This can be written as
$$T' = \begin{bmatrix} -\lambda_h & \lambda_h \\ \lambda_c & -\lambda_c \end{bmatrix} T$$
This is a first order linear system of two scalar ODEs with constant coefficients. It can be solved by the matrix exponential technique: $T(t)=e^{At} T(0)$ where $A$ is the coefficient matrix in the system. The matrix exponential can be computed by the eigendecomposition: if $A=P D P^{-1}$ for a diagonal $D$ then $e^{At}=P e^{Dt} P^{-1}$. Here $(e^{Dt})_{ii}=e^{D_{ii}t}$ and $(e^{Dt})_{ij}=0$ for $i \neq j$.
Doing this eigendecomposition symbolically is a bit ugly/tedious but is possible in the $2 \times 2$ case (especially here, where one of the eigenvalues is always zero).
Following the operator formalism, if you apply $D+\lambda_h$ to the second equation and substitute in the first equation, you get
$$(D+\lambda_h)(D+\lambda_c) T_c - \lambda_c (D+\lambda_h) T_h = (D+\lambda_h)(D+\lambda_c) T_c - \lambda_c \lambda_h T_c = 0.$$
This is now a single scalar ODE for $T_c$, which you can try to solve and then back-substitute. However, you'll find that the roots of the characteristic polynomial of this equation are just the eigenvalues of $A$, so this is really just carrying out the algebra that the eigendecomposition does "for you" by hand.