2/26 VA report: pancytopenia, B12 deficiency

Thanks to Ling Wu for presenting an excellent firm report today: a 51 year old Sudanese man with subacute fatigue and weight loss, found to have pancytopenia and Coombs-negative hemolysis and ultimately found to have severe B12 deficiency.

Broadly speaking, pancytopenia results from:

  1. Bone marrow infiltration/replacement: malignancies, infection (TB, fungi), myelofibrosis
  2. Bone marrow aplasia: nutritional deficiencies, infection (HIV, Parvo B19), immune destruction, medications
  3. Cell destruction or sequestration: DIC, TTP, MDS, hypersplenism

Pancytopenia has a huge differential which can be looked up on UpToDate, so I won’t put it here.

How to work up new pancytopenia? Things to consider:

  • CBC w/ differential, BMP, LFTs (look for hemolysis/jaundice)
  • Peripheral blood smear
  • PT/PTT/INR, fibrinogen
  • B12/folate/iron studies
  • Infectious w/u: HIV, HBV/HCV, EBV or CMV as the history dictates, Parvo B19
  • Thorough medical review, consider autoimmune workup

Things like flow cytometry and bone marrow biopsy may need to be considered if the initial workup is unrevealing.

This patient was ultimately diagnosed with B12 deficiency.

Screen Shot 2018-02-26 at 1.51.47 PM.png

Normal mechanisms and defects of B12 absorption. The vitamin B12 (Cbl) released from food protein by peptic action is bound to haptocorrin (HC) in the stomach and travels to the duodenum, where pancreatic proteases digest the HC, releasing Cbl to bind to intrinsic factor (IF). The IF-Cbl complex binds to a specific receptor in the distal ileum (the cubam receptor) and is internalized, eventually released from lysosomes, and transported into the blood. NEJM 2013.


See the NEJM article below for more comprehensive information on causes of B12 deficiency.

Screen Shot 2018-02-26 at 1.44.02 PM.png

Clinical manifestations of B12 deficiency. NEJM 2013.


  • The B12 assay is imperfect: although an extremely low level (<100) is usually associated with clinical deficiency, such low levels are rare
  • Up to 50% of tests have either false positive or false negative values
  • Moral of the story: don’t use the lab’s lower limit of normal to reassure yourself that there’s no B12 deficiency. If they have compatible signs/sx despite a value above the LLN, they may still require supplementation.


  • The body’s daily requirement of B12 is 2.4 μg
  • Severe deficiency may require injected B12, sometimes lifelong if it is for treatment of pernicious anemia
  • Neither injection nor oral therapy are very efficiently absorbed
  • An increase in the reticulocyte count should be seen in 1 week and correction of megaloblastic anemia in 6-8 weeks


  1. Stabler S. Vitamin B12 deficiency. NEJM 2013.

1/29 VA report: clots clots clots clots clots clots

Thanks to Malcolm Kearns (not Lil Jon as the title might suggest) for giving an excellent talk on DVTs, related syndromes, and their management.

First, we talked about medications that can cause edema:

Screen Shot 2018-01-31 at 4.00.41 PM.png

Courtesy of Malcolm Kearns

The patient in this case was diagnosed with an SVT. But which veins are considered superficial?

Generally, superficial vein thrombi (SVTs) can be managed conservatively (compression stockings, NSAIDs, hot packs). But there is a role for anticoagulation even with SVTs IF:

  1. The SVT is associated with a DVT
  2. It is located near a junction with a deep vein (saphenofemoral or saphenopopliteal junction)
  3. There is a known or strongly suspected hypercoagulable state (cancer, etc), or the SVT is extensive or recurrent

This patient had repeat imaging that now showed DVTs. It’s important to know that treatment of DVTs can vary depending on where they are:

Screen Shot 2018-01-31 at 4.25.19 PM.png

Other reasons to anticoagulate isolated distal DVTs include:

  • Unprovoked DVT
  • D-dimer >500 ng/ml
  • Inpatient status
  • Immobility
  • Persistent or irreversible risk factors (malignancy etc)
  • Extensive thrombus
  • Symptomatic

We touched on several studies that studied the use of IVC filters for prevention of DVT/PE. Very briefly, the bottom line for some of them:

  • PREPIC (1998)
    • Groups: pts w/ proximal DVT, high risk for PE, placed on UFH vs LMWH w/ or w/o IVC filter
    • All patients were anticoagulated
    • Bottom line: IVC filters significantly reduced PE incidence at 12 days and 8 years BUT they also significantly increased DVT incidence at 2 years and 8 years
  • PREPIC 2 (2015)
    • Group: pts w/ acute PE, high risk for recurrence, placed on A/C alone or A/C + IVC filter
    • Bottom line: IVC filters did not reduce symptomatic PE at 6 months

In general, the take away is that IVC filters are not a panacea and should really be considered in patients in whom a recurrent PE would be catastrophic (likely due to poor cardiopulmonary reserve).

