The Changing Face of Acute Epiglottitis

Epiglottitis (also called supraglottitis) is more common in children, right?

You would be surprised. Since the introduction of the conjugate Haemophilus influenzae B vaccine in the early 1990s, the incidence of acute epiglottitis in children has fallen. As a result, epiglottitis is now 3 times more common in adults as in children.

Haemophilus influenzae remains the most commonly identified agent in both adults and children, but other bacteria such as S.aureus, streptococci, and viruses have been implicated. In immunosuppressed subjects such those with HIV, Candida and Pseudomonas can cause acute epiglottitis.

Typically, epiglottitis presents with sore throat, anterior tenderness over the neck in the region of the hyoid, drooling, dysphagia and odynophagia. While the triad of drooling, dysphagia and (respiratory) distress- 3D- is considered pathognomonic, respiratory distress is commoner in children because of the small size of the supraglottis. Symptoms can progress rapidly in children with respiratory obstruction and death within 12 hours of onset of symptoms, and therefore the paediatrician must have a high index of suspicion.

Children with epiglottitis look anxious, breath through an open mouth, with the neck hyperextended and the chin thrust forward. They often assume a posture where they bend forward, arms stretched, splinting their trunk, much as a subject with emphysema would (the tripod sign).

Stridor occurs in a minority of patients. Cough and hoarseness are not usual features and should raise suspicion of an alternative diagnosis such as croup in children, or laryngitis in adults. However, the voice can be muffled.

The most important clue on examination is the presence of sore throat with a normal appearing pharynx. Direct or indirect Laryngoscopic examination should be postponed, particularly in children as it carries a risk of respiratory arrest, and in both adults and children, should be carried out in a setting where emergency intubation is possible. Attention to airway takes primacy in all subjects, regardless of age, if epiglottitis is suspected.

An alternative, non-invasive way of diagnosing epiglottitis is through the lateral neck X-ray, which shows an enlarged epiglottis- called the thumb sign, along with swollen ary-epiglottic folds, often in association with straightening of the curvature of the normally slightly lordotic cervical spine (Figure 1).

Figure 1. The Thumb Sign in Acute Epiglottitis


Treatment should comprise a combination of 3rd generation cephalosporin such as cefotaxime or ceftriaxone with an anti-staph agent such as clindamycin or vancomycin.

In subjects who have developed epiglottitis despite immunisation, or whose immunisation history is not known, or in subjects who have had epiglottitis in the past and now present with another invasive infection possibly due to H.influenzae B such as cellulitis, osteomyelitis, meningitis or septic arthritis, it is easy to check for functional antibodies to H.influenzae B. If the titres are suboptimal, the subject should receive a single dose of Menitorix vaccine- a combination of conjugate vaccine against H.influenzae B and Meningococcus C.

Varicella & Shingles Vaccines, DMARDs & Biologics

Varcella vaccine and Zoster (Shingles) vaccine are both comprised of the same strain- Oka- but are not the same vaccine. While the varicella vaccine (Varivax®, Varilix®) contains 1350 Plaque forming units (PFU), the zoster vaccine (Zostavax®) contains 18,700-60,000 PFU, i.e. at least 14 times more virus. The varicella vaccine needs to be given in two doses 4-8 weeks apart, while only one dose of the Zoster vaccine need be given.


In Rheumatology, the varicella vaccine is indicated in subjects who are about to start immunosuppression and are non-immune. Although, a history of chicken-pox is a reliable indicator that the subject will have protective antibodies, it is customary to check the varicella serology before starting immune-suppressive treatment. It is estimated that 90% of adults in the UK will have had either clinical or sub-clinical exposure to varicella growing up and will be immune- i.e. have protective levels of antibodies to varicella in the serum.


In the minority of patients with auto-immune disease that do not have protective antibodies, 2 doses of the varicella vaccine are administered 4-8 weeks apart. This should ideally be done 4 weeks before starting DMARDs.


