Gynecological Graft-Versus-Host Disease Mimics Cervical Intraepithelial Neoplasia in Pap Smear– Single Center Prospective Study

The true incidence of gynecological graft-versus-host disease (gynGvHD) is difficult to determine because symptoms may result from hypoestrogenism, and predefined histopathological diagnostic criteria do not exist. We previously reported that gynGvHD was diagnosed in 29% female allogeneic hematopoietic cell transplantation (HCT) recipients and high rates (60%) of abnormal pap smear results were observed in that population (Klasa L. et al. Annals of Hematology, 2020). To determine the incidence of gynGvHD and to verify the frequency of false positive pap smear results, we conducted a prospective analysis of patients who underwent HCT at our institution.

STUDY POPULATION

The study population included 111 female HCT recipients transplanted between 2018-2023 and who survived to the time of the first gynGvHD evaluation. From further analyses we excluded 3 patients suffering from human papillomavirus (HPV) related disorders before HCT, 18 monitored by local physicians, 10 without follow up, 2 virgins, and 4 patients shortly after 2^nd HCT. The final cohort consisted of 74 (67%) women supervised by one dedicated, trained gynecologist.

The median age at HCT was 44 (16-69), and the main indication for HCT was acute myeloid leukemia/myelodysplastic syndrome (46%) and acute lymphoblastic leukemia (24%). Matched related (MRD), unrelated (MUD) or haploidentical donors were used in 22 (30%), 43 (58%) and 9 (12%) cases respectively, and most of the patients received myeloablative conditioning regimen (76%). The main immunosuppression consisted of cyclosporine with methotrexate (MRD), with addition of anti-thymocyte globulin (MUD), mismatched transplants received post-transplant cyclophosphamide based GvHD prophylaxis. After discharge from the transplant unit, unless contraindicated, all patients received vaginal estrogen therapy twice weekly up to 18 months after HCT, and hormone replacement therapy if they were younger than 45. GynGvHD diagnosis was based on morphological changes detected in physical examination (lichen planus like, vaginal scarring, stenosis, labial agglutination, erosions, ulcers). Pap smear was collected during the first visit and repeated at 6-12 months intervals depending on the result. Patients with abnormal results were qualified for diagnostic colposcopy with histopathological examination of obtained tissue.

RESULTS

With the median time of 36 (8-77) months of follow up, gynGvHD was diagnosed in 16 (21%) patients, with score 1, 2 and 3 in 4 (25%), 3 (18%) and 9 (56%) cases respectively. Clinical improvement was observed in 44% cases after local steroid therapy, but only in 1 patient with score 3. Other manifestations of chronic GvHD were documented in 40 (54%) cases (mild 40%, moderate 38%, severe 22%) There was a significant association between oral (93%), ocular (50%), liver (43%) involvement and gynGvHD (p<0.05).

An abnormal Pap smear was found in 29 patients (39%), of which 17 (59%) had high-grade intraepithelial lesions (HSIL). In colposcopy-guided histopathological examination, HPV related neoplasia was confirmed in only 2 cases (3% of the study cohort), with no signs of gynGvHD. Abnormal cytology was observed in 11 out of 16 (68%, p<0.05) women diagnosed with gynGvHD. This association was not observed in patients without any gynGvHD signs (18 out of 58 cases, 31%, p=NS). The median time to the first smear collection did not differ between groups: 5 (2-35 months) vs 6 (2-21months), and no correlation was observed between the time of collection, age at HCT, type of conditioning and abnormal result (p=NS). Histopathological features of mucosal or skin GvHD were suggested in 5 cases of cervical biopsy, including 3 patients without any clinical signs of gynGvHD.

CONCLUSIONS

In female HCT recipients receiving topical estrogens to alleviate symptoms resulting from mucosal atrophy, gynGvHD occurred in 21% of cases. Early initiation of local treatment should be applied to ensure proper tissue healing, unless contraindicated. Compared to the general population where the false-positive rate in cytology reaches 10%, pap smear seems to perform poorly in HCT population, with false positive results detected in 40% of the cases. Additional tests including colposcopy with histopathology and HPV molecular testing should be performed, especially in female patients with gynGvHD.