INTRODUCTION. KEY WORDS: squamous cell carcinoma, lymph node metastasis, tongue, tumor budding, tumor depth

Transcription

1 ORIGINAL ARTICLE Histologic assessment of tumor budding in preoperative biopsies to predict nodal metastasis in squamous cell carcinoma of the tongue and floor of the mouth Mai Seki, DMD, 1,2 * Takaaki Sano, MD, PhD, 1 Satoshi Yokoo, DMD, DMSc, 2 Tetsunari Oyama, MD, PhD 1 1 Department of Diagnostic Pathology, Gunma University Graduate School of Medicine, Maebashi, Gunma, Japan, 2 Department of Stomatology and Maxillofacial Surgery, Gunma University Graduate School of Medicine, Maebashi, Gunma, Japan. Accepted 12 September 2015 Published online 23 November 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI /hed ABSTRACT: Background. In squamous cell carcinoma (SCC) of the tongue and the floor of the mouth (FOM), it is important to predict lymph node metastasis, including occult metastasis, before operating. The purpose of this study was for us to determine practical histopathologic parameters as predictive factors for lymph node metastasis in preoperative SCC biopsy specimens. Methods. We examined 91 cases of SCC for conventional histopathologic assessment and a new factor, tumor budding, and their relationship with lymph node metastasis. Results. Significant factors via univariate analysis (p <.01) were budding (score 3) and tumor depth (3 mm) and these were associated with lymph node metastasis. Moreover, both budding and tumor depth significantly correlated with relapse-free survival; however, evaluating biopsy specimens often proved inaccurate for predicting true tumor depth of cancer invasion. Conclusion. Tumor budding using immunohistochemistry for cytokeratin should be added to routine histologic assessments as a new criterion factoring into the decision as to whether neck dissection is indicated. VC 2015 Wiley Periodicals, Inc. Head Neck 38: E1582 E1590, 2016 KEY WORDS: squamous cell carcinoma, lymph node metastasis, tongue, tumor budding, tumor depth INTRODUCTION In squamous cell carcinoma (SCC) of the tongue and the floor of mouth (FOM), 30% to 40% of all patients eventually develop detectable lymphatic metastasis at the time of the first operation, and patients with no clinical lymph node metastases do not always portend a good prognosis, as evidence shows that 20% to 40% of cases already have occult metastasis at presentation. 1 3 It is difficult to predict lymph node metastasis, including occult metastases, before tumor resection and neck dissection, but careful histopathologic evaluation of SCC biopsy specimens for differentiation and mode of invasion is a commonly used method. Grading the mode of invasion, which is defined by the infiltrative growth pattern at the tumor invasion front, is known as the modified Jakobsson criteria. 4,5 Infiltrative growth patterns are classified into 4 groups: grade 1, well-defined borderlines; grade 2, cords with less marked borderlines; grade 3, groups of cells with no distinct borderlines; and grade 4, diffuse invasion. Grade 4 is further divided into cord-like type and widespreadinvasion (The Yamamoto-Kohama mode of invasion). 6 In the aforementioned article, grade 1 is specific for oral SCC, as it is not present in other organs, and grades 1 *Corresponding author: M. Seki, Department of Diagnostic Pathology, Gunma University Graduate School of Medicine, Showa-machi, Maebashi, Gunma , Japan. m @gunma-u.ac.jp and 2 correspond to infiltrative pattern (INF)a, whereas grades 3 and 4 correspond to INFb and INFc, respectively. Many investigators have recommended elective neck dissection because of the high incidence of occult metastases in patients with clinically node-negative SCC and the poor survival rates of patients who undergo neck dissection at a period of time after the initial surgery. 