For subjects who are already on DMARDs, it may not be too late to give the varicella vaccine. If the subject is on <0.4 mg methotrexate per kg body weight per week (equivalent to 25 mg weekly in a 60 kg subject), or receiving <3 mg azathioprine daily, or <1.5 mg 6-mercaptopurine daily, the varicella vaccine may still be administered. Similarly, those on low dose steroids (<20 mg daily for less than 2 weeks) are considered eligible. These are expert (level 3, not based on RCTs or case control studies) recommendations made by the Advisory Committee on immunization Practices (ACIP) based in the USA, an organisation that celebrates its 50th year in 2014. This is a body that has shaped immunisation guidelines worldwide. The Zoster or Shingles vaccine is recommended for subjects between 70 and 80 years in the UK. This year, the vaccine is being rolled out to subjects who are 70, 78 or 79 years old as of 1 September 2014, but not to other subjects. This is because it is quite expensive- each dose costs ~£100. A prior history of chickenpox or immunity to varicella is not considered necessary by the ACIP for giving the shingles vaccine. However the UK guidelines by the Joint Committee on Vaccination and Immunisation advice that such history be established before giving the shingles vaccine for fear of causing varicella in virus-naïve subjects late in life. The zoster vaccine provides only 50% protection to subjects from shingles, compared with no immunisation. While vaccinated subjects may still develop shingles, the disease is often attenuated. Live virus vaccines should never be given to subjects on biologics or pregnant women. Further, if the subject is about to receive, or has received rituximab, the vaccine must be given either 4 weeks before giving rituximab or 6 months after giving rituximab. With other biologics, varicella or zoster vaccines can be given 4 weeks before starting the biologic or approximately 5 half-lives after completing the biologic. What if a subject on immunosuppression is non-immune and has had a “significant” exposure to varicella or shingles? Significant exposure is defined as “face to face” contact or being in the same room or in the same 4-6 bedded bay. Exposure in more open wards or subjects living in a different part of the house or those exposed in a larger area such as a classroom are treated on a case to case basis. For such purposes, exposure to anybody with chickenpox is considered significant. Exposure to immunocompetent subjects with localised shingles in a covered area such as the thoracolumbar area is not considered significant. However exposure to shingles in an exposed area such as ophthalmic shingles is considered significant. Further, any exposure to shingles developing in an immunosuppressed subject, whether covered or not, is considered significant, as such subjects shed the virus at a much faster rate than immunocompetent subjects. Vulnerable subjects with such significant exposure are treated with varicella zoster immunoglobulin (VZIG). Currently, VZIG is prepared from pooled plasma accumulated from donors outside the UK. Donors from UK are not accepted for this purpose because of concerns about vCJD. Donors are screened for hepatitis B, hepatitis C and HIV, and the collected plasma is also screened for RNA and DNA from these viruses. Subjects who are considered eligible for VZIG are non-immune subjects with “significant exposure” as follows: 1. Immunosuppressed subjects- on biologics, high dose steroid therapy (>20 mg prednisolone for 4 weeks or longer), subjects with leukaemia, lymphoma, myeloma or generalised solid cancers, subjects who have had chemotherapy or have inherited disorders of immune deficiency. (Hopefully, most of these subjects would have been screened for varicella early in their illness trajectory).
2. Pregnant women in the first 20 weeks of pregnancy. Exposure in this period carries a risk of “congenital varicella syndrome” in the foetus, comprising microcephaly, maldeveloped limbs, skin lesions and other features.
3. In women who have been exposed to varicella during the last 7 days of their pregnancy or during the first 7 days after giving birth, the neonate must receive VZIG. During this period, the neonate is at risk of disseminated varicella. The risk lessens considerably after the first 7 days of life.



However, if the mother develops shingles rather than varicella during this period, VZIG need not be given to the neonate as it would have received pre-formed antibodies from the mother.



Prophylaxis still required if the neonate is exposed to chickenpox or shingles from another source apart from the mother in the first 7 days of life.



The dose of VZIG is as follows:
0-5 years- 500 mg
6-10 years- 1000 mg
11-15 years- 1500 mg
>15 years- 2000 mg


VZIG is given intramuscularly. If the subject has a bleeding diathesis and cannot receive IM injection, IVIG can be given in a dose of 0.2 mg/kg.


