Prognostic nomogram for patients with invasive endocervical adenocarcinoma incorporating HPV E6/E7 mRNA in situ hybridization and clinical characteristics

Background: Human papillomavirus (HPV) E6/E7 mRNA in situ hybridization (HPV E6/E7 RNAscope) appears to be a sensitive and specic method. We aimed to examine the diagnostic and prognostic utility of this technique in endocervical adenocarcinoma (ECA) and build a useful prognostic nomogram model using this approach. Methods: The model was constructed from a retrospective study of 200 patients with ECA who had undergone surgery at Sun Yat-sen University Cancer Center between 2010 and 2014. The model's predictive eciency and discriminative capability were dened by a concordance index (C-index) and calibration curve. Results: The overall sensitivity and specicity of HPV E6/E7 RNAscope for distinguishing HPV-associated adenocarcinoma (HPVA) from non HPV-associated adenocarcinoma (NHPVA) in the whole cohort were 75.8% and 80%, respectively, compared to 60.2% and 90.0% for p16 immunohistochemistry, 80.5% and 80.0% for HPV DNA, and 63.3% and 50% for HPV genotype. The independent factors derived from multivariable analysis of the whole cohort to predict overall survival (OS) were age, lymphovascular invasion (LVI), lymph node involvement (LNI), and HPV E6/E7 RNAscope, which were all incorporated into the nomogram with (nomogram B) or without (nomogram A) FIGO stage and treatment. The C-index of nomogram A for predicting OS was 0.825 (p = 0.002 and p < 0.001), which was signicantly higher than the C-index for FIGO stage (0.653) and treatment (0.578). No signicant difference occurred between nomograms A (0.825) and B (0.836). Furthermore, a risk stratication system was established to accurately stratify patients with ECA into two subgroups with signicantly different prognosis. Conclusions: HPV E6/E7 RNAscope is highly specic for ECA, and the proposed nomogram using HPV E6/E7 RNAscope showed more accurate outcome in terms of prognosis in patients with ECA.


Introduction
Endocervical adenocarcinoma (ECA) accounts for 15-20% of all cervical carcinomas. Studies have reported that ECA is increasing in incidence, and that it often occurs in young women (1). Recent studies have reported that the 5-year overall survival (OS) rate is 10-20% lower in ECA than in squamous cell carcinoma (2). Additionally, the International Federation of Gynecology and Obstetrics (FIGO) staging system is the gold standard for predicting outcomes in patients with ECA, and concurrent chemoradiotherapy is the standard treatment for advanced ECA. However, patients with ECA that have the same FIGO stage can have marked heterogeneities of outcome (3,4). According to the latest World Health Organization (WHO) Classi cation of Tumors of Female Reproductive Organs, the most frequent type of ECA is usual-type, which is human papillomavirus (HPV)-related (5). A new classi cation-the International Endocervical Adenocarcinoma Criteria and Classi cation (IECC)-categorizes endocervical adenocarcinomas (ECAs) based on morphological features linked to etiology (i.e., HPV infection), resulting in the separation of ECAs into HPV-associated (HPVA) and non-HPVA (NHPVA) types (4).
Comparing HPVA with NHPVA types reveals essential differences in tumor behavior and patient survival, with signi cantly worse clinical outcomes for patients with NHPVA-type tumors (6,7). It follows that a classi cation based on pathogenesis and other clinicopathological factors may be more clinically informative and reproducible than the current FIGO scheme. Hence, to complement the FIGO staging system, researchers must ascertain better prognostic determinants.
The validity of this new classi cation is supported by HPV status and, at present, by limited clinical data (8,9). In the present study, the following HPV tests were used: PCR, genotype, RNAscope, and immunochemistry (IHC) against p16 protein. Plasma HPV DNA levels can be determined using PCR, providing a potential marker for cervical cancer (10). The other assays can be conducted on formalinxed, para n-embedded (FFPE) samples. Immunostaining for p16 is a cost-effective marker of viral infection (11). It is highly sensitive, but lacks speci city (11). PCR for HPV genotyping is the gold-standard assay to diagnose active HPV infection (12). Chromogenic in situ hybridization against RNA can identify HPV in situ via microscopic observation (13). The technique is expensive, but can be used on FFPE samples (14). In situ detection of HPV E6/E7 mRNA using the RNAscope also appears to be a sensitive and speci c method, and the present study aimed to investigate the potential prognostic utility of this technique in ECAs.
Numerous disagreements have arisen regarding the traditional staging system, which is completely based on the anatomical data. By stratifying pattern C ECAs into subgroups based on LVI and LNI status, clinicians could better determine treatment in patients with pattern C tumors (15

