Written by Professor Valerie Lund
Thank you to the Semon Committee and Professor Gleeson for the honour of giving this lecture and Peter (Clarke) for his very kind introduction.
I was very lucky to have the benefit of a copy of Felix Semon’s autobiography in our library and of the Donald Harrison book and also to have had some time on a recent trip to Easter Island to consider this topic.
Felix Semon, of course, was born in Berlin 163 years almost to the day, and having trained in Germany came to London on a post-graduate Fellowship via Vienna and Paris. Here he visited six teaching hospitals and began what can only be described as love/hate relationship with Morell McKenzie who offered him a six-month clinical assistant post, following which Semon joined the staff.
Here we see him leaving Wimpole Street with his wife after an afternoon call on one of his celebrity clientele. He very much cultured Royal connections but also founded the London Laryngological Society in 1893 and this RSM section of Laryngology & Rhinology in 1907. Interestingly, in 1903 he also attempted to merge the Royal National Throat, Nose and Ear with Golden Square and the Royal Ear, which actually took another eighty years to achieve.
Mostly, he is best known for his now largely discredited law, which was based on experiments done by Victor Horsley and a single case report.
He didn’t actually have much interest in the nose apart from a condition which he described as “nasal reflex neurosis” which affected upper class women of a certain age, producing nasal symptoms associated with the menopause and depression but which could be successfully treated with the reassurance of a cheerful prognosis!
He also experienced acute maxillary sinusitis at first hand but went to Berlin rather than have it treated in London and even then would not allow the sinus to be washed out. Following this he claimed it was the first case he had ever seen.
He did express some opinion when he was the President of the newly formed section of Laryngology and Rhinology in 1888 on the use and abuse of local treatments in the upper respiratory tract, and in 1901 made a further attack on the practices of his colleagues who naturally took a rather dim view when he stated that, “exaggeration for the truly astonishing nature have been made by some doctors to pressurise patients in to surgery.”
Since the first Semon lecture was given in 1913 there have been 88 lectures, of which only 10% have been rhinologic, and you see below some of the recent well known recipients.
So, in considering whether we are over-treating the nose let us start with allergic rhinitis. This has been comprehensively looked at by the WHO group, headed up by Jean Bousquet and there is a lot of data on the effect of this disease worldwide and estimating the huge economic burden due to its effect on quality of life and the cost of treatment. However, for every medical therapy either in seasonal or perennial allergic rhinitis we do at least have level 1 evidence to support utilisation and these have been assimilated into a step-wise approach to treatment and subjected to cost-benefit analysis supporting the use of the most effective drugs and reducing the costs of treating co-morbidities, such as asthma.
The importance of validating such guidelines, however, was recognised over a decade ago when we were able to show that in general practice a group of over 200 patients treated according to the guidelines did better in terms of total symptom scores and quality of life as compared to those treated however the GP wished.
So, we can say in the medical treatment of allergic rhinitis we are not guilty of over-treating the nose. Indeed, we could even be accused of under-treating the condition.
When it comes to the management of rhinosinusitis, there have been three iterations now of the European Position Paper on this topic with the regular revisions reflecting the increasing number of good quality trials that are being performed in this area and EPOS covers both acute and chronic disease.
We know that acute rhinosinusitis is extremely common, affecting virtually all adults and children each year and represents a significant number of GP consultations and, in particular, costs for antibiotics. However, there is a well recognised transition from the common cold to a bacterial infection where symptoms increase after five days or persist after ten, and are characterised by discoloured discharge, severe local pain and fever.
However, true acute bacterial rhinosinusitis represents a very tiny proportion of the condition as a whole.
When we look at sales of antibiotics throughout the European Union we can see that there are surprising differences between some countries such as France at the top and the Netherlands at the bottom, and this is almost completely replicated when we look at prevalence of anti-microbial resistance.
So why do we use antibiotics? Well, because we believe acute rhinosinusitis is a bacterial disease and if we do not treat it, it can lead to severe life threatening complications. However, in one of the recent Cochrane reviews looking at 57 studies that compared antibiotic with placebo there was only a slight statistical difference in favour of antibiotics and their conclusion was that the small benefit gained may be over-ridden by the negative effects of antibiotics, both on the patient and on the population as a whole. Indeed, in the meta analysis published by Young and colleagues in the Lancet one needs to treat 15 patients in order to cure 1.
So from this we can conclude that acute rhinosinusitis is sometimes a bacterial disease but antibiotics have very little effect.
Do they actually stop complications? Well, if we compare the data from between France and Netherlands, countries with vastly different antibiotic usage, we can see that there is no indication that complications occur more frequently in the Netherlands than in France. Indeed, in a one year prospective study of all cases of acute complicated bacterial rhinosinusitis done through the BRS we were able to show that there were similar complication rates occurring irrespective of whether or not the patient had had antibiotics, as you can see from this table here.
