[  Site Map  ]

Medical Papers

The Neutralisation Technique

Dr David L. J. Freed, MB, MD, MIBiol

For Debate

The efficacy of neutralisation is no longer a subject for serious debate in our circle or among those who practice evidence-based medicine. I reviewed and defended the method in 1987 (1). Since I started using it in 1985 I have experimented with certain variations and my practice now departs in several particulars from the conventional Miller method (2) (how ironic that already we can start describing Miller's method as "conventional"!) My purpose is to invite debate, among those who already neutralise, on the optimal way to do it.

The first question concerns allergen concentrates ("mother liquors"), which can be bought from either Eaton Laboratories (01344-53919), Miles Laboratories (001-509-482-1732; allow for time difference of 8 hours) or from homoeopathic pharmacists. Food extracts can be bought from the same suppliers or home-made (3). Keith Eaton and I have discussed this at length elsewhere (3) but I will summarise the unanswered questions here.

1)  For how long should we extract the solid starting material?Seven days, for completeness of extraction, as KKE does, or five minutes, as DLJF does in order to minimize chemical deterioration?

2)  Should the liquid extract be filtered (KKE) to maximise purity, or deliberately left turbid (DLJF) in order not to lose insoluble moieties that might be allergenically relevant?

3)  Should mother liquors be stored cold (KKE) or protected from light (DLJF)? 4)  Should extracts be defatted?Should food extracts be from raw or cooked food (for those normally eaten cooked). Remember, Kstner (of Prausnitz-Kstner fame) was exquisitely allergic to cooked fish but not at all to raw.

5)  What shall we do about preservative?

6)  And how, oh how, can we standardize the extract so we know how much, and what, we've got in it?

These and other questions are discussed in reference (3), written 5 years ago. We still have no answers. Most of the questions are answerable by research and it's about time we did it.

Now to questions of technique.
I explain herewith my current procedures in the hope that someone will comment. Those who do should please cite evidence to what they say.

You cannot put your number one dilution (1/100 w/v) into anyone's skin because of the glycerol (added to the mother liquor to stabilise proteins). It makes an artificially hard weal. So I start with the second dilution. For foods I usually inject dilutions 2 to 6 in a row, simultaneously, starting at the top of the arm and working downwards. An intra dermal injection of 50 µl raises a little lump (weal) about 7mm across. The next row then starts at the bottom of the deltoid area and works up, parallel to the first, to reduce the likelihood of two Type I reactions (see below) close together (which would interfere with each other). Very occasionally you find a skin that produces strong reactions to phenol or glycerol. Since phenol and glycerol are present in every injection, you need to allow for these when you come to read the other skin tests. It also does no harm to neutralise the patient for phenol and glycerol, although I am not sure whether it does any good.

After having done (say) ten parallel rows of injections (ie.ten different substances) I go back and inspect the first. Miller (2) lays down very precise criteria and times for reading the weals, and the beginner should stick to those, but I have had superior results in practice, especially in multi-sensitive patients, by being more flexible. Instead of waiting for exactly ten minutes, I wait until the softest weals in the row have almost disappeared into the skin, and that usually takes 15-25 minutes. The softest weal is then the neutralising dilution, and I administer a further injection of that dilution as soon as I am sure.

Frequently you find two or more weals that go flat at the same rate, and I find it pays to palpate them very carefully to try to find the softest. It does not matter where in the row the softest weal is, that is the neutraliser.

In Miller's original method the dilutions are injected singly, and you wait ten minutes before injecting the next, in order to allow symptoms to develop or abate. The development of symptoms was believed to have diagnostic significance, but in fact is unreliable (4). Neither symptoms nor skin reactions tell you reliably whether or not the patient is allergic to that substance. The neutralization technique is a treatment method pure and simple, NOT a diagnostic test. One advantage of injecting multiple substances and dilutions simultaneously is that you rarely provoke symptoms. Another is that you then have the ability to compare the weals with the others of that row.

The symptoms I have seen during testing are either (a) asthma, during pollen or dust-mite testing, or (b) acute hyperventilation, during any testing. Both resolve rapidly once you inject the neutralising dilution.
I keep adrenaline handy, but have never had to use it (see below).