Immunohistochemistry (IHC)
A tissue microarray consisting of the ECA and adjacent non-tumorous tissue was constructed using a tissue array instrument (Minicoreexcilone; Minicore, UK). For each patient, hematoxylin and eosin (H&E) stained slides were examined, and at least two areas from different regions were marked for sampling. Each tissue core with a diameter of 1.0 mm was punched from the marked elds and re-embedded. FFPE ECA sections were dewaxed in xylene and graded alcohols, hydrated, and washed in PBS. After pretreatment in a microwave oven, endogenous peroxidase was blocked using 3% hydrogen peroxide in methanol for 20 minutes. This was followed by avidin-biotin blocking using a kit (DAKO, Germany). The Real-time PCR for HPV DNA HPV DNA viral load were routinely evaluated using quantitative PCR before therapy, as described in previous studies (16,17).

HPV Genotype
Tumor tissue for PCR testing was removed in tumors that were not represented in tissue microarrays, as described in a previous study (18 The stained slides of each specimen were examined using the RNAscope scoring system, as described in previous studies (19,20). High-risk HPV subtypes were evaluated on all ECAs in the tissue microarray that had su cient tissue to allow scoring (n = 200). HPV uorescent in situ hybridization (FISH) was performed using a chromogen and results were classi ed into ve degrees based on the following scoring guidelines: score 0, no staining or less than one dot in every ten cells (visible at 40x magni cation); score 1, 1-3 dots per cell (visible at 20-40x magni cation); score 2, 4-10 dots per cell, with very few dot clusters (visible at 20-40x magni cation); score 3, > 10 dots per cell, with < 10% of positive cells having dot clusters (visible at 20x magni cation); score 4, > 10 dots per cell with > 10% of positive cells having dot clusters (visible at 20x magni cation). Cases with RNAscope score ≥ 1 were identi ed as positive.

Cut-off values of variables
Among the patients enrolled, the following variables were recorded: age, treatment (Surgery with or without chemotherapy/radiotherapy), HPV DNA, HPV subtype, HPV E6/E7 RNAscope, histological type, tumor size, differentiation, LVI, invasion level of the uterine cervix, LNI, parametrium invasion, surgical margin, MMR status, p16, Ki-67, and FIGO stage. Speci cally, 1 pg/mL of serum HPV DNA levels, as detected by PCR, was determined as the cutoff level. The best cutoff values were determined using Xtile (21), with the following results: age, 37 years; RNAscope

Characteristics and survival in patients enrolled
We enrolled 200 patients with ECA for analysis, comprising 185 with HPVA and 15 with NHPVA (Table 1).
The correlations among serum HPV DNA level, p16 IHC, HPV PCR, and HPV E6/E7 RNAscope was also studied (Table S2). A signi cantly higher proportion of ECA cases showed positivity for p16 IHC, HPV PCR, and HPV E6/E7 RNAscope in subgroups with HPV DNA levels that differed by 10-fold. In patients with serum HPV levels of 1-100 pg/mL, the positivity for p16 IHC was similar to that for HPV PCR (60%).
In patients with serum HPV levels of 100-10,000 pg/mL, the positivity for HPV PCR was similar to that for HPV E6/E7 RNAscope (93.8%). The positivity for HPV E6/E7 RNAscope for the diagnosis of ECA was much higher than that of p16 IHC and HPV PCR (Fig. S4A). Our results also showed that HPV E6/E7 RNAscope was positive in almost all cases that showed p16 IHC positivity (88.5% of all cases, 89.5% of HPVA cases), but that more cases were negative by IHC and serum levels (Fig. S4B). In different HPV subtypes, the positivity of RNAscope for the diagnosis of ECA was much higher compared to that of p16 IHC, but lower compared to HPV DNA (Fig. S4C). The HPV subtypes in patients with HPVA and NHPVA are listed in Table S3. In cases with RNAscope scores of 3 and 4, the positivity for HPV DNA was higher than that for p16 IHC and HPV PCR (Fig. S4D).
Kaplan-Meier analysis showed signi cant diversity (Fig. 2B-E). Furthermore, the ROC curve for OS revealed that our model was higher than that of the FIGO stage and treatment (Fig. 3A). Comparison of e ciency between our nomogram and conventional systems.
The 1-, 3-, and 5-year OS rates between the two risk groups were 94.2%, 84.3%, 70.3%, respectively, in the low-risk group and 71.4%, 46.4%, 10.7%, respectively, in the high-risk group (Table 4). Moreover, signi cant differences in OS were noted between patients with stage I/II ECA and those with III/IV ECA (Fig. 5). Each risk subgroup represented a distinct prognosis, and our system accurately separated OS in the two subgroups.

