Home

German Journal of Psychiatry

ISSN 1433-1055

A Naturalistic Study of Response in Melancholia to Transcranial Magnetic Stimulation (TMS)

S. Pridmore, M. Rybak, Y. Turnier-Shea, P. Reid P., R. Bruno, D. Couper


Royal Hobart Hospital and University of Tasmania

Hobart, Tasmania, Australia

 

Corresponding Author: Prof. S. Pridmore, M.D., Department of Psychological Medicine, Royal Hobart Hospital, Hobart, Tasmania, Australia, 7000, Telephone: +61 3 6222 8804, Fax: +61 3 6234 7889, e-mail: S.Pridmore@utas.edu.au


Abstract


Background: Twenty two consecutive patients with 24 major depressive episodes (DSM-IV) and melancholia (CORE system criteria) which had failed pharmacotherapy were treated with TMS. Method: A naturalistic study featuring prospective assessment with Montgomery Åsberg Depression Rating Scale (MADRS) and Self-rating Depression Scale (SDS) by non-blind clinicians. Twelve to 14 TMS treatments were given over a 14-16 day period. Results: 21 episodes were brought into remission with respect to both major depressive episode and melancholia. Mean CORE, MADRS and SDS all improved significantly (p<0.0001, paired t-test). Time from treatment to relapse or follow-up was 20 weeks. Conclusion: TMS may have a place in the treatment of major depressive episode and melancholia. There is a danger of under treating and a need to explore the effects of higher numbers of stimuli over longer periods of time (German Journal of Psychiatry, 1999, 2:13-21).

Key Words: Electroconvulsive therapy - Transcranial magnetic stimulation - Melancholia - Major Depressive disorder

 


INTRODUCTION

The potential role of TMS is psychiatry has been reviewed by George et al. (1996) and ourselves (Kirkcaldie et al., 1997). Pascual-Leone et al. (1996) reported a double-blind, placebo controlled trial of TMS in 17 individuals with psychotic depression who had failed to respond to medication. TMS was given for 5 consecutive days. Treatment sessions were 20, 10 s trains of 10 Hz (i.e., 2000 stimulations per day) at 90% of motor threshold (MT), separated by 60 second rest periods, and applied to various sites over the head. Ratings were made with Hamilton Depression Rating Scale (HDRS) and Beck Depression Inventory (BDI). Analysis of variance of HDRS and DBI scores showed a significantly greater reductions with active treatment to the left dorsolateral prefrontal cortex (LDLPFC), but with no other experimental position. There was relapse over the following two weeks.

George et al. (1997) reported a placebo-controlled crossover trial of TMS to the LDLPFC of 12 patients treated 10 times over two weeks. Threshold was set at 80% MT. The stimulus frequency was 20 Hz and 20 two strains (i.e., 800 stimulations per day) were given over 20 minutes. Active treatment was associated with a mean HDRS score decrease of 5.25 points and placebo with an increase of 3.33 points. This treatment effect was statistically significant, but the authors called for studies to determine whether alternative stimulus variables could produce a more robust effect.

TMS has been found to be free of permanent side-effects and safety guidelines have been published (Pascual-Leone et al., 1993; Pascual-Leone & Wassermann, 1996). As TMS has been shown to have an antidepressant effect and is apparently free of permanent side-effects, it has been made available at the Royal Hobart Hospital as a clinical option, to fully informed, consenting patients suffering major depressive episodes, who have failed to respond to pharmacological intervention and for whom ECT is the next treatment option.

 

METHODS

Twenty two consecutive patients who presented with 24 major depressive episodes (DSM-IV) and melancholia [scoring >8 on the CORE system (Parker & Hadzi-Pavlovic, 1996)] were offered TMS. The CORE system is a dimensional construct for rating psychomotor disturbance in depressed patients and is reported to be a marker of the neuropathological process underpinning. All cases had failed to respond to at least one standard antidepressant at maximal doses for at least one month and ECT was considered to be the next treatment option. These patients were advised that this was a novel treatment. They were fully informed and signed consent to treatment with TMS. Particular care was taken to exclude any individual with a past history of significant head injury, epilepsy or intracranial metal appliances.