Other interesting takeaways:

  • There is no evidence to suggest that SCDs can mobilize a clot in someone who already has a DVT or SVT
  • There is also no evidence that exercise in someone with a PE can lead to recurrent PE
  • The Choosing Wisely Campaign strongly recommends against doing a hypercoagulability workup in patients with a known cause of DVT. In general, these should be left to the outpatient setting, ideally when patients are off anticoagulation

Ultimately this patient was diagnosed with phlegmasia cerulea dolens, an uncommon condition characterized by extensive and dramatic ileofemoral DVT. The clot can extend into capillaries and prevent blood from leaving the leg, leading to compartment syndrome, hypovolemic shock, and gangrene!

1/3 PPMC Report: New Insights on Treatment of NSCLC

Thank you to Dr. Ciunci for joining us to discuss the case of a 44 year old female smoker who presented with subacute cough and scant hemoptysis, found to have a new diagnosis of stage IV NSCLC adenocarcinoma.

We discussed that when lung malignancy is suspected, initial staging usually involves a PET scan as well as a brain MRI as patients often have intracranial metastases at the time of diagnosis. If imaging is suspicious for malignancy, then tissue biopsy is indicated to confirm the diagnosis. Tissue can be obtained via several routes, from endoscopic bronchoscopy biopsy to a more surgical approach with mediastinoscopy, or biopsy of other organs as indicated. If there is suspicion of metastases to distant organs, it is usually preferable to attempt to biopsy these lesions — if they are positive for metastatic lung cancer, that helps with staging and is important to ensure that there is not a separate process occurring in the organ with mets. The only exception may be with bony metastases — bone biopsies are not amenable to molecular testing, which is important to guide treatment.

Lung cancers are broadly divided into small cell (approximately 15%) and non-small cell (85%); NSCLC is subdivided based on histology. Regarding NSCLC, the traditional treatment for stage IV disease had been a combination chemotherapy regimen using a platinum-based compound. Recently, certain targetable mutations have been discovered in a minority of tumors, so it is important to always test all non-squamous NSCLC (and some squamous, like nonsmokers and young patients) for these mutations, including EGFR, ALK, and ROS1. If a mutation is detected, then the patient can be treated with a specifically-targeted chemotherapy and expect an increased progression-free survival.

More recently, immunotherapy has also entered the mix and has been amazingly effective in the right circumstances. It is now suggested that squamous NSCLC or other NSCLC that does not have a targetable mutation get tested for PDL1 levels, and if >50% then pembrolizumab can be considered as a first line agent! Here are some figures from a recent NEJM article about this topic.



Our patient had some palliative radiation and since she did not have a targetable mutation was started on a standard chemotherapy regimen — but had progression of disease. She had high PDL1 expression and was started on pembrolizumab with an excellent response, and has continued to do well for over a year!

Also for lung cancer, always remember to consider early involvement of palliative care — this has been associated with better outcomes for patients (see first reference below for more information).


Temel et al. Early Palliative Care for Patients with Metastatic Non–Small-Cell Lung Cancer. N Engl J Med 2010; 363:733-742August 19, 2010DOI: 10.1056/NEJMoa1000678

Reck, M and Rabe, KF. Precision Diagnosis and Treatment for Advanced Non-Small-Cell Lung Cancer. N Engl J Med 2017; 377:849-861August 31, 2017DOI: 10.1056/NEJMra1703413



HUP Report 12/14: Renal Infarcts and Sumatriptan

For this case, we discussed a 44 year old female who presented to the ED with flank pain, and on CT scan was found to have a unilateral renal infarct.  She was discharged with no clear diagnosis and then re-presented a few days later with bilateral infarcts.  On review of the scans, it was found that she had bilateral renal artery clots and further history revealed that she was taking increasing doses of sumatriptan for migraines.  The most pertinent teaching points were:

  1. Not all flank pain is a UTI, and having a “positive UA” in the absence of infectious does not denote a urinary tract infection.
  2. Renal infarcts are a rare presentation, but are most commonly due to Afib. Published case series of patients with renal infarcts has yielded the below differential (although this is by no means all-inclusive), and standard work-up includes TTE, halter monitor, and renal artery imaging.Capture
  3. Hypercoagulability work-ups are rarely indicated because they rarely change management.  They are even more complicated if a patient is anticoagulated.  I compiled a list of tests that are the most high yield to send if a patient is anti-coagulated and a work-up must be done.  This is a very unique situation, and should be approached with caution.Capture
  4. Triptans rarely cause non-trigeminal vasoconstriction, but have been implicated in a number of case reports of vasospasm in various organ systems (STEMI, CVA, Mesenteric Ischemia, Splenic Infarct).  This should remain on your differential when evaluating patients who are taking Triptans for migraine.

