Sunday, 14 December 2008

December Case: The Long Awaited Answers

Dear Bloggers

Thank you for waiting. This case is very difficult and it would be the kind of grey-case that one would be given in the UK Membership examination (MRCP). In such exams, only the relevant data is provided with much focus on the history and physical examination with some basic radiology and salient, albeit selective laboratory data to give hints but not to lead the doctor directly to the answer. To give all the studies to you would have made the case more easy to answer but would not have been such a challenge. Nevertheless, from the information provided, it should at least have given you some clues as to what was happening in this patient.

Dr Keita Tatsuno, Department of Infectious Disease, Tokyo University, Japan has kindly sent in his answers to the case as follows:

Very interesting and difficult case. I try to answer, but please forgive me not in order. First, I guess the valve is mitral valve. Because the pansystolic murmur suggest the regurgitation. If the murmur were associated with ejection, the climax of sound would be at mid contracting phase. And the radiation to the left axilla fit in the direction from left ventricle to left atria.

Second, the subsequent loss of consciousness suggested another etiology than MSSA bacteremia. Usually, it might be typical pattern to suspect the embolic event in central nervous system, especially in acute infective endocarditis case. Of course, still there is some possibility of too early CT imaging study and need to check the MRI image. But in this case, it is not typical for emboli that the neurological abnormality ranged from loss of conscious to decreased muscle tone. I want to think such etiology as low blood glucose, hyponatremia, drug, hypothyroidism and so on, which affect all of the central nervous system including reticular activating system.

I want to check up these problems primarily, i.e. sodium, calcium, pH (arterial blood gas), glucose, serum creatinine. Especially, I am curious about sodium level. Because you mentioned she had very high urine osmolality. These sound suggesting SIADH.

Dr Stein, Florida, USA, has kindly provided answers to the case as follows:

1. Valve: Mitral-regurgitation
2. Hypoperfusion of kidneys and brain from hypovolemia and or relative hypotension [pseudo-normalization in hypertensive pt] from distributive (vasodilatation) near shock from sepsis. Cold extremities and modest tachycardia may be marker for vasoconstriction. If tent sign = extensor skin remaining up when lifted up, that is unreliable in elderly who have + sign from loss of elastic tissue. Axillary sweat absence may be indeterminate for volume status in septic patient. Another possibility is septic emboli from mitral valve to brain such as vertebral artery system to mid-brain system. Also electrolyte abnormalities such as hyponatremia may account for deterioration. She entered in hemocontrated and hypovolemia state. High urinary osmolality might indicate SIADH. However SIADH can only be diagnosed in normo-volemic state.
Another aspect is the acute IE on assumed diseased mitral valve causing acute valve perforation and acute pulmonary edema. Rarely hypothyroid and adrenal insufficiency may co-exist or exacerbate from the IE septic state. 3. Neuro involvement includes diffuse higher cortical dysfunction as found in hypoperfusion, sepsis and metabolic derangements all of which this patient had. I teach not to use the term `Dolls eyes positive.` What does this mean in simple English?: dolls eyes present(+) which is normal brain stem response or dolls eyes abnormal which is brain stem dysfunction(+sign). Hence Positive/+ has 2 opposite meaning. To avoid both inter/intra/transcultural confusion, only normal(present) or abnormal(absent) is required to avoid your Brit and US confusion. Do not depend on `tradition` to avoid this pitfall.

4. Fluid challenge/resuscitation. Follow hourly urine output increases with bolus and high DIV infusion guarding for heart failure, follow cardiac ECHO for valve perforation. Also repeat brain CT if suspect brain septic emboli.

5. See above

Hirotaka Kato, a 6th-year-student of Kumamoto University has provided an answer to the case.

Q1. Which valve is affected by the confirmed endocarditis?
Mitral valve. Radiation to axilla and becoming louder during expiration are suggestive.

