Independent validation of the prognostic significance of invasion patterns in endocervical adenocarcinoma: Pattern A predicts excellent survival
Introduction
Cervical cancer is the fourth most lethal cancer in women worldwide [1]. Adenocarcinoma (AC) is the second most common histological subtype, comprising approximately 15–25% of all cervical cancer cases [2,3]. Contrary to squamous cell carcinoma (SCC), incidence rates of AC are rising, especially in young women and in western countries [4,5].
Cervical cancer is clinically staged according to the FIGO (International Federation of Gynaecology and Obstetrics) classification [2,3]. Tumour size (horizontal spread) and depth of invasion (DOI) are important microscopic/morphological parameters in assigning FIGO stage. The choice of treatment of cervical cancer patients primarily depends on the FIGO stage and the presence of lymph-vascular invasion (LVI) [3]. Stage IA tumours, without LVI, are treated more conservatively by cone biopsy or simple hysterectomy alone, whilst larger tumours or tumours invading the cervix more deeply (stage IB), are treated by radical hysterectomy or trachelectomy with pelvic lymph node sampling in the form of a lymphadenectomy or sentinel lymph node biopsy [3]. The latter concerns major surgery associated with high morbidity and infertility. However, in cervical AC, tumour size and especially the DOI are poorly reproducible measurements, because AC often presents in large exophytic masses, in which it is difficult to distinguish the in situ component (AIS) from the invasive tumour [2,6].
Therefore, a potentially more reproducible pattern-based classification-system for usual type endocervical AC was proposed [[7], [8], [9], [10], [11]]. This system classifies AC into three categories, based on the destructiveness of stromal invasion. Pattern A tumours consist of well-demarcated glands, and are without destructive stromal invasion and without LVI. Pattern B tumours present focally destructive stromal invasion, arising from pattern A glands spanning no >5 mm contiguously. And pattern C tumours show diffuse destructive stromal invasion with an associated desmoplastic response. LVI may be present in Pattern B and C [11]. Pattern C tumours were associated with higher FIGO stages, lymph-node metastasis, disease recurrence and tumour associated death, whilst pattern A tumours lacked lymph-node metastasis and showed no disease recurrence, independent of tumour size or DOI [7,9,12]. A more conservative approach for AC patients with pattern A morphology was suggested [9,13].
Usual type endocervical AC is caused by a persistent infection with carcinogenic Human Papillomavirus (HPV) [14]. However, the progression from initial HPV infection to invasive cervical cancer is a complex and multifactorial process in which various immunological, molecular and genetic alterations play a role [[15], [16], [17]]. The Cancer Genome Atlas (TCGA) evaluation of cervical cancer concluded that over 70% of cervical cancers exhibited genomic alterations in either one or both the PI3K/Akt- and TGFβ-pathways [15]. However, the few AC included in TCGA analysis were not classified by pattern of invasion. We have reported on the mutation status in different histological subtypes, and found PIK3CA mutations occurring more frequently in SCC than AC (25% vs. 11%, p = 0.025), whereas KRAS mutations occurred more frequently in AC than SCC (24% vs. 3%, p < 0.001) [17]. However, also in the latter study, patterns of invasion were not assigned. Recent work by Hodgson et al. revealed a higher prevalence of oncogene and tumour suppressor gene abnormalities in the PI3K/Akt-pathway in pattern B and C tumours compared to pattern A tumours [18]. Although this study consisted of a relatively small (n = 20) and heterogeneous (FIGO stage IB-IVB) cohort of AC, with only a short follow-up period (mean follow-up 38 months), it provided the first molecular support for the proposed pattern-based classification [18].
In the present study, based on a large cohort of n = 82 usual type endocervical AC with mature follow-up data, we aimed to independently validate the clinical significance of the pattern-based classification-system and we investigated whether the patterns of invasion are associated with p53 expression, mismatch repair (MMR) deficiency or somatic hot-spot mutations in genes of the PI3K/Akt-pathway.
