There are almost no epidemiologic studies about HPV 18 variants carried out in Spain and even though national HPV prevalence is low, it cannot be forgotten that this genotype together with genotype 16 cause 70% of cervical cancer cases.
Many authors confirm that distribution of HPV variants is related to geographic or race distribution [6, 17] and therefore, Spain should expect predominance of European variant, followed by African and Asian-amerindian variants. Our results show concordance with this stating: 25 European variants (56.8%), 10 African (22.7%) and 5 Asian-amerindian variants (11.4%).
HPV 18 has been associated with both recurrent lesions with very bad clinical prognosis  and benign lesions . This fact may reflect the oncogenic potential difference among variants. Hecht et al  identified a HPV 18 variant with lower oncogenic potential due to its absence in cervical cancer but presence in 40% of intraepithelial lesions. Villa et al  suggested that non-European HPV 18 variants persisted more frequently and were more associated with pre-invasive lesions. Since then, most studies confirm that different variants of the same genotype differ in their pathogenic characteristics and therefore, nucleotide substitutions may play an important role. Our study results show concordance with these statements as African variants and variants where most of specific-african substitutions were detected were the only type of variants detected in H-SIL.
Most nucleotide changes reported in our study have been previously described and some of them are of particular importance. In URR, the mutation A41G is located in the Sp-1 binding site and isolates with this nucleotide variation have shown to have an increased transcriptional activity . Variations in positions 41 and 104 modulate Sp1 and YY1 activities and are associated to a higher activity of the E6/E7 promoter. Patients with T104C substitution are less likely to present tumour recurrence . Other nucleotide changes like T7651C, A7658C and C7726T also lay within transcription factor binding sites.
Substitutions C287G, C6842G and T7592C were found in all our isolates. Variation C6842G has been previously reported as error in the original sequence , and H. Arias–Pulido et al. sequenced the original reference HPV 18 plasmid (provided by E-M de Villiers, Deutsches Krebsforschungzentrum, Germany) and observed the substitution T7592C , so it is considered as a sequencing error in the original HPV 18 reference sequence report.
Furthermore, ten “new” nucleotide variations have been detected and two of them were non-synonymous and lead to amino acid changes (T318C and C3617T, Tir/His and Ser/Leu, respectively).
Knowledge on HPV variants and their nucleotide variability is essential for three main reasons: i) nucleotide variations may interfere with the viral oncogenic potential, ii) host cellular immune response can be different when there are substitutions in the amino acids on the viral capsid which may be relevant for the vaccination, iii) HPV infections with a variant may not give immunological protection against a subsequent infection with other variant of the same genotype.
Nowadays, HPV variants recombination has already been described and it is more often found since coinfection with more than one HPV type prevalence is not a unusual finding .
In our study we detected 2 recombinant variants (4.5%) which might have been missed or wrong classified if only amplifying one genomic region. Furthermore, non-recombinant samples as LSCM or LSM3 showed specific-african substitutions in some regions (for example E6) whereas they would be classified as european variants when only analyzing nucleotide variation in URR. Therefore, characterizing more than one genomic region may be essential in order to detect recombination and classify HPV variants properly.
We amplified URR and E6, E7 and L1genes from at least 93% of samples. However, when characterizing E4 region, we were only able to amplify 35 samples. E4 gene is generally disrupted during DNA integration into the host genome and this disruption may explain the inability to amplify E4 gene in some of our samples.
In conclusion, data and knowledge on geographic HPV intratypic variants distribution might help to establish a data base about the diversity and pathogenicity of different HPV variants, which may help to design and optimize diagnostics protocols in order to reduce the disease.