Q2. What is the likely etiology of the other cardiovascular clinical signs and the loss of consciousness?
#1 hypovolemia
#2 sepsis
From the first physical exam, hypovolemia came up to my mind first. Tachycardia, slightly low blood pressure, cold fingers and dry membranes indicate that. Despite of IV fluid + antibiotics administered, the patient has been worsening, however. To me, high Hb/Ht and a high urine osmolality keep the possibility of hypovolemic state. And also, sepsis should be considered in this case, I think.

Q3. Which part of the nervous system has been affected?
Given areflexia and hypotonia, the cerebellum or neuromuscular junction or muscles can be considered.
In this patient who is likely to have the DIC state (or CNS infection possibly), the cerebellar involvement is most likely.
Neuromuscular junction (Lambert-Eaton) is also considered, but less likely now because malignancy cannot unify the symptoms.

Q4. What would be your next test?
1 physical exam. - vital sign + Is there now no edema?, cold extremities?, dry membrane?, no skin rash?
2 Blood test - Cr, BUN, Na, K, Ca, Glucose, NH3, Blood smear

Q5. What might be the cause of the neurological condition present?
Head CT and CSF analysis didn't show any abnormalities! I would like to list two as the possible causes.
1. septic or hypovolemic shock leading to hypoxia in the brain especially the cerebellum
2. shock + paraneoplastic syndrome.

Dr Yosuke Ebisu of Kameda Hospital, Chiba, also provided a reply to the case.

I read your blog. It is interesting case!! I think like that.
(Problem lists)
#1 fever
#2 obtundation
#3 Cheyne-Stokes respration
#4 oliguria
#5 splinter hemorrhage
# endocarditis caused by MSSA
Now you have treated it with fluoroxacillin and gentamycine. After that the patient became gradually obtundation,and her urine output decreased.
Then I think three passibilities.First,it is heart failure caused by destruction of the mitral vulve.It is because HF cause reduction of effective blood flow and then oliguria occur. Then the blood flow of medulla reduce, it cause Cheyne-stokes respiration.
Second,renal toxicity of gentamycin.It cause oliguria and if it became severe,the patient may have uremia. Then her consciousness will be impaired.
Third, septic emboli. It cause cerebral infarction. According to your blog, her extremities were flaccid and doll's eye movement was intact. So I suspect infarction of medulla.(Cheyne-Stokes respiration support this.)
-check heart sound(S3,S4),JVD and edema
-echo cardiogram: check the valve.(If the valve is destroyed we should think operation soon.)
-blood test,especially Cre and BUN.(If renal function decrease, we shoud change the drug or it's dose.)
-Head MRI

Here is the answer from Professor Matsumura, Kanazawa University of Medicine.

Thank you for showing me an interesting case.
I will try to make my hypothesis.

Questions: 1) From the physical examination, which valve is affected by the
confirmed endocarditis?

This case showed classic features of acute infective endocarditis. There
were several splinter haemorrhages in the nails and Janeway lesions on the
Cardiac examination disclosed Levine II/VI pansystolic murmur in the APEX
area radiating to the left axilla accentuated on expiration.
Vegetation must be on the mitral valve.

2) From the history and first physical examination, what is the likely
aetiology of the other cardiovascular clinical signs (excluding those of
endocarditis) and the subsequent loss of consciousness?

This patient had high fever. I suspect that this patient had volume
depression and pure water deficit following findings. Cold finger and toes,
Pulse 100/min regular, BP 100/80, JVP not raised, dry skin, dry mucous
membranes, tenting sign positive, no axillary sweating.

3) With the follow-up neurological examination in mind which part of the
nervous system has been affected?