7 Recently, it is recommended that supraomohyoid neck dissection should only be performed for selected patients with high risk of lymph node metastasis. Supraomohyoid neck dissection is known to occasionally detect occult metastases in the dissected lymph nodes of patients with clinically node-negative SCC. 8 If it is not possible to predict lymph node metastasis and to stratify high-risk patients for lymph node metastasis using a preoperative biopsy specimen, prophylactic supraomohyoid neck dissection should be performed only in selected patients at the same time as the primary tumor resection. Tumor budding is defined as the presence of either isolated single cells or small-cell clusters scattered in the stroma ahead of the invasive tumor front. 9,10 Tumor budding describes 5 or fewer cancer cells lacking gland-like structures or pearl formation. 9,10 Tumor budding may represent malignant features, such as cellular discohesion and active invasion. The presence of tumor budding is considered a characteristic of aggressive cancer. According to many studies, grading based on tumor budding is more valuable than conventional histologic grade, infiltrative growth pattern, E1582 HEAD & NECK DOI /HED APRIL 2016

2 TUMOR BUDDING IN SCC OF THE TONGUE TABLE 1. Clinical summary of included patients. Operation (no. of patients) Pathologic lymph node metastasis at first operation, no. (%) Late lymph node metastasis after first operation, no. (%) Died of cancer, no. (%) Tumor resection 1 neck dissection (33) Supraomohyoid neck dissection: prophylactic neck dissection (23) 5 (22) 2 (9) MRND: due to clinical lymph node metastasis (10) 10 (100) 0 (0) Tumor resection only (58) 9 (16) 1 (2) Total (91) 15 (16) 9 (10) 3 (3) Abbreviation: MRND, modified radical neck dissection. and lymphovascular invasion to predict the metastatic potential of many cancers. 9,10 However, few studies have ascertained whether tumor budding is valuable as a predictive factor of prognosis or lymph node metastasis in oral SCC In the present study, we aimed to detect practical histopathologic parameters that are predictive for lymph node metastasis in preoperative biopsy SCC specimens of the tongue and FOM. Next, we evaluated conventional histopathologic assessment and a new factor, tumor budding, and their relationship with lymph node metastasis. MATERIALS AND METHODS This retrospective study was performed in the Gunma University Hospital of consecutively tumor-resected patients diagnosed with SCC of the tongue and FOM between 2009 and Among a total of 107 patients, including both cn0 and clinically node-positive patients, we analyzed 91 cases of the tongue (n 5 81) and FOM (n 5 10) for which preoperative biopsies that could be evaluated histopathologically for invasive front of tumors were performed within the hospital. A summary of all patients is shown in Table 1. The mean age of the patients was 66 years (range, years). Fifty-two patients were men and 39 were women. Fortyseven patients received preoperative chemotherapy and 44 patients had no chemotherapy. Thirty-three patients were initially treated with local excision of the tumor and neck dissection. Patients with clinically detectable lymph node metastasis at the time of initial presentation (n 5 10) underwent modified radical neck dissection (MRND) of level I to level V with preservation of nonlymphatic structures. These 10 patients were also positive for lymph node metastasis by histopathologic examination. In addition, 23 patients who showed histopathologic INFc on preoperative biopsy underwent supraomohyoid neck dissection at level I, level II, and level III as they were at high risk of lymph node metastasis despite not having clinically detectable lymph node metastasis. 