The scan is flawed. The operator set the focus to be too sharp for the specified number of scan lines. If at all possible, have the scan redone rather than trying to fix it after the fact -- the results will be much better.
If re-doing the scan is impossible, there are several ways to handle this type of problem in post:
The easiest solution is to make a couple of copies of the original. Displace one by 1/3rd the line separation in one direction, displace the other by 1/3rd the line separation in the opposite direction and set the opacities of the layers for equal contributions to the overall image, ie, 1, 1/2, and 1/3 as you go up the layer stack.
Another approach is to run the image through scanning software which has the ability to estimate the line spacing and then minimize it. This is commonly needed when the original art was produced by a half-tone / line screen process. I know that some of the Silverfast products include this.
Another approach is to use "Neat Image" noise reduction software on it. Even though Neat Image is not advertised as being able to reduce this sort of artifact, if trained correctly, it does quite a good job -- much better than Topaz DeNoise, or the NR built into even the newest versions of ACR and LR.
The final approach is only if you are mathematically inclined: Run the image through a Fourier Transform plugin, remove the huge peaks that correspond to the line spacing and its harmonics, and run the result through an inverse Fourier filter.
HTH,
Tom M
PS - The clone or patch tools are not appropriate for problems like this that pervade the entire image.
