Controlled and uncontrolled or open studies of the CCBs in affective illness are reviewed in Table 6–3. Initial open and blind studies of the phenyl-alkylamine L-type CCB verapamil were positive in the affective disorders, particularly in the treatment of acute mania. However, some preliminary con-trolled data are negative (Janicak et al. 1998); these data are highly subject to a type II error with the design used, the relatively small numbers of patients randomly selected for verapamil and placebo, and the associated relatively high placebo response rate in acute mania observed in many controlled studies Figure 6–2. Graph showing sustained prophylactic efficacy of carbamazepine combination treatment in patients (N = 24) with increasing primary affective ill-ness morbidity before institution of treatment. Numbers within bars are numbers of patients.
of this type (Bowden et al. 1994b). Bowden et al. (1994b) had to use substan-tial sample sizes to demonstrate significant effects of valproate compared with those of placebo. Also, in this analysis (Bowden et al. 1994b), the effect of lithium versus placebo on the manic syndrome was nonsignificant, presum-ably because of the smaller number of patients randomly selected to that arm compared with the other groups. Thus, this largest controlled study of lith-ium in acute mania to date illustrates the perils of the use of this type of paral-lel design in initial or exploratory studies of the potential efficacy of Figure 6–3. (Top) Life chart of a patient with affective disorder showing a progressive loss of efficacy of carbamazepine. (Bottom) Preclinical data showing pro-gressive loss of efficacy of carbamazepine in an amygdala-kindled rat.
psychotropic agents in acute mania or depressive syndromes.
To help avoid these study design confounds, we have used other statistical designs such as mirror-image strategies and B-A-B-A (off-on-off-on) designs.
These designs are subject to statistical verification within individual subjects and reconfirmation of responsivity by the appropriate mood assessments in different phases of active treatment. Thus, once efficacy has been estab-lished in a subgroup of patients, clinical focus can shift rapidly from demon-stration of acute efficacy to determination of whether a subgroup of patients with specific clinical or biological markers predicts this response. Although these strategies are not widely used in the psychiatric community, and have been criticized, they would appear to markedly accelerate the potential de-velopment of new agents, bringing them into a focus for clinical therapeutics.
That is, once efficacy has been unequivocally established in a small group of well-characterized individuals studied with systematic blind designs, the per-centage response rate and identification of potential clinical and biological markers of response can then be ascertained in other designs:
Once a true process or effect has been established as having occurred in one person it can reasonably be assumed, or inferred, that there will be other per-sons as well in which the process or effect will occur. (Chassen 1992, p. 177)
Moreover, for patients who show a wide range of cycle frequencies (as is typical for bipolar illness), this strategy is likely to avoid many of the pitfalls associated with a high rate of “placebo” response, actually attributable to the natural course of illness and highly predictable on the basis of systematic ret-rospective and pret-rospective life charting (Post et al. 1988; Squillace et al.
1984).
We have adopted such a strategy in the exploration of the psychotropic and antimanic effects of the dihydropyridine L-type CCB nimodipine in pa-tients with rapid, ultrarapid, and ultra-ultrarapid (or ultradian) cycling bipo-lar illness. Nimodipine has important differences from verapamil and diltiazem in pharmacokinetics (better central nervous system [CNS] entry [D. D. Freedman and Waters 1987] and fewer drug interactions [Ketter et al. 1995]) and a host of biochemical and pharmacological differences (Table 6–4). In the initial work of Pazzaglia et al. (1993), response in five of nine pa-tients in the group evaluated was, in many instances, reconfirmed in a B-A-B-A design. A total of 10 of 30 (30%) patients responded to blind nimodipine monotherapy (Pazzaglia et al. 1998). P. J. Goodnick (1995) has also reported positive nimodipine response with open trials in two patients with ultradian cycling bipolar illness.
Table 6–3. Calcium channel blockers in affective illness Open studies Results Blind studies Results
Verapamil Verapamil
Giannini et al. 1987 20 M equal to Li
Patterson 1987 1/1 M Giannini et al. 1989 10 M equal to Li, better than valproate Pollack and
Rosenbaum 1987
1/1 UP Dubovsky et al. 1985 1/1 Ma
Deicken 1990 1/1 BP Dose et al. 1986 7/8 M Hoschl et al. 1992 ?/4 BP-Dep
?/7 UP Garza-Trevino 1990 17 M equal to Li Garza-Trevino et al. Janicak et al. 1998 3/17 vs 2/15 placebo
Nimodipine Nimodipine
Brunet et al. 1990 6/6 M Pazzaglia et al.
1993, 1998
7/23 BP
Manna 1991 12 Mb 0/4 UP
(continued)
P. J. Goodnick 1995
2/2 BP McDermut et al.
1995
9/10 UP Montenegro et al.