7 In 5 of these 23 patients (22%) with prophylactic supraomohyoid neck dissection, lymph node metastasis was positive via histopathologic examination. The remaining 58 patients who were clinically node-negative were initially treated surgically with local excision of the tumor, including a sufficient safety margin without neck dissection. The postoperative observation period ranged from 4 months to 5 years. Nine of the 58 patients had late occurring lymph node metastasis in the neck region within the first 2 years after operation, and these patients subsequently underwent MRND and positive lymph node metastasis was confirmed by histopathologic examination. Of the total 91 patients, 3 died with accompanying neck recurrence during follow-up. Informed consent was obtained from the patients to use their remaining surgical specimens after pathologic diagnosis for this investigational study, which was approved by the Institutional Review Board for Clinical Research at Gunma University Hospital. Histopathologic variables All histopathologic variables were evaluated from biopsy specimens, and both budding score and tumor depth were also evaluated using a part of the resection specimens for reference. According to the World Health Organization criteria, tumors <2 cm in the greatest dimension were defined as T1, whereas tumors >2 cm were defined as T2 or greater. 15 Histologic grading of SCC differentiation was divided into 3 groups: (1) welldifferentiated (grade 1); (2) moderately differentiated (grade 2); and (3) poorly differentiated (grade 3). 15 Cancer growth characteristics and mode of invasion was classified using the INF classification. 16 With respect to tumor INFs into the surrounding tissue, INFa is defined as a tumor that displays expanding growth with a distinct FIGURE 1. Microscopic representative figure for tumor budding measurements in squamous cell carcinoma with hematoxylineosin staining. The circles represent an approximate field of a single 320 objective lens along the edge of the invasive tumor front. Tumor budding was counted in a field in which budding foci were most intense within a biopsy specimen. HEAD & NECK DOI /HED APRIL 2016 E1583

3 SEKI ET AL. FIGURE 2. Microscopic findings of tumor budding in squamous cell carcinoma (SCC) with hematoxylin-eosin staining (A, B, and C) and immunostaining for cytokeratin (D, E, and F). Tumor budding (arrow heads) in the invasive front of the SCC are clearly shown in the immunohistochemistry stained slides (D, E, and F). Three examples of tumor budding in infiltrative pattern (INF)a cases (A and D), INFb cases (B and E), and INFc cases (C and F). border from the surrounding tissue, INFb is defined as a tumor that shows an intermediate pattern between INFa and INFc, whereas INFc is defined as a tumor that displays infiltrative growth with no distinct border with the surrounding tissue. Tumor depth of invasion (mm) was measured from the surface of the tumor to the deepest point of an invasive tumor in all biopsies and parts of resected specimens. Cancer cells within a D2-40 positive vessel structure were considered lymphatic invasion. Lymphatic invasion was positive if one or more foci were present, and 0 if they were absent. Elastic fibers, includ- ing the elastic lamina of blood vessel walls, were stained by Victoria blue. Cancer cells within Victoria bluepositive vascular structures were considered blood vessel TABLE 2. Positive and negative predictive values of different cutoff points for tumor budding and tumor depth. Cutoff value Budding, score Depth, mm Lymph node metastasis 1/ND 1 PPV (%) Lymph node metastasis 2/ND 2 NPV (%) 15/29 15/26 15/24 11/16 8/12 6/9 4/7 4/ /4 7/7 9/9 13/17 14/21 15/24 15/26 15/ /22 6/9 4/ /11 15/24 15/ Abbreviations: lymph node metastasis 1/2, lymph node metastasis positive/negative; ND 1, cases subjected to nodal dissection; PPV, positive predictive value; ND2, cases not subjected to nodal dissection; NPV, negative predictive value. E1584 HEAD & NECK DOI /HED APRIL 2016 FIGURE 3. The results of receiver operating characteristic (ROC) curves is displayed with variable cutoff numbers for budding score and tumor depth (mm) as a predictive factor for lymph node metastasis in 33 resected cases. Tumor budding 3 or tumor depth 3 mm were the best cutoff values as 100% sensitivity was obtained, although the highest specificity was not seen at this cutoff point.

