Posted by Jonathan on March 10, 2003, at 21:58:56
In reply to Bowden: Guest expert on bipolar disorders, posted by Dr. Bob on March 4, 2003, at 8:57:29
... or do only unipolar depressives respond? or (making the question more general) What percentage response to Li augmentation (LiAug) of an antidepressant (AD) would you expect in populations of patients reliably diagnosed (a) unipolar and (b) 'soft' bipolar with depressive episodes so much more frequent than (hypo)manic that prophylaxis of the latter is not an issue?
If the percentage responses for these two populations are significantly different, then (non-)response to LiAug would be an aid to a notoriously difficult diagnosis. According to "Bartos", any patients now diagnosed as predominantly-depressed soft bipolar have a history of years of inappropriate treatment, when they were misdiagnosed as unipolar because the first hypomanic episode needed for bipolar diagnosis either had not yet occurred or had not been recognised as such by their doctor.
According to my psychiatrist, LiAug is tried only as a last resort on patients for whom all available classes of AD have failed (i.e. on a group for whom success rate of any AD without LiAug = 0%). Despite this selection of patients who are least likely to respond to anything, he claims an astoundingly high success rate of 57%. (I'm in the UK: the US figure may be different for various reasons including your higher diagnosis rate of BP2.)
Until recently I assumed that most of these LiAug responders, although (mis-)diagnosed as unipolar, are really undiagnosed BP2, BP3 or cyclothymics whose first hypomanic episode either has not yet occurred or was not identified as such. Such patients are likely to have a much higher frequency of depressive than of (hypo)manic episodes (otherwise the episode enabling a bipolar diagnosis would already have occurred), so antidepressant-induced cycling will probably present as a depressive episode soon after starting any antidepressant. Lithium augmentation would appear to succeed for this group by suppressing AD-induced cycling.
However, a few months ago I read your review paper, Clinical correlates of therapeutic response in bipolar disorder, J. Affective Disorders 67 (2001) 257-265, in which you say "Elated mania is quite responsive to lithium, but such patients are likely to suffer from worse depressive symptomatology during subsequent maintenance treatment with lithium." (Section 6, last paragraph, p. 260, col. 1); Slide 16 of your recent Grand Rounds presentation confirms the same phenomenon using a different experimental source. Combined with the well-known observation that Li on its own is an effective antidepressant for unipolar patients (e.g. Souza FG & Goodwin GM (1991) Lithium treatment and prophylaxis in unipolar depression: a meta-analysis. Br. J. Psychiatry 158: 666-675) this difference in the effects of Li on unipolar and bipolar patients suggests that, *if* the differential response is maintained in the presence of an AD, then only unipolar depressives would be expected to respond to LiAug, while bipolar depression would respond better to the AD alone than with lithium: the opposite of the previous paragraph's apparently plausible conclusion!
No doubt the truth is somewhere between these two simplistic and extreme views. Perhaps someone has performed a retrospective study in which patients' (non-)response to LiAug a number of years ago is matched with their present diagnosis as bipolar or unipolar, the latter being assumed to correct any misdiagnosis at the time of treatment? A couple of percentage response figures to plug into Bayes's Theorem would be ideal!
Dr Bowden, I am looking forward very much to hearing your views on this question, not least because of their possible implications for my own diagnosis. I recently started lithium augmentation of a tricyclic NRI, lofepramine, after four years trying various ADs without or with only ephemeral success. My current diagnosis is atypical depression, which according to Benazzi (Prevalence of bipolar II disorder in atypical depression, Eur. Arch. Psychiatry Clin. Neurosci. (1999) 249: 62-65) implies prior probabilities of about 2/3 bipolar and 1/3 unipolar.
Thanks for reading this, and for a fascinating and informative presentation.
Jonathan.
poster:Jonathan
thread:205791
URL: http://www.dr-bob.org/babble/20030310/msgs/207881.html