Discussion
The present study was the rst to describe a prognostic nomogram model for invasive ECA in a large Chinese series, reporting HPV E6/E7 RNAscope using a new RNA ISH assay that recognizes 18 high-risk HPV types. The most important results were as follows: (1) the validity of the HPVA and NHPVA categories were supported by p16 immunophenotyping and HPV status; (2) the high-risk HPV probe set is more sensitive and speci c than p16, HPV DNA PCR, and HPV genotype for recognizing HPVA; (3) we established a prognostic nomogram model based on pathogenesis that included age, LVI, HPV E6/E7 RNAscope, and LNI, and the model had greater predictive e ciency than the current conventional staging system; (4) based on our model, patients could be divided into low-risk and high-risk groups.
HPV E6/E7 RNAscope measurement is not routinely available in most hospitals and the methodology is not globally standardized. In this regard, our ndings suggest that clinicians should increase HPV E6/E7 RNAscope measurement in clinical practice. About 94% of adenocarcinomas are associated with human papillomavirus (HPV), particularly with strains 16 and 18 (5), consistent with our results (Table S3). In the present study, 87% (160/184) of HPVA samples overexpressed p16 or were HPV + by RNAscope, validating the IECC criteria and consistent with a previous study (9,22). In terms of the IECC criteria, the HPV RNAscope is more speci c for clinical practice, and it is more robust than p16 IHC, HPV PCR, and HPV genotype to identify HPVA. Despite this, occasional p16 and/or HPV-HPVAs were identi ed. When outlying cases were excluded from statistical analysis, 25 of the 184 HPVA cases were p16-and HPVpositive. All HPVA cases in the present study were con rmed on review. The specimens were more than 5 years, may not optimal tissues for the performance of HPV RNAscope and p16. A recent study reported that p16-and HPV-positive, usual-type ECA might represent unusual morphological variants of gastrictype carcinoma (23). Furthermore, rare HPV genotypes not included in the RNA ISH probe set may be responsible for negative HPV results. There are precedents in the literature for p16-and HPV-associated neoplasia. Speci cally, p16-positive invasive squamous carcinomas of the cervix have been reported (24), as has methylation-induced inactivation of the p16 gene and allelic loss of p16 (25,26). Other studies have reported that the HPV genome can be differentially expressed between primary tumors and metastases, as well as among different disease sites, suggesting that some "HPV-related," usual-type adenocarcinomas may have had detectable HPV in multiple sections from the primary carcinoma and that all metastatic sites should been tested (22,27,28).
Clinicopathological variables have prognostic signi cance and presently impact clinical management (29,30). Depending on variables such as LVI and desire for fertility may be considered (31). Depending on variables such as LVI and desire for fertility may be considered (32), so histological evaluation of risk would be bene cial at the time cervical samples are taken. The National Comprehensive Cancer Network guidelines recommend chemotherapy or radiotherapy as the standard treatment for patients with advanced ECA in clinical practice. Hence, the performance of nomograms must be examined separately in patients treated using these therapies. The C-indices of nomograms A and B for predicting OS in treated patients were signi cantly higher than those of conventional classi cation, indicating that nomograms still have signi cant clinical value in patients with ECA treated using chemotherapy or radiotherapy. With the addition of the FIGO stage and treatment into nomogram B, the added values of these parameters over nomogram A were 0.836 and 0.825, respectively, across the entire population. There were no signi cant differences between nomograms A and B, probably because the number of patients with advanced-stage ECA was small.
The present study had several shortcomings. Firstly, there may have been selection bias. Secondly, we lacked data on the impact of our nomogram on other prognoses, such as disease-free survival (DFS) prediction. Thirdly, the in situ measurement of HPV RNA using RNAscope must still be globally standardized. Fourthly, our sample size was relatively small and collected from one center. To verify our results, more samples should be collected from other institutes. Lastly, it remains unclear whether our nomogram can be applied to advanced-stage patients (stages III and IV).
In summary, we built a portentous nomogram for prognosis in the patients with ECA. It was more accurate than the conventional FIGO staging system. The model is simple, economical, and helpful for clinical decision making.     The calibration curves, decision curve analysis for predicting patient overall survival (OS). The calibration curves for predicting OS at 1, 3, and 5 years in all cases. (A&B) Nomograms A and B, with model-predicted OS plotted on the x-axis and actual OS plotted on the y-axis. Closer alignment with the diagonal line represents better estimation. (C) Decision curve analysis for 1-, 3-, and 5-year survival predictions.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.