Abstract
The 5th edition of the WHO Classification of Tumours (Digestive System) recognizes a new subtype of colorectal adenocarcinoma, called adenoma-like adenocarcinoma. In this study we sought to determine its clinicopathologic associations, and how it compares to adenocarcinoma, of no special type (NOS). We retrospectively reviewed all available archival slides of stage I-III colonic adenocarcinoma resection specimens at our institution from 2013 to 2016. 91 cases were classified as adenoma-like adenocarcinoma, and 251 cases were adenocarcinoma, NOS. Of the adenoma-like adenocarcinoma cases, a majority (65 cases, 71%) were composed exclusively of adenoma-like features, designated as pure adenoma-like adenocarcinoma, while in the rest, the component of adenoma-like morphology was more than 50% but less than 100%, designated as mixed adenoma-like adenocarcinoma. Compared to adenocarcinoma, NOS, adenoma-like adenocarcinoma cases were significantly associated with the absence of tumor budding (p<0.001), absence of an immature/myxoid desmoplastic reaction (p<0.001), presence of intraepithelial tumor associated lymphocytes (iTILs) (p=0.006), fewer lymph nodes involved (p<0.001), fewer tumor deposits (p=0.042), lower pT stage (p=0.047), lower pN-stage (p<0.001), and consequently pTNM prognostic group (p<0.001), as well as better recurrence-free survival (RFS) on univariate analysis compared to adenocarcinoma, NOS cases (p=0.026) but not on multivariate analysis. However, mixed adenoma-like adenocarcinoma had a worse RFS compared to pure adenoma-like adenocarcinoma (HR: 1.639, 95%CI: 0.494–5.437). Our findings support the importance of distinguishing this new subtype of colorectal adenocarcinoma, but also raises the question whether mixed adenoma-like adenocarcinoma cases should be included in this category, and if so, whether 50% is an appropriate cutoff, as currently defined by WHO.
Keywords: adenoma-like adenocarcinoma, colorectal cancer, colorectal adenocarcinoma, tumor budding, intraepithelial tumor associated lymphocytes, desmoplastic reaction
1. INTRODUCTION
The 5th edition of the World Health Organization (WHO) Classification of Tumours (Digestive System) has recognized a new subtype of colorectal adenocarcinoma, called adenoma-like adenocarcinoma[1]. Over the last few decades, several publications attempted subclassifying colorectal adenocarcinoma based on a distinctive histologic appearance, such as villous tumor[2,3], invasive papillary adenocarcinoma[4] and villous adenocarcinoma[5]. Based on the morphologic descriptions, it appears that some of them could be referring to this subtype that is now termed adenoma-like adenocarcinoma. Yao T. et al.[3] noted a well-differentiated morphology with less p53 protein expression in 17 villous carcinoma cases, but unfortunately no follow-up data was available. A concomitant study with 20 cases noted that these colorectal adenocarcinoma cases with a villous morphology were associated with fewer lymph nodes involved[2]. Another report of this new subtype of colorectal carcinoma was made by Palazzo J.P. et al.[4] in their description of two cases designated as invasive papillary adenocarcinoma. In their limited number of cases, an aggressive behavior was noted, with one patient dying of disease[4]. A more comprehensive review of 36 cases demonstrated an association with a favorable prognosis and raised the importance of awareness of this entity, given its diagnostic challenge due to its well-differentiated morphology[5]. In 2016, the term adenoma-like adenocarcinoma was introduced by Gonzalez RS et al.[6] In their study, they noted frequent KRAS mutations and a lower rate of metastasis in this group of tumors.
Currently, adenoma-like adenocarcinoma cases are defined according to WHO as an “invasive adenocarcinoma in which ≥ 50% of the invasive areas have an adenoma-like aspect with villous structures, with a low-grade aspect.”[1] The tumor is also associated with a pushing border and with minimal desmoplastic reaction[1]. In this study, we present the largest cohort of this entity, and sought to validate the clinicopathologic implications. For this purpose, we used a large comparison group of colonic adenocarcinomas of no special type (adenocarcinoma, NOS).
2. MATERIALS AND METHODS
2.1. Patient Population
Approval from the Institutional Review Board of Washington University School of Medicine was obtained prior to initiating the study. A retrospective search was performed for all colonic adenocarcinoma resection specimens in our database from 2013 to 2016. Cases with clinical or pathologic stage IV disease according to the American Joint Committee on Cancer (AJCC) 8th Edition[7] were excluded, to appropriately assess for recurrence-free survival (RFS). Rectal carcinomas were also excluded, since majority of these cases receive neoadjuvant therapy and could alter morphologic assessments. A total of 372 cases of colonic adenocarcinoma stage I-III were identified, whose archival slides were available for review. Of those 372 cases, 30 cases were excluded from the study since they represented a special subtype of colorectal carcinoma (22 cases of mucinous adenocarcinoma, 7 cases of medullary carcinoma, and one case of serrated adenocarcinoma). The tumor location was recorded as right-sided when tumor involved anywhere from the cecum to the proximal two thirds of the transverse colon, and as left-sided, from the distal third of the transverse colon up to the rectum. The relevant clinical information was obtained from the electronic medical chart.
2.2. Histopathologic Characteristics
All the tumor slides were reviewed by one pathologist in training (IAG) and one GI pathologist (DC) simultaneously in a multiheader microscope, blinded to the clinical information. The findings recorded were based on consensus. The specimens were grossed according to our institutional protocol which consists of one tumor section submitted per centimeter of the greatest tumor dimension, and if the tumor is 3 cm or less, the entire tumor is submitted for histopathologic evaluation.