4 TUMOR BUDDING IN SCC OF THE TONGUE TABLE 3. Correlation of lymph node metastasis in 33 cases with clinical and pathologic variables. Lymph node metastasis Negative (18) Positive (15) Univariate analysis (p <.01) Adjusted OR 95% CI Age, y < Sex Male Female 7 4 Location Tongue FOM 1 4 Tumor size T T2/3/ Preoperative chemotherapy Tumor grade / Tumor depth, mm < < INF a/b c Lymphatic invasion Blood vessel invasion Budding score <3 9 0 < Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; FOM, floor of the mouth; INF, infiltrative pattern. invasion. Blood vessel invasion was positive if one or more foci were present, and 0 if they were absent. Tumor budding describes a cancer cell cluster comprising fewer than 5 cells or isolated single cells without a distinct structure and appearing as a bud near a large cancer nest (Figures 1 and 2). 9,10 The number of tumor buds was evaluated using a slide immunostained for pancytokeratin (AE1/AE3) in all biopsies and some of the resection specimens. The field with the greatest number of budding foci was selected within a slide, and the number of budding foci was then counted in 1 field using a 320 objective lens; budding was assessed along the edge of the tumor invasion front and reported as a budding score for each case (see Figure 1). Immunohistochemistry All sections were deparaffinized using xylene and rehydrated ethanol, and treated with 0.3% hydrogen peroxidase in methanol for 30 minutes to block endogenous peroxidase activity. As primary antibodies, mouse monoclonal anti-d2-40 antibody (D2-40, 1:200 dilution; DAKO, Kyoto, Japan), and mouse monoclonal anticytokeratin antibody (AE1/AE3, 1:100 dilution; DAKO) were used. Antigen retrieval was performed by boiling for 30 minutes in 0.01 mol/l sodium citrate buffer (ph 6.0) for cytokeratin (AE1/AE3) or by boiling for 30 minutes in 0.01 mol/l sodium phosphate-citrate buffer (ph 8.0) for D2-40. The slides were incubated with primary antibodies at 48C overnight and then reacted with a Simple Stain MAX-PO MULTI (Nichirei Bioscience) for 30 minutes at room temperature. Specific antigen antibody reactions were visualized with 0.2% diaminobenzidine tetrahydrochloride and hydrogen peroxide. The sections were counterstained using Mayer s hematoxylin. Statistical analysis In the univariate analysis using the chi-square test, p <.01 was considered statistically significant; for all other statistical analyses, including the Kaplan Meier method and the logrank test, p <.05 was considered statistically significant. RESULTS Determination of cutoff values for tumor depth and tumor budding First, we examined the relationships between all clinicopathologic findings in preoperative biopsy specimens HEAD & NECK DOI /HED APRIL 2016 E1585

5 SEKI ET AL. FIGURE 4. Scatter plot of the budding score (A) and tumor depth (B) values between biopsy and resected specimens in each patient without preoperative chemotherapy. and lymph node metastasis in 33 cases, of which 15 cases were positive for lymph node metastasis via histologic examination of specimens for neck dissection at the first operation. In the 33 specimens, tumor depth (mm) and budding score were counted, and then the positive predictive value (PPV) and negative predictive value (NPV), and a receiver operator characteristic (ROC) curve with variable cutoff numbers for tumor depth (area under the curve ; 95% confidence interval [CI] ) and budding score (area under the curve ; 95% CI ) for lymph node metastasis were measured and displayed (Table 2 and Figure 3). As lymph node metastasis in oral SCC after first tumor resection without neck dissection is considered to lead to a lethal outcome, and total neck dissection is a highly invasive treatment, we thought that the predictive factors for determining whether to perform a lymph node dissection should have 100% of NPV and as high a PPV as possible. From this perspective, a cutoff point of budding 3 andadepth 3 mm offered 100% NPV, and had a relatively high value for PPV (Table 2 and Figure 3). Moreover, we found that when the tumor budding 3 and the depth 3 mm criteria were used, 100% sensitivity and high specificity for lymph node metastasis were obtained in the ROC curve (see Figure 3). Thus, we used tumor budding 3 and depth 3 mm as cutoff values in biopsy specimens for subsequent analyses. Relationships between clinicopathologic variables and lymph node metastasis in preoperative biopsies We set cutoff values for tumor size (1, or 2, 3, and 4), tumor differentiation grade (1, or 2, and 3), tumor depth (<3 or3 FIGURE 5. Histogram of the distribution of tumor budding score and infiltrative pattern (INF) status on lymph node metastasis (LNM) and late lymph node metastasis in a total of 91 biopsy specimens. The open circle, the closed yellow circle, and the closed blue circle correspond to 1 case negative for lymph node metastasis, 1 case positive for lymph node metastasis, and 1 case of late lymph node metastasis, respectively. E1586 HEAD & NECK DOI /HED APRIL 2016