Acceptees were further assessed using the Montgomery-Åsberg Depression Rating Scale (MADRS; Montgomery & Åsberg, 1979) and the Self-rating Depression Scale (SDS; Zung, 1965). After 12-14 treatments over 14-16 days, diagnostic interviews and assessments were repeated. Diagnosis was determined by the first author, in collaboration with the treating psychiatrist. Ratings were performed by nursing staff and registrars trained in the use of the instruments but not directly involved in providing TMS. Sex, age, previous ECT history, current medication and side effects of treatment were recorded. All were inpatients and remained under the care of their treating psychiatrist, being referred to the first author for TMS.

Treatment was with a MagStim Super Rapid stimulator with a double 70 mm taped coil. Treatment sessions were 25, five s trains of 10 Hz (i.e., 1250 stimualtions per day) at 90-100 % MT, separated by 25 s rest periods, applied to the LDLPFC (five cm anterior an in a parasagittal plane from the point of maximum stimulation of the right thumb muscles). Treatment was given daily five or six days per week. Twelve to 14 treatments were given, according to response.

Wherever possible, medication was ceased during treatment. Where this was not possible, medication was reduced and patients were maintained on that which would be used prophylactically in the event of remission. With one exception, no new medications were commenced in the month before TMS. The exception was episode 7 and a very low dose of an antidepressant was commenced two weeks before TMS was started. In those cases in which no medication was ceased during treatment, a prophylactic medication, either a mood stabiliser or antidepressant, was commenced on the completion of the course.

Remission of the major depressive episode was defined as achievement of a mental state which did not meet the diagnostic criteria according to DSM-IV. Remission of melancholia was defined as the falling of the CORE score to below 8. Improvement on the MADRS was defined as a 50% reduction in the initial assessment figure. Improvement in the SDS, which has a base level score of 20, was defined as a 20% reduction in the initial assessment figure.

Information regarding subsequent course was obtained through the treating psychiatrists. They were asked to categorise the patient as either in remission or relapsed. Where there was no relapse the time to follow-up was recorded, where there was relapse, the time to relapse was recorded.

Statistical analysis was performed using the paired t-test using a standard statistical package.

RESULTS

Twenty two consecutive patients (20 females and two males), with an mean age of 52.5 years (range 29 to 78) suffering 24 major depressive episodes and melancholia were offered and accepted TMS. Details of age, sex, diagnosis, ECT history and medication are presented on Table 1.

A 67 year old female (episode 8) who was not eating and drinking satisfactorily initially, failed to improve in the short term. TMS was ceased and she went on to ECT. She had a protracted course (12 treatments) and made a gradual recovery. A 78 year old female (episode 19) who had failed to make a good recovery with ECT during an episode of major depression six months earlier, failed to respond to 14 TMS sessions and was transferred to a psychogeriatric facility and received ECT. A 70 year old female who was perplexed, eating and drinking poorly and incontinent initially, gradually deteriorated and required ECT.

Remission was achieved in 21 illness episodes, for both the major depressive episode and the melancholia. Diagnoses and scores before and after TMS are presented in Table 2. The group CORE score moved from 19.0 to 6.0 (paired t-test p<0.0001, mean change -13.0, 95% confidence interval (-16.6, -9.4)). In 19 episodes there was improvement in MADRS. The group MADRS score moved from 38.5 to 16.8 (p<0.0001, -21.7, -16.8)). In 13 episodes there was improvement on the SDS. The group SDS scores moved from 60.3 to 45.2 (p<0.0001, -15.0 (-20.6, -9.5)).