The tumor morphology was reviewed and categorized according to the 5th edition of the WHO Classification of Tumours (Digestive System)[1]. Recognition of adenoma-like adenocarcinoma was made based on the tumor description provided by WHO, where the invasive carcinoma consisted of an adenoma-like morphology, with low-grade cytologic features, and no to little surrounding desmoplastic reaction[1] (Figure-1). We subclassified these cases as pure, when the tumor was entirely composed of adenoma-like morphology, and mixed, when another component of adenocarcinoma was admixed, but at least 50% of the tumor showed adenoma-like features, in accordance with the WHO definition (Figure-2). The percentage of this tumor component was based on eyeballing, and considering, all tumor slides of each case. In the mixed category, a percentage of the adenoma-like morphology was assigned to each case. Other special types of colorectal adenocarcinoma as defined by the WHO were excluded from the study.
Figure-1:
Representative photomicrographs of separate cases of pure adenoma-like adenocarcinoma cases (a – f, including a pT3 tumor in 1b).
Figure-2:
Representative photomicrographs of separate cases of mixed adenoma-like adenocarcinoma (a – d, with adenoma-like areas on the right, and conventional invasive areas on the left).
Pathologic tumor stage, nodal involvement, tumor deposits, lymphovascular invasion (LVI) and perineural invasion (PNI) were verified and recorded based on the guidelines published in the current CAP protocol (v4.1.0.0). Tumor grading was recorded as low-grade and high-grade based on the least differentiated component and not based on overall percentage of gland formation (according to the latest WHO classification system). Tumor budding (TB) was recorded as present when single tumor cells or clusters of 4 or fewer tumor cells were identified at the tumor edge[8]. TB was also classified according to number of buds when present, into three categories, 0 – 4 (low), 5 – 9 (intermediate), and ≥10 buds (high) per 0.785 mm2. Cases containing 4 or more intraepithelial lymphocytes infiltrating tumor, per high-power field, were designated as being positive for intraepithelial tumor infiltrating lymphocytes (iTILs)[9-12]. The presence and type of desmoplastic reaction at the invasive tumor edge was noted, and classified as immature/myxoid, intermediate and mature desmoplastic reaction according to the proposed assessment by Ueno et al[13–15] and recently validated by our group[11]. The expression pattern of DNA mismatch repair proteins (MMR) were obtained from the pathology reports. Retained MMR (pMMR) was defined as any nuclear expression of MSH2, MSH6, PMS2, and MLH1 in the tumor cells, and MMR loss (dMMR) as complete absence of nuclear expression of any protein in the tumor cells.
2.3. Statistical Analysis
The clinical characteristics were summarized using descriptive statistics. Overall survival (OS) was defined as the years from the date of surgery to death. Alive patients were censored at the last follow-up. RFS was defined as the years from the date of surgery to recurrence. Alive patients without recurrence were considered as a competing event at the last follow-up. The macro %findcut was used to find a best cutpoint of the percentage of adenoma-like adenocarcinoma for the 26 mixed cases. This approach gave the largest difference of log-rank test statistics between subjects in the two groups. Recurrence free probabilities were calculated using cumulative incidence curves using the Fine and Gray approach. Differences between the different categories were determined by the Fine and Gray approach. Cause-Specific Analysis of competing risks was used to evaluate the relationship of the variables of interest for RFS analysis. The variables with p<0.25 from univariate models were considered in the multivariable model. The stages included pathological N stage, T stage, and pTNM AJCC prognostic stage grouping. Given the possible correlation among these, the prognostic stage grouping had the highest priority. The final multivariable model was built using the backward stepwise selection approach to identify all significant risk factors. Factors significant at a 10% level were kept in the final model. All statistical tests were two-sided using an α = 0.05 level of significance. SAS Version 9.4 (Cary, NC) was used to perform all statistical analyses.
3. RESULTS
3.1. Patient and Tumor Characteristics: Adenoma-like Adenocarcinoma
The mean age of diagnosis for adenoma-like adenocarcinoma was 67.9 (± 13.1) years and presented evenly in men (46 patients, 51%) and women (45 patients, 49%), and 88% of the patients were Caucasian (Table-1). In 17 patients (19%), a prior history of another neoplasm was noted, with breast cancer being the most common (4 patients). Four patients (4%) had a prior diagnosis of Lynch syndrome. Most of the cases (57 patients, 63%) were located in the right colon, with 47% of the cases located in the cecum and ascending colon. In the left sided tumors, the most common location was the sigmoid colon (20 cases, 22%). Most of the cases were resected via right hemicolectomy (45 patients, 49%) followed by sigmoidectomy (11 patients, 12%) and extended right hemicolectomy (8 patients, 9%). Other procedures included extended left hemicolectomy, subtotal colectomy and total abdominal colectomy. All cases had negative proximal, distal and radial surgical resection margins. None of the cases received neoadjuvant therapy, but 13 patients (15%) received adjuvant chemotherapy.
Table-1.