So, we can conclude that whilst acute rhinosinusitis can lead to severe complications they usually occur early in disease and antibiotics do not seem to prevent them, and thus we can
conclude that we have been guilty of overuse of antibiotics of the treatment of this condition, which is a source of worldwide concern, as recognised by Margaret Chan, the director of WHO earlier this year, when in our lifetime we may have to face a world without antibiotics. The stromatolites and cyanobacteria that make up the white edge of this lake in Patagonia have been here for some 500 million years and will probably be here long after our demise.
However, there is some reason for optimism with studies from France and the Far East showing that increasingly EPOS criteria are being used both in the diagnosis and treatment of acute bacterial rhinosinusitis with antibiotics being reserved for the more severe cases.
When we look at rhinosinusitis we can see that approximately 10% of the European population have been shown to suffer from this condition, with or without nasal polyposis, though there are some regional differences. Again, this is a condition that costs a great deal of money both in terms of treatment, making it one of the most expensive of chronic diseases, both in the direct and indirect costs of absenteeism and presentism.
We rely heavily on symptoms to make the diagnosis reported in secondary care with endoscopy and/or CT scanning, but what is normal when it comes to symptoms?
We have some normative data on the number of times it is normal to sneeze or blow the nose, the range of mechanical obstruction and levels of qualitative smell both for age and sex. Even nosebleeds occur quite commonly in the normal population with nearly a quarter of women experiencing some bleeding from the nose at some point during a twelve month period. However, there have been some large studies such as the RCS of England Sinonasal Audit, looking at patients both with or without nasal polyps and confirming that nasal obstruction, problems with the sense of smell and nasal discharge feature highly in these populations.
Based on an increasing number of randomised controlled trials we can support the use of medical therapies and wrestle them into algorithms, of which I am notoriously fond, prompting referral of patients if they fail. Then, following appropriate diagnosis and further medical therapy, a proportion of patients will be considered for surgery, prior to which a CT scan would normally be done.
So what then constitutes an abnormal CT scan? The problem is about 20% of normal people have minor mucosal changes, probably related to a viral cold in the last six weeks or so and the average Lund-Mackay score has been estimated to be somewhere between 3.4 and 5.1 in an adult population.
What about the state of anatomic variants in the aetiology of these conditions? Well, if we look in the literature the frequency range is considerable, with one or more being found in about 2/3 of patients, but they are no more common than in the normal population, are not associated with higher opacification and are not correlated with symptoms scores. So, there is no causal link and therefore no indication to remove them per se, except perhaps for access.
The same applies to septal deviation which is said to be present in anything between 13% and 96% of the population, depending on the method of diagnosis. Once again, there is no correlation overall between the presence of a deviation and the development of CRS and in the absence of a causal link, again there is no indication to straighten the septum, except perhaps for access.
Sometimes it is claimed that septal surgery is necessary for access. However, I would like to point out that in the last 27 years I have done over 7,000 endoscopic procedures on the nose, and I have had to do septal surgery for access in fewer than 20 cases.
The last incidental finding I would like to mention is the maxillary mucosal cyst, which is asymptomatic and found on a 1/3 of normal scans.
I think that my conclusion would be that these entities are often blamed but rarely responsible. However, with appropriate patient selection there is good evidence that surgery is of benefit to patients with chronic rhinosinusitis. Our study, which compared medical and surgical treatment showed considerable benefit in both arms in all the objective and subjective parameters, including improvement in the lower respiratory tract in those with asthma, but did not demonstrate a difference between the two groups, reinforcing the need for adequate prior medical therapy.
The National Comparative Audit looked at the 3,000 patients in what we might call the real world and relied particularly on the SNOT-22 to assess improvement, which indeed it did and
continued to show significant improvement over the five year follow-up with a high effect size as shown in this graph here.
On the face of it we are not guilty of overtreating chronic rhinosinusitis in patients who have failed appropriate medical therapy except when we look in a bit more detail at the audit which rather surprisingly showed that the Lund-Mackay score in over a fifth of patients was within the normal range and that a smaller, though nonetheless significant number of patients had a normal SNOT-22 score prior to surgery. As a ten point change in the SNOT-22 is required to detect a change in symptoms, a patient who scores under 20 is rather unlikely to benefit from treatment. We also have good studies now to support the use of long term medical therapy in maintaining that benefit.
This is particularly vexing when we see surgery for the treatment of chronic sinusitis featuring on the lists of that most irritating acronym, Procedures of Limited Clinical Effectiveness. More information of which can be found within this volume of joy.