6 TUMOR BUDDING IN SCC OF THE TONGUE FIGURE 6. Histogram of the distribution of tumor budding score and tumor depth status on lymph node metastasis (LNM) and late lymph node metastasis in a total of 91 biopsy specimens. The open circle, the closed yellow circle, and the closed blue circle correspond to 1 case negative for lymph node metastasis, 1 case positive for lymph node metastasis, and 1 case of late lymph node metastasis, respectively. mm), INF (a, b, or c), and budding score (<3 or3), and all variables were assessed used the chi-square test for a potential relationship with lymph node metastasis. The relationships between all clinicopathologic findings in preoperative biopsy specimens and lymph node metastasis are indicated in Table 3. Significant factors via univariate analysis (p <.01) in the 33 cases were depth and tumor budding, whereas the remaining factors were not associated with lymph node metastasis (Table 3). Correlations (p <.05) between tumor size and depth, and depth and budding were present. Comparison of tumor depth and tumor budding between biopsies and resected specimens For the 44 cases without chemotherapy before resection, we compared the budding score and tumor depth (mm) of the biopsies with resected specimens from each patient (Figures 4A and 4B). For budding score, a good correlation (r ) between biopsies and resected specimens was observed and each value in both specimens in the same patient was nearly equal (Figure 4A). By contrast, it was found that values for tumor depth (mm) measured on biopsies and resected specimens were substantially different, although some correlation (r ) was observed (Figure 4B). Comparison between tumor budding and invasion depth or infiltrative pattern status on lymph node metastasis Tumor budding score and INF or tumor depth in lymph node metastasis status is summarized as a scatter plot in Figures 5 and 6. It is clearly shown in Figures 5 and 6 that all lymph node metastasis cases and cases with late occurring lymph node metastasis during the follow-up period had tumor budding 3 on biopsy specimens. It should be noted that many cases with tumor budding 3 even in the INFa or INFb group (11 of 14) showed lymph node metastasis or late lymph node metastasis (see Figure 5), and that a few cases with tumor budding 3 but with tumor depth <3 mm (4 of 12) were late occurring lymph node metastasis (see Figure 6). FIGURE 7. Kaplan Meier plots of overall survival in all 91 patients according to tumor budding (<3 or 3) (A) and tumor depth (<3 or 3 mm) (B) statuses. The difference in overall survival rates were both statistically significant (p <.05) based on tumor budding and tumor depth. HEAD & NECK DOI /HED APRIL 2016 E1587

7 SEKI ET AL. FIGURE 8. Cases without neck dissection at first tumor resection (58 cases) and those that had relapse-free survival throughout the follow-up period according to tumor budding (<3 or3) (A) and tumor depth (<3 or3 mm) (B) statuses. The differences in relapse-free survival rates were both statistically significant (p <.05) based on tumor budding and tumor depth. Tumor budding and tumor depth as a predictive value for overall and relapse-free survival Three of the 91 patients died of lymph node metastasis and no patients died of other diseases in this study. With respect to using tumor budding and tumor depth to predict patient prognosis, Kaplan Meier plots of overall survival in all patients according to tumor budding status (<3 or3; Figure 7A) and tumor depth (<3 or3 mm; Figure 7B) were generated. The difference in survival rates were statistically significant (p <.05) according to tumor budding and tumor depth. Finally, among only cases without initial lymph node dissection (58 cases), we predicted relapse-free survival according