Patient and Tumor Characteristics
| All N = 342 |
ACA, NOS N = 251 |
AA N = 91 |
P | |
|---|---|---|---|---|
| Age, mean ± SD | 65.2 ± 12.6 | 64.2 ± 12.3 | 67.9 ± 13.1 | 0.009 |
| Gender, n, % | ||||
| Male | 179, 52% | 133, 53% | 46, 51% | 0.690 |
| Female | 163, 48% | 118, 47% | 45, 49% | |
| Race, n, % | ||||
| Caucasian | 288, 84% | 208, 83% | 80, 88% | 0.417 |
| African American | 45, 13% | 35, 14% | 10, 11% | |
| Other1 | 9, 3% | 8, 3% | 1, 1% | |
| History of other cancer, n, % | ||||
| No | 293, 86% | 219, 87% | 74, 81% | 0.166 |
| Yes | 49, 14% | 32, 13% | 17, 19% | |
| Lynch syndrome, n, % | ||||
| No | 336, 98% | 249, 99% | 87, 96% | 0.045 |
| Yes | 6, 2% | 2, 1% | 4, 4% | |
| Location, n, % | ||||
| Right colon | 204, 60% | 147, 59% | 57, 63% | 0.498 |
| Left colon | 138, 40% | 104, 41% | 34, 37% | |
| Adjuvant chemotherapy, n, % | ||||
| No | 244, 72% | 167, 67% | 77, 86% | <0.001 |
| Yes | 97, 28% | 84, 33% | 13, 14% | |
| Tumor grade, n, % | ||||
| Low | 294, 86% | 206, 82% | 88, 97% | <0.001 |
| High | 48, 14% | 45, 18% | 3, 3% | |
| Lymphovascular invasion, n, % | ||||
| No | 217, 63% | 143, 57% | 74, 81% | <0.001 |
| Yes | 125, 37% | 108, 43% | 17, 19% | |
| Perineural invasion, n, % | ||||
| No | 289, 85% | 204, 81% | 85, 93% | 0.006 |
| Yes | 53, 15% | 47, 19% | 6, 7% | |
| Tumor budding, n, % | ||||
| Low (0 - 4) | 272, 80% | 184, 73% | 88, 97% | <0.001 |
| Intermediate (5 - 9) | 31, 9% | 30, 12% | 1, 1% | |
| High (≥10) | 39, 11% | 37, 15% | 2, 2% | |
| Desmoplastic reaction n, % | ||||
| Immature/myxoid | 66, 19% | 58, 23% | 8, 9% | <0.001 |
| Intermediate | 133, 39% | 109, 44% | 24, 26% | |
| Mature | 143, 42% | 84, 33% | 59, 65% | |
| Intraepithelial tumor infiltrating lymphocytes, n, % | ||||
| No | 228, 67% | 178, 71% | 50, 55% | 0.006 |
| Yes | 114, 33% | 73, 29% | 41, 45% | |
| Tumor deposits, n, % | ||||
| No | 303, 89% | 217, 86% | 86, 95% | 0.038 |
| Yes | 39, 11% | 34, 14% | 5, 5% | |
| Number of tumor deposits, mean ± SD | 0.4 ±2.1 | 0.5 ±2.4 | 0.2 ±0.9 | 0.042 |
| MMR status, n, % (n = 298) | ||||
| Retained | 245, 82% | 187, 84% | 58, 77% | 0.201 |
| Lost | 53, 18% | 36, 16% | 17, 23% | |
| BRAF gene, n, % (n=38) | ||||
| Wild type | 23, 61% | 18, 64% | 5, 50% | 0.473 |
| V600E | 15, 39% | 10, 36% | 5, 50% | |
| KRAS gene, n, % (n=28) | ||||
| Wild type | 16, 57% | 16, 64% | 0 | 0.067 |
LVI and PNI were present in 17 cases (19%) and 6 cases (7%), respectively. 3 cases were classified as high-grade based on extensive cribriforming in the least differentiated area of the tumor. Intermediate and high TB was identified in 1 (1%) and 2 (2%) cases, respectively, all of which were identified in mixed adenoma-like adenocarcinoma. Only 8 cases (9%) showed an immature/myxoid desmoplastic reaction, and most of the cases (59 cases, 65%) had a mature stromal reaction. As expected none of the pure adenoma-like adenocarcinoma show either tumor budding or a desmoplastic reaction. iTILs were identified in 41 cases (45%). In only 5 cases, tumor deposits were identified, with a mean number of tumor deposits of 0.2 (± 0.9, range: 0 - 6). The majority of the cases were pT3 stage (65%) followed by pT2 (23%), pT1 (8%) and pT4 (4%), and most of the cases were pN0 (80%) followed by pN1 (12%) and pN2 (8%). The mean number of lymph nodes involved was 0.5 (± 1.3, range: 0 – 6) with a mean number of lymph nodes examined being 20.6 (± 8.6). Forty-nine cases (54%) were pTNM prognostic stage groups II, followed by group I (26%) and group III (20%). MMR proteins were retained in 58 cases (77%). Among dMMR cases, the more common proteins lost were PMS2 and MLH1 (14 cases, 83%). Molecular studies were available only in a limited number of patients, with BRAF gene mutation (V600E) reported in 5 tumors, and KRAS gene mutation (codon 12) in three tumors.
3.2. Patient and Tumor Characteristics: Adenocarcinoma, NOS
A total of 251 adenocarcinoma, NOS cases were included. The mean age of diagnosis was 64.2 years (± 12.3) with 53% of the cases presenting in men (Table-1). In 32 cases (13%) a prior history of another neoplasm was noted with breast cancer being the most common (9 patients). 83% percent of the patients were Caucasian. Two patients (1%) had a prior diagnosis of Lynch syndrome. The cases were fairly evenly distributed to the right side (59%) and left side (41%) with the most common specific location being the sigmoid colon (73 cases, 29%). The other two most common locations were the ascending colon (51 cases, 20%) and the cecum (48 cases, 19%). The two most common surgical procedures were right hemicolectomy (111 cases, 44%) and left hemicolectomy (60 cases, 24%). All the cases had negative proximal, distal and radial surgical resection margins. None of the cases received neoadjuvant therapy, but 84 cases (33%) received adjuvant chemotherapy with FOLFOX being the most common regimen (45 cases, 54%).