Section snippets
Case selection
All patients with usual type endocervical AC FIGO stage IB1 – IIA1, who underwent a radical hysterectomy or trachelectomy with pelvic lymphadenectomy as primary treatment at the Leiden University Medical Centre between January 1990 and December 2011, were included in this study. All cases were used according to the Code of Conduct for Proper Secondary Use of Human Tissue, established by the Federation of Dutch Medical Scientific Societies [19]. Clinical and histological parameters were
Patient samples
In total 82 patients with usual type endocervical AC were included in this study; 52 with molecular data from a prior data set, published by our group [17]. Eighteen (22%) tumours were classified as pattern A, 30 (37%) as pattern B, and 34 (41%) as pattern C. Baseline characteristics are summarized in Table 1 for the total cohort, and for pattern A, B, and C separately. When comparing the clinicopathological parameters between the various invasion patterns, there was a significant association
Discussion
In this retrospective, single-centre, cohort study we independently validated the pattern-based classification-system as proposed by Diaz de Vivar et al. [7] on a large cohort of usual type endocervical AC (n = 82). Patients with pattern A tumours showed excellent long term recurrence-free and disease-specific survival. In combination with the absence of both lymph node metastasis and parametrial invasion in these tumours, our data support the suggestion by Park and Roma for a less radical
Conflict of interest statement
The authors have nothing to disclose.
Financial support
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author contributions
Study concept by CdK, GF, TB, and EJ; study design by VS, BB, TB, and EJ; data acquisition by VS, DS, BB, GF, and TB; quality control of data and algorithms by VS, CdK, EJ; data analysis and interpretation by VS, BB, GF, and TB, statistical analysis by VS and EJ; manuscript preparation by VS; manuscript editing by CdK, TB, EJ; manuscript review by VS, DS, BB, CdK, GF, TB, and EJ.
References (26)
- et al.
Cervical cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up
Ann. Oncol.
(2017) Glandular lesions of the uterine cervix
Mod. Pathol.
(2000)- et al.
New pattern-based personalized risk stratification system for endocervical adenocarcinoma with important clinical implications and surgical outcome
Gynecol. Oncol.
(2016) - et al.
Pattern classification of endocervical adenocarcinoma: reproducibility and review of criteria
Mod. Pathol.
(2016) - et al.
Pattern based classification of endocervical adenocarcinoma: a review
Pathology
(2018) - et al.
Prevalence of human papillomavirus DNA in various histological subtypes of cervical adenocarcinoma: a population-based study
Mod. Pathol.
(2005) - et al.
Genomic abnormalities in invasive endocervical adenocarcinoma correlate with pattern of invasion: biologic and clinical implications
Mod. Pathol.
(2017) - et al.
Global cancer statistics, 2012
CA Cancer J. Clin.
(2015) - et al.
Tumours of the uterine cervix
- et al.
Increasing incidence of invasive and in situ cervical adenocarcinoma in the Netherlands during 2004–2013
Cancer Med.
(2017)
Trends in the incidence of cervical cancer and severe precancerous lesions in Denmark, 1997–2012
Cancer Causes Control
Invasive endocervical adenocarcinoma: proposal for a new pattern-based classification system with significant clinical implications: a multi-institutional study
Int. J. Gynecol. Pathol.
Invasive endocervical adenocarcinoma: a new pattern-based classification system with important clinical significance
Am. J. Surg. Pathol.
Cited by (23)
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
2019, PathologyCitation Excerpt :Tumours are classified according to the status (presence or absence) and extent of destructive stromal invasion rather than tumour size and DOI. In line with previous studies,15–17 we found that this classification was easy to learn. Pattern A tumours lack aggressive behaviour, because patients do not have LVSI, LN metastasis, and >10 mm DOI.
Cervical Glandular Neoplasia: Classification and Staging
2019, Surgical Pathology ClinicsCitation Excerpt :Unlike the Hodgson study, 2 (20%) pattern A tumors harbored KRAS mutations as well as 5 (29%) pattern B and 5 (20%) pattern C. The remaining pattern A tumors (8) showed no other mutations, whereas patterns B and C showed additional mutations in PIK3CA, PTEN, CDKN2A, PPP2R1A, and FBXW7.69 The data suggest that pattern A tumors are indeed biologically distinct from patterns B and C with a possible tumor progression model in which mutations accumulate as the tumor invades the surrounding stroma.69,70
- 1
Present address: University of Milan, Faculty of Medicine and Surgery, Milan, Italy.
- 2
Authors contributed equally.