A subject receiving VZIG will not respond to the varicella vaccine if given concurrently. If VZIG has to be given within 4 weeks of giving the varicella vaccine, it is advisable to re-administer the vaccine after 3 months. VZIG also interferes with immunity to other viral vaccines with the exception of yellow fever, something that should be kept in mind for travellers.


There is no need to give VZIG in any subject who has protective levels of varicella antibodies in blood. The titre of such antibodies is not increased significantly by giving VZIG.


It is worth mentioning that many subjects will develop varicella despite receiving the VZIG. Only 50% of subjects are fully protected from chickenpox by VZIG. However, it is thought to attenuate the severity of chickenpox if it does develop.


VZIG must be given ideally within 7 days, and certainly within 10 days of exposure to provide protection. If the subject’s immune status is not known, it is possible to send off the serology and receive a report within a week under current arrangements. Therefore, in such cases, VZIG can still be given.


It is occasionally possible for immunosuppressed subjects to develop varicella despite having protective antibodies in blood. In such subjects, it is thought that a deficit in cellular immunity contributes to development of chickenpox.


Treatment with acyclovir should be promptly administered to immunosuppressed subjects who develop chickenpox or disseminated zoster despite immunisation or in non-immune immunosuppressed subjects who develop varicella despite receiving VZIG.

Non-pharmacological Therapies in Atrial Flutter and Atrial Fibrillation

Atrial flutter and atrial fibrillation often co-exist, i.e. occur in the same patient at different times, but have different anatomical substrates in the heart. Atrial flutter is due to a macro-re-entrant circuit arising from the junction of the tricuspid valve annulus and inferior vena cava, while early atrial fibrillation is due to spontaneous depolarisation arising from around the pulmonary veins, most commonly from the ostium of the left superior pulmonary vein.

The commonest form of atrial flutter is due to a counter-clockwise macro-re-entrant current at the cavo-tricuspid isthmus. On the ECG, this is reflected in predominantly negative deflections in the inferior leads and in V5 & V6, and positive flutter waves in V1, as illustrated in the figure below.

Less common is clockwise atrial flutter, where the flutter waves have positive deflections in the inferior leads and lateral precordial leads, but are negative in V1, as shown below.

In practice, the management of anti-clockwise and clockwise flutter does not vary, as they both arise from the cavo-tricuspid junction. However, when flutter arises from other areas of the heart, it is called atypical flutter and the site of origin will have implications in terms of management, if radio-frequency ablation (RFA) is considered.

Atrial flutter differs in important ways from atrial fibrillation, apart from its site of origin. Flutter is a relatively unstable rhythm, and often reverts to sinus rhythm, particularly when the underlying contributing factors such as pulmonary embolism, hypoxia or ischaemia are addressed. It may also convert spontaneously into atrial fibrillation. Secondly, flutter waves are far more likely to be conducted across the AV junction compared with fibrillatory waves. Thus, the most common rate of atrial flutter is 300/minute, while the ventricular rate is 150/minute, i.e. 2:1 conduction. Flutter waves may also be conducted in a 3:1, 4:1 or 5:1 ratio, but higher ratios, in the absence of negatively chronotropic medications, indicates disease of the AV junction. Even ratios such as 2:1 or 4:1 are far more common than odd ratios such as 3:1 or 5:1. On the other hand, atrial fibrillatory waves occur at rates between 400-600, but the ventricular rate is usually between 90-170. Lesser ventricular rates usually connote AV junction disease, while higher rates may be seen in thyrotoxicosis, sympathetic overdrive, parasympathetic withdrawal or with bypass tracts.

Unlike atrial fibrillation, persistent atrial flutter is remarkably resistant to pharmacological manipulation. Most such cases should be considered for cardioversion following appropriate anti-coagulation, or if this is not possible or not acceptable, for RFA or overdrive atrial pacing. The cavo-tricuspid isthmus is relatively easy to access for RFA, and the procedure is usually safe, with very few complications.