Types of Reaction

The commonest reaction is no reaction. The skin swells initially with the volume of fluid injected, to give the weal, then gradually shrinks back to its normal configuration. There is often some reddening that resolves within a few minutes; this signifies nothing. If the first (no 2) dilution gives the softest weal, that is the neutraliser.

Sometimes you get a type I reaction. This is the classic IgE-histamine type of reaction, with weal and flare, and frequently occurs when testing atopics. As always, the softest weal is the neutraliser, provided there is no visible erythema at 10-15 minutes. The first dilution that fails to give a weal-and-flare is not necessarily the neutraliser; you may find that one or two dilutions further on.

With type I reactions to pollens and Dermatophagoides you often get the so-called "hourglass" phenomenon. The first few dilutions (the strongest ones) give classic weal-and-flare reactions, diminishing as the dose gets smaller and dwindling to nothing at (say) the sixth dilution. But then as you go weaker still, say the eighth and ninth dilutions, the type I reactions come back again, and can be indistinguishable from those at the stronger doses. Very rarely, as you get weaker still, the reactions will dwindle away again, then come back again as you get into the 13th and 14th dilutions (a two-waisted hourglass), to finally disappear at say the sixteenth.

This is a fascinating phenomenon and I wish I understood the mechanism, but I don't. Unfortunately many doctors, having heard the phenomenon described but not having seen it, look no further for an excuse to condemn the whole neutralisation phenomenon as obvious rubbish and its practitioners as mad. This is a shame for their patients who are thus denied a useful therapy because of the doctor's prejudice, and a terrible illustration of how prejudice can blind us to what lies literally in front of our eyes. The important thing is not so much to understand the mechanism, but to know what to do in practice. In this case the important thing is to get the right neutraliser, and in this choice you are obviously limited to the non-reactogenic dilutions. You may have a choice, say, between dilutions 6,7,11,12,15 and 16, and you make that decision in the usual way, by feeling for the softest. Sometimes you just cannot decide, and then you have no choice but to issue both candidates in separate bottles, and instruct the patient to try them each, on separate weeks.In practice, if you cannot decide which is best, both will usually work equally well.

Miller (2) insists that the best neutraliser is the first non-reactogenic dilution (counting from the strong end of the row). I disagree. I find I can often do better by moving one or more dilutions further along, looking for the softest (obviously this option is only available to you if you inject all the relevant dilutions simultaneously, or the differing ages of the weals will confound you). I have not done a formal trial to validate this practice, and that obviously must be done, but I have tried both candidate neutralisers, alternately, in dozens of patients and have almost always achieved better desensitisation using the criterion of touch and ignoring the order.

Type III reactions are occasional sequelae of type I reactions. They are slower, bigger, and more painful. They do not have the sharply demarcated edges characteristic of type I reactions or classic Miller-type positives, but rise in a gentle slope from the surrounding skin.They are maximal at 6-8 hours. They are due to intra dermal complement activation and recruitment of neutrophils to the site. They are rare, and give no useful information beyond the fact that there was once a type I reaction at that spot.

By contrast, type IV reactions must be anxiously searched for as they can sometimes change your treatment. Type IV reactions are also known as delayed-type hypersensitivity (DTH) or tuberculin-type reactions, and are exactly similar to a positive Mantoux test. They are hot, red and painful, not very swollen, and are not seen until the following day at the earliest (sometimes not for a few days).
You occasionally see a DTH reaction at a dilution which the previous day looked like the perfect neutraliser, and you must obviously alter your therapy accordingly, or the patient will get local inflammation every time he uses that injection and will not neutralise properly. With antigens that frequently do this (microbial extracts) I always get the patient to come back the following day so that I can check for DTH's, and I mark the skin with indelible pen on the first day so that I know which dot is which. It is essential to keep accurate records when skin-testing, so that if an unexpected DTH shows up the following day, you can identify it.If you are uncertain whether one of yesterday's spots is a DTH or a bruise, press it with a glass slide or tumbler. A DTH will blanch, a bruise will not (this is an old dermatologist's trick for examining lupus vulgaris).

