Not sure if this is Reactome's bug or a feature.  A human gene matches to both human and mouse pathways.

library(reactome.db)
library(AnnotationDbi)
pwthws <- as.list(reactomeEXTID2PATHID['55697'])[[1]]
mapIds(reactome.db, pwthws,"PATHNAME","PATHID")

                                                         1430728 
                                      "Homo sapiens: Metabolism" 
                                                         5991024 
                                      "Mus musculus: Metabolism" 
                                                          556833 
           "Homo sapiens: Metabolism of lipids and lipoproteins" 
                                                         5991065 
           "Mus musculus: Metabolism of lipids and lipoproteins" 
                                                         1483257 
                         "Homo sapiens: Phospholipid metabolism" 
                                                         5991243 
                         "Mus musculus: Phospholipid metabolism" 
                                                         1483255 
                                   "Homo sapiens: PI Metabolism" 
                                                         5992197 
                                   "Mus musculus: PI Metabolism" 
                                                         1660516 
"Homo sapiens: Synthesis of PIPs at the early endosome membrane" 
                                                         5992226 
"Mus musculus: Synthesis of PIPs at the early endosome membrane" 
                                                         1660514 
         "Homo sapiens: Synthesis of PIPs at the Golgi membrane" 
                                                         5992224 
         "Mus musculus: Synthesis of PIPs at the Golgi membrane" 
                                                         1660517 
 "Homo sapiens: Synthesis of PIPs at the late endosome membrane" 
                                                         5992222 
 "Mus musculus: Synthesis of PIPs at the late endosome membrane" 

Indeed, on Reactome's site http://www.reactome.org/content/detail/R-HSA-1806219 there are both human and mouse pathways for this gene.  I know this question probably belongs to the Reactome's mailing list (not sure if it exists) rather Bioconductor's, but I thought I'd share this observation anyway.

As you mention yourself in Reactome itself it is mentioned in both a human and a mouse pathway. The package is meant to reflect the Reactome database so it is at least not a bug in the package.

For specific questions about Reactome I would like to refer to help@reactome.org.