LVI was identified in 43% of the cases and PNI in 19%. Majority of the cases (86%) were low-grade (exclusion of tumor invasive front with TB and poorly differentiated clusters were performed during review, based on the new WHO recommendation). Low TB was present in 184 cases (73%), intermediate TB in 30 cases (12%) and high TB in 37 cases (15%). A myxoid/immature desmoplastic reaction was noted in 58 cases (23%). iTILs were identified in 73 cases (29%). Tumor deposits were seen in 34 cases (14%) with a mean number of tumor deposits of 0.5 (± 2.4, range: 0 - 34). The majority of the cases were pT3 (57%) followed by pT2 (20%), pT4 (16%) and pT1 (7%). 145 cases (48%) had no nodal disease (pN0), 70 cases (28%) were considered pN1 and 36 cases (14%) pN2. The mean numbers of lymph nodes involved were 1.4 (± 2.8, range: 0 – 23) with a mean number of lymph nodes examined being 19.9 (± 7.9). 106 cases (42%) were pTNM prognostic stage group III followed by group II (36%) and group I (22%). MMR proteins were retained in 187 cases (84%). Among dMMR cases, the more common proteins lost were PMS2 and MLH1 (28 cases, 73%). Limited molecular information was available. BRAF (V600E) gene mutation was reported in 10 of 28 tumors (36%), and KRAS (codon 12) gene mutation was reported 9 of 25 tumors (36%).
3.3. Clinicopathologic Comparisons Between Adenoma-like Adenocarcinoma and Adenocarcinoma, NOS
Adenoma-like adenocarcinoma and adenocarcinoma, NOS cases were similar in gender, race, and a prior history of cancers. However, adenoma-like adenocarcinoma cases presented at an older age compared to adenocarcinoma, NOS cases (67.9 vs. 64.2 years, p = 0.009). Adenoma-like adenocarcinoma cases were more likely to have a history of Lynch syndrome, and presence of iTILs compared to adenocarcinoma, NOS (p = 0.006), although there was no significant difference between the groups in overall MMR expression. They were significantly less likely to have poor differentiation, LVI, PNI, TB, myxoid/immature desmoplastic reaction, and tumor deposits. Adenoma-like adenocarcinoma cases were also significantly associated with lower pT-stage, pN-stage, number of lymph nodes involved, and pTNM prognostic stage group. Consequently, they were significantly less likely to receive adjuvant chemotherapy (p<0.001). There was no significant difference in tumor location, BRAF or KRAS status between adenoma-like adenocarcinoma and adenocarcinoma NOS cases. The details are recorded in Table 1.
3.4. Clinicopathologic Comparisons Between Pure and Mixed Adenoma-like Adenocarcinoma
A total of 91 cases of adenoma-like adenocarcinoma of which 26 (29%) were considered as mixed adenoma-like adenocarcinoma. There was no difference in age or gender between pure or mixed adenoma-like adenocarcinoma, and patients of African-American descent were more commonly mixed adenoma-like adenocarcinoma (23% vs 6%, p = 0.038) (Table-2). There was no preference for right or left colon between pure vs mixed adenoma-like adenocarcinomas (p = 0.732). No significant difference was identified in LVI and PNI. All pure adenoma-like adenocarcinoma showed low TB compared to 12% of the mixed cases which had intermediate or high TB (p = 0.021). None of the pure cases showed a myxoid/immature desmoplastic reaction in contrast to mixed cases, among which 31% had a myxoid/immature desmoplastic reaction (p < 0.001). iTILs was significantly associated with pure cases (p = 0.028). There was no difference in pathologic T- or N-stage, or in pTNM prognostic groups between pure and mixed cases.
Table 2.
Clinicopathologic characteristics of pure and mixed adenoma-like adenocarcinoma.
| Clinical and pathologic characteristics |
Pure AA, N = 65 |
Mixed AA, N = 26 |
P |
|---|---|---|---|
| Age, mean ± SD | 67 ± 13.4 | 68 ± 12.6 | 0.661 |
| Gender, n, % | |||
| Male | 30, 46% | 16, 62% | 0.185 |
| Female | 35, 54% | 10, 38% | |
| Race, n, % | |||
| Caucasian | 60, 92% | 20, 77% | 0.038 |
| African American | 4, 6% | 6, 23% | |
| Othera | 1, 2% | 0 | |
| Location, n, % | |||
| Right colon | 40, 62% | 17, 65% | 0.732 |
| Left colon | 25, 38% | 9, 35% | |
| Tumor grade, n, % | |||
| Low | 64, 98% | 24, 92% | 0.195 |
| High | 1, 2% | 2, 8% | |
| Lymphovascular invasion, n, % | |||
| No | 56, 86% | 18, 69% | 0.077 |
| Yes | 9, 14% | 8, 31% | |
| Perineural invasion, n, % | |||
| No | 61, 94% | 24, 92% | 1.000 |
| Yes | 4, 6% | 2, 8% | |
| Tumor budding, n, % | |||
| Low (0–4) | 66, 100% | 23, 88% | 0.021 |
| Intermediate (5–9) | 0 | 1, 4% | |
| High (≥10) | 0 | 2, 8% | |
| Desmoplastic reaction, n, % | |||
| Immature/myxoid | 0 | 8, 31% | <0.001 |
| Intermediate | 15, 23% | 9, 35% | |
| Mature | 50, 77% | 9, 35% | |
| Intraepithelial tumor-infiltrating lymphocytes, n, % | |||
| No | 31, 48% | 19, 73% | 0.028 |
| Yes | 34, 52% | 7, 27% | |
| Tumor deposits, n, % | |||
| No | 62, 95% | 24, 92% | 0.622 |
| Yes | 3, 5% | 2, 8% | |
| Number of tumor deposits, mean ± SD | 0.2 ± 0.9 | 0.2 ± 0.6 | 0.586 |
| MMR status, n, % | |||
| Retained | 42, 76% | 16, 80% | 1.0 |
| Lost | 13, 24% | 4, 20% | |
| pT stage, n, % | |||
| pT1 | 7, 11% | 0 | 0.192 |
| pT2 | 15, 23% | 6, 23% | |
| pT3 | 39, 60% | 20, 77% | |
| pT4 | 4, 6% | 0 | |
| pN stage, n, % | |||
| pN0 | 54, 83% | 19, 73% | 0.264 |
| pN1 | 8, 12% | 3, 12% | |
| pN2 | 3, 5% | 4, 15% | |
| Number of LNs involved, mean ± SD | 0.4 ± 1.1 | 0.8 ± 1.8 | 0.518 |
| Total number of LNs examined, mean ± SD | 20.8 ± 9.1 | 19.9 ± 7.1 | 0.755 |
| pTNM prognostic stage group, n, % | |||
| I | 19, 29% | 5, 19% | 0.439 |
| II | 35, 54% | 14, 54% | |
| III | 11, 17% | 7, 27% | |
NOTE. Percentages might not total to 100% owing to rounding. Bold value indicates significant p-value.