1985 Ban et al. 1990 87 on nimodipine
88 on placebo
1/1 M Eckmann 1985 14/17
Diltiazem Isradipine
Caillard 1985 5/5 M Pazzaglia et al. 1993, 1998
Note. BP = bipolar disorder; D/C = discontinue; Dep = depression; Li = lithium;
M = mania; SA = schizoaffective; RBD = recurrent brief depression; UP = unipolar.
aDrug-induced hypomania.
bLithium and nimodipine combination in prophylaxis is better than either drug alone.
Table 6–3. Calcium channel blockers in affective illness (continued) Open studies Results Blind studies Results
Nimodipine Nimodipine
Table 6–4. Differential effects of dihydropyridine L-type calcium
Patients + + NA Brandt et al. 1988
Larkin et al. 1991 de Falco et al. 1992 Block
cocaine-induced:
a. Hyperactivity + + 0 Pani et al. 1990
Rossetti et al. 1990
de Jonge et al. 1993 Mogilnicka et al. 1987,
1988 Positive in learned
helplessness model
+ + 0 Geoffroy et al. 1988
P. Martin et al. 1989 de Jonge et al. 1993 Like antidepressants:
Much more detailed analysis of the statistical methodology used with this design was reported in the study of McDermut et al. (1995), again docu-menting the unequivocal statistically and clinically significant effects of nimodipine monotherapy in a patient with refractory depression and previ-ous ultrarapid cycling (Figures 6–4A to 6–4C). This ultrarapid cycling course reemerged during the period of placebo administration and was again sup-pressed with active nimodipine therapy. Remarkably, when a blind crossover to the maximal tolerated daily dose of 320 mg of phenylalkylamine CCB verapamil was attempted, mood instability rapidly reemerged. Even at maxi-mally tolerated doses, this patient was unable to maintain her former degree of clinical improvement. When verapamil was replaced with nimodipine on a third double-blind trial, response was evident. In an attempt to see whether this type of response to the dihydropyridine L-type CCB would show cross-responsivity to another dihydropyridine CCB, the patient was switched in a blind fashion to isradipine (21 mg/day), a successful transition that resulted in discharge on monotherapy.
This case report (Figures 6–4A to 6–4C) (McDermut et al. 1995) of selec-tive response to dihydropyridine CCBs but not a phenylalkylamine CCB is of considerable interest in relationship to the patient’s history of nonre-sponsivity to multiple tricyclic antidepressants, the selective serotonin reuptake inhibitors, lithium, carbamazepine (the patient developed drug-induced hepatitis on carbamazepine and was unable to be evaluated), alprazolam, trazodone, and phenelzine. This suggests that patients with re-fractory mood disorders may have differential responses to various CCBs and that nonresponse to one CCB does not preclude response to another CCB, particularly if the other CCB is from a different category (Table 6–3).
Brunet et al. (1990), in an open study, reported positive antimanic effects of nimodipine in six patients with acute mania. Our results showed a much lower response rate. However, our patients were much more refractory by history, were treated in a tertiary referral research center, and generally showed a higher incidence of rapid, ultrarapid, and ultradian cycling patterns than in more traditional studies. Our results, however, are consistent with those of Manna (1991), who reported equal long-term efficacy of nimodipine and lithium monotherapy and greater efficacy on a combination of the two drugs than on either drug alone.
Although to date we have studied only 30 patients in a double-blind fash-ion on nimodipine (Pazzaglia et al. 1998), given the methodology involved and the consistency of clinical response observed in the small subgroup (n = 10) of refractory patients, together with the larger double-blind litera-ture on verapamil and other CCBs, it would appear that CCB therapy should
receive greater systematic study and clinical experimentation, particularly in treatment-responsive and -refractory patient subgroups. Other data support this view, such as Dubovsky and Franks’ (1987) observations that nonresponders to lithium predict nonresponders to verapamil. Most of the responders to nimodipine in our study were lithium-nonresponsive in the past, however, and the relationship of these two different L-type CCBs to ul-timately responsive clinical subtypes remains to be further delineated.
Dubovsky (1995), moreover, noted that the CCBs show relatively greater safety in pregnancy, thus offering another potential alternative to lithium and the teratogenic mood-stabilizing anticonvulsants valproate and carba-mazepine. Furthermore, in light of the adjunctive role for nimodipine de-scribed by Manna (1991) and in our studies of the combination of Figure 6–4A. Prospective daily life chart ratings of a patient’s mood as de-termined by nurses blind to medication status. Boxes above the horizontal line (euthymia) represent mild, moderate, or severe mania. Boxes below the line repre-sent mild, moderate, or severe depression. A complete record of all the patient’s medications for each day of the study is presented below each life chart line. Num-bers indicating the day of the patient’s hospitalization are in 2-month intervals.
nimodipine and carbamazepine, the use of nimodipine and related agents in augmentation therapy remains a promising area for clinical exploration.