Your approach is entirely the wrong way around. Putting meta-data (e.g. pseudo-indices, subject IDs, parameter values, etc.) into variable names is a sign that you are doing something wrong.
Accessing variable names dynamically is one way that beginners force themselves into writing slow, complex, obfuscated, buggy code that is hard to debug. Read this to know why:
You would be much better off loading the data into one array (e.g. a cell array or a structure) and then using a simple loop. For example:
P = 'absolute/relative path to where the files are saved';
S = dir(fullfile(P,'*.txt'));
for k = 1:numel(S)
S(k).data = dlmread(fullfile(P,S(k).name));
end
Now all of the file data will be contained in the non-scalar structure S. You can trivially loop over the elements of S and process each file's data. Or you can get the filenames and filter for whichever ones you want, which makes it easy to combine the matrices exactly as you asked in your question, e.g.:
F = {S.name};
[~,F] = cellfun(@fileparts,F,'uni',0); % optional
X = ~cellfun(@isempty,regexp(F,'ccc$')); % or use ENDSWITH
ccc = cat(1,S(X).data); % if the matrices have compatible sizes
You need to understand how to use comma-separated lists:
