Unripe Akee Fruit & Pivalic Acid Generating Antibiotics- Mimics of Carnitine Deficiency

In the 1950s, researchers noticed that people in Jamaica, who ate the unripe fruit of Akee tree, usually in the cold season, developed a constellation of symptoms comprising vomiting, hypoglycaemia and altered sensorium. The putative component was therefore named "hypoglycin" and later identified as methylenecyclopropane acetic acid (MCPA). It turned out that MCPA inhibited several acyl-coA dehydrogenases, namely those pertaining to isovaleryl-coA, glutaryl-coA and isobutyryl co-A. These fatty acids are complexed to carnitine and therefore lead to secondary carnitine deficiency. Plasma acylcarnitines relevant to these fatty acids are thus raised, accompanied by an increase in urinary dicarboxylic organic acids such as ethylmalonic acid and suberic acid, presumably formed by w-oxidation of the involved fatty acids, given that beta-oxidation is blocked. Thus, subjects eating unripe Akee fruit had features mimicking several disorders that shared defective beta-oxidation of fatty acids as their underlying aetiology, such as medium chain acyl-coA dehydrogenase deficiency and isovaleric acidemia (both of which are screened for in the UK through a heel prick test 5 days after birth in all neonates as part of the national newborn screening programme).

A similar picture can be seen in subjects taking prolonged courses of certain antibiotics (mostly for prophylaxis against recurrent infections). These antibiotics, namely pivmecillinam (widely used in the UK for UTI), pivampicillin (only used in Denmark), and cefditoren pivoxil (Spectracef) all have pivalic acid as one of their metabolites. Pivalic acid is excreted as pivaloylcarnitine, and thus can lead to secondary carnitine deficiency, and the familiar symptoms of nonketotic hypoglycaemia, vomiting and abdominal pain. Prolonged courses of these antibiotics is therefore inadvisable.

How to Detect Inadvertent Placement of Pacemaker Lead in Left Ventricle

Occasionally, the pacing lead meant for the right ventricle (RV) can be inadvertently placed in the left ventricle (LV) instead. There are several ways this can happen- through inadvertent transarterial catheterisation rather than transvenous, through an existing PFO or ASD, or through puncture of the interventricular septum.

Such misplacement of the ventricular lead is not benign. LV leads are thrombogenic and can be a nidus for thrombus formation and embolisation, leading to stroke.

So how can you tell?

ECG and Xray with a LAO view are the most useful.

Take a look at the following ECG. The top ECG reflects pacemaker lead inadvertently placed in the left ventricle, while at the bottom, the lead has been re-sited in the right ventricle.

Not unexpectedly, the precordial leads reflect a RBBB pattern rather than LBBB. While the RBBB pattern can persist in V1 and V2, even after repositioning the lead in the right ventricle(through open surgery- the LV lead should never be "pulled out"), V3 is the key. V3 will always be positive when the lead is in the LV, and will always be negative, when the lead is in the RV. With LV placement of lead, the frontal plane axis is usually between 0 and -90 degrees, but is often in the "northwestern" quadrant, as in this case.

This picture of apparent RBBB, suggesting a wrongly placed lead in the LV, can sometimes be mimicked even if the lead is in its rightful place in the RV. Again, V3 is the key. Even if V1 and V2 are upright (positive), V3 will always be negative with a correctly placed RV lead. The ECG can be "corrected" by placing the precordial leads one interspace lower, as in the following figure:

The lateral view on Xray is useful when the ventricular lead is in LV rather than RV. The lead is more posterior than when it is in the RV and curves away from the vertebral column, as shown in the following reference:

https://www.mdedge.com/ccjm/article/155202/cardiology/detecting-and-managing-device-leads-inadvertently-placed-left/page/0/1

Of course, such aberrant placement can be confirmed by echocardiography, but should really be picked up on the post-placement ECG.

Reference:

https://www.mdedge.com/ccjm/article/155202/cardiology/detecting-and-managing-device-leads-inadvertently-placed-left/page/0/1

The Girl with Raised CK & High Plasma Acylcarnitines

Over the years, I have got into the habit of requesting plasma acylcarnitines in subjects with raised CK. As a rheumatologist, one sees one's share of subjects with high CK. Most of them have muscle pain, some have fatigue and exercise intolerance. Largely they look well, and investigations are unrewarding.

