Monday, 27.03.2017

Comment by Malcolm Lader

Lithium Controversy - Malcolm Lader’s Comments

I was honoured to receive an invitation from Tom Ban to contribute to this interesting topic. I am gratified to join so many distinguished elder colleagues. I conducted lithium clinics at the Maudsley for many years but these were untypical, tertiary referrals and usually severe and/or treatment refractory.

Too many points have been raised to discuss in detail but I hope my observations are not too destructive, and add light rather than heat on the issues.

Some of the debate revolves around diagnosis or at least what passes for it in much of medicine, psychiatry in particular. The Oxford English Dictionary, among others defines diagnosis as the identification of the nature of an illness or other problem by examination of the symptoms. It then implies that treatment can follow. In many areas of medicine, the diagnosis is a convenient shorthand description. In psychiatry, the DSM criteria have over the years provided a reasonable consensus but based in the past on clinical and epidemiological observations rather than on any evidence-based science. When it comes to treatment, very little rigorous evidence is extant, only conclusions drawn from artificially constructed clinical trials meeting certain criteria - defined samples, randomised allocation to treatment, double-blind procedures, acceptable outcome criteria and appropriate controls are paramount. Despite this, there is a demoralising uniformity about the outcomes in most of psychiatry – a third of the sample respond to dummy treatment, another third to the test procedure, leaving a third unresponsive. Unfortunately, the predictability of such an outcome in each individual remains too imprecise for clinical use.

I trained under Aubrey Lewis and was his assistant in the final years of his sway at the Maudsley. We often discussed diagnosis. He was sceptical and much preferred to formulate a patient in terms of her or his problems. The treatments were consequently practical but empirical. With respect to lithium, we would show the uncertain nature of diagnosis in the affective disorders by making a coruscating analysis of the symptoms of a patient diagnosed by our trainees as unipolar and uncovering mood changes that could be interpreted as above the normal range, thus raising the possibility of a bipolar II diagnosis. Michael Shepherd, with whom I also trained, was a past master of this but he had a very wide overview of “normality”.

The practical consequence was that the diagnosis was relegated to a secondary place where treatment implications arose. My own scepticism persisted and I was not obsessive at matching diagnosis to treatment. I would much rather attempt a trial of therapy. But it had to be a properly organised trial with observations before, during and after treatment. I accumulated a plethora of N=1 trials but also some treatment successes despite an unpromising prognosis.

Is the mirror design used by Mogens Schou essentially a compilation of N=1 trials of therapy? Mogens, whom I met many times, was an unassuming and warm person. We discussed the issue of the design of his prophylactic trial amicably and concluded that the design had strengths and weaknesses. The classic controlled clinical trial was but one approach. In the absence of reasonably precise prognostic indicators, it could lead to misleading results if the refractory population was too great. We agreed that in the uncertainties that plagued the field, a more practical approach was a series of trials-of-therapy and then an attempt to identify who was likely to continue to respond.

Notwithstanding, it is important in a study attempting to establish efficacy for prophylaxis that gold-standard criteria are followed as far as possible. Thus, for a remedy for an acute episode, a false positive due to natural remission will expose the patient to a relatively short period of treatment with the attendant risks. In trials of prophylaxis, a false positive with an apparent response with suppression of episodes, i.e., when natural remission occurs, will be followed by indefinite treatment. This is unless the prescriber is willing to try a period of no treatment to see whether relapse – that is more episodes occur a trial-of-no-therapy. This is especially problematic when a medication with high risks, such as lithium, is involved. In my medico-legal practice – still ongoing – I documented over 50 cases of irreversible lithium neurotoxicity, usually, but not always, associated with poor management (KORES, B. and LADER, M.H.,1997).

I was also sceptical as to whether a real prophylactic effect was being shown. Strictly speaking, a prophylactic is a medication or a treatment designed and used to prevent a disease from occurring. But the complicating factor was that lithium has a therapeutic effect in treating actual episodes of affective disorder, particularly mania. If so, then the episodes were not being prevented but rather suppressed early in their evolution. But did it matter? If the burden of illness was being lightened, this was a worthwhile therapeutic effect.

In the earlier history of lithium use, as expounded by Sam Gershon, it is clear that the development of a reliable and valid method for monitoring serum lithium concentrations was the key advance. It transformed a crude therapy with an unacceptable risk/benefit ratio into a crude therapy which could be monitored. Done properly, such monitoring prevented almost all the toxicity associated with its use. Nevertheless, careful clinical observation was still necessary to obviate some subtle toxicities. I am less convinced that serum monitoring facilitated the establishment of a therapeutic range within which efficacy could be demonstrated.  I have seen patients doing well on low levels and doing badly despite putatively adequate levels. Also I would often need to combine mood stabilisers in a medieval type of polypharmacy.

We should remain humble in our acclamation of our modern psychopharmacological therapies. Many are symptomatic remedies, even those for anxiety, panic and schizophrenia. With respect to the affective disorders, we have all seen gratifying major improvements in treating episodes of affective disturbance and apparently attenuating further attacks. We have also been baffled by treatment-resistant patients. If only we could predict who would do what!


Kores B, Lader MH. Irreversible lithium neurotoxicity:  An overview. Clinical Neuropharmacology 1997; 20, 283 - 99.


Malcolm H. Lader

April 16, 2015