Since $n$ is a real number, for convenience denote $m = 2 n$.
If $m$ is a positive, even integer, instead repeatedly applying the reduction formula in the problem statement together gives
$$\int \tan^m x \,dx = \sum_{k = 1}^{\frac m2} \frac{(-1)^{\frac m2 + k}}{2 k - 1} \tan^{2 k - 1} x - x + C ,$$
and if $m$ is a positive, odd integer, applying the reduction formula and using the special case $\int \tan x \,dx = \log |\sec x| + C$ gives
$$\int \tan^m x \,dx = \sum_{k = 1}^{\frac{m - 1}2} \frac{(-1)^{\frac{m - 1}2 + k}}{2 k} \tan^{2 k} x + (-1)^{\frac{m - 1}2} \log |\sec x| + C .$$
More generally, if $n$ is a positive rational number, $\frac pq$, then the substitution $v = (\tan x)^{\frac 1q}$ rationalizes the integral, giving
$$q \int \frac{v^{p + q - 1} \,dx}{1 + v^{2 q}},$$
so in antiderivative has a closed form in terms of elementary functions, though in practice explicit formulae are complicated even for small $q > 1$; see Quanto's answer to another related question for a compact summation formula for the case $p = 1$.
If $m$ is a negative integer, say, $-M$, then we are evaluating $\int \cot^M x \,dx$, for which another reduction formula is available. More generally, if $m$ is a negative rational number, $-\frac PQ$, then substituting $w = \cot^{\frac1Q} x$ again rationalizes the integral.
More generally, if $m$ is not rational, the antiderivative cannot be expressed in closed form in terms of elementary functions alone, but it can be expressed in terms of the ordinary hypergeometric function, ${}_2F_1$, or the Lerch transcendent, $\Phi$. Substituting $u = \tan x$ gives
$$\int \tan^{m} x \,dx = \int \frac{u^{m} \,du}{1 + u^2} .$$
For $|u| < 1$, expanding into a series gives
$$\int \sum_{k = 0}^\infty (-1)^k u^{m + 2 k} \,du = \sum_{k = 0}^\infty \int u^{m + 2 k} \,du = \sum_{k = 0}^\infty \frac{u^{m + 2 k + 1}}{m + 2 k + 1} + C .$$
(Provided that $m$ is not a negative integer,) rewriting this series using special functions and back-substituting gives that
$$\boxed{\int \tan^{2 n} x \,dx = \frac{\tan^{m + 1} x}{m + 1} {}_2F_1\left(1, \frac12 + \frac m2; \frac32 + \frac m2; -\tan^2 x\right) = \frac{\tan^{m + 1} x}2 \Phi\left(-\tan^2 x, 1, \frac12 + \frac m2\right)} .$$ Cf. Robert Israel's answer to a similar question.
(Even though the series in $u$ converges only for $|u| < 1$, i.e., $|\tan x| < 1$, these expressions are antiderivatives for $\tan^m x$ wherever they are defined.)