There were few side effects. One individual who had difficulty with sleep reported daytime drowsiness following the first treatment. The same night there was a marked improvement in sleep pattern. Daytime drowsiness following treatment persisted, becoming progressively less and ceasing after the sixth session. Some patients admitted, on specific questioning, that stimulation could be uncomfortable, especially when the coil was applied too far forward or laterally, but none withdrew from treatment because of discomfort. Only one individual suffered headache (which persisted for from half to two hours) after treatment. No medication was required. The majority of patients were retained on antidepressant medication, but there were no seizures. Fourteen of the 17 patients who had a history of ECT found TMS to be a more agreeable. The other three could not remember the ECT experience to make the comparison. None preferred ECT to TMS. There were no instances of induced mania or hypomania.

For the 21 remissions, at the time of data collection the period from final TMS to follow-up ranged from 45 to 16 weeks. Eight patients had relapsed. The average time from last TMS treatment to relapse or follow-up was 20 weeks.

 

 

Table 1. Details of 22 patients with 24 episodes of mood disorder who were treated with TMS

MDD; major depressive disorder; BD I; bipolar disorder I; * same patient. # same patient;

A, B, C indicates presence of personality disorder refers to the relevant cluster.

 

Episode Sex Age Axis I Axis II Previous ECT Medication

1

F

75

BD I

 

Y

nil

2

M

42

MDD

 

N

sertraline 200mg

3

F

57

BD I

 

Y

imipramine 100mg, doxepin 100 mg, thioridazine 100 mg

4*

F

34

MDD

 

N

fluoxetine 40 mg, thioridazine 50 mg, trifuloperazine 6 mg

5

F

57

MDD

 

Y

doxepin 100 mg , Various medications for rheumatoid arthritis

6

F

56

BD I

 

Y

Li 1.25 g, valproate 1.25 g

7#

F

36

MDD

C

Y

venlafaxine 75 mg

8

F

67

MDD

 

Y

venlafaxine 150 mg, thioridazine 250 mg, temazepam 20 mg

9

F

49

MDD

C

Y

paroxetine 20 mg, temazepam 20 mg

10

F

74

MDD

 

Y

venlafaxine 75 mg, haloperidol 5 mg

11

F

54

BD I

 

Y

Li 0.5 g, carbamazepine 1.2 g, clonazepam 4 mg

12

F

29

MDD

B

N

nil

13#

F

36

MDD

C

Y

venlafaxine 75 mg

14

F

47

MDD

 

N

nil

15

F

31

MDD

 

Y

nil

16

F

33

MDD

 

N

Li 2 g, Carbamazepine 0.8 g, paroxetine 40 mg

17

F

61

MDD

B

Y

trimipramine 150 mg, lorazepam 3 mg

18

M

73

MDD

 

Y

nefazodone 200mg, temazepam 10 mg

19

F

78

MDD

C

Y

venlafaxine 225 mg

20

F

70

MDD

 

Y

sertraline 200 mg

21

F

77

MDD

 

Y

nil

22*

F

33

MDD

A

N

fluoxetine 40 mg, thioridazine 50 mg, trifluoperazine 6 mg

23

F

29

MDD

 

N

sertraline 100 mg

24

F

63

MDD

 

Y

paroxetine 20 mg

 

 

Table 2. Response of 24 episodes of mood disorder to a course of TMS

MDE; major depressive episode (DSM-IV); Melan*; melancholia (CORE system score 8 or above)

*same patient. # same patient.

