2021 Update on Stroke Diagnosis and Management (Josephson, 1/27/2021)

If you have never heard Dr. Andy Josephson (neuro-hospitalist extraordinaire from UCSF) give a lecture, you are missing out. I was blessed to have Dr. Josephson as a teacher during medical school, and he gives lectures at the UCSF Hospital Medicine Conference every year. Always high high yield. Silver linings of pandemic life-- Grand Rounds speakers can speak from anywhere in the world. 

Here is a link to a recording of his presentation: https://youtu.be/QWkQLJfRNBo. For those of you who prefer the written word, here are the highlights:

• 15% strokes are hemorrhagic, 85% strokes ischemic 
• the ONLY way to tell them apart is a non-contrast head CT
• Time of onset=last time seen normal 
• DON'T ask When did your stroke start? 
• Rather ask When was the last time you were completely well?

UCSF "Stroke Protocol" for evaluation of acute stroke symptoms

• Non-contrast CT of the head (to r/o hemorrhage stroke)
• CT Angiogram (from aortic arch to the top of the head)
• CT perfusion study (CTP), in image above areas red are infarcted/dead, green areas are at risk but could be salvaged
• Post contrast CT of the head

Speed matters: time is brain; for every 15 minutes earlier administration of tPA,

• significantly lower in house mortality
• significantly lower rates of intracranial hemorrhage (ICH)
• significantly more independent ambulation at discharge
• significantly higher rate of discharge to home

2021 Acute Stroke Timeline

• 0-4.5 hours since last seen normal--> IV tPA 
• 0-6 hours--> mechanical embolectomy for all with large vessel occlusion (LVO), done in interventional suite, requires transfer from SSRRH
• 6-24 hours-->  mechanical embolectomy for SOME (based on CT perfusion)--> this is relatively NEW

Mechanical Embolectomy

• In 2015, there were 5 major trials in NEJM that show embolectomy helped if done within 6 hours of symptom onset
• CT Angiogram is used to determine if there is a LVO (many have in small blood vessel that will respond to tPA but cannot be removed)
• 2018 DAWN and DEFUSE3 studies showed benefit in certain population up to 24 hours out using CTP (a lot of green but very little red, then the clot should be removed)
• now CTP helps us decide who can benefit from clot extraction

Everyone within 4.5 hours should get tPA

If you are within 6 hours, you should get embolectomy

If you are within 6-24 hours, CTP should drive the decision of who should be transferred for embolectomy

Standard work up for Large Vessel Stroke

• Cardioembolic: atrial fibrillation, clot in heart, paradoxical embolus (telemetry, TEE with bubble)
• Aortic arch (TEE with bubble)
• Carotids (CTA, ultrasound, MRA, angiogram)
• Intracranial vessels (CTA, MRA, angiography)

TEE is superior to TEE for: LA appendage, R to L shunt, examination of the aortic arch. TEE finds additional findings in 52% and changes management in 10%. However, most hospitals still do TTE first and then only TEE in limited cases.

Atrial Fibrillation can be missed if you don't monitor long enough

• Standard care is EKG, 48 hours of telemetry
• Everyone with stroke with unclear etiology should get long term cardiac event monitoring (21-30 days): 15-20% of patients with cryptogenic stroke otherwise unexplained had afib detected, clearly changes managemetn, probably cost effective

Should we anticoagulate in stroke? Nope, unless a fib, rare hypercoagulable states (e.g. APLS)

What about PFO? 1/4-1/5 the population has a PFO (true in stroke patients too)

Really only a VERY large PFO make people at risk for stroke. 3 large trials <2017 showed that closing PFOs didn't give any benefit. In 2017, 3 trials found that a SUBGROUP of patients with stroke and PFO would benefit from closure. Don't close all of them!

Consider PFO closure if patient

1.  <60 years AND 
2. You can be sure that PFO is most likely etiology after a thorough work up AND the qualifying event is a stroke (not a TIA) that appears embolic (not lacunar). Probably only large PFOs or those with an ASD
3. Moderate or Large PFO only

What about heparin? Nope. No indication for heparin in stroke.