This is the most difficult question of this patient.
I would make problem list.
#1 Altered mental status, deep coma
#2 Doll's Eye sign positive
#3 Tone diffusely decreased
#4 Areflexia
#5 Cheyne-Stoke respiration
#6 Anorexia
#7 Oliguria
#8 Pure water deficit and volume depression (Hb 18, haematocrit 0.5 and a
urine osmolality of 1000)

I have to differentiate the cause of #1 altered mental status first.
Mnemonic of altered mental status is AIUEOTIPS,“あいうえおちっぷす” in

A:Alcohol, not likely
E:Encephalopathy(hypertensive, hepatic), not likely, Endocrinopathy
(adrenal, thyroid), not likely, Electrolytes(hypo/hyper‐Na, K, Ca, Mg)
O:Opiate , not likely or other Overdose, not likely,decreased O2
(hypoxia, CO intoxication

T:Trauma, not likely, Temperature(hypo/hyper)
I:Infection(CNS, sepsis, pulmonary)
P:Psychogenic, not likely, Porphiria, not likely
S:Seizure, not likely, Shock, Stroke, not likely, SAH, not likely

In case with endocarditis, septic emboli is the common cause of neurological
problem. But head CT disclosed no abnormalities. #2 Doll's Eye sign positive
indicates that the brain stem of this patient was intact.
#3 Tone diffusely decreased and #4 areflexia indicated PNS and/or muscle
problem, not the problem of upper motor neuron. #5 Cheyne-Stoke respiration
can indicate bilateral forebrain disease or diffuse bihemispheric disease,
Cheyne-Stokes respiration is now known to occur with unilateral hemispheric
and brainstem infarcts (UpTo Date).
It is difficult to point out the one part of neuro-anatomical problem of
this patient.
Hypo/hyper‐glycemia, uremia, or hypo/hyper‐Na, K, Ca, Mg should be

4) What would be your next test to confirm your suspicions?
I want to know following test.
Na, K, Cl, Ca, Mg, BUN, Cr, glucose, serum OSMO, blood gas analysis.

5) What might be the cause(s) of the neurological condition present?
My most likely diagnosis is hypernatremia, hyperosmolality. Serum sodium
must be high more than 160 meq/L. And BUN and serum creatinine must be high
level too.

Izumi Nakayama, 6th year medical student of Osaka University named the correct diagnosis below.

.... I tried this month's case on your blog. It is the first time to send you my answer.

My answer for the December case...

Questions: 1) From the physical examination, which valve is affected by the confirmed endocarditis?
1. Mitral regurgitation sound. (it is common but accentuation on expiration is unusual for the left side heart murmur, isn't it? )

2) From the history and first physical examination, what is the likely aetiology of the other cardiovascular clinical signs (excluding those of endocarditis) and the subsequent loss of consciousness?
Her dry skin, dry mucous membranes indicate she had been dehydrated.
I'm sorry but I don't know 'the tenting sign'. What is it like?
1. brain stem infarction. small thrombi from mitral vulve.(dehydration increases the risk of thromboembolism.)
2.central pontine myelinolysis (although chronic hyponatremia is not likely because she had been dehydrated. I don't know CPM can occur in rapid adjustment of dehydration or hypernatremia.).
3) With the follow-up neurological examination in mind which part of the nervous system has been affected?
positive Doll's sign and generalized weakness indicate the lesion either in the midbrain or in the pontine.

4) What would be your next test to confirm your suspicions?
MRI including brain stem. repeat neuro exam. plasma sodium concentration, plasma osmolarity, .
5) What might be the cause(s) of the neurological condition present?
same as question 2.

My impression is like above. First I came up with the infarction. But I think it cannot explain her gradual declining.

Many thanks for the excellent opinions sent in this month.

Now, let's go through the answers.

1) From the physical examination, which valve is affected by the confirmed endocarditis?

This patient has a pansystolic murmur in the left apex which radiated to the axilla. It is accentuated by expiration which makes it a left sided cardiac murmur. By definition, pansystolic murmur in the aforementioned location and attendant features would suggest mitral regurgitation.

2) From the history and first physical examination, what is the likely aetiology of the other cardiovascular clinical signs (excluding those of endocarditis) and the subsequent loss of consciousness?

The other cardiovascular signs include cold hands and feet (suggesting peripheral vasoconstriction), tachycardia, low blood pressure, tenting of the skin and absent axillary sweating are all consistent with hypovolaemia e.g. as a result of dehydration.