Abbreviations: AA, adenoma-like adenocarcinoma; MMR, mismatch repair; LN, lymph node; SD, standard deviation.
Includes Asian and Hispanic.
3.5. Histopathologic Features of Adenoma-like Adenocarcinoma in Biopsies
The prior biopsies were available for review in 23 cases (35%) of pure adenoma-like adenocarcinoma. 11 cases (46%) were interpreted as an adenoma only, even though 3 of these cases had multiple (1-3) biopsies, all of which were also considered to represent adenomas. In these biopsies, no high-grade dysplasia was identified. The remaining 12 cases were diagnosed as well-differentiated adenocarcinoma arising in an adenoma. Two of them were repeat biopsies after prior diagnoses of adenoma. Histologic analyses of these positive biopsies on review yielded four clues to diagnosis of adenocarcinoma in this group of cases. 5 of them featured evidence of deeper invasion in an oriented specimen (Figure-3a), 5 of them featured glands surrounded by altered stroma (dense eosinophilic) instead of lamina propria (Figure-3b), 3 of them had adjacent large vessels, one with thrombus (Figure-3c) and others with associated ulceration, and only one showed focal desmoplastic reaction while most of the other fragments resembled a tubulovillous adenoma (Figure-3d).
Figure-3:
Four separate biopsy patterns diagnostic of adenocarcinoma, in the setting of adenoma-like adenocarcinoma. Deeper infiltration of well-differentiated invasive glands into collagenous but not desmoplastic stroma, in a large oriented biopsy (a). Villiform low-grade neoplastic epithelium with underlying collagenous stroma, not lamina propria (b). Low-grade neoplastic epithelium associated with vascular thrombosis (c). Area of focal desmoplastic reaction, within other fragments that resembled tubulovillous adenoma (d).
3.6. Adenoma-like Adenocarcinoma and Recurrence-free Survival
Adenocarcinoma, NOS cases had a worse RFS compared to adenoma-like adenocarcinoma, when using the current WHO definition (HR = 2.07, 95% CI: 1.09 – 3.94) (Figure-4a). Also, pure adenoma-like adenocarcinoma had a longer RFS compared to the rest of the cases (HR = 0.43, 95% CI: 0.19 – 0.94) (Figure-4b). Importantly, mixed adenoma-like adenocarcinoma had a worse RFS compared to pure adenoma-like adenocarcinoma cases (HR = 1.64, 95% CI: 0.49 – 5.44), and a better RFS compared to adenocarcinoma, NOS cases (HR = 0.68, 95% CI: 0.26 – 1.79) (Figure-4c). For the 26 mixed cases, a macro %findcut method was used to find the largest difference of log-rank test statistics from OS between two groups. Using this method describe by Contal and O’Quigley[16] the best cutpoint was found to be 70%. Using the cutpoint of 70% cases with more than 70% of pure morphology were associated with a better RFS to those below 70% (p = 0.002)
Figure-4:
Risk of disease recurrence plots. Adenoma-like adenocarcinoma compared to colonic adenocarcinoma, NOS cases (a). Pure adenoma-like adenocarcinoma compared to all other cases (b). Comparison of pure adenoma-like adenocarcinoma, mixed adenoma-like adenocarcinoma and colonic adenocarcinoma, NOS cases (c).
On univariate Cause-Specific Analysis of competing risks, age (p = 0.081), adenoma-like adenocarcinoma (p = 0.035), TB (p = 0.004), iTILs (p = 0.002), LVI (p < 0.001), PNI (p = 0.006), tumor deposits (p < 0.001), myxoid/immature desmoplastic reaction (p = 0.002), pT-stage (p < 0.001), pN-stage (p < 0.001) and pTNM prognostic groups (p < 0.001) were associated with RFS (Table-3). On multivariable Cause-Specific Analysis of competing risks, the absence of iTILs (HR = 2.558, 95%CI: 1.38 – 4.83) was associated with a worse RFS, and lower pTNM prognostic groups (HR = 0.075; 95%CI: 0.023 – 0.241) was associated with a better RFS.
Table-3.
Cause-Specific Analysis of Competing Risks Models of Recurrence-Free Survival
| MMR status, n, % | |||
| Retained | 42, 76% | 16, 80% | 1.0 |
| Lost | 13, 24% | 4, 20% | |
| pT stage, n, % | |||
| pT1 | 7, 11% | 0 | 0.192 |
| pT2 | 15, 23% | 6, 23% | |
| pT3 | 39, 60% | 20, 77% | |
| pT4 | 4, 6% | 0 | |
| pN stage, n, % | |||
| pN0 | 54, 83% | 19, 73% | 0.264 |
| pN1 | 8, 12% | 3, 12% | |
| pN2 | 3, 5% | 4, 15% | |
| Number of LNs involved, mean ±SD | 0.4 ±1.1 | 0.8 ±1.8 | 0.518 |
| Total number of LNs examined, mean ±SD | 20.8 ±9.1 | 19.9 ±7.1 | 0.755 |
| pTNM prognostic stage group, n, % | |||
| I | 19, 29% | 5, 19% | 0.439 |
| II | 35, 54% | 14, 54% | |
| III | 11, 17% | 7, 27% |
Abbreviations: AA - adenoma-like adenocarcinoma; MMR - mismatch repair
proteins; LN - lymph node
includes Asian and Hispanic
Percentages might not total 100% due to rounding
When only the TNM prognostic stage group II cases were compared between these two groups, there was no difference in survival in terms of RFS (p=0.57), supporting the fact that the diagnosis of adenoma-like adenocarcinoma is not an independent prognostic factor but the difference in RFS is primarily due to differences in T-stage, N-stage, TB, and other prognostic parameters rather than related to intrinsic tumor type.