12/5 HUP Report: An unusual case of TTP

Today we discussed an unusual case of TTP, and along the way we reviewed the differential for hemolytic anemia and Evans Syndrome in particular. Thank you to our faculty expert, Dr. Cuker, for some amazing teaching! Our patient was a young female previously treated for melanoma who developed thrombocytopenia and hemolytic anemia. There are different frameworks in which to classify hemolytic anemias. Today we thought about immune vs. non-immune mediated processes (which is not perfect, and some etiologies can fall within both categories!). Here is some of the differential that we discussed:


Direct antibody test can be used to determine if there is antibody or complement bound to the patient’s RBCs. In this test, the patient’s RBCs are separated out from the serum and incubated with anti-human IgG; if there is antibody or C3 bound to the patient’s RBCs, then agglutination will occur, resulting in a positive test result. Indirect antibody testing looks at the patient’s serum to test for antibodies to RBCs – the serum is incubated with other RBCs (ie not from the patient), and then anti-human IgG  is added. If the serum contains anti-RBC antibodies, they will bind the cells and then agglutination will occur when anti-human IgG is added. (Figure adapted from Wikipedia.org)


Our patient was initially thought to have Evans Syndrome, which is a condition that involves at least two immune-mediated cytopenias, most commonly thrombocytopenia and hemolytic anemia. Corticosteroids are usually the first line of treatment, especially if the patient is presenting with anemia. If the patient is presenting with thrombocytopenia as the primary concern, then IVIg can be considered. These patients typically have relapsing courses, and this can be really difficult to treat.

Our patient initially did really well with steroids, but then had recurrence of thrombocytopenia and hemolytic anemia, this time accompanied by abdominal pain and a peripheral smear with many schistocytes, which had not previously been seen. She was ultimately diagnosed with TTP, which is a form of thrombotic microangiopathy caused by low ADAMTS13 level. In healthy patients, ADAMTS13 cleaves ultralarge Von Willebrand Factor multimers, so when activity is low these multimers can persist. This causes platelet activation and aggregation, which can lead to thrombosis and ischemia to end-organs, and a hemolytic anemia from shearing.


Treatment for TTP has made a dramatic change in mortality of this condition. Patients with suspected TTP should be treated immediately! Plasma exchange should be initiated as soon as possible, along with corticosteroids. There is also growing evidence to support use of rituximab in all patients without contraindications; this can be removed by plasma exchange, so should be dosed following sessions when possible.

References: please see citations underneath figures









Introduction to Thalassemia

Today we had a few special guests at Presby – Alex filled in for Hari, and (more importantly) Dr. Farzana Sayani gave us a awesome case study of Thalassemia.  Here is the summary of the high points of the talk:

Definitions and Diseases

(note: I will use small letters (ex: a, b) in place of corresponding Greek letters)

HbA = a2b2 (normally 96-98% of adult hgb)

HbA2 = a2d2 (normally 2-3.5% of adult hgb)

HbF = a2g2 (normally < 1% adult hgb)



  • Underproduction of a chains, thus relative over-abundance of b chains.
  • There are 2 a alleles (a1 and a2), and there are 2 of each chromosome, so a total of 4 a alleles

Silent Carrier = One alleles deleted (-a/aa)

  • Most Common, typically seen in African Americans (25% of all AAs are silent carriers)
  • CBC findings: Normal hgb, low-normal MCV (high 70’s)
  • Hgb electrophoresis: normal
  • DNA analysis: uncovers single a gene deletion


Double Deletion = 2a alleles deleted (aa/– or a-/a-), can be both on the same chromosome or on opposite chromosomes which only matters for prenatal counselling not phenotype

  • Typically seen in SE Asians
  • CBC findings: Hgb slightly low or normal (11-13), MCV low (~70)
  • Hgb electrophoresis: normal
  • DNA analysis: uncovers 2 a gene deletions


Hemoglobin H Disease = 3a alleles deleted (-a/–), but the real story is that all of the excess b globulin makes its own tetramers called HgH, but it’s not as good at its job as normal hgb

  • HgH hemolyzes more and doesn’t bind O2 very well, HgH inclusion bodies can even be seen on smear. Patients are often transfusion dependent
  • CBC findgins: hgb low (think 7’s), MCV low (think < 70)
  • Hgb electrophoresis: Can see many different patters of HgA vs. HgA2, but the big point is there is HgH