The European Federation of Neurological Societies recommends that CK levels above the following thresholds be investigated:

Caucasian female- 325 U/l
Caucasian male- 503 U/l
Afro-Caribbean female- 600 u/l
Afro-Caribbean male- 1200 U/l

The girl in question was Caucasian, in her early twenties, and was referred with a diagnosis of ?Fibromyalgia. As physicians all over the world will be aware, this has become a popular and somewhat lazy diagnosis in almost anybody with pain, particularly in primary care. However, this particular girl had had a few admissions with vomiting and abdominal pain. Looking back through one such episode, I found that her CK was ~1500 and 800 on 2 occasions, but had not been repeated.

So I repeated her CK- it was normal. A few days later, however, her plasma acylacarnitines were reported as at least moderately raised for all fatty acid lengths- short chain, medium chain and long chain fatty acids.

So what did she have?

The most likely diagnosis in an adult with raised CK and increase in plasma acylcarnitines of all lengths is Multiple Acyl Co-A Dehydrogenase Deficiency (MADD). This is a condition that is quite responsive to riboflavin in 98% of cases and is therefore important not to miss.

MADD is just one of a number of lipid storage myopathies- the others being Neutral Lipid Storage Disorder Myopathy (NLSD-M), Primary Carnitine Deficiency (CD), and Carnitine Palmitoyl Transferase-II (CPT2) deficiency. More about those others later.

MADD is autosomal recessive, caused by deficiency of one of 3 enzymes- Electron Transfer Flavoprotein Dehydrogenase (ETFDH), or the two isoforms of ETF itself- A&B. Ninty-three percent of cases are due to ETFDH deficiency. When the deficiency is severe, such as with 2 null alleles, you get the most severe form of the disease, with congenital abnormalities (Type I) or without (type II). Such subjects present as neonates with metabolic decompensation such as vomiting, metabolic acidosis, non-ketotic hypoglycaemia, and hepatic failure with hyperammonemia, as seen in Reye's syndrome.

Type 3, with delayed onset presentation in teenage or adulthood, is the form most of us will come across. In 2014, Grunert reported that 350 such cases had been published (of all types), but it is likely that a vast majority of cases with milder phenotypes remain undiagnosed and unpublished.

Most subjects with delayed onset MADD have chronic muscle pain, muscle fatigue, subjective proximal muscle or neck weakness and exercise intolerance. A small minority have episodes of rhabdomyolysis. Subjects can remain asymptomatic. A minority present with acute metabolic decompensation with abdominal pain, vomiting, and acidosis, more commonly seen in types 1 & 2, often after episodes of stress such as fever, infection, fasting, pregnancy or labour. Chronic & acute forms can co-exist in 20% of cases.

The screening test is plasma (or serum) acylcarnitines, which will show a rise across all fatty acid lengths, i.e short, medium and long chain fatty acids (C6-C18). Corroboration can be had by measuring urinary organic acids (spot sample in non-acidified container), specifically that of glutaric acid, 2-hydroxy glutaric acid, ethylmalonic acid, 3-hydroxyisovaleric acid, adipic, suberic and sebacic acids and glycine conjugates (acylglycines). These will be raised. Hence MADD is also known as type-2 glutaric aciduria. Since the plasma acyl-coA is all bound to carnitine, serum free carnitine will be low. However, some subjects have raised plasma acylcarnitines and high urinary organic acid excretion only during periods of stress.

The confirmatory test is sequencing of the ETFDH gene, looking for mutations or null alleles. This will yield the diagnosis in 93% of cases. Muscle biopsy will show lipid containing vacuoles adjacent to normal looking mitochondria on Oil Red-O stain, but does not contribute anything extra if molecular techniques are available.