In subjects who already have a pacemaker or a pacing line in situ, such as after cardiac surgery, overdrive atrial pacing is safe and offers a third non-pharmacological alternative.

Unlike with flutter, RFA in atrial fibrillation is technically much more difficult. This is because most fibrillatory waves arise from the ostia of the pulmonary veins, which must be included in any ablatory procedure (pulmonary vein isolation). RFA with atrial fibrillation commonly takes one of two forms- either segmental- limited to the pulmonary vein ostia with electrical mapping, or the technically easier circumferential RFA, which simply encircles the pulmonary veins. With the latter, special intra-procedure electrical mapping is not required. Atypical flutter, such as that arising from the left atrium, alluded to above, may be a complication of RFA for atrial fibrillation.

Access to the pulmonary veins requires septal puncture. There is a risk of causing a pericardial effusion and a small risk of causing an atrio-oesophageal fistula, which is often fatal. Thus, RFA in atrial fibrillation is often resorted to only after all pharmacological options have failed. Further, RFA only works for paroxysmal atrial fibrillation. Once fibrillation has become chronic, fibrillatory waves arise from all over the left atrium, and sometimes the right atrium, and is no longer amenable to RFA. Such cases can still be treated with a surgical procedure such as the Maze Procedure. The current version is Maze III. For obvious reasons, this is only practical when the patient is undergoing open heart surgery for another indication such as CABG or valve replacement. Subjects with paroxysmal atrial fibrillation who fail to respond to the Maze procedure may still respond to RFA.

When rhythm control is not possible or not practical, and drug therapy has failed, a non-pharmacological strategum for rate control in atrial fibrillation is to ablate the AV node and pace the ventricles. In most cases, this would comprise pacing the right ventricle alone. However, in subjects who otherwise fulfil the criteria for cardiac resynchronisation therapy, such as those with NYHA Class II, III or ambulatory IV heart failure, an ejection fraction <35% and QRS> 120 ms, a CRT or CRT-D device may be considered with good haemodynamic results.

It is worth mentioning that the risk of thrombo-embolism remains high soon after RFA in both flutter and fibrillation, and most experts would recommend continuing anticoagulation for at least 3-6 months. Apart from the fact that most recurrences occur during this period, it is cautionary that the period immediately following reversion to sinus rhythm, whether achieved through electrical or pharmacological means, is associated with a high risk of thrombo-embolism.

MR Spectroscopy and Brain Tumours

Intracranial space occupying lesions can be brain neoplasms, abscesses or infarcts. Primary brain neoplasms mainly arise from glial tissue and are thus called gliomas. The 3 principal varieties of gliomas are astrocytomas, oligodendrogliomas and ependymomas. Gliomas can sometimes be difficult to differentiate from infective lesions such as abscesses. MR spectroscopy can help differentiate these lesions by measuring the relative concentration of metabolites in and around lesions.

The two most useful metabolites are choline and N-acetyl aspartate (NAA). Choline is a component of cell membranes and is thus increased in areas of rapidly dividing cells such as tumours, and is decreased where there is cell destruction such as in abscesses. Choline levels are unchanged in infarcts.

NAA is a metabolite of glutamate, and is produced by neurons. It will be decreased whenever there is paucity of neurons due to destruction. Thus NAA levels are reduced in areas of gliomas, abscesses or infacts.

Some investigators use the ratio of choline:NAA to predict lesions that may be gliomas. A choline:NAA ratio>2 is thought to be specific for gliomas. While the ratio may be high in brain metastases, in the case of primary brain tumours, the ratio remains high in watershed areas around the glioma, as these tumours infiltrate beyond their radiologically apparent borders, while metastases are truly cicumscribed.

Other metabolites such as lactate or creatine can also be measured by MR spectrscopy. Lactate is increased when their is anaerobic activity, and is thus expressed in tumours, infarcts as well as abscesses.

These principles are illustrated in the figure.