4. DISCUSSION
The 2019 WHO classification recognized adenoma-like adenocarcinoma as a new subtype of colorectal carcinoma, which is characterized by a distinct low-grade morphology mimicking adenoma, but with a pushing invasion into deeper colonic tissue.[1] To our knowledge, our study reports the largest cohort of adenoma-like adenocarcinoma cases in colon, and also presents a comprehensive comparison with adenocarcinoma, NOS cases. Carcinomas of other special types were not included in this study. Adenoma-like adenocarcinoma cases in our cohort were identified in an older population than adenocarcinoma, NOS (p=0.009), were significantly associated with a better RFS compared to adenocarcinoma, NOS and, were inversely associated with known pathologic poor prognosticators in colorectal carcinoma such as LVI, PNI, tumor deposits, lymph node metastasis, higher pathologic tumor stage, TB[8,17-20] and immature/myxoid desmoplastic reaction[13,14,21]. Adenoma-like adenocarcinomas also showed higher incidence of intraepithelial tumor infiltrating lymphocytes than adenocarcinomas, NOS (p=0.006). However, the presence of adenoma-like adenocarcinoma morphology was not significantly associated with RFS on multivariate analysis. Nonetheless, these findings validate a separate category of colorectal adenocarcinoma for these cases, as has been incorporated by the 5th edition of the WHO Classification of Tumours (Digestive System)[1].
In our study, adenoma-like adenocarcinoma was not only significantly associated with absence of established poor pathologic prognosticators compared to adenocarcinoma, NOS but also with novel and newly recognized prognostic parameters such as TB, immature/myxoid desmoplastic reaction, and iTILs. TB is a relatively recently established poor pathologic prognosticator in colorectal carcinoma[8,11,17,20,22]. Similarly, the association of immature/myxoid desmoplastic reaction has been associated with a poor prognosis in colorectal cancer[11,13-15,21]. Adenoma-like adenocarcinoma cases were associated with absence of TB and immature/myxoid desmoplastic reaction, with none of the pure adenoma-like adenocarcinoma showed any evidence of these features. Contrary to TB and immature/myxoid desmoplastic reaction, TILs have been reported to be associated with a better RFS[11,23-25]. Adenoma-like adenocarcinoma cases in our cohort were significantly associated with iTILs compared to the adenocarcinoma, NOS cases (p = 0.006). Additionally, iTILs were significantly associated with RFS on univariate as well as multivariate analysis.
The WHO has suggested a cutoff of 50% of adenoma-like morphology to be present in the tumor, to be classified as adenoma-like adenocarcinoma[1]. In the initial description of these cases by Yao T. et al.[3] cases were defined macroscopically by having 80% or more of the tumor surface composed by a "shaggy" or "velvety structure," and no microscopic percentage cutoff was used. Later on, Loy T.S. and Kaplan P.A. in their study assigned a score to the cases depending on the percentage of villous architecture: 1 - 25% (score 1), 26 - 50% (score 2), 51 - 75% (score 3) and 76 - 100% (score 4); cases with score 3 and 4 were designated as villous adenocarcinomas[5]. In their cohort, 1 patient out of 35 died due to disease; however, this patient presented with multiple synchronous colorectal adenocarcinomas. Unfortunately, recurrence-free survival information was not provided.
In our study, the adenoma-like adenocarcinoma cases were divided into pure (100% adenoma-like morphology) and mixed (≥50% but less than 100% adenoma-like morphology). Pure adenoma-like adenocarcinoma had a significantly better survival compared to mixed adenoma-like adenocarcinoma, and adenocarcinoma, NOS, combined (p = 0.03). Using a three-tier system, a trend in differences of survival is seen between the three groups (p = 0.076), with the mixed group showing intermediate prognosis. The statistical significance is probably not reached due to the low number of mixed cases. Nonetheless, this raises the question whether mixed cases should be classified into the adenoma-like adenocarcinoma subgroup, or should be left in the adenocarcinoma, NOS category. If mixed cases are to be considered in adenoma-like adenocarcinoma, this also raises the question whether 50% is an appropriate cutoff. Our result suggests a 70% cutoff using the macro %findcut method, however, some limitations of this method includes a low number of mixed cases, and a range of 1-99% in all cases should be used for obtaining an unbiased percentage cutoff. Our study did not incorporate interobserver reproducibility since the cases were reviewed by two pathologists simultaneously in a multiheader microscope, and the percentages assigned for the purposes of this study was determined by consensus. While it was not difficult to identify the adenoma-like adenocarcinoma components in the resection specimens, we can foresee that in clinical practice, the percentage determination and appropriate classification might pose significant challenges for mixed adenoma-like adenocarcinoma cases.