Anti-platelets are the mainstay of stroke management

• ASA 50mg-1.5gm equal long term efficacy, so go with 81mg (always fine)
• Aggrenox (effective but rarely used due to $$, headache, BID)
• Clopidogrel is effective
• Of note, there is NO long term benefit in dual anti-platelet therapy (DAPT)-- that is, combination with ASA (should stop long term DAPT)
• BUT DAPT may have benefit for short term post minor stroke x 21 days
• Ticragelor can also be used in combination with ASA for short-term, but study shows more bleeding

Permissive Hypertension: 220/120 unless IV tPA then max is 185 systolic x 24 hours

when to stop is controversial, start htn meds before discharge (~72 hours) and aim for normo-tension over a matter of weeks. Choose thiazides and ACE-I first (best secondary prevention)

TIA vs. Stroke (up to 30% of TIA have infarct on MRI)

• Conceptually same disorder, same workup same treatment
• TIA should be aggressively worked up  because of risk of stroke is highest during the 7-14 days after a TIA
• Aggressive TIA treatment with ASA/statin/evaluation: you can reduce risk of stroke in short-term by 80% 

Prenatal Genetic Screening and Diagnosis (Shaffer, 1/20/2021)

Many thanks to Dr. Brian Shaffer, Sutter West Bay Medical Group Perinatologist, who gave an informative Grand Rounds this week on Prenatal Genetic Screening and Diagnosis. Prenatal Screening is a BIG topic that is ever evolving with improving gene sequencing technology and innovation, necessitating important (and potentially challenging) decision-making conversations with women early in their prenatal care.

Just a reminder that there are important differences between genetic screening tests (e.g. first and second trimester blood screens, nuchal translucency ultrasound, and even our standard fetal survey) and prenatal diagnostic tests (e.g. chorionic villous sampling, amniocentesis and cord blood in terms of their positive predictive values as well as their risks. The new-ish kid on the block is non-invasive prenatal testing/screening (aka NIPT, NIPS), which is a screening test and NOT diagnostic. 

Of note, ACOG and other guiding bodies  recommend that a discussion of the risks, benefits, and alternatives of ALL prenatal screening and diagnostic testing should occur with ALL patients, ideally at first visit, regardless of maternal age, including non-invasive prenatal screening (NIPS) and diagnostic testing "even in low risk pregnancy" (e.g. CVS, amniocentesis)

Dr. Shaffer gifted us with a set of probing questions to conduct these conversations about prenatal screening with patients. This can be a really challenging conversation, and it is helpful to consider the language we use to frame the conversation. Consider starting with:

If your pregnancy was affected with a genetic disorder or chromosome abnormality, would you want to know?

• What would the information mean to you?
• Would an abnormal result or prenatal diagnosis ruin the experience of the pregnancy or provide valuable guidance?
• Some people want more information to help them (and healthcare providers) prepare for the birth of a child with special needs
• Some might choose to terminate a pregnancy with a diagnosed condition
• Some couples do not want this information during pregnancy

There is a spectrum of congenital disease that can affect a pregnancy (as depicted in image below), and no ONE screening test can provide results about the entirety of this spectrum. Really, in fact, a combination of screening tests aims to cover this range (e.g. ultrasound is best to evaluate structural malformation vs. a quad screen designed to detect some autosomal disease, NIPT is excellent for down syndrome but gives no information regarding spina bifida). 

The menu of screening tests we currently offer includes: 

1. Nuchal translucency ultrasound (NT),
2. First Trimester screening (serum),
3. Integrated/sequential screening (first and second trimester), 
4. Quad screen, and 
5. Cell free DNA (often referred to as NIPS or NIPT).  

Each of these tests has benefits and limitations as well as variable sensitivities for different conditions. See table below for some of these stats. For example, for Down Syndrome (DS), the NT alone has a 70% sensitivity as compared to the NT plus serum markers which brings that to 90-95%. 

However, also note that because these conditions are quite rare, the positive predictive value (PPV) is only 5-10%. That is only 5-10% of positive screens will actually correlate with a abnormality in the fetus. And the impact of false positives is real.

What are the implications of this 5% PPV?

Well, let's take an example: with the prevalence of DS in the US, 1/16 women with a positive screen will have a child with DS. 15/16 will have a normal fetus. And if the abnormal screen leads to more invasive diagnostic testing (CVS, amnio)--> that equates to 570 losses of normal fetuses. Is this an acceptable tradeoff? Up to each woman to decide. And up to us to counsel those women.

What should we know about NIPS?

NIPS is a pretty awesome technology that is able to identify cell-free DNA segments (from the placenta) in maternal serum. It can be done from 10-30 weeks EGA and has a a very high positive predictive value in "high risk pregnancies" (defined as women with advanced maternal age/AMA). In younger women-- lower risk women-- while the sensitivity and specificity remain the same, the PPV is decreased because the prevalence of these chromosomal abnormalities are just so much rarer. 

NIPS screens for Trisomy 21 (Downs), Trisomy 18, and Trisomy 13, as well as sex chromosome abnormalities (XXX, XXY, XY) and some other microdeletion and microduplication disorders.