Once diagnosed, it is important to start these subjects on riboflavin in a dose of 200 mg daily. In most reported case series, riboflavin has been co-prescribed with either carnitine (3g daily) or co-enzyme A, but not both. Many subjects respond within a month, with normalisation of raised CK alongside clinical improvement of pain, fatigue and muscle power, but others may need longer.

Other lipid storage myopathies have different phenotypes. Neutral Lipid Storage Disorder Myopathy (NLSD-M) presents in adults with myopathic features, but in addition is accompanied by cardiomyopathy, fatty liver, transaminitis, high tryglycerides & VLDL, and type 2 diabetes mellitus. There may be parental consanguinity. The putative defect is in the PNPLA2 gene, with a deficit in the enzyme Adipose Triglyceride Lipase (ATGL), which catalyses the first step in the breakdown of intracellular triglycerides. Leucocytes show characteristic tryglyceride accumulation, called "Jordan's anomaly". The only effective treatment is medium chain fatty acids (these enter the mitochondria directly for beta-oxidation, without needing to be conjugated to carnitine).

Primary carnitine deficiency (CD) presents early in life, can be fatal, but is treatable if diagnosed. It presents with myopathy, cardiomyopathy, nonketotic hypoglycaemia and Reye's syndrome type hepatic failure with hyperammonemia. The defect is in the Carnitine Organic Cation Transporter gene, OCTN2. Treatment is Carnitine, in a dose of 3g daily.

CPT2 deficiency is seen in young adults or teenagers, who present with a characteristic history of recurrent rhabdomyolysis, often after exercise, fasting or fever. Some subjects develop renal failure due to myoglobinuria. Unlike the other lipid storage myopathies described above, fat accumulation in the form of lipid droplets is largely absent in muscle biopsy samples on Oil Red-O staining. Recently, bezafibrate has been shown to improve lipid shuttle in these patients in fibroblasts by activating PPAR-gamma, but clinical trials have failed to show benefit.

Why Does My Friend's House in Gurgaon Get More Sunshine on Winter Mornings?

A friend asked me why his house, located in Gurgaon, India, received more morning sunlight in winter, and more evening sunlight in summer. He looked up the angles of sunrise & sunset in summer & winter.

Here's the sunrise and sunset angles for Gurgaon that was posted.

Sunrise : 63* NE
Sunset : 297* NW

Winters :

Sunrise : 116* SE
Sunset : 244* SW

I will deal with the direction, then the angles. Notice that the sun is depicted as rising and setting in the North East and North West respectively in summer and in the South East and South West in Winter. It doesn't change direction. It is simply that in summer, the sun rides much higher in the sky than in winter. As India is in the Northern hemisphere, the position of the sun is higher, i.e. more Northerly, than in winter, when it sits more to the South. That is to say, it follows a much bigger arc, which intuitively makes sense, as the sunrise to sunset time, i.e. the total hours of sunlight, are much longer in summer. That explains the NE, NW in summer and SE, SW in winter.

Now the degrees. This confirms what I said earlier- the arc traversed in summer is much bigger- 297-63=234 degrees than in winter- 244-116= 128 degrees. The sun sits lower, i.e. more to the South in the Northern hemisphere, hence the arc is flatter and lower.

While this explains why the sun should hang around for longer in summer evenings, it certainly doesn't explain why you enjoy sunnier mornings in winter. I looked into this as well. The workings are fascinating and made lovely reading. I would be happy to share if there is any interest at all.

It is commonly believed that the winter solstice in the Northern hemisphere is the day with the most delayed sunrise and the earliest sunset. Indeed, the play of seasons and longer-shorter day length in summer-winter is a function of the earth's inclination relative to it's geometric axis. This angle of inclination ( or "declination" as it's called) is about 23.4 degrees, so that the North pole always faces the sun in summer, while the South pole faces the sun in winter. The inclined axis doesn't change direction. It always faces the Polaris, which is why it's as "Constant as the Northern Star".