Episode

MDE Before

MDE After

Melan* Before

Melan* After

CORE Before

CORE After

MADRS Before

MADRS After

SDS Before

SDS After

1

Y

N

Y

N

16

7

32

16

64

55

2

Y

N

Y

N

21

0

33

0

58

21

3

Y

N

Y

N

19

1

24

4

58

29

4*

Y

N

Y

N

29

2

42

13

61

47

5

Y

N

Y

N

37

6

50

23

71

53

6

Y

N

Y

N

23

2

24

5

47

30

7#

Y

N

Y

N

20

3

39

18

56

46

8

Y

Y

Y

Y

30

30

45

45

67

67

9

Y

N

Y

N

12

6

39

25

68

63

10

Y

N

Y

N

23

7

41

25

50

56

11

Y

N

Y

N

25

1

52

20

55

34

12

Y

N

Y

N

12

0

36

8

64

27

13#

Y

N

Y

N

11

5

36

5

52

47

14

Y

N

Y

N

14

0

38

17

62

20

15

Y

N

Y

N

14

5

41

10

73

67

16

Y

N

Y

N

12

0

38

6

73

38

17

Y

N

Y

N

22

6

55

13

72

57

18

Y

N

Y

N

16

3

34

15

52

49

19

Y

Y

Y

Y

24

24

40

37

57

54

20

Y

Y

Y

Y

18

23

42

53

60

60

21

Y

N

Y

N

16

3

30

8

40

22

22*

Y

N

Y

N

16

4

36

9

65

46

23

Y

N

Y

N

13

4

39

14

66

50

24

Y

N

Y

N

14

3

39

14

55

47

DISCUSSION

These observation are from our clinical practice and they were collected prospectively. This was not a controlled study, however, each individual could be seen as their own control, as each had failed to respond to medication and was about to be offered ECT. Only 3 of 24 episodes went on to ECT.

There is no doubt that TMS may be strong placebo effect. This is a modern, expensive and dramatic (with an apparently magical ability to cause the hand to twitch) technique. Further, our staff are enthusiastic and the doctor spends 15 to 30 minutes with the patient each day and is attentive to both the "new" machine and the patient. We minimise, to the best of our ability, the communication of our enthusiasm. The staff who performed the testing were not directly involved in the treatment.

Raskin et al. (1970) and Fairchild et al. (1986) found that placebo response is much less likely in endogenous than in non-endogenous depression, and Zimmerman & Spitzer (1989) stated that "melancholics are particularly unresponsive to placebo". All our patients had major depressive episode (DSM-IV) and melancholia (CORE system criteria), thus, it is unlikely that these results simply represent a placebo response. Downing & Rickles (1978) and Fairchild et al. (1986) reported that individuals who score highly on self rating scales are unlikely to respond to placebo. The threshold for clinical depression in the SDS is a raw score of 50 or greater (Zung, 1965). In the present study of 24 episodes, 22 were characterised by SDS scores of 50 or greater, also making difficult the total discounting of these results on the basis of placebo effect.

A question has been raised about our reason for selecting a 50% reduction in the MADRS but only a 20% reduction in the SDS as our threshold for "improvement". This was because the way we scored the SDS, which has 4 possible responses to 20 questions. We scored the response categories as one to four, which meant that a totally asymptomatic individual would score 20. The mean SDS score of our patients at intake was 60.3 and accordingly a reduction in baseline score by 20% was designated. We could find no published guidelines on the definition of improvement in the SDS. A 50% reduction in MADRS is widely accepted as indicating improvement.

These results suggest a more prolonged remission than that found by Pascual-Leone et al. (1996) and a stronger antidepressant effect than that found by George et al.(1997). This may be explicable as placebo effect. It may also be that the placebo effect was only partially responsible and the facts that more of our patients took concurrent medication and that we gave a greater total number of stimuli were at least partially responsible for our beneficial results.

Pascual-Leone et al. (1996) reported that all patients received nimodipine and "some" continued to receive antidepressant medication. George et al. (1997) reported that three quarters of their patients were without antidepressants. In contrast, more that three quarters of our patients continued to receive antidepressants. Conca et al, (1996) conducted a controlled study of depressed patients in which one group received antidepressant medication plus TMS and the other received antidepressant medication alone. Their stimulus frequency, number and site was different, nevertheless, they showed a statistically significantly greater reduction in symptoms over the first two weeks in the double treatment group and recommended TMS as an "add-on therapy". The results reported here may support that view.

Pascual-Leone et al. (1996) reported that the beneficial effects of TMS lasted for only two weeks. In contrast, in our work the remission period was 20 weeks. However, length of remission must reflect, to a large extent, the prophylaxis employed. Ours was a group of people with serious mental illness and on completion of the TMS course, all were on pharmacological prophylaxis. We lengthened the course of TMS treatment in the hope of lengthening remission. Whereas Pascual-Leone et al. (1996) treated people for one week, our treatment extended into the third week. This decision was based on the clinical observation that antidepressant medication and ECT take at least two weeks to come to full effect.