to tumor budding (<3 or3; Figure 8A) and tumor depth (<3 or 3 mm; Figure 8B) status during follow-up to determine if these factors could be predictive for lymph node metastasis. Only budding was significant on univariate analysis (p <.01) in the 58 cases, whereas depth was not significant (Table 4). Moreover, the difference in budding status and tumor depth on relapse-free survival was statistically significant (p <.05; Figures 8A and 8B). Of note, none of the cases with a budding score <3 (n 5 45) had late occurring lymph node metastasis, and all cases with late occurring lymph node metastasis had a budding score 3, and these cases developed late lymph node metastasis within 20 months of local excision (Figure 8A). DISCUSSION Previous reports have shown that the incidence of late cervical lymph node metastasis is 14% to 26% after initial treatment within 2 to 5 years for patients with clinically negative cervical lymph nodes Therapeutic strategies for oral SCC treatment in our institute, including deciding whether neck dissection is indicated, depends on clinical findings and histopathologic findings of biopsy specimens. Briefly, in our institutions, cases with no cervical lymph node metastases with INFa or INFb infiltrative growth patterns on biopsy specimens undergo primary tumor resection only, and all cases that have clinically evident cervical lymph node metastases also undergo MRND. In cases classified as INFc, except cases in which it is difficult to perform supraomohyoid neck dissection owing to problems related to a patient s general medical condition, tumor resection and supraomohyoid neck dissection are performed as a precaution, according to previously reported guidelines. 9 In our experience, however, some cases classified as INFa or INFb have late occurring lymph node metastasis during the early period after tumor resection. Thus, we believe that the criteria for supraomohyoid neck dissection based on infiltrative growth pattern as a predictive factor for lymph node metastasis used in our institution may not be sufficiently specific or sensitive. Therefore, there is a need to detect cases with poor prognoses lurking in welldifferentiated carcinomas and classified as INFa and INFb. Sentinel lymph node biopsy (SLNB) may represent an alternative to elective neck dissection for the staging of patients with early SCC of the oral cavity. 21 It can be used as the sole staging tool for the majority of patients. In addition, in our hospitals, SLNB has been performed in a subset of patients with SCC of the tongue. It is possible that combining the SLNB with histologic assessment of tumor budding in preoperative biopsies may improve detection rates and sensitivity for occult lymph node metastasis in clinically node-negative patients. In this study, we evaluated the relationship between lymph node metastasis and many histologic factors to establish the best practical factor predictive for lymph node metastasis in SCC of the tongue and FOM. The present study confirmed the significant impact of tumor depth and tumor budding on lymph node metastasis via univariate analysis. In the tumor budding and tumor depth analyses related to relapse-free survival in 58 patients, all of whom had no clinical lymph node metastasis at presentation and underwent only tumor resection without prophylactic supraomohyoid neck dissection, Kaplan Meier plots in patient groups defined by intensity of tumor budding (budding <3 or3) or tumor depth (<3 or 3 mm) were both statistically significant. E1588 HEAD & NECK DOI /HED APRIL 2016

8 TUMOR BUDDING IN SCC OF THE TONGUE TABLE 4. Correlation of late lymph node metastasis in 58 cases with clinical and pathologic variables. Late lymph node metastasis Negative (49) Positive (9) Univariate analysis (p <.01) Adjusted OR 95% CI Age, y < Sex Male Female 24 4 Location Tongue FOM 5 0 Tumor size T T2/3/ Preoperative chemotherapy Tumor grade / Tumor depth, mm < INF a/b c 5 3 Lymphatic invasion Blood vessel invasion Budding score < < Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; FOM, floor of the mouth; INF, infiltrative pattern. On budding score and tumor depth, a good correlation between biopsy and resected specimen evaluation for tumor budding was observed. Tumor depth measurements with resection specimens were, however, significantly higher than biopsy specimens and, thus, the latter could not be used to accurately determine depth of cancer invasion. As biopsy specimens are usually taken from peripheral sites, which are compatible with a shallow part on the edge of cancer tissue, it is difficult to pick the deepest site of cancer invasion in the center of the tumor. A large number of oral SCC studies correlating tumor depth with lymph node metastasis have been reported, and several cutoff values, such as 2, 3, 4, 5, or 7.5 mm, have been introduced. 