If you think geometrically
(using the strategy of the answer I referenced in the comment:
Lowest kinetic energy of particle for which reaction is possible (invariant mass)),
one need not "transform" to the COM-frame and back.
Draw an energy-momentum diagram.
Your problem in this threshold case is akin to a problem involving the doppler effect [Two timelike vectors with a common tail, whose tips are joined by a lightlike vector.].
When you say
reference frame in which both proton and meson have zero velocity after the collision.
that means that the 4-momenta of the final products are parallel (collinear).
[That is, what is generally not-parallel is, in the threshold case, parallel. What would have been a quadrilateral with three timelike vectors (two of which have equal mangitude) and a lightlike vector, you instead have a triangle with two timelike vectors (one with the invariant-mass $m_p+m_{\pi}$) and a lightlike vector.]
The Doppler factor (Bondi k-factor) $k=\exp\theta$
can be expressed in terms of the proton mass and the invariant mass, $(m_p+m_{\pi})$. The $\theta$ is the rapidity [Minkowski-angle] so that velocity $v=\tanh\theta$ and time-dilation factor $\gamma=\cosh\theta$.
With $k$ known, the rapidity-angle $\theta$ can then be used to solve for an unknown leg of Minkowski right triangle using hyperbolic-trig functions of $\theta$.
It's like solving for an unknown component in a free-body diagram. Try it!
[Using $m_p=938$ and $m_\pi=135$, I get 144.715.]
Looking at your answer, it's close... hint: think Doppler effect.