In our series we noted a much higher prevalence of this entity compared to what was reported by Gonzalez R.S. et al[6] of 3.5%. This could be due to differences in racial prevalence (84% of our patient population is Caucasian, most of the remainder being African American, with only 3% belonging to other races). The racial profile may be significantly different in other hospitals, but an objective comparison cannot be made since the racial distribution of the previous studies were not reported. The higher prevalence can also be related to exclusion of stage IV and rectal adenocarcinoma cases from our cohort. Since adenoma-like adenocarcinoma cases in our cohort had a lower pT and pN stage, and significantly lower LVI, PNI, tumor deposits, and TB compared to adenocarcinoma, NOS, including stage IV cases would most likely have lowered the percentage of these cases. The incidence of this group of tumors involving the rectum is unfortunately not known. Adenoma-like adenocarcinoma in our cohort was more commonly located in the right colon (63%). Similar findings were seen in Yao T. et al.[3] and Gonzalez R.S. et al.[6], as 58% and 63% of their cases were located in the right colon, respectively. In contrast, 61% of the cases in the study by Loy T.S. and Kaplan P.A. were located in the sigmoid colon or rectum[5]. Combining our current cohort with the prior reported cases, a total of 181 cases has been reported of which 101 cases (55.8%) were located in the right colon[3-6]. Therefore excluding rectal adenocarcinomas from our cohort could also explain a higher percentage of this special group of tumors.
In summary, we present the largest cohort of adenoma-like adenocarcinoma cases and provide a comprehensive comparison to a large cohort of adenocarcinoma, NOS cases. Adenoma-like adenocarcinoma cases were significantly associated with a better RFS compared to adenocarcinoma, NOS cases, and showed significantly less association with poor pathologic prognosticators. These tumors, on the other hand, were associated with iTILs, which is an independent good prognosticator on multivariate analysis in this cohort. Our study also highlights a better survival for pure adenoma-like adenocarcinoma compared to mixed adenoma-like adenocarcinoma.
| Univariate Analysis |
Multivariate Analysis | ||
|---|---|---|---|
| P | HR (95% CI) | P | |
| Age | 0.081 | 1.032 (1.011 – 1.053) | 0.003 |
| Gender | 0.190 | ||
| Tumor location | 0.984 | ||
| Adenoma-like adenocarcinoma | 0.035 | ||
| Tumor budding | <0.001 | ||
| Intraepithelial tumor infiltrating lymphocytes | 0.002 | 0.010 | |
| Yes | 1 | ||
| No | 2.332 (1.221 – 4.453) | ||
| Lymphovascular invasion | <0.001 | ||
| Perineural invasion | 0.006 | ||
| Tumor deposits | <0.001 | ||
| Stromal reaction | <0.001 | <0.001 | |
| Immature/myxoid | 3.085 (1.629 – 5.843) | ||
| Intermediate | 0.932 (0.466 – 1.864) | ||
| Mature | 1 | ||
| MMR expression | 0.744 | ||
| pT-stage | <0.001 | ||
| pN-stage | <0.001 | ||
| pTNM prognostic stage group | <0.001 | <0.001 | |
| I | 0.118 (0.035 – 0.394) | ||
| II | 0.349 (0.196 – 0.622) | ||
| III | 1 | ||
Abbreviations: HR - Hazard ratio; CI - Confidence interval
Acknowledgment:
This work was supported by the Department of Pathology and Immunology, Washington University School of Medicine; REDCap is supported by Clinical and Translational Science Award (CTSA) Grant [UL1 TR000448].
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosure:
None of the authors have any relationship with, or financial interest in, any commercial companies pertaining to this article.
This project was partially presented at the United States and Canadian Academy of Pathology (USCAP) 2020 annual meeting as a poster presentation (Abstract #1400). This manuscript is not under consideration for publication elsewhere. We appreciate your review of our work and look forward to hearing from you.
REFERENCES
- [1].WHO Classification of Tumours. Digestive System Tumours. Fifth Edit. Lyon, France: WHO Classification of Tumours; 2019. [Google Scholar]
- [2].Rubio CA, Yanagisawa A, Kato Y. Histologic phenotypes of colonic carcinoma in Sweden and in Japan. Anticancer Res 1998;18:2649–56. [PubMed] [Google Scholar]
- [3].Yao T, Kajiwara M, Kouzuki T, Iwashita A, Tsuneyoshi M. Villous tumor of the colon and rectum with special reference to roles of p53 and bcl-2 in adenoma-carcinoma sequence. Pathol Int 1999;49:374–82. 10.1046/j.1440-1827.1999.00881.x. [DOI] [PubMed] [Google Scholar]
- [4].Palazzo JP, Edmonston TB, Chaille-Arnold LM, Burkholder S. Invasive Papillary Adenocarcinoma of the Colon. Hum Pathol 2002;33:372–5. 10.1053/hupa.2002.32228. [DOI] [PubMed] [Google Scholar]
- [5].Loy TS, Kaplan PA. Villous Adenocarcinoma of the Colon and Rectum: A Clinicopathologic Study of 36 Cases. Am J Surg Pathol 2004;28:1460–5. 10.1097/01.pas.0000141394.64707.02. [DOI] [PubMed] [Google Scholar]
- [6].Gonzalez RS, Cates JMM, Washington MK, Beauchamp D, Coffey RJ, Shi C. Adenoma-Like Adnocarcinoma: A Subtype of Colorectal Carcinoma with Good Prognosis, Deceptive Appearance on Biopsy, and Frequent KRAS Mutation. Histopathology 2016;68:183–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Amin MB, Edge S, Greene F, Byrd DR, Brookland RK, Washington MK, et al. , editors. AJCC Cancer Staging Manual. 8th Editio. Switzerland: Springer International Publishing; 2017. [Google Scholar]
- [8].Lugli A, Kirsch R, Ajioka Y, Bosman F, Cathomas G, Dawson H, et al. Recommendations for reporting tumor budding in colorectal cancer based on the International Tumor Budding Consensus Conference (ITBCC) 2016. Mod Pathol 2017;30:1299–311. 10.1038/modpathol.2017.46. [DOI] [PubMed] [Google Scholar]
- [9].González I, Goyal B, Xia MD, Pai RK, Ma C. DNA mismatch repair deficiency but not ARID1A loss is associated with prognosis in small intestinal adenocarcinoma. Hum Pathol 2019;85:18–26. 10.1016/j.humpath.2018.10.013. [DOI] [PubMed] [Google Scholar]
- [10].Royal College of Pathologists of Australasia. Colorectal cancer structured reporting protocol 2016. www.rcpa.edu.au/Library/Practising-Pathology/Structured-Pathology-Reporting-of-Cancer/Cancer-Protocols/Gastrointestinal/Protocol-colorectal-cancer (accessed March 17, 2020).
