ovarian endometrioma in 2014, European Society of Human Reproduction and Embryology (ESHRE) recommended for ovarian cystectomy instead of drainage and coagulation of endometriosis in cases of surgical treatment, since ovarian cystectomy can reduce endometriosis-associated pain and recurrence rate effectively [22]. Fertility is affected by presence of endometrioma [23], while after IVF overall pregnancy rates are unaffected [24]. Any surgical intervention to remove endometrioma may be associated with decrease ovarian reserve and possible recurrence [25]. At present, no consensus has been reached on the timing of surgery in young women; whether surgery should be delayed in infertile women planning IVF is still debated [26].

Possible complications with non-surgical approach are:

1) Difficulties during oocyte retrieval

2) Progression of endometriosis

3) Missing an occult early-stage malignancy

4) Risk of development of a pelvic abscess or rupture of the endometrioma

5) Follicular fluid contamination with endometrioma content

Most common mode of treatment is surgical which is laparoscopic cystectomy, benefit is reduced pain symptoms and recurrence. Excision of endometrioma in comparison with drainage of endometrioma with or without ablation of pseudo capsule is associated with better outcome and higher pregnancy rates [27]. However, ovarian cystectomy can lead to decreased ovarian reserve or due to excessive coagulation can be reason [28]. So to attain follicular development, increased amounts of gonadotropins are needed [29]. Laparoscopic cyst fenestration and ablation of the cyst capsule is another alternative method which improve pelvic pain and result in high patient satisfaction but high recurrence. That’s why opinion shifted towards more conservative approach. In 2013, The European Society of Human Reproduction and Embryology guideline suggested that surgery should be considered only if size of endometrioma is >3 cm, to improve access to follicles or pain [30]. Size play an important role in decrease ovarian reserve before surgery and difficulty for complete removal in case of superficial destruction as well as damage to ovary in case of surgical excision. Post-operative medical treatment markedly reduces the recurrence rate of endometrioma [31]. Therefore, long-term medical treatment to prevent recurrence is routinely recommended [32]. Post-operative medical treatments including oral contraceptives GnRH agonists, and progesterone commonly used to suppress possible residual lesions due to the oestrogen-reducing effects [33]. However, each of these treatments has reported adverse effects. Post operative medication needs, or efficacy was not studied in women aged 40 year or more.

Medical treatment used for treatment of endometrioma include Oral contraceptive pills, progestins, gonadotropin-releasing hormone agonists [34] as well as aromatase inhibitors are helpful to reduce size, symptoms, and post-surgery recurrence [35]. However, problem is reappearance of symptoms after stopping medical treatment [36]. To reduce recurrence after aspiration another promising method is sclerotherapy [37]. It involves injecting into cyst cavity a sclerosing agent which can be either removed by washing or left within cyst. It is thought that it will work by causing inflammation and fibrosis causing destruction of epithelial lining of cyst and at the end will cause obliteration of cyst [38]. It has been shown that sclerotherapy is cost effective method for endometrioma but not widely used [39]. Pain improved in 68- 96% independent from duration of ethanol inside endometriotic cyst. Compared to laparoscopic cystectomy, with sclerotherapy number of oocytes retrieved during IVF treatments was higher but no difference in pregnancy rates after sclerotherapy and untreated cases. Sclerotherapy was found to be safe with possible complication of transient abdominal pain. After sclerotherapy difference in the recurrence rate in studies can be due to variation in selection criteria (cyst size and number of cysts), technique used (sclerosing agent, concentration, installed volume, and retention time), duration of ethanol inside the endometrioma and the followup time. Risk of unexpected malignancy with typical features of endometrioma has been found in 1% in patients [40]. Other factors will influence the decision in an asymptomatic patient like the rate of growth, the age of patients, personal and family history of breast and ovarian malignancies [41].

Other alternative is phytotherapeutic options obtained from plants or herbal preparations some of them work by influencing apoptosis, epigenetic factors, angiogenetic processes, cell survival, oxidative stress and oestrogen modulation [42]. During course of fertility treatment, endometrioma often present a clinical dilemma due to uncertainty regarding decision of either to operate or manage conservatively while balancing possible risk of surgery on ovarian reserve. So far guidance available from either small and/ or retrospective controlled studies. Surgery does not improve the results of IVF treatment [43], but a sequential use of surgery and IVF in those that do not conceive spontaneously probably results in slightly higher cumulative pregnancy rates [44]. There may be spilling of chocolate fluid of endometrioma in peritoneal cavity in women undergoing IVF. This fluid may not induce endometriosis but is adhesiogenic [45]. Considering the risk of ovarian damage during surgery and the excellent results of IVF, actual guidelines [46] therefore have concluded that if IVF indicated then should not undergo surgery if size of endometrioma is ˂ than 3-4cm. Surgical treatment of endometriomas prior to IVF is widely practiced, [47] although debatable on its effect and need. To date, there has been no evidence that surgical treatment improves reproductive outcome of women treated with the use of ART, no difference in the clinical pregnancy rate and the number of oocytes retrieved from women who had surgical treatment compared with those with intact endometrioma. Cancellation rate and number of retrieved oocytes were comparable. After surgical treatment of endometrioma there is lower antral follicle count and higher doses of gonadotrophins required for ovarian stimulation.

Women of advanced reproductive age, asymptomatic, those with reduced ovarian reserve, bilateral endometriomas or a history of prior ovarian surgery may benefit from proceeding directly with IVF, as Ovarian reserve may be compromised further after surgery. In case of symptomatic women, large endometrioma, intact ovarian reserve, suspicious features of cyst on radiological investigations or with clinical features surgery may be considered. There is risk of Infertility and Premature Ovarian failure after treatment of endometrioma in very young women. Pathophysiology and manifestation of endometriomas in adolescents may be different than adult women [48]. The diagnosis of endometriosis in adolescents is often delayed due to several factors. Regarding early diagnosis followed by surgical removal of endometriomas in the adolescent population, currently no original studies are present as fertility is a major concern as well as future recurrence. New concept is early treatment instead of postponing surgery to prevent adhesions, ovarian damage and recurrence. Considering that the endometriosis in cystic ovarian endometriosis is only superficial, a superficial destruction by electro surgery, CO2 laser or alcohol should be sufficient before the development of more lesions and size of endometrioma is getting more or symptoms becoming more severe or concern of fertility arises. With these concepts, the use of THL in women with infertility should be reconsidered. Transvaginal hydro-laparoscopy (THL) [49,50] offers a minimal invasive procedure for early diagnosis and treatment of small endometrioma up to a diameter of 20 mm not seldom these small endometriotic cyst are missed at routine vaginal ultrasound examination in approximately 50% of the cases.

It is always surprising after opening of such small cysts to see the pronounced presence of inflammation and neo-angiogenesis, a signature for the aggressiveness of the disease in these early stages. Due to concern of ovarian reserve early stages treatment using ablative technique with a bipolar 5Fr probe causes a minimal trauma and a lower risk for recurrences [51]. In absence of suspicious radiological, clinical features chances of missing an occult malignancy in an endometrioma is extremely low and surgery is not advised. But in later life risk of developing ovarian cancer can be a concern with the lifetime probability increasing from 1% to 2% in the presence of an endometrioma [52].

Conclusion

Need clear guidelines for when to treat, when to stay conservative and if need treatment then what mode of treatment out of available options should be used keeping in view risk of recurrence, reduced ovarian reserve and fertility concerns. All available options have their own benefits and risks. So in current circumstances we need to decide either to stay conservative or treatment depending on symptoms or need for intervention.

For more Articles on: https://biomedres01.blogspot.com/