Lung Entrapment & Trapped Lung- The Role of Pleural Fluid Manometry

Malignant pleural effusions are common in clinical practice. In my specialty, I often see Rheumatoid related pleural effusions. These are often self limited, and may response to NSAIDs or steroids. However, not infrequently, they can be large and cause dyspnoea or chest pain, requiring aspiration.

One of the principal concerns while aspirating large effusions, whether they be malignant or inflammatory to dryness, is the risk of reperfusion related pulmonary oedema or RPO. This is often heralded by chest pain. The risk of RPO is increased in subjects where the lung fails to re-expand fully, as the pleural pressure in these subjects is more negative than they would otherwise be.

There are two situations where the lung fails to re-expand- the first is acute, because of an inflammatory visceral pleural peel around a segment of the lung, or because of increased elastic recoil of the lung due to lymphangitis carcinomatosa, or due to associated endobronchial obstruction stopping the lungs from re-expanding. The accepted term for this acute phenomenon is "lung entrapment". The pleural fluid in such cases is directly related to the underlying malignancy or serositis, and is an exudate. Such cases, if symptomatic, may be dealt with via a tunnelled pleural catheter.

The second situation is when the inability of the affected portion of the lung to re-expand is because of a fibrotic peel around the visceral pleura due to remote inflammation. Here, there is chronic negative pressure in the pleural cavity, and the effusion is the consequence of this negative pressure. This condition is called "trapped lung". The effusion is a transudate, and requires no treatment.

Pleural fluid manometry measures the relationship between intrapleural pressure and the volume of pleural fluid. Normal intrapleural pressure is slightly negative~3-5 cm H2O at functional residual capacity (FRC), allowing the lung to re-expand after expiration. When pleural fluid accumulates, intrapleural pressure rises. In the case of transudates such as with heart failure or hepatic hydrothorax, or exudates where the lung is able to re-expand, the pleural pressure falls steadily during aspiration until the normal slightly negative pressure of -3-5 cm H2O is reached at FRC.

The relationship between change in pleural pressure and pleural fluid volume is described as pulmonary elastance or Pel. The unit for Pel is cm H2O/litre.

In the case of lung entrapment, the pressure falls gradually during the first part of aspiration, but when more than 2000 ml is taken out, the pressure drops rapidly, giving rise to a steep curve in its later half. Pel is usually higher than 19 cm H2O/l.

With a trapped lung, the ambient pleural pressure is negative to start with, unlike the other two situations. In addition, as pleural fluid is withdrawn, the intrapleural pressure drops steeply, without an intervening flat trajectory. Thus, Pel tends to be very high, >25 cm H2O per litre.

These three situations are represented in the figure.

Thus, by observing the relationship between intrapleural pressure and pleural fluid volume during manometry, one can a)predict the chances of successful pleurodesis following aspiration of malignant effusions. One study found that Pel>19 cm H2O/litre during the first 500 ml of aspiration predicted unsuccessful pleurodesis (by failing to appose the visceral and parietal pleura) b) predict whether there is lung entrapment or trapped lung. and c) guard against RPO, as these typically tend to occur at pleural pressures more negative than -40 cm H2O. Some authors stop aspirating at intrapleural pressures below -20 cm H2O, or at the first sign of chest pain.

Reference:
Feller-Kopman D. Should Pleural Manometry Be Performed Routinely During Thoracentesis? Yes. Chest.2012;141:844-845.

ECG Problem

This ECG, from an elderly man with chronic heart failure, is relatively unchanged over at least 6 months. What did the echocardiogram show?

Courtesy: ECG Wave-Maven

http://ecg.bidmc.harvard.edu/maven/mavenmain.asp

The Curious Case of Sickle Cell C

Unusual observations in Medicine sometimes have very simple explanations. Take the case of sickle cell C disease, for example.