However, the earth also revolves around the sun. This orbit is slightly elliptical, so that in December, the earth is closest to the Sun, called Perihelion, while in June, it is at it's furthest, called aphelion. Due to a law of planetary motion, called Kepler's second law, the earth moves faster in its orbit at perihelion, i.e. in December, than at aphelion, i.e. in June. From the vantage point of somebody sitting in the Sun, the earth will appear to be moving almost 7% faster in December therefore*. From the earth however, the Sun's movement through the sky will appear delayed. Thus, towards the end of the year, the Sun is "late". Both sunrise and sunset are later than would be predicted by the earth's declination alone. The reverse occurs in June, albeit less accentuated. That is to say, the sun is "early". Both sunrise and sunset occur earlier than would be expected from the earth's declination alone.

Thus, there are two solar phenomena determining the sunrise-sunset cycle. The first is due to the earth's declination, and can be called the "geometric effect". The second is due to the effect of the elliptical axis, and is called the "clock effect".

In practical terms, what this means is that in December, the delayed sunrise from the geometric effect is augmented by the clock effect, while the earlier sunset from the geometric effect is countered by the clock effect. Thus, say at 40 degrees North, sunset starts to occur later in the day as early as Dec 8, rather than Dec 21, while sunrise starts to occur earlier much later- around Jan 6. The effect is around 16 minutes each way.

The reverse occurs in June. As the sun is "early", sunrise occurs earlier than expected even before June 21, while the peak in late sunset is delayed until July.

It's worth saying here that the Clock effect is invariant, regardless of latitude, while the Geometric effect is more pronounced the further away from the equator you go. At the equator, because the Clock effect predominates, sunset starts occurring later and later as early as November, while it's February before sunrise starts occurring earlier. Timewise, the Clock effect is least pronounced around the equinoxes, and most pronounced around the solstices.

Practically, this means that closer to the equator (Gurgaon 28 degrees North), the Geometric effect is minimal, and the Clock effect occurs for most of the variation in sunrise and sunset. At the equator, this amounts to only around 30 minutes between summer and winter.

In more northerly latitudes ( say Nottingham 53 degrees North), the Geometric effect becomes more pronounced, accounting for the very short days in winter.

Now back to my friend's query. Why are the mornings sunnier in winter? The clock effect should not contribute more than 16 minutes to a change in the time of sunrise, but the problem with invoking it is that sunrise occurs later due to Clock effect in winter, chiming in with the Geometric effect. Hence that doesn't explain the winter morning Sun. The most likely explanation is therefore either less cloud cover in Winter or more likely a South East facing window, which captures most of the sunshine that's available. This latter explains why South facing gardens are prized in more northerly countries such as the UK.

* Kepler's second law - sometimes referred to as the law of equal areas - describes the speed at which any given planet will move while orbiting the sun. The speed at which any planet moves through space is constantly changing. A planet moves fastest when it is closest to the sun and slowest when it is furthest from the sun. Yet, if an imaginary line were drawn from the center of the planet to the center of the sun, that line would sweep out the same area in equal periods of time. For instance, if an imaginary line were drawn from the earth to the sun, then the area swept out by the line in every 31-day month would be the same. (http://www.physicsclassroom.com/class/circles/Lesson-4/Kepler-s-Three-Laws)

Why are West Coasts Warmer in the Northern Hemisphere?

The Coriolis effect diverts winds in the Northern hemisphere to the right, and in the Southern hemisphere to the left. Thus, in the northern hemisphere, winds blowing from north to south will be diverted westwards, while winds blowing from south to north would be diverted eastwards. The reverse will happen in the Southern hemisphere.

The Coriolis effect is more pronounced the further you go from the equator, i.e most pronounced at the poles, and negligible at the equator (which is why you almost never get cyclones at the equator itself. Cyclones rarely occur in the latitudes between 8 degrees north and 8 degrees south).

So far, so good, but this is all in the books, therefore not novel. I tried to think of a practical application of this that cannot be gleaned from the books. If one puts the two things above together, one can hypothesise that winds blowing southwards from the North pole would be directed westwards, and thus hit east facing coasts, while winds blowing northwards from the equator would be directed eastwards, and thus hit west facing coasts. It should follow therefore, that for large landmasses, equilatitudinal places should be warmer on the West coast than the East coast.

But are they?

