LAUSR

OBJECTIVE Evaluate the accuracy of tissue sampling techniques for the diagnosis of adenomyosis. DATA SOURCE Systematic Review via MEDLINE and the Cochrane Library searches. METHODS OF STUDY SELECTION Review performed utilizing the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, utilizing MeSH terms and keywords including "Adenomyosis/diagnosis" or "Adenomyosis/pathology" or "Myometrium/pathology" and "Biopsy" or "Hysteroscopy" or "Laparoscopy". Articles initially screened by title and abstract to include pertinent studies with reference lists cross-referenced to find additional studies. Articles related to the diagnosis of uterine malignancy or studies in which tissue sampling was obtained via excisional surgical procedures were excluded from review. TABULATION Fourteen studies were identified describing tissue sampling techniques for the purpose of diagnosing adenomyosis, with a total of 1909 patients, from 12 different countries, involving 6 different continents. Tissue sampling techniques were categorized based on (1) biopsy approach as either intra-uterine and extra-uterine, and (2) techniques that were validated or not validated with a confirmatory hysterectomy pathology. INTEGRATION AND RESULTS Overall, there was significant heterogeneity in the tissue sampling techniques including intra-uterine sampling obtained via hysteroscopic biopsy or resection and extra-uterine tissue sampling obtained via needle biopsy by a percutaneous, transvaginal, laparoscopic or ex-vivo approach. Sensitivity of these techniques varied greatly based on technique, tissue sampling location and the number of biopsies obtained, and was as low as 22.2% via an ultrasound guided transvaginal biopsy of suspicious uterine lesions with 4 biopsies per patient and was as high as 97.8% via a laparoscopic guided myometrial biopsy of suspicious uterine lesions with 10 biopsies per patient. Specificity for the identified tissue sampling techniques were more homogeneous ranging from 78.5% - 100.0% for all methods identified. The positive predictive value and negative predictive value ranges were 75.9% - 100.0% and 46.4% - 80.0% respectively amongst all tissue sampling techniques identified with confirmatory hysterectomy pathology. CONCLUSION Due to the heterogeneity of the tissue sampling techniques, diverse patient populations, and significant conflicting recommendations, no conclusive recommendation on the optimal tissue sampling technique can be made. However, it is of the authors opinion that it would be reasonable to limit uterine tissue sampling for confirmatory diagnosis of adenomyosis in those patients with a suspicion of adenomyosis based on both symptom profile and pelvic ultrasound, where a planned diagnostic laparoscopy for either infertility or pelvic pain has already been contemplated and scheduled, and where the confirmatory results may be of clinical benefit in discussing the prognosis of post-operative recurrent symptoms and guide any future treatment recommendations.