Results return as:

LOW risk

INCREASED risk (including borderline)

Other (including no call, maternal chromosomal conditions, and inconclusive/atypical)

A "No call" (aka test failure) occurs in 1-8% of patients, often due to low fetal fraction. Low fetal fraction may be due to too young gestational age, as well as maternal conditions (e.g. SLE) and women on lovenox. Can also be low in obese women (<1% in women <60kg, >50% in women >160mg). Can also be low in placentas with certain aneuploidies (particularly Trisomy 18 and triploidy). NIPS cannot be used for paternity or single gene deletions.

This table from Society Maternal and Fetal Medicine summarizes NIPS key points nicely

Dr. Shaffer cautioned us that there is no "best test"-- up to 2% of of pregnancies with a conventional positive screen had a detectable chromosomal abnormality not detectable on NIPS.

What about ultrasound (including NT and 2nd trimester screening ultrasound)?

• Between 11 through 13+6 EGA
• NT<3.5mm is considered normal
• If  an abnormal NT is detected, recommendation include 1) formal genetic counseling 2) invasive testing (CVS) as well as 3) fetal echocardiogram

What about advanced paternal age (APA)?

• in men >40, associated with small increased risk for de novo autosomal dominant disorders
• possible increase in behavioral issues including schizophrenia and autism
• structural malformation
• fetal growth restriction and preterm birth

Expanded carrier screening vs. Ethnicity Based Screening

Over the last decade, we have been offering women "ethnicity based screening'-- e.g. offering CF, Tay Sachs, hemoglobinopathies based on a maternal report of heritage. There is evolving argument in the literature that we are very unaware of our genetic heritage AND that there may be benefit to offering all women expanded carrier screening for conditions that might impact pregnancies and pregnancy outcomes. As expanded carrier screening cost come down, and as our capacity to test for a large number of conditions simultaneously, the argument is that more information in the hands of the mother can be helpfu. However, the risk is that we currently have no uniform or standard best practices on these ECS panel, and such testing may give us information on late onset genotypes (e.g. BRCA1) that parents don't really want to be thinking about.

In summary,

All screening and diagnostic testing are optional in pregnancy and women should ALWAYS have the right to decline.

Remember that a screening/anatomy ultrasound is still a screening test (and women should know this)

If a patient wants screening, they should have the most accurate screening tests available

Providers should be able to explain what conditions we are screening for, explain false positives and detection rates, and have a plan for results if they are anything other than "normal"

Any questions: contact Dr. Shaffer at shaffebl@sutterhealth.org

Pulmonary Hypertension (Wang, 1/13/2021)

Dr. Helena Wang gave an excellent Grand Rounds this week on Pulmonary Hypertension. 

You know when you have read about a particular topic a hundred times and you still don't quite grasp even the most basic of concepts? Well, that is pretty much how I have felt about pulmonary hypertension for the last decade. I get all tangled up in the RVSP and the strange WHO Classifications and have never quite been able to solidify my illness script for pulmonary hypertension. 

Well, Dr. Wang, thank you. I feel like I am finally here. Pulmonary hypertension is such an interesting intersection between the lungs and the heart and a good opportunity to review basic physiology too! As one of our residents said to me yesterday after the talk, " I feel like hers is a presentation I want to watch again and again." Me too (and I did!). For those of you who want the written summary, here goes:

First a refresher on Pulmonary Function Tests (PFTs):

• The FEV1/FVC Ratio should be used to determine if a patient has an OBSTRUCTIVE lung disease. 
• Remember that an FEV1/FVC <70% is consistent with an obstructive disorder (e.g. asthma, COPD, etc)
• Total Lung Capacity (TLC) is used to decide if a patient has a RESTRICTIVE lung disease
• 80-120% is considered normal 
• Patients with restrictive lung disease will have a TLC <80%
• Diffusion capacity for carbon monoxide (DLCO) provides information on the efficiency of gas transfer from alveolar air into the bloodstream
• for these defects in gas exchange, ddx includes pulmonary hypertension and liver disease

Second, reminds the pulmonologist, don't forget to pay attention to the right side of the heart in the echocardiogram!!  Dr. Wang urged us to scroll down and look at the text. Is the RV big? What is the velocity of the regurgitant jet? What is the TAPSE?

