Histologic Evaluation of Malignant Polyps and Low-Stage Colorectal Carcinoma

Evaluation of margin status begins with proper gross examination of the specimen. For intact polyps, the resection margin or polyp stalk should be identified and inked. Larger polyps should then be sectioned perpendicular to the resection margin/stalk margin. Transanal excision specimens are ideally received pinned and oriented by the surgeon to allow proper evaluation of all margins. If received oriented, the specimen can be differentially inked and handled similarly to a skin ellipse (Figure 1, A).²⁴  Margin status cannot be reliably evaluated for fragmented specimens, and this should be conveyed in the pathology report.²⁵ 

The status of the margin (positive or negative) should be reported, as well as distance of tumor to margin in negative cases. The true margin of resection is defined as the free edge of the submucosal connective tissue with cautery artifact. Tumor present at the true margin or less than 1 mm from the resection margin is an indication of incomplete excision and a risk factor for tumor recurrence. As such, patients with positive or close margins often require further surgical intervention (Figure 1, B through D).^8,25 

Tumor grade has been regarded as an independent risk factor for nodal involvement. The grading system for CRC is based on the percentage of gland formation. Well-differentiated adenocarcinomas exhibit gland formation in more than 95% of the tumor, moderately differentiated adenocarcinomas in 50% to 95% of the tumor, and poorly differentiated adenocarcinomas exhibit less than 50% gland formation (Figure 2, A and B).²⁶  Poorly differentiated adenocarcinomas account for 5% to 10% of all CRC cases and are associated with a greater incidence of adverse outcome.^{1,8,13–16,19,25} 

Tumor budding should be considered independently from tumor grading and is discussed in more detail below.²⁷  Additionally, the above-mentioned grading system is applicable only to conventional adenocarcinoma. Special histologic variants such as medullary carcinoma and mucinous adenocarcinomas should not be graded according to these criteria. Medullary carcinomas have solid growth but are typically microsatellite unstable and associated with a better prognosis. Histologic grading systems for mucinous adenocarcinomas have not been standardized, but prognosis of these tumors is typically related to microsatellite instability status.^26,28,29 

Several studies* have also shown the presence of lymphovascular invasion to be an independent predictor of lymph node metastases. Small vessel lymphovascular invasion is defined as tumor within an endothelial-lined channel without identifiable surrounding smooth muscle or elastic lamina, and may represent invasion of lymphatic channels, capillaries, or postcapillary venules.²⁶  We do not routinely attempt to differentiate the type of small vessel invasion. Small vessel lymphovascular invasion should be differentiated from retraction artifact (Figure 2, C and D). Retraction artifact is usually seen within the main tumor mass, whereas lymphatic invasion is seen away from the main tumor mass.²⁵  In challenging cases, immunohistochemical stains for D2-40 or CD34 can be used to help differentiate retraction artifact from true lymphovascular invasion.^8,16 

In comparison, large vessel or venous invasion is defined as tumor within an endothelial-lined channel with surrounding smooth muscle or elastic lamina and should be reported separately from small vessel lymphovascular invasion. Although extramural venous invasion is known to be a poor prognostic factor and associated with liver metastases, the significance of intramural venous invasion is uncertain.²⁶ 

Various authors have proposed different criteria to evaluate the extent of submucosal invasion in malignant polyps. Kikuchi et al⁶  proposed a 3-tiered system for sessile polyps, defined as sm1 = slight submucosal invasion from the muscularis mucosa to the depth of 200 to 300 μm, sm2 = intermediate invasion, and sm3 = carcinoma invasion near the inner surface of the muscularis propria. In this study, the incidence of lymph node metastasis was 0%, 5%, and 25% in sm1, sm2, and sm3, respectively, and sm3 was an independent risk factor for lymph node metastasis. Subsequent studies have supported these findings.⁹ 

Haggitt et al⁴  proposed a classification for pedunculated malignant polyps as follows: level 1, carcinoma invading head of polyp; level 2, carcinoma invading to the level of neck of polyp; level 3, carcinoma invading any part of polyp stalk; and level 4, carcinoma invading submucosa of underlying bowel wall below the polyp stalk but above the muscularis propria. Patients with level 4 invasion were significantly more likely to have an adverse outcome.

