Language improvements after TMS plus modified CILT: Pilot, open-protocol study with two, chronic nonfluent aphasia cases

Abstract

Purpose: The purpose of this study was to investigate: 1) the feasibilty of administering a modified CILT (mCILT) treatment session immediately after TMS; and 2) if this combined therapy could improve naming and elicited propositional speech in chronic, nonfluent aphasia.

Methods: Two chronic stroke patients with nonfluent aphasia (mild-moderate and severe) each received twenty minutes of rTMS to suppress the right pars triangularis, followed immediately by three hours of mCILT (5 days/week, 2 weeks). (Each patient had received TMS alone, 2-6 years prior.) Language evaluations were performed pre- TMS+mCILT, and post- at 1-2 months, and 6 or 16 months.

Results: Both patients showed significant improvements in naming pictures, and elicited propositional speech at 1-2 months post- TMS+mCILT. The improved naming was still present at 6 months post- TMS+mCILT for P2; but not at 16 months post- TMS+mCILT for P1.

Conclusions: It is feasible to administer mCILT for three hours immediately after a TMS session. It is unknown if the significant improvements in naming pictures, and elicited propositional speech were associated with the second series of TMS, or this first series of mCILT, or a combination of both. A larger, sham controlled clinical trial is warranted.

Keywords: TMS; aphasia; constraint-induced language therapy; speech therapy; stroke rehabilitation.

Figures

Fig. 1

A) and B): T1-weighted structural MRI scans, with axial, and L and R reconstructed lateral views for each patient. White triangle on the R lateral view marks the area of cortex (R PTr) where 1 Hz rTMS was used to suppress that cortical area, in each case. C) Application of rTMS with the figure 8-shaped, hand-held rTMS coil, using the neuronavigation system, Brainsight (Rogue Industries, Montreal).

Fig. 2

mCILT treatment session setup. An opaque screen was placed on the table between the clinician and the patient. There was eye contact only above the screen. There was a window to pass the card back and forth (arrow). The patient was expected to name the picture on the card. No gestures, writing or sound effects were permitted.

Fig. 3

Performance on language outcome measures pre- and post- TMS+mCILT (black bars) for P1 (mild-moderate nonfluent aphasia) on A) BNT, and B) BDAE subtest, Tools/Implements. Previous scores are also shown when intervention was TMS alone (gray bars). *=+2 SD above Baseline for each treatment intervention series (TMS alone, or TMS+mCILT). See also Table 3.

Fig. 4

Elements of propositional speech (cookie theft) pre-and post -TMS+mCILT (black bars) for P1 (mild-moderate nonfluent aphasia) for A) total number of narrative words, and B) number of different nouns. Previous scores are also shown when intervention was TMS alone (gray bars). *=+2 SD above Baseline.

Fig. 5

P1 (mild-moderate nonfluent aphasia), total percent of pictures named correctly (without cueing) during each mCILT treatment day; “Always-named” at entry; “Sometimes-named”; and “never-named”. Note, P1 continued to correctly name pictures never named at entry, even for the lower frequency-of-occurrence words. By days 9 and 10, P1 had successfully named his maximum set of 360 color pictures at least once without cueing. mCILT therapy then focused on sentence production.

Fig. 6

P1 (mild-moderate nonfluent aphasia), A) performance on naming by level of difficulty (frequency of occurrence) by decile, during mCILT (triangles). Performance on naming by level of difficulty (frequency of occurrence) by decile, at entry (circles). B) The black diamonds show the difference between percent of color pictures named correctly during mCILT intervention by decile, and the percent of color pictures named correctly at entry by decile (slope = 3.72, r = 0.721, p = 0.019). Thus, the differences in naming performance during mCILT (by decile), and at entry (by decile), increased as the level of difficulty increased. This suggests a beneficial effect of the TMS+mCILT intervention in naming performance across the deciles, despite the increasing level of difficulty.

Fig. 7

Performance on language outcome measures pre- and post- TMS+mCILT (black bars) for P2 (severe nonfluent aphasia) on BDAE subtests: A) Action naming; B) Tools/Implements; and C) Single Word Repetition. Previous scores are also shown when intervention was TMS alone (gray bars). *=+2 SD above Baseline. See also Table 4.

Fig. 8

Elements of propositional speech (cookie theft) pre-and post -TMS+mCILT (black bars) for P2 (severe nonfluent aphasia) for A) total number of narrative words, and B) number of different nouns. *=+2 SD above Baseline. Her phrase length, pre-and post -TMS alone, was only 0-1 word (not plotted here).

Fig. 9

P2 (severe nonfluent aphasia), total percent of pictures that were named correctly (without cueing) during each mCILT treatment day; “always-named” at entry; “sometimes-named”; and “never-named”. Data were incomplete for day 1. Note, P2 continued to correctly name pictures never named at entry, even for the lowest frequency-of-occurrence words. No color pictures in the “always-named” category from entry testing were available for presentation during days 5–10.

Fig. 10

P2 (severe nonfluent aphasia), A) performance on naming by level of difficulty (frequency of occurrence) by decile, during mCILT (triangles). Performance on naming by level of difficulty (frequency of occurrence) by decile, at entry (circles). B) The black diamonds show the difference between percent of color pictures named correctly during intervention by decile, and the percent of color pictures named correctly at entry by decile (slope = 4.61, r = 0.71, p = 0.034). Thus, the differences in naming performance during mCILT (by decile), and at entry (by decile), increased as the level of difficulty increased. This suggests a beneficial effect of the TMS+mCILT intervention in naming performance across the deciles, despite the increasing level of difficulty.