Expanded study on the risk of lymphovascular space invasion and lymph node metastasis of endocervical adenocarcinoma using Pattern Classification: a single-centre analysis of 213 cases

Abstract

A new three-tiered Pattern Classification system for usual-type endocervical adenocarcinomas (U-EACs) recommends using tumour invasive patterns rather than depth of invasion (DOI) and horizontal spread to categorise tumours. Tumours categorised by Pattern Classification are associated with lymph node (LN) metastasis and adverse outcomes. The aim of this study is to further explore the potential of Pattern Classification in surgical pathology practice. A total of 213 consecutive cases [201 U-EACs and 12 gastric-type adenocarcinomas (GACs)] diagnosed between 2006 and 2017 was retrospectively analysed. Clinicopathological data included age at diagnosis, DOI measurement, the status of lymphovascular space invasion (LVSI) and LN metastasis, and the number of LVSI foci, dissected and metastatic LNs. Immunostaining for CD34 and D2-40 was performed to identify LVSI in 14 challenging cases. Overall, mean age at diagnosis was 51 years (range 23-75). LVSI and LN metastasis occurred in 128 (60.1%, 128/213) and 42 (20.5%, 42/205) cases, respectively. Also, 28 (13.1%), 21 (9.9%), and 164 (77.0%) patients had pattern A, B, and C tumours, respectively. Patients with pattern C tumours had the oldest age at diagnosis (p=0.007), the highest incidence of LVSI and LN metastasis, and the highest DOI (p<0.001). Due to a highly heterogeneous growth pattern, pattern C U-EACs were stratified into four subgroups: C1, C2 and C3 corresponded to solid, extensive linear destructive, and band-like lymphocytic infiltrate growth patterns, respectively, and C4 included diffuse destructive, confluent, micropapillary and mixed growth pattern. C2 and C3 subgroup tumours had lower incidence of LVSI (20% and 40%, respectively) than the other two subgroups (p<0.001). None of the patients with C2 and C3 subgroup had LN metastasis and ≥3 LVSI foci. All GACs belonged to pattern C and had deeper stromal invasion (p=0.008), higher incidence of LN metastasis (p=0.001), and larger quantity of LVSI foci (p=0.008) and metastatic LNs (p=0.004) than those of pattern C U-ECAs. Number of LVSI foci were moderately positively correlated with LN metastasis status (p<0.001, γ=0.489) or number of metastatic LNs (p<0.001, γ=0.409). Our study further supports that Pattern Classification is a system easy to follow, which has a strong correlation to LVSI and an effective predictability for LN metastasis. Extensive linear destructive and band-like lymphocytic infiltrate growth patterns in pattern C U-EACs need to be recognised, as they behave less aggressively than that for the other growth pattern subgroups. Our study supports that Pattern Classification can be routinely applied to guide therapies for patients with U-EACs.