What is interesting, if we look at numbers of ENT procedures and procedures on the nose and sinuses, whilst there have been some fluctuations over the last decade they are as nothing compared to the inexorable rise of procedures overall in the NHS since the turn of the century.
And we can see overall that when it comes to the septum and indeed the turbinates there has been a small but consistent reduction since 1999 for both these procedures shown in red and green respectively.
So, when it comes to septal surgery who are we operating on and why? I have already mentioned the number of deviated septa varies considerably in the literature, but one thing is clear – that there is a poor correlation with the sensation of nasal obstruction as evidenced by my own nose.
Unfortunately, when patients undergo septal surgery only about ¾ are satisfied with their surgery at 6 months, and this recent study showed that the satisfaction rate does not necessarily correlate well with improved breathing.
My good friend, Mats Holmstrom highlighted the need for objective measures in selecting patients for septal surgery. How many of you in this audience would undertake grommet
insertion without an audiogram? It is hardly surprising that patients who have a normal airway to start with do rather poorly when surgery is undertaken to improve it, and we now have a range of simple, cheap and reliable instruments, such as nasal inspiratory peak flow or nasal spirometry to inform that decision, particularly when we know that inflammation is often interpreted by the patient as nasal congestion or nasal blockage, which will improve with endoscopic sinus surgery despite there being no effect whatsoever on actual air flow. And if anyone asks you what the difference to an SMR and a septoplasty is, then according Tony Bull, it’s 80 guineas.
And here you see a graph of my figures showing that I do two or three septal procedures a year, almost all at the behest of the ophthalmologist for placement of DCR Jones’ tubes.
So, what about the turbinates then? Equally, poorly represented on my graph. But often combined with septoplasty due to compensatory hypertrophy. But this recent study confirms many older studies from the 1990s showing that compensatory change will improve with time if the patient can be persuaded to wait, so it may be worth you and the patient resisting the temptation of attacking the turbinates, bearing in mind this may increase the haemorrhage and adhesion rates significantly.
Of course, the methodology of turbinate reduction is many and various, and in analysing the studies there is no gold standard with very few randomised trials, certainly no placebo controlled trials and generally very short follow-up.
As a summary, one can say that those techniques which remove the most tissue have the longest effect but also the greatest morbidity.
As to when one would undertake this surgery, as always, the mantra of ‘after failure of adequate medication’ is invoked, but following this and working with my colleague Glenis Scadding over the last thirty years, I have undertaken a total of forty turbinectomies on patients with intractable nasal obstruction with perennial non-allergic rhinitis.
So, I think we can say that with some evidence, that rather more septoplasty and inferior turbinate reduction is undertaken in this country than is absolutely necessary, and in the absence of an event, such as significant trauma, we should question why a patient who has had a septal deviation throughout their life presents now with complaint of nasal blockage, which is more likely to relate to the present inflammation than the mechanical obstruction, and therefore resist the temptation to automatically put them down for septoplasty, turbinate reduction and endoscopic sinus surgery.
So, finally, I would like to say a few words about facial pain and nasal surgery. We all know that patients focus on facial pain and interpret it as being caused by sinusitis. They come to the Friday afternoon, clutching their piece of paper and a large print out of any papers you might have written in the last twenty-years, and you take the paper and thank them very much, and put it in the notes, but then they give you another piece and say “doctor, you don’t understand. I really do have this terrible, terrible pain which is coming from my sinuses” and these drawings are often colour coded.
And then finally, they give you one that has purple on it and purple, as you know is a very bad prognostic sign in rhinologic surgery, and from that moment on you can guarantee that there is very little that you can do.
And yet the paradox is that many patients, particularly those with nasal polyps, have minimal symptoms yet maximal disease and will often present with only nasal obstruction.
So, bearing this in mind, my final message is ‘surgeon beware of the patient with facial pain’, even if they have a large osteoma. Pain is not generally a feature of sino-nasal tumours, and although many patients with osteoma complain of headache, unless there is actual obstruction of the sinus, as in a case like this, it is unlikely to be the cause, per se.
So there has been no more important time to critically examine what we do and why we do it, and as the government scrutinise variations of behaviour around the country and we grapple with the minefield of commissioning, I can conclude that there is evidence both for and against the proposition that we are over treating the nose.
But I would like to make a plea for careful patient selection, whatever we do, using simple assessment tools and utilising simple outcome measures as there has never been a greater need to phenotype out patients and provide evidence that they are getting value for money. This will be one of the planks for revalidation and will underpin the push from government to provide individual outcome data to the public.
Overexploitation of resources led to the demise of the Rapa Nui of Easter Island.
Many sinonasal conditions have severe effects on patients and their quality of life, which is underestimated by the politicians and primary care, but unless we refine our patient selection and persist in doing operations without good evidence of their need, we will lay ourselves open to criticism and bad things will follow.