Hemoglobin Barts = all 4 a’s are deleted (–/–) which leads to Hydrops Fetalis, there is no a chains so g chains make tetramers that are called HbBarts.  But wait, why don’t we just get HgH again?!  It’s because the b globulin isn’t made until the baby is a bit older, at this stage the baby makes g chains instead.  All babies die within the first few days of birth (from heart failure from high output 2/2 severe anemia), but if they somehow lived long enough, yes they would make HgH instead of HgBarts.  Oh, and of course CHOP is doing some crazy stuff with in-utero transfusions, so maybe babies will start living long enough for exactly this scenario…


  • Similar story to above. Underproduction of b chains, thus relative over-abundance of a chains.
  • There are 2 b alleles, some mutations reduce production and some stop production all together

Minor/Trait = One normal b, and one abnormal

  • CBC findings: slightly low hgb (10-13), low MCV (60’s)
  • Hgb electrophoresis: presence of HbA2 (remember, that’s a2d2 so doesn’t require b), maybe some HbF but < 5%

Major = many possible genetic possibilities, but classically is a loss of function of both B genes

  • CBC Findings: Severe anemia usually transfusion dependent, MCV low (60’s)
  • Hgb electrophoresis: the point is there is ZERO HbA (remember, the one that’s a2b2) because there are no b to put in. Should have a lot of HbF and some HbA2

Intermedia = Essentially any combination of genotypes that has any CBC findings, but the point is that the don’t have absent HbA so can’t be called Major.  They may or may not require intermittent transfusions depending on the actual gene alterations.


Taking Care of Thalassemia Patients

  • Big issue is frequent transfusions leading to iron overload (usually goal hgb 9-10)
  • Many patients also need genetic counseling


Iron Overload

  • Most common complication is hypogonadism
  • High mortality is actually driven by heart failure 2/2 iron deposition (not high output 2/2 anemia)
  • Patient’s get frequent MRIs of heart and liver to monitor therapy, and see a ton of specialists to mitigate a number of complications
  • Treatment works, and patients can meet targets that adequately manage iron overload and leads to very long lives
  • Here is an outline of the common iron chelators:
    Property Desferoxamine Deferiprone Deferasirox
    Route of administration Subcutaneous, intravenous Oral tablets Oral suspension/Film coated tablet
    Remove liver iron +++ ++ +++
    Remove cardiac iron ++ +++ ++
    Reverse iron related cardiac dysfunction ++ +++
    Adverse effects Skin reaction



    Bone changes







    Rise in creatinine

    Elevated liver enzymes

    GI bleeding

    Laboratory Monitoring Weekly: CBC

    Monthly: CBC, liver enzymes

    Monthly: CBC, liver enzymes, creatinine, urinalysis

11/21 HUP Report: Pembrolizumab-Induced Myocarditis

Today we discussed a case that demonstrates an interesting side effect of immunomodulatory chemotherapy. Thank you to Dr. O’Quinn for joining us! The case we discussed today was a young female patient being treated for ovarian cancer with pembrolizumab, who presented with positional chest pain and was found to be in ventricular tachycardia with elevated cardiac enzymes. Her workup included a left heart catheterization that was negative for coronary artery disease, and a CT chest that was negative for pulmonary embolus.

One point that we discussed was the role for contrast to enhance echocardiograms. The use of contrast can be helpful to allow the cardiologists to get a more accurate estimate of LV ejection fraction and assess for presence of regional wall motion abnormalities; in this case the patient had a TTE without contrast that was difficult to interpret, and the addition of contrast provided a study consistent with reduced LV ejection fraction and moderately decreased RV systolic function.

Ultimately this patient’s presentation was thought to be myocarditis secondary to pembrolizumab toxicity. Pembrolizumab is a PD1 inhibitor, and many side effects of this therapy are related to activation of the immune system.


Cardiac side effects of this therapy are relatively rare, but have been reported in case series with presentations similar to our patient. Histologic studies show that pathology involves a lymphocytic infiltrate causing inflammation in cardiac tissue, which can manifest with arrhythmias and heart failure.

The primary treatment for autoimmune side effects of PD1 inhibitors is usually steroids. This patient was treated with corticosteroids and improved initially, but was unable to wean. Next steps in management were discussed with multiple institutions and departments within our own institution, as this is a newly described entity with limited experience. Ultimately, she was given infliximab and did really well!



Varracchi et al. Cardiotoxicity of immune checkpoint inhibitors. ESMO Open. 2017 Oct 26;2(4):e000247. doi: 10.1136/esmoopen-2017-000247.