11,22 25 These tumor depths seem to correlate well with lymph node metastasis when the assessments are performed on resection specimens, including whole cancer tissue. It is, however, unreasonable to use tumor depth measured on biopsy specimens as a predictor of cervical lymph node metastasis. In colorectal cancer, infiltrative growth pattern is not included as an independent predictive factor for lymph node metastasis because it is associated strongly with several other parameters. 26 Tumor budding is recognized by the latest World Health Organization textbook as an additional prognostic factor in colorectal cancer. 27 Recently, in many types of other carcinomas, such as lung, pancreatic, esophageal, laryngeal, nasopharyngeal, and oral, tumor budding has been discussed 11 14,28 32 and it has been identified as a poor indicator even in some SCCs, such as the tongue, larynx, esophagus, and in lung cancer ,28,30,31 Tumor budding is a cancer growth form representing differentiation and a dissociation of cancer cells that can be regarded as an initial phase of invasion preceding vascular invasion. Compared to tumor budding, infiltrative growth pattern is defined mainly based on features on the border between the cancer and mesenchyme in a low-power light microscopy field without immunohistochemistry, whereas tumor budding should be evaluated at the invasive front under a 320 or 340 objective lens. Considering that tumor buds are very small, because they represent a small cluster of cells or sometimes even a single cell that has separated from a tumor mass, it is sometimes difficult to detect tumor budding using conventional hematoxylineosin stained specimens. Therefore, in addition to hematoxylin-eosin stained slides, investigators should examine tumor budding by keratin immunohistochemistry, 33 which makes these buds much easier to detect. HEAD & NECK DOI /HED APRIL 2016 E1589

9 SEKI ET AL. Our evaluation method for budding score involved using a 320 objective lens along the edge of the invasive tumor front and to count the highest budding foci. However, different scoring methods have been presented previously and they include using a 340 objective lens, and using the average or highest values as the final score It remains to be determined which evaluation methods for determining a budding score should be standardized among pathologists. 34 According to our new criteria, budding 3 should be added to the general histologic panel for pathologists, and prophylactic supraomohyoid neck dissection to prevent lymph node metastasis should be considered when tumor budding exceeds this value. The findings of the present study require verification in other cohorts, including a large number of patients in a large multicenter study. REFERENCES 1. Beenken SW, Krontiras H, Maddox WA, Peters GE, Soong S, Urist MM. T1 and T2 squamous cell carcinoma of the oral tongue: prognostic factors and the role of elective lymph node dissection. Head Neck 1999;21: Pimenta Amaral TM, Da Silva Freire AR, Carvalho AL, Pinto CA, Kowalski LP. Predictive factors of occult metastasis and prognosis of clinical stages I and II squamous cell carcinoma of the tongue and floor of the mouth. Oral Oncol 2004;40: Kelner N, Rodrigues PC, Bufalino A, et al. Activin A immunoexpression as predictor of occult lymph node metastasis and overall survival in oral tongue squamous cell carcinoma. Head Neck 2015;37: Jakobsson PA, Eneroth CM, Killander D, Moberger G, Mårtensson B. Histologic classification and grading of malignancy in carcinoma of the larynx. Acta Radiol Ther Phys Biol 1973;12: Anneroth G, Hansen LS. A methodologic study of histologic classification and grading of malignancy in oral squamous cell carcinoma. Scand J Dent Res 1984;92: Yamamoto E, Miyakawa A, Kohama G. Mode of invasion and lymph node metastasis in squamous cell carcinoma of the oral cavity. Head Neck Surg 1984;6: Umeda M, Yokoo S, Take Y, Omori A, Nakanishi