So I randomly picked three pairs of locations, on East and west coasts, all reasonably to the north (so that the Coreolis effect would be prominent) to compare their average temperatures. By comparing places located at sea level, altitude is largely eliminated as a factor. Here are the results.

Arviat, located at 61 North on the Nunavut peninsula on the East Coast of Canada, has an average annual temp of -9 degree C. Hooper Bay, 61 North, located on the west coast of Alaska, USA, has an average annual temp of -1.7 degree C.

Portland, Maine, 43.66 North, on the Eastern seaboard, has an average annual temp of 7.4 degree C. Newport, Oregon, 44.63 North, on the West coast, has an average annual temp of 10.7 degree C. Note that Newport is actually a degree north of Portland.

To avoid local variations, I then took two places virtually miles apart, facing each other across the Bering Strait at around 65 North. Lorino, located on the East coast of Russia, has an average annual temp of -6 degree C. Teller, Alaska, on the West coast has an average annual temp of -4.9 degree C.

In the Southern hemisphere, things should reverse, and the East coast should be warmer. And it is. Two South American towns located at around 36.5 South-Concepcion, on the West coast of Chile, has an average annual temp of 12 degree C, while Santa Teresita, on the East coast of Argentina, averages 15.2 degree C.

Moving to Australia, we consider Sydney on the East coast, averaging 18.5 degree C annually, compared with Bunbury on the West coast, which averages 16.8 degree C. Both are around 33-34 degrees South.

Interesting, isn't it? Although the pairings were completely random, I deliberately left out places in the extreme south of South America, as it's very mountainous down one coast.

Permissive Haplotypes Are the Fertile Ground for Disease Causation

Fully 95% of subjects with essential polycythaemia and 50-60% of those with Essential thrombocythemia or Primary Myelofibrosis have a JAK2 V617F mutation. What is less well known is that this particular mutation arises preferentially in subjects with a certain haplotype- the so called 46/1 or GGCC haplotype, present in 50% of the general population. In the other 50%, myeloproliferative neoplasms (MPN) are distinctly uncommon.

In around 150 SNP haplotypes drawn from Caucasians in a British population, haplotype 46 and haplotype 1 were almost identical, except for a single base. These two haplotypes, together designated as 46/1 are found ("tagged") in 2 SNPs- rs 12343867 and rs 12340895. When the base present in either of these 2 SNPs is guanine rather than cytosine (GC or GG rather than CC), the V617F mutation is far more likely to arise (around 3.7 times more frequently than if the predisposing haplotype was not present). Why this happens is not known. MPN arising in non-46/1 haplotypes is not phenotypically different to that arising in 46/1 haplotypes.

A similar situation can be seen with Facioscapulohumeral dystrophy (FSHD), except that here the effect of the predisposing haplotype is even more dramatic. FSHD is caused by the accumulation of a toxic protein in muscle cells called DUX4. The DUX4 gene is present next to a microsatellite region called D4Z4 on chromosome 4q35 and is ordinarily not transcribed. Most subjects have 11-100 D4Z4 repeats of CpG and GpC nucleotide base sequences. Subjects with FSHD have only 1-10 such repeats in the D4Z4 region ("contraction" of D4Z4 region). The D4Z4 microsatellite region can be identified by digestion of DNA with a restriction endonuclease called EcoR1.

However D4Z4 is not exclusive to chromosome 4. This microsatellite sequence is also located on chromosome 10. However, here the contraction of the D4Z4 region does not lead to FSHD.

Here's the important bit. Simply having a contracted D4Z4 region on 4q35 will not lead to FSHD. For that to happen, you need a permissive haplotype located just next to the D4Z4 region called 4qA 161 (or rarely, some other 4qA haplotypes). If the person involved has a haplotype containing 4qB rather than 4qA or a non-permissive 4qA haplotype, FSHD will not develop.

Why does this happen? 4qA 161 is responsible for generating the polyadenylate (polyA) tail at the 3' end of the DUX4 mRNA that stabilises it and leads to its translation. Without the polyA tail, the mRNA would be unstable and would be destroyed.

Interestingly in procaryotes such as bacteria, the polyA tail serves the exact opposite function. Polyadenylated mRNA are tagged for destruction.