Evolutionary pressures have led to the existence of several mutants of the beta chain of haemoglobin. Thus Hb S, Hb C and Hb E all have mutations on the beta chain. For example, in Hb S, glutamic acid is replaced by valine in position 6, while in Hb C, lysine is substituted in the same position. Early on, epidemiologists noticed that these mutated haemoglobins were found in areas with high prevalence of falciparum malaria. For example, Hb C is found in Western Africa, Hb E is present in around 60% of subjects in the Indian subcontinent, and Hb S is widely prevalent in Africa. These variants have evolved because heterozygotes with Hb S, C or E are resistant to severe infestation with P.falciparum, and thus provide a survival advantage in these geographical locations.

In the normal adult, two beta chains combine with two alpha chains to form the complete globin chain (alpha2-beta2) and thus constitute the most abundant form of haemoglobin present in adults, known as Hb A. While the beta chain has only one gene, the alpha chain is coded by two genes. Thus, the alpha chains have 4 different alleles across the two chromosomes.

Heterozygotes with the sickle haemoglobin (sickle cell trait) have one normal allele producing the beta chain, and one mutant allele producing Hb S. Since each allele produces an equal amount of Hb A and Hb S, you'd expect an equal (50% each) proportion of Hb S and Hb A in subjects with sickle cell trait. Yet, this is not so. On haemoglobin electrophoresis, these subjects have 50-60% Hb A, and only 35-45% Hb S [the rest being contributed by Hb A2 (alpha2-delta2) and Hb F (alpha2-gamma2)]. Why does this happen?

As it happens, the reason beta chains and alpha chains join so harmoniously is because they carry an almost equal, and importantly, opposite electrical charge. Beta chains carry a negative charge of -2.5 coulomb (C), while alpha chains carry a positive charge of +2.4 C, thus ensuring electroneutrality (almost) when they combine.

However, the beta chain mutants are less negatively charged than the native beta chains. Thus, they combine less effectively with the alpha chain to form Hb S, C or E. This is why, in heterozygotes, instead of a 50-50 split, Hb A produced by the normal allele predominates over the variant haemoglobin. Thus, subjects with sickle cell trait have ~55% Hb A, and 40% Hb S, while heterozygotes for Hb E, have~ 70% Hb A and only 30% Hb E. This also explains why such heterozygotes are not anaemic. Subjects with sickle cell trait can only be picked up on electrophoresis, while heterozygotes with Hb E are only revealed by microcytosis with an absence of iron deficiency.

The principle is further illustrated in subjects with Hb SC disease. Here both alleles of the beta chain are mutant- one is producing Hb S, the other Hb C. As these two beta chain mutants have roughly equal charge (and thus affinity for the alpha chain), they are present in roughly equal concentration on electrophoresis~45-50% each. There is no normal beta chain to compete with.

A similar phenomenon occurs in sickle cell beta thalassaemia. As you may know, the defect in beta chain production in beta thalassaemia may be only partial (denoted as beta thal+) or severe (denoted as beta thal 0). Despite the deficit in production of normal beta chains, subjects with sickle cell beta(+) thalassaemia still have Hb A comprising around 30% of the total Hb in RBC, the other 70% being Hb S, as even in diminished quantities, the available normal beta chains combine more efficiently with alpha chains than the mutated beta chain found in Hb S. Thus, these subjects have a less severe phenotype than those with sickle cell beta (0) thalassaemia, who can't produce any Hb A.

This principle can be put to good use in the diagnosis of newborn subjects (with carrier parents) with one of the mutated beta chains. While Hb F is the predominant haemoglobin in newborns, the proportion of Hb A and Hb S will vary depending on homozygosity, heterozygosity and the co-existence of beta thal (+) trait. Thus, newborn with sickle cell disease will have a FS (F>S) pattern at birth, subjects with sickle cell trait will have a FAS (F>A>S)pattern, while a FSA (F>S>A)pattern at birth is diagnostic of sickle cell beta (+) thalassaemia.

Finally, a correction. In my post on hereditary spherocytosis, I had said that I did not know of any other condition that caused a high MCHC. This is incorrect. Subjects with Hb AC or Hb SC have RBC that are prone to dehydration due to a chloride channel defect, a condition known as xerocytosis. Due to loss of water, the RBC have a high MCHC, which might be the only clue to diagnosis in subjects with Hb AC.