If you don't mind using matrices, you can combine:
$$a_{k+1} = 7~a_{k} - 10~a_{k-1}$$
with
$$a_{k} = a_{k}$$
To get:
$$\begin{align}
a_{k+1} &= 7~a_k - 10~a_{k-1} \\
a_{k} &= 1~a_k + 0~a_{k-1}
\end{align}$$
$$\begin{align}
\begin{bmatrix} a_{k+1} \\ a_{k} \end{bmatrix}
%
&= \begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix} \begin{bmatrix} a_{k} \\ a_{k-1}\end{bmatrix}
%
\\ &=
\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}
\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}
\begin{bmatrix} a_{k-1} \\ a_{k-2}\end{bmatrix}
%
\\ &=
\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}
\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}
\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}
\begin{bmatrix} a_{k-2} \\ a_{k-3}\end{bmatrix}
%
\\ \vdots
%
\\ &=
\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}^k
\begin{bmatrix} a_{1} \\ a_{0}\end{bmatrix}
%
\end{align}$$
The eigenvalues of $\begin{bmatrix} 7 & -10 \\ 1 & 0 \end{bmatrix}$ are $5$ and $2$, so for some matrix $P$:
$$\begin{bmatrix} a_{k+1} \\ a_{k} \end{bmatrix}
%
= P~\begin{bmatrix} 5 & 0 \\ 0 & 2 \end{bmatrix}^k~P^{-1} \begin{bmatrix} a_{1} \\ a_{0}\end{bmatrix}$$
So for some numbers $X$ and $Y$:
$$a_{k+1} = X~5^k + Y~2^k$$
You can use the initial conditions $a_0 = a_1 = 2$ to find $X$ and $Y$:
$$2 = X~5^1 + Y~2^1$$
$$2 = X~5^0 + Y~2^0$$
To get:
$x = \frac{-2}{3}$ and $y = \frac{8}{3}$
to get the total solution:
$$a_{k+1} = -\frac{2}{3}~5^k + \frac{8}{3}~2^k$$
If you aren't comfortable with matrices, and just want "how do I do this" rather than "why does this work", then I recommend Claude's answer.