- [11].González IA, Bauer PS, Liu J, Chatterjee D. Intraepithelial Tumor Infiltrating Lymphocytes are Associated with Absence of Tumor Budding and Immature/Myxoid Desmoplastic Reaction, and with Better Recurrence-Free Survival in Stage I-III Colorectal Cancer. Histopathology 2020. 10.1111/his.14211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].González I, Bauer PS, Chapman WC, Alipour Z, Rais R. Clinicopathologic determinants of pathologic treatment response in neoadjuvant treated rectal adenocarcinoma. Ann Diagn Pathol 2020;45:151452. 10.1016/j.anndiagpath.2019.151452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Ueno H, Kanemitsu Y, Sekine S, Ishiguro M, Ito E, Hashiguchi Y, et al. Desmoplastic Pattern at the Tumor Front Defines Poor-Prognosis Subtypes of Colorectal Cancer. Am J Surg Pathol 2017;41:1506–12. 10.1155/2014/828131. [DOI] [PubMed] [Google Scholar]
- [14].Ueno H, Konishi T, Ishikawa Y, Shimazaki H, Ueno M, Aosasa S, et al. Histologic Categorization of Fibrotic Cancer Stroma in the Primary Tumor is an Independent Prognostic Index in Resectable Colorectal Liver Metastasis. Am J Surg Pathol 2014;38:1380–6. 10.1097/PAS.0000000000000232. [DOI] [PubMed] [Google Scholar]
- [15].Ueno H, Jones AM, Wilkinson KH, Jass JR, Talbot IC. Histological categorisation of fibrotic cancer stroma in advanced rectal cancer. Gut 2004;53:581–6. 10.1136/gut.2003.028365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Contal C, O’Quigley J. An application of changepoint methods in studying the effect of age on survival in breast cancer. Comput Stat Data Anal 1999;30:253–70. [Google Scholar]
- [17].Graham RP, Vierkant RA, Tillmans LS, Wang AH, Laird PW, Weisenberger DJ, et al. Tumor Budding in Colorectal Carcinoma: Confirmation of Prognostic Significance and Histologic Cutoff in a Population-based Cohort. Am J Surg Pathol 2015;39:1340–6. 10.1097/PAS.0000000000000504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Landau MA, Zhu B, Akwuole FN, Pai RK. Site-specific Differences in Colonic Adenocarcinoma: KRAS Mutations and High Tumor Budding Are More Frequent in Cecal Adenocarcinoma. Am J Surg Pathol 2018;42:351–8. 10.1097/PAS.0000000000001004. [DOI] [PubMed] [Google Scholar]
- [19].Petrelli F, Pezzicam E, Cabiddu M, Coinu A, Borgonovo K, Ghilardi M, et al. Tumor Budding and Survival in Stage II Colorectal Cancer: a Systematic Review and Pooled Analysis. J Gastrointest Cancer 2015;46:212–8. [DOI] [PubMed] [Google Scholar]
- [20].Wang LM, Kevans D, Mulcahy H, O’Sullivan J, Fennelly D, Hyland J, et al. Tumor Budding is a Strong and Reproducible Prognostic Marker in T3N0 Colorectal Cancer. Am J Surg Pathol 2009;33:134–41. 10.1097/PAS.0b013e318184cd55. [DOI] [PubMed] [Google Scholar]
- [21].Ueno H, Jones A, Jass JR, Talbot IC. Clinicopathological significance of the “keloid-like” collagen and myxoid stroma in advanced rectal cancer. Histopathology 2002;40:327–34. 10.1046/j.1365-2559.2002.01376.x. [DOI] [PubMed] [Google Scholar]
- [22].Eriksen AC, Sørensen FB, Lindebjerg J, Hager H, dePont Christensen R, Kjær-Frifeldt S, et al. The Prognostic Value of Tumor-Infiltrating lymphocytes in Stage II Colon Cancer. A Nationwide Population-Based Study. Transl Oncol 2018;11:979–87. 10.1016/j.tranon.2018.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Berntsson J, Svensson MC, Leandersson K, Nodin B, Micke P, Larsson AH, et al. The clinical impact of tumour-infiltrating lymphocytes in colorectal cancer differs by anatomical subsite: A cohort study. Int J Cancer 2017;141:1654–66. 10.1002/ijc.30869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Markl B, Wieberneit J, Kretsinger H, Mayr P, Anthuber M, Arnholdt HM, et al. Number of Intratumoral T Lymphocytes Is Associated With Lymph Node Size, Lymph Node Harvest, and Outcome in Node-Negative Colon Cancer. Am J Clin Pathol 2016;145:826–36. 10.1093/AJCP/AQW074. [DOI] [PubMed] [Google Scholar]
- [25].Turksma AW, Coupé VMH, Shamier MC, Lam KLH, De Weger VA, Belien JAM, et al. Extent and Location of Tumor-Infiltrating Lymphocytes in Microsatellite-Stable Colon Cancer Predict Outcome to Adjuvant Active Specific Immunotherapy. Clin Cancer Res 2016;22:346–56. 10.1158/1078-0432.CCR-13-2462. [DOI] [PubMed] [Google Scholar]