So why does contraction of the microsatellite D4Z4 region lead to de-repression of the toxic DUX4 mRNA? It is thought that the full length D4Z4 region keeps the DUX4 gene in a "compressed" state. In FSHD, the briefer D4Z4 sequence allows uncoiling of the DUX4 gene and its subsequent transcription, providing of course the polyA tail from the permissive 4qA 161 haplotype was also available.

Men Should Receive Lifelong Anticoagulation Following Unprovoked DVT or PE

The current standard of care is to treat subjects who have a provoked venous thromboembolism (VTE-either DVT or PE) with anticoagulation for 3 months. In those who have an unprovoked VTE (no history of leg fracture, leg in cast, immobility for > 3 days, major surgery within 3 months or diagnosed cancer) for 6 months. However, many subjects in this latter group have "catch-up" thromboembolism once anticoagulation is stopped.

Two observations should prompt a change in current practice.

Firstly, several studies, particularly two papers published in 2008 and 2016 from Canada, have shown that men run a far higher risk of recurrent DVT/PE over a follow up period of around 5 years after stopping anticoagulation following an index event. (7.6% annually in men v 2.8% in women). The average risk of repeat VTE in the first year is around 10% when both genders are combined. In the second year, this falls to 5%. Subsequently, the risk remains constant at around 3% per year. Over 8 years, the cumulative risk of recurrent VTE across both genders is ~30%. For men, this risk is higher at around 40%.

Women are not completely safe, however. Among women, using a risk stratifier called HERDOO2 (explained soon), women with 0 or 1 risk factors had a very low risk of recurrent VTE after stopping anticogulation, while women with 2 or more risk factors, had a risk of recurrent VTE only slightly lower than men (approximately 1% annually in low risk women and 6% in high risk women).

HERDOO2 includes HER-Hyperpigmentation, pretibial Edema, and Redness, all indicators of post thrombotic syndrome measured at ~6 months after the index DVT or PE in either leg. D is serum D-Dimer of >250 ug/ml, measured while on anticoagulation, the first O is Obesity (BMI>30) and the 2nd O is for Old age (>65 years).

HERDOO2 should only be used to risk stratify women. All men are at a higher risk of recurrent VTE.

Across all risk groups, continuing anticoagulation reduces the risk of a repeat VTE episode by around 80%.

What about mortality benefit? In low risk women, the risk of dying from major haemorrhage while on anticoagulation is around the same as the risk of dying from recurrent VTE if anticoagulation is stopped after 6 months. However, in men and high risk women, continuing anticoagulation reduces the risk of death from repeat VTE from 1.4% over 8 years to less than 0.3%. The risk of dying from a major bleed while on anticoagulation over the same period is 1.1%.

The second observation comes from two recent papers that show that in subjects receiving newer oral anticoagulants (NOACs), maintenance treatment with low dose NOAC (10 mg with rivaroxaban or 2.5 mg BD with apixaban daily) has the same benefit in terms of VTE prevention as maintenance with full dose NOAC (20 mg of rivaroxaban or 5 mg BD of apixaban daily) with similar or lower risk of bleeding as placebo (in the first study) or aspirin (in the second study), although the risk of bleeding in general was low across all treatment groups. All subjects received full dose NOAC for the first 6-12 months after the index VTE.

There is thus a strong case for maintaining men and high risk women with unprovoked VTE on lifelong low dose apixaban or rivaroxaban after an initial 6-month period of full dose anticoagulation. Current practice needs to change.

References.
1. Rodger MA, Scarvelis D, Kahn SR, et al. Long-term risk of venous thrombosis after stopping anticoagulants for a first unprovoked event: a multi-national cohort. Thromb Res 2016;143:152-158
2. Agnelli G, Buller HR, Cohen A, et al. Apixaban for extended treatment of venous thromboembolism. N Engl J Med 2013;368:699-708
3. Weitz JI, Lensing AWA, Prins MH, et al. Rivaroxaban or aspirin for extended treatment of venous thromboembolism. N Engl J Med. DOI: 10.1056/NEJMoa1700518