• A first hint of pulmonary hypertension on a TTE comes on apical 4 chamber view where the R ventricle shows turbulent flow back into the R atrium
• RVSP (right ventricular systolic pressure) is an estimate reported on most echo reports and can be very challenging for the Echo tech to calculate
• First they measure the velocity of tricuspid regurgitant jet (velocity >3 is a clue that there may be pulmonary hypertension present)
• then they guestimate of CVP (0, 5, 10) based on compressibility of IVC
• NOTE: a calculated RVSP>25 is considered HIGH right sided pressure
• The R ventricle is often very plump and dilated under high pressure (so. . .a big chubby RV on echo is another clue that you might have pulmonary hypertension)
• Tricuspid Annular Plane Systolic Excursion (TAPSE): RV contracts in a twisting motion with each contraction as RV shortens. Normal excursion is 1.5cm. 
• If RV is volume overloaded and dilated, cannot squeeze as well, not shortening, TAPSE drops <1.5cm on echo

And finally, right heart catheterization is gold standard for measuring R sided pressures, but really this should be done after you have done a full evaluation (see below) once you are pretty certain you have primary pulmonary hypertension (WHO Group 1)

WHO Classification of Pulmonary Hypertension

• WHO Class 1: Pulmonary ARTERIAL Hypertension (it's like going from a 6 lane highway down to a 2 lane highway--> pressure goes up, fluid backs up)
• WHO Class 1 has the worst outcome: 5 year survival 61%
• only once other causes of pulmonary hypertension are identified and treated
• Seen in HIV (regardless of viral load, duration), Connective tissue disease (e.g. scleroderma), portal hypertension
• All other WHO classes are actually pulmonary VENOUS hypertension (something downstream to the lungs causing poor forward flow, back up of blood, then high pressure on R side of heart). These include
• WHO 2: L heart disease
• WHO 3: bad lung disease or hypoxia
• WHO 4: chronic thromboembolic pulmonary hypertension (small, little tiny chronic)
• WHO 5: potpourri

Once Pulmonary Hypertension is identified, the next question is DOES the patient have pulmonary hypertension that is primary or secondary

Ddx pulmonary arterial hypertension (WHO Group 1)

• idiopathic
• heritable (no current gene therapies)
• drug and toxin induced (fen fen, methamphetamines including stimulant medications prescribed)
• connective tissue disease (scleroderma)
• HIV
• portal hypertension
• congenital heart disease
• Schistosomiasis (#1 cause of pulmonary hypertension outside the US)

Ddx WHO Group 2: a bad heart

• left heart disease: LV dysfunction (systolic and diastolic), valvular dz, outflow obstruction

Ddx WHO group 3: bad lungs

• asthma and COPD (need PFT: spirometry, lung volumes, diffusion capacity)
• ILD (need HIGH resolution CT chest)
• sleep disordered breathing (episodic hypoxia at night causes vasospasm), need sleep study
• chronic high altitude
• developmental lung dz

A quick interruption to review the three types of CT chest you might consider to evaluate the lungs:

• CT angiogram (aka CTA) is a very quick scan from feet to head chasing the dye, itskips very low bases and apices of lungs and should never be used to assess for interstitial lung disease
• A High Resolution CT Chest: much thinner cuts (2mm) than a CTA starting at apices and going all the way to the bases. 
• Very high detail of parenchyma
• Ground glass opacities can be atelectasis vs. ILD--> flip when prone to see if atelectasis resolves while doing high res scan
• CT w/wo contrast (e.g. nodule) 
• can see 4 generations of branches of bronchial tree, but keep in mind that there are 17, so you cannot see filling defects in the small peripheral vessels

Ddx WHO group 4: chronic thromboembolic pulmonary hypertension (itty bitty clots on VQ scan)

• need VQ scan to get the level of detail in microvasculature

Ddx WHO group 5: potpourri

• hematological disorder (myeloproliferative)
• systemic disorders (sarcoid, LAM--> chylothorax)
• glycogen storage disorder
• thyroid disorders
• chronic renal failure
• mechanical (fibrosing mediastinitis, tumoral obstruction)

Right heart catheterization

Once you have maximally treated all WHO 2-5 diagnoses, treat as aggressively as possible (e.g. diastolic dysfunction, COPD/asthma, chronic clots, sleep apnea, etc), then you repeat ECHO and see if there is still elevated pressure on R side of heart--> NOW, patient should get a right heart cath to confirm exactly what the pressure is.

• On A R heart cath, PA pressure normal <25/15
• Mean PA pressure normal is <25
• Pulmonary hypertension = mean pressure >25

• Gold standard is also to do a vasodilator challenge during R heart cath to see if there is reversible spasm in the pulmonary arteries? (inhaled nitrous oxide, epoprostenol, or adenosine--> look for decreased pressure)
• if PA pressure decreases by 10 mmg HG, ending pressure <40, no drop in cardiac output
• less than 10% of patients have + response to vasodilator
• calcium channel blockers are treatment of choice for those who do have vasospasm

When to treat Pulmonary Arterial Hypertension?