Practical use of the Kikuchi and Haggitt systems may be difficult in polypectomy specimens, as muscularis propria is not typically present to gauge the level of invasiveness. For this reason, we typically provide a quantitative rather than qualitative depth of invasion. Absolute depth of invasion can usually be easily obtained on properly oriented specimens using a micrometer or digital slide software on scanned slides. In specimens where the muscularis mucosa is intact, invasion is measured from the level of the muscularis mucosa. When the muscularis mucosa is obliterated, invasion can be measured from the surface of the polyp (Figure 3, A and B).⁸  Submucosal invasion 1 mm or deeper shows a strong increase in risk of lymph node metastases (relative risk, 5.2; 95% CI, 1.8–15.4).¹ 

Peritumoral tumor budding is defined as single tumor cells or clusters of 4 or 5 cells at the invasive front of the tumor and should be distinguished from poorly differentiated clusters (clusters of >5 tumor cells) (Figure 3, C).^26–28  It is thought to be a manifestation of epithelial-mesenchymal transformation where cells transition to a more motile and invasive phenotype.³⁰  Several studies have demonstrated the prognostic significance of tumor budding. In pT1 tumors, tumor budding is significantly associated with an increased risk of lymph node metastases, and in stage II CRC it is associated with worse disease-free survival.^11,31–38 

According to College of American Pathologists guidelines for synoptic reporting on CRCs, tumor budding is not a required element but is recommended for all colorectal adenocarcinomas arising in polyps as well as stage I and II CRC cases.²⁶  Recently the International Tumor Budding Consensus Conference issued a standardized scoring system and made recommendations for reporting tumor budding in CRC. According to the guidelines, tumor buds should be counted on hematoxylin-eosin–stained slides. Most outcome data for tumor buds are based on hematoxylin-eosin assessment, and this method is also more cost-effective. Immunohistochemistry for cytokeratin can be used to identify tumor buds obscured by inflammation, but the final count should be performed on hematoxylin-eosin stain only. Tumor budding is assessed in one hot-spot field measuring 0.785 mm² (typically ×20 objective lens) at the invasive front of the tumor. Ten separate fields should be scanned along the invasive front of tumor at medium power (×10 objective) to identify the hot spot. In a malignant polyp there may not be 10 fields of tumor for evaluation, in which case the whole tumor front would be examined. The International Tumor Budding Consensus Conference proposed a 3-tiered grading system for reporting tumor buds: 0 to 4 buds, low budding (Bd1); 5 to 9 buds, intermediate budding (Bd2); and 10 or more buds, high budding (Bd3). Both the tumor bud grade and absolute number of tumor buds should be provided in the pathology report.^26,27  Using this system, intermediate to high budding is associated with lymph node metastasis in pT1 cancers and high tumor budding is associated with an increased risk of recurrence and mortality in stage II cancers (Figure 3, D).^{1,10,11,19–21,39} 

Caution must be taken in assessing tumor buds in certain histologic variants of CRC. Clusters of cells suspended in pools of mucin should not be counted as tumor buds in mucinous and signet-ring cell carcinomas. In medullary carcinoma, discohesion of tumor cells secondary to inflammation may mimic tumor buds, and poorly differentiated clusters of tumor should not be counted in micropapillary carcinomas. In cases where an accurate tumor bud count cannot be performed, the tumor bud count and grade should be reported as “cannot be assessed” with a note explaining the reason. Additionally, tumor budding should not be reported in rectal cancer resections after neoadjuvant therapy, as there are insufficient data on its prognostic significance.²⁷ 

When evaluating malignant colorectal polyps, pathologists can assess several histologic features that are associated with adverse outcome, including margin status, tumor grade, presence of lymphovascular invasion, depth of submucosal invasion, and tumor budding. When this information is provided to our clinical colleagues, patients can appropriately be triaged to undergo additional surgical resection with lymph node dissection as necessary.