Pascual-Leone et al. (1996) reported a greater improvement on the HDRS than did George et al. (1997). This may have been because of the intensity and number of stimuli given per day. Pascual-Leone et al used 90% MT and administered 2000 stimuli, while George et al. (1997) used 80% and administered only 800 stimuli per day. Our strong response maybe because we used 90-100% and administered 1250 stimuli per day, which was then sustained for in excess of two weeks. [The total stimulations provided - Pascual-Leone et al. (1996) gave 5000, George et al. (1997) gave 8000 and we gave an average of 16250.]

While the open study is always in danger of finding false positives, the tightly structured blind trial is in danger of under treating and missing leads. Through working with TMS in depression in a naturalistic manner, we have come to believe that TMS can and perhaps should be used in combination with antidepressants and that better results are achieved with greater numbers of stimuli. We have used only 10 Hz and do not have a preference for any particular frequency. We believe that it is can be necessary to continue treatment for up to four weeks. The findings in this report suggest exploration of the effects of higher numbers of stimuli over longer periods of time.

REFERENCES

Conca A, Koppi S, Konig P, Swoboda E, Krecke N. Transcranial magnetic stimulation: a novel antidepressive? Neuropsychobiology 1996;34:204-7.

Downing R, Rickles K. Predictors of response to amitriptyline and placebo in three outpatient treatment settings. J Nerv Ment Dis 1978;156:109-29.

Fairchild C, Rush A, Vasavada N, Giles DE, Khatami M. Which depressions respond to placebo? Psychiatry Research 1986;18:217-26.

George M S, Wassermann E, Kimbrell T, Little J, Williams W, Danielson A, Greenberg B, Hallett M, Post R. Mood improvement following daily left prefrontal repetitive transcranial magnetic stimulation in patients with depression: a placebo-controlled crossover trial. Am J Psychiatry 1997;154:1752-6.

George M S, Wassermann E, Post R. Transcranial magnetic stimulation: a neuropsychiatric tool for the 21st century. J Neuropsychiatry Clin Neurosci 1996 8, 373-382.

Kirkcaldie M, Pridmore S, Reid P. Bridging the skull: electroconvulsive therapy (ECT) and repetitive transcranial magnetic stimulation (rTMS) in psychiatry. Convulsive Ther 1997;13:83-91.

Montgomery S, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry 1979;134:1165-71.

Parker G, Hasdi-Pavlovic D. Melancholia: a disorder of movement and mood. New York: Cambridge University Press, 1996.

Pascual-Leone A, Houser C, Reese K, Shotland L, Grafman J, Sato S, Valls-Sole J, Brasil-Neto J, Wassermann E, Cohen L, Hallett M.. Safety of rapid-rate transcranial magnetic stimulation in normal volunteers. Electroencephalogr Clin Neurophysiol 1993;89:120-30.

Pascual-Leone A, Rubio B, Pallardo F, Catala MD. Beneficial effect of rapid-rate transcranial magnetic stimulation of the left dorsolateral prefrontal cortex in drug-resistant depression. Lancet 1996;348:233-38.

Pascual-Leone A, Wassermann, E. Repetitive transcranial magnetic stimulation : applications and safety considerations. Advances in Occupational Medicine and Rehabilitation. Aggiornamenti in Medicina Occupazinale E Riabilitazione. Vol 2, No. 2, May-August 1996, pp 105-116. Pavia, Fondazione Salvatore Maugeri Edizioni.

Raskin S, Schulterbrandt J, Reatig N, Differential response to chlorpromazine, imipramine and placebo: a study of subgroups of hospitalised depressed patients. Arch Gen Psychiatry 1970;23:164-73.

Zimmerman M, Spitzer R. Melancholia: from DSM-III to DSM-III-R. Am J Psychiatry 1989;146:20-8.

Zung W. A self-rating depression scale. Arch Gen Psychiatry 1965;12:63-70.