There is NO number for pressure goals. Look at patients functional class, 6 minute walk, and Echo

Treatment goals:

• Functional class: symptoms at rest (class IV), normal (class I): goal is 2 or less
• 6 minute walk test: goal is 400mg in 6 minutes ("please walk as far as you can in 6 minutes")
• BNP: can be elevated in R heart strain, goal is normal
• Echo: goal is RV not fat and chubby, decompressed and slender, TAPSE moving nicely

Pulmonary Hypertension Therapy

• Endothelin pathway: block (bosentan (black box liver toxicity, no longer used), ambrisentan, macicentan: side effects: edema, check LFTs, watch for rare malignant facial edema
• Nitric oxide pathway: enhance (sildenafil TID, tadalafil QD, no lab monitoring, can get blue green colorblindness, rare complication nonarterici anterior ischemic optic neuropathy (sudden blindness one eye, reversible when stop), riociguat (TID, more systemic hypotension)
• Prostacyclin pathway: enhance (potent, short half life and unstable)
• epoprostenol: IV infusion, tubing connected to a pump, best mortality data, very labor intensive
• treprostinil: IV vs. sq (like insulin pump) and inhaled QID (takes a lot of time, breathing treatment like nebulizer, 20-30 minutes per treatment
• iloprost: inhaled 6-9 times/day
• selexipag: oral BID, no labs but unrealistic uptitration protocol

Guidelines for treatment are based on functional class. The ultimate goal is minimal symptoms with regular activity

If functional class IV-->  needs prostacyclins right away

If functional class III-->  can start with orals (e.g. tadalafil and macicentan)

Any questions? Email or staff message Dr. Helena Wang at: wanghl@sutterhealth.org

Augmented Intelligence in Primary Care (Mendoza 1/6/2021)

Many thanks to Dr. Emmanuel Mendoza for his Grand Rounds tour of the current state of technology in primary care and the emerging role of artificial intelligence (or as he reframed as "augmented intelligence") in primary care.

Dr. Mendoza reminded us that the transition in medicine to electronic medical records (EMR) has happened intensively and quickly over the last decade, leading to a great change in how we deliver healthcare and affecting our experience as clinicians caring for patients. Dr. Mendoza recommends three books on this transition for your pleasure reading: Deep Medicine, The Patient Will See You Now, and The Digital Doctor (see image below).

While the transition to the  EMR was touted as something that would be helpful and advantageous to physicians and healthcare delivery, many providers have felt professionally burdened and by the EMR. 

• 70% of clinicians report that the EMR increases their work hours
• 70% of clinicians believe that the EMR is contributing to burn out
• 70% of clinicians say that the EMR takes time away from patients

Are machines going to replace doctors?

Dr. Mendoza thinks not. He thinks, in fact, that computers and AI and machines should be thought of as tools that are designed to enable us to practice better medicine. Physicians are still needed because medicine is not just a set of algorithms. This includes a wide range of evolving AI technologies:

• Image processing

e.g. benign vs. malignant lesions

e.g. radiology (cxr reading)

e.g. diabetic retinopathy

• Patient engagement

via APPS

via Health trackers

via E-health programs

Dr. Mendoza covered a long list of different technologies and AI programs, and he reminded us that many of these technologies are works in progress; some will be raging successes and are more effective and more applicable than others. In fact, many will ultimately fail. Be aware of where the product is in its natural evolution (see image below).

Dr. Mendoza encouraged us to engage with the technology and explore them as they are being developed. Here are a few programs/apps/ideas for you to try in your spare time:

Technology you should play round with!

1. Patient symptom checker:  Symptomate: https://symptomate.com/

2. Preventive health via Facebook Preventive health: https://preventivehealth.facebook.com/

3. Chronic disease management via Omada: https://www.omadahealth.com/

4. Clinical Decision support: https://www.ibm.com/products/micromedex-with-watson

5. Medical education: FOAM: https://litfl.com/foam-free-open-access-medical-education/

6. Diagnosis: Human Diagnosis project: https://www.humandx.org/

7. EMR Voice Assistant: https://www.epic.com/epic/post/hey-epic-tell-voice-assistant-clinicians

8. Digital assistant for physicians: https://www.suki.ai/

Finally, Dr. Mendoza encouraged us to check out the AAFP innovation lab, where you can sign up to trial many of these technologies and give direct feedback.

https://www.aafp.org/family-physician/practice-and-career/managing-your-practice/health-it/innovation-lab.html