The tenting sign is procured by lifting a piece of skin with the examiners finger tips to produce what looks like an upside down V sign or what is commonly referred to as a 'tent'. In a patient with normal skin elasticity, if there is normal hydration, the skin should immediately retract to its normal flat position. In the presence of dehydration, the skin takes longer to return to the normal position producing a positive sign. However, in the elderly who have loss of elastic tissue, the sign can be falsely positive. Although the tenting sign is usually performed on the dorsum of the hand or forearm (extensor surface), the best place to check for tenting sign in the elderly is the inner thigh.

3) With the follow-up neurological examination in mind which part of the nervous system has been affected?

This patient had a completely normal nervous system examination on admission. However, the follow-up examination produced a picture that should have been interpreted as generalised flaccidity, generalised areflexia and indeterminate Babinski suggesting either an acute brainstem or diffuse cerebral injury. However, with the presence of Cheyne-Stoke respiration and positive Doll's eye sign* (abnormal), a brainstem lesion would be highly suspected. However, as we ascertain from the physical examination, the pupillary reflexes are normal suggesting that it does not completely involve the midbrain. That leaves the possibility of the medulla oblongata and pons being affected.

*note: I take heed of Dr Stein's advice and agree about positive / negative aspect that can cause confusion in respect of the Doll's Eye sign. It is probably better to describe what is seen so that everyone can understand whether the sign is present or not rather than as I originally described it. Apologies. Thank you Dr Stein for the very constructive advice.

With a normal CT brain examination, cerebral oedema would seem less likely and a significant structural brain lesion in the cerebral hemispheres would be ruled out, as would bleeding. It is unfortunate that the opening pressure of the CSF was not measured in this patient which is the usual way for understanding if raised intracranial pressure is truly present rather than just relying on a seemingly normal CT scan.

Infarction might be less easy to rule out especially in acute neurological deterioration and particularly when concerning the brainstem and cerebellum which are difficult to image accurately on CT. As this patient has confirmed endocarditis, one would naturally suspect multiple embolic infarction although other diagnoses should also be entertained such as global ischaemia, metabolic or endocrine causes as rightly mentioned by several of the above physicians and students.

4) What would be your next test to confirm your suspicions?

An MRI of the brain and brainstem with the addition of vascular views.

Although measuring the serum osmolality is useful, the physical examination tells us that this patient already has a high osmolality and moreover, it can also be calculated from the serum electrolytes. I see that formerly measuring it might be useful as a guide to know how quickly it is recovering but in reality, one follows the patient's conscious level, hydration status and basic serum electrolytes to know that.

Measuring the glucose (hypoglycaemia / hyperglycaemia) would be correct -- it was normal.

An ABG would be reasonable and in hypovolaemic states, one might see a rise in the serum lactate (lactic acidosis) and renal impairment would also contribute to a metabolic acidosis. Urine ketones would be useful to establish if there was a DKA present (not in this case) and remember that starvation can raise the ketones.

5) What might be the cause(s) of the neurological condition present?

The patient has confirmed endocarditis. Hence, an embolic source leading to brainstem infarction might realistically be the cause of the brainstem abnormality. However, we are told that the patient has history, physical examination and basic lab studies consistent with severe dehydration.
In respect of Dr Stein's comments, he is correct in saying that SIADH can only be diagnosed in a patient who is normo-volaemic. Hence, with concentrated urine being present in the context of hypovolaemia / dehydration, then this would be Appropriate ADH secretion rather than inappropriate secretion.

Severe dehydration is a risk factor for hypernatraemia and haemoconcentration. Hypernatraemia can result in brainstem abnormalities such as Central Pontine Myelinolysis (as can hyponatraemia with rapid correction) and haemoconcentration can cause cerebral vein thrombosis.

What happened next?

The MRI scan of the brain is shown below. As can be seen, the brainstem shines like a beacon as do the basal ganglia. This patient indeed had severe pre-renal failure with severe uncorrected hypernatraemia (max 170mmol/L), high BUN and creatinine for several days due to dehydration, which when corrected, was done so over 3-4 days, making the rare diagnosis of Severe Central Pontine Myelinolysis as the top diagnosis here.