Dangers of Testing

Obviously the most worrying danger is anaphylaxis or acute severe asthma. The first of these is rare, but you had better be alert to it and know how to cope with it. The only drug fast enough to make any difference to anaphylaxis is subcutaneous adrenaline. By the time your steroid injection has started to work the patient will be either recovered or dead. Make sure you keep up-to-date with your intubation and cut-down skills, and have the necessary equipment ready (not forgetting to keep an eye on the state of the laryngoscope batteries). Acute asthma is not so rare, especially when testing atopics; again, make sure you can cope. In practice, neither condition normally evolves so rapidly that you can't find and administer the neutralizer before it gets serious, and once you have administered that the reaction usually abates quite quickly - nevertheless, be watchful in case it doesn't. In my 25 years of allergy practice I have never yet had to resuscitate anyone from acute anaphylaxis, but maybe next week I will. All this is how I advise newcomers. Actually I myself keep adrenaline but nothing else. I do not possess oxygen, laryngoscope or cut-down set, and I have not had any practice with them for a quarter-century. I never use "tri-salts". I just rely on getting the neutraliser in quickly enough to avert disaster. The first one of us who kills a patient during testing will probably cause the whole panoply of conventional medicine to descend upon us in wrathful majesty and ban the lot of us.It's a responsibility.

Far more common than allergic reactions is acute hyperventilation, especially when the dose last administered in any row is a lot more dilute than the neutraliser (that is, you are in the under dose phase). If you surreptitiously watch the front of the patient's chest, you will soon learn to spot the tell-tale rise-and-fall of the chest wall, interspersed with sighs, often before the patient himself is aware of symptoms.Here, of course, is a snag to my practice of injecting several rows of substances simultaneously; if you do cause hyperventilation, you don't know which one has done it.It doesn't matter, though. Just proceed seriatim and give neutralisers for everything as soon as you have identified them. I rarely have to ask a patient how he feels; I usually tell him.

I once had a patient who started chronic hyperventilation soon after the end of the testing programme, and was so scared of the treatment that he never used it. He was one of my failures, though for an unusual reason!


Once you have worked out the various neutralizing dilutions ("end-points") for everything that you want to cover, you can combine them into a multivalent "vaccine" (better termed "neutragen" to avoid confusion). Remember that when you mix one volume with four others, it is now further diluted to 1/5 of what it was. The way to overcome this difficulty is to use a volume (say 0.5 ml) from the bottle 25 time stronger, and mix that with equal volumes of 24 other ingredients, each of which is also 25 times stronger than the final strength you want.Thus, if your patient neutralises, say, on chicken 3 and pork 4, you would take a volume each of chicken 1 and pork 2. If arithmetic makes you panic, learn the rules by heart.

For youngsters, the neutragen works across mucous membranes as well as by injection. It can be used in the form of sublingual drops, nasal drops and eye drops, depending on which organs are involved. The older the patient, the less useful this route is. In practice, for anyone over 20, subcutaneous injections are best. Initially it is safest to administer the injection(s) daily, then after a month or two the patient can gradually reduce the frequency until desensitisation is complete and no further neutragen is needed. Virtually all patients can be taught to self-inject using insulin syringes. Expect relief of symptoms to begin within days (sometimes hours), but some cases take weeks to respond.

It is quite handy to keep a dog in the clinic.
When you have neutralised a dog-sensitive asthmatic, send him out into the corridor to play with the dog for ten minutes. That serves to convince both of you that your method works. But don't try it with cats; they are too immunologically individualistic and you may have to neutralise for that patient's personal cat.

Parting Warning!

Beware of the temptation to use skin-testing for diagnostic purposes. It is quite true that a floridly positive skin-test makes it probable that the patient is genuinely allergic to that substance, and it is also true that if skin-testing provokes systemic symptoms that often means the same thing, but false-positive and (more dangerous) false-negatives are frequent.
Treat the patient, not the skin-tests.


1)Freed DLJ (1987)
The provocation-neutralisation technique.
in (ed) Dobbing J: 'Food Intolerance'
Bailliere Tindall, Eastbourne, 151-184.

2)Miller JB.(1972)
'Food Allergy: Provocative Testing and Injection Therapy',
Charles C Thomas, Springfield Illinois.

3)Eaton KK, Freed DLJ (1991)
the Preparation of allergen extracts for environmental medicine.
'Environmental Medicine'
8: 101-105.

4)Jewett DL, Fein G, Greenberg MH (1990)
'A double-blind study of symptom provocation
to determine food sensitivity'.
New England Journal of Medicine. 323: 429-33.

# Back to top of page

Return to Medical Information Index

Return to the Home Page