As already stated in the comments, the best way to do this is probably to do the manipulations before plotting. Here's a way how to do it:

import pandas as pd
from matplotlib import pyplot as plt
import numpy as np

countries = [
        'Albania',
        'Brazil',
        'Denmark',
        'France',
        'Mexico',
        'Nigeria',
        'Spain',
        'Germany',
        'Finland',
    ]

#the full dataframe
df = pd.DataFrame(
    data = {'country': countries, 'value' :np.random.rand(len(countries))},
    ).sort_values('value', ascending = False)

#the top 5
df2 = df[:5].copy()

#others
new_row = pd.DataFrame(data = {
    'country' : ['others'],
    'value' : [df['value'][5:].sum()]
})

#combining top 5 with others
df2 = pd.concat([df2, new_row])

#plotting -- for comparison left all countries and right 
#the others combined
fig, axes = plt.subplots(nrows = 1, ncols = 2, figsize = (9,4))
df.plot(kind = 'pie', y = 'value', labels = df['country'], ax = axes[0])
df2.plot(kind = 'pie', y = 'value', labels = df2['country'], ax = axes[1])
axes[0].set_title('all countries')
axes[1].set_title('top 5')
plt.show()

The result looks like this.

Hope this helps.