The T2 'Flare' MRI pictures below demonstrate central pontine plus extra-pontine involvement in the basal ganglia, the cerebellum and its peduncles. Thank you Dr Y.

Given the fact that the CSF examination was normal goes against a CSF bound infectious aetiology for cause of the brainstem problem. Moreover, embolic phenomena were considered but the arterial views of the MRA were within normal limits. In addition, the changes of the brainstem and basal ganglia are symmetrical with no diffuse cerebral involvement making symmetrical, selective embolic impaction very unlikely. Of course, a brainstem glioma should be considered but with the over-riding history of this patient, the rapid deterioration and a previously normal CT scan makes this diagnosis somewhat remote.

Hypernatraemia is a rare cause of central pontine myelinolysis. It usually occurs in the context of hyponatraemia with the over-rapid correction of the sodium.

After a Medline Search, please see the following interesting cases below:

J Intensive Care Med. 2006 Nov-Dec;21(6):372-6.Click here to read Links

Osmotic demyelination and hypertonic dehydration in a 9-year-old girl: changes in cerebrospinal fluid myelin basic protein.

University of Missouri School of Medicine, Columbia, MO 65212, USA.

A 9-year-old girl was admitted for the treatment of hyper-natremic dehydration. Her history was significant for psychogenic polydipsia, hyponatremia, and a renal concentrating defect. She presented with a 2-day history of altered mental status, ataxia, lethargy, fever, nausea, vomiting, and diarrhea. Meningitis was ruled out. Over the course of her illness, slow rehydration was maintained with a gradual decrease (10 mEq per 24 hours) of the serum sodium. Despite this care, she developed quadriparesis, and magnetic resonance imaging performed on day 6 of her illness was consistent with osmotic demyelination (central pontine myelinolysis). To rule out an excessively rapid correction of hypernatremia as the etiology of the problem, a myelin basic protein was measured in the cerebrospinal fluid that had been obtained on hospital day 1. The myelin basic protein was 649.50 ng/mL (normal, 0.07-4.10 ng/mL). The current literature is presented regarding the postulated pathogenesis of central pontine myelinolysis and suggested therapies, previous reports of central pontine myelinolysis in children are reviewed, and the potential role of myelin basic protein in its diagnosis is discussed.

Presse Med. 1999 Jun 12;28(21):1112.Links

[Central pontine myelinolysis following correction of hypernatremia in an aged patient]

[Article in French]

Rinsho Shinkeigaku. 1994 Nov;34(11):1130-5.Links

[A case of central pontine myelinolysis and extrapontine myelinolysis during rapid correction of hypernatremia]

[Article in Japanese]

Third Department of Internal Medicine, Yamanashi Medical College.

A 69-year-old woman was admitted because of severe dehydration due to anorexia. Consciousness disturbance was found to be due to severe abnormalities of serum electrolyte balance, but recovered quickly by correcting the hyperosmolality. While the initial serum sodium value of 186 mEq/L was corrected to 139 mEq/L in 5 days, locked-in syndrome, bilateral hand tremor and tetraparesis appeared. Brain magnetic resonance imaging (MRI) revealed symmetrically high signal intensity areas on T2-weighted images and low signal intensity areas on T1-weighted images in central part of pons and bilateral middle cerebellar peduncles. One and a half month later, these neurologic symptoms were improved and the MRI abnormalities also disappeared. Auditory brain stem responses which showed prolongations of III to V wave peak to peak latency at the onset returned to normal. It is noted in this case that central pontine myelinolysis (CPM) and extrapontine myelipolysis (EPM) appeared during the period of rapid correction of hypernatremia. Although it is known CPM and EPM are caused by hypernatremia or the rapid correction of hyponatremia, there has been reported only one case of CPM and EPM after rapid correction of hypernatremia. According to the hypothesis of Norenberg, rapid rise in serum sodium may cause CPM and EPM, but if CPM and EPM are caused by the rapid correction of hypernatremia in this case, CPM and EPM may be caused by another pathogenesis of the disorder.

As can be appreciated from the above case reports, it can occur in all age groups. In its severe form, it can result in tetraparesis, pseudobulbar palsy and hence, a locked-in syndrome. There is a high risk of death in such patients; 5-10% beyond 6 months.

This problem occurs due to release of myelotoxic compounds resulting in destruction of myelin sheaths with the preservation of neurones and axons. Such patients can be found to have a raised Myelin Basic Protein in the CSF.

Hypernatraemia and Hyponatraemia

Severe hypernatraemia can be avoided- The patient presented with signs of hypovolaemia which worsened during the hospital admission, and as a result, appropriate fluid management to ensure the return of normal circulation and electrolyte balance would have gone some way to reverse this situation.

There is a commonly held view that patients should not be over-hydrated because of the fear of causing heart failure. This is an unfortunate opinion. Patients with mismanaged circulation die from renal failure and may develop problems such as coagulation abnormalities e.g. DVT, PE, cerebral and renal vein thrombosis and hypernatraemia. Signs of dehydration / hypovolaemia and pre-renal impairment suggests a generalised tissue hypo-perfusion state and therefore, it tells us that the patient needs more circulation and hence, more water and / or blood.

Of course, over vigorous hydration can in some instances result in pulmonary oedema especially if there is an impairment of cardiac or renal function. However, pulmonary oedema is much more manageable than renal failure. Fluid overload responds to drugs such as furosemide, nitroglycerine and if need be, CPAP can be used to push the fluid out of the lungs. In the worst case scenario a patient with pulmonary oedema and or renal failure can utilities haemofiltration or dialysis respectively (if necessary and if available). Of course, diuretics will only work well if the kidneys were previously normal and have a sufficient circulation and hence we come back to the use of appropriate fluid to replenish the renal blood flow.

In such a patient, central venous pressure monitoring / JVP measurement would be beneficial to help guide the fluid administration plus repeated examination of the chest for crepitations (crackles).

Conversely, if the patient has developed a severe hypernatraemia or hyponatraemia, then the correction of the sodium should be done more slowly e.g. 10-12 mmol/24 hours. However, if the patient is shocked and in pre-renal failure such fluid might have to be administered more quickly. However, the type of fluid given is very important. Opting for 5% dextrose in severe hypernatraemia can cause rapid shifts in Na-H20 and should initially be avoided. Most practitioners would start with normal saline (0.9%) which is still nevertheless hypo-osmotic when compared to the serum hyper-osmotic state. This leads to a more gradual reduction in Na level. If the serum Na does not start to correct despite the use of normal saline then half-normal saline (0.45%) can be considered but regular monitoring of Na should be ensured so that an over-rapid correction of the Na does not occur.

So remember, it is not only over-rapid correction of hyponatraemia that can cause central pontine myelinolysis, by also the development of hypernatraemia or its over-rapid correction too !

This brings me back to a salient point that I have mentioned on numerous occasions over time, examine your patients! List the problems of the history, physical examination, labs and radiology. Aim to interpret the features and group them together and try to form a hypothesis for different potential scenarios e.g. IE -> embolic infarction, brainstem signs in dehydrated/hypovolaemic patient -> ? hypernatraemia ? hypoxaemia ? cerebral vein thrombosis etc... When doing the physical examination, look for the signs of dehydration (as above), check the urine output (place a Foley catheter and measure the hourly urine production) and serum renal function and use enough fluids to correct the circulation.

Hypernatraemia with raised BUN and Creatinine usually signifies a reduced oral intake of fluid or that the patient is not producing or reacting to Vasopressin (Diabetes insipidis). The latter two are unusual causes but should be investigated if the circulation fails to improve with vigorous fluid correction. The clue that this was dehydration rather than diabetes insipidis was the highly concentrated urine suggesting appropriate ADH activity whereas in the latter disease, the urine is dilute because of failure to concentrate the urine. This could not have been consistent with SIADH as described above.

Thank you ever so much for such great contributions to this month's case. It was certainly a tough one.

Moreover, I look forwards to more challenging case to present to you into 2009 and beyond. Please keep coming back to the blog when you get the chance.