K, Shimada K. Lymph node metastasis in squamous cell carcinoma of the oral cavity: correlation between histologic features and the prevalence of metastasis. Head Neck 1992;14: Thiele OC, Seeberger R, Flechtenmacher C, Hofele C, Freier K. The role of elective supraomohyoidal neck dissection in the treatment of early, node-negative oral squamous cell carcinoma (OSCC): a retrospective analysis of 122 cases. J Craniomaxillofac Surg 2012;40: Ueno H, Mochizuki H, Hashiguchi Y, et al. Risk factor for an adverse outcome in early invasive colorectal carcinoma. Gastroenterology 2004;127: Prall F. Tumour budding in colorectal cancer. Histopathology 2007;50: Xie N, Wang C, Liu X, et al. Tumor budding correlates with occult cervical lymph node metastasis and poor prognosis in clinical early-stage tongue squamous cell carcinoma. J Oral Pathol Med 2015;44: Wang C, Huang H, Huang Z, et al. Tumor budding correlates with poor prognosis and epithelial-mesenchymal transition in tongue squamous cell carcinoma. J Oral Pathol Med 2011;40: Almangush A, Bello IO, Keski S antti H, et al. Depth of invasion, tumor budding, and worst pattern of invasion: prognostic indicators in early-stage oral tongue cancer. Head Neck 2014;36: Marangon Junior H, Rocha VN, Leite CF, de Aguiar MC, Souza PE, Horta MC. Laminin-5 gamma 2 chain expression is associated with intensity of tumor budding and density of stromal myofibroblasts in oral squamous cell carcinoma. J Oral Pathol Med 2014;43: Barnes L, Eveson JW, Reichart P, Sidransky D, editors. World Health Organization Classification of Tumors, Pathology and Genetics of the Head & Neck. Lyon, France: IARC Press; Izumo T, Kirita T, Ariji E, et al. General rules for clinical and pathological studies on oral cancer: a synopsis. Jpn J Clin Oncol 2012;42: Okamoto M, Nishimine M, Kishi M, et al. Prediction of delayed neck metastasis in patients with stage I/II squamous cell carcinoma of the tongue. J Oral Pathol Med 2002;31: Kurokawa H, Yamashita Y, Takeda S, Zhang M, Fukuyama H, Takahashi T. Risk factors for late cervical lymph node metastases in patients with stage I or II carcinoma of the tongue. Head Neck 2002;24: Goto M, Hasegawa Y, Terada A, et al. Prognostic significance of late cervical metastasis and distant failure in patients with stage I and II oral tongue cancers. Oral Oncol 2005;41: Yuasa Nakagawa K, Shibuya H, Yoshimura R, et al. Cervical lymph node metastasis from early-stage squamous cell carcinoma of the oral tongue. Acta Otolaryngol 2013;133: Alkureishi LW, Ross GL, Shoaib T, et al. Sentinel node biopsy in head and neck squamous cell cancer: 5-year follow-up a European multicenter trial. Ann Surg Oncol 2010;17: Balasubramanian D, Ebrahimi A, Gupta R, et al. Tumour thickness as a predictor of nodal metastases in oral cancer: comparison between tongue and floor of mouth subsites. Oral Oncol 2014;50: Gonzalez Moles MA, Esteban F, Rodriguez Archilla A, Ruiz Avila I, Gonzalez Moles S. Importance of tumour thickness measurement in prognosis of tongue cancer. Oral Oncol 2002;38: Kumar T, Patel MD. Pattern of lymphatic metastasis in relation to the depth of tumor in oral tongue cancer: a clinico pathological correlation. Indian J Otolaryngol Head Neck Surg 2013;65(Suppl 1): Wallwork BD, Anderson SR, Coman WB. Squamous cell carcinoma of the floor of the mouth: tumour thickness and the rate of cervical metastasis. ANZ J Surg 2007;77: Akishima Fukasawa Y, Ishikawa Y, Akasaka Y, et al. Histopathological predictors of regional lymph node metastasis at the invasive front in early colorectal cancer. Histopathology 2011;59: Hamilton SR, Nakamura SI, Bosman FT, et al. Carcinoma of the colon and rectum. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, editors. World Health Organization Classification of Tumors, Pathology and Genetics of Tumors of the Digestive System. Lyon, France: IARC Press; pp Masuda R, Kijima H, Imamura N, et al. Tumor budding is a significant indicator of a poor prognosis in lung squamous cell carcinoma patients. Mol Med Rep 2012;6: Karamitopoulou E, Zlobec I, Born D, et al. Tumour budding is a strong and independent prognostic factor in pancreatic cancer. Eur J Cancer 2013;49: Brown M, Sillah K, Griffiths EA, et al. Tumour budding and a low host inflammatory response are associated with a poor prognosis in oesophageal and gastro-oesophageal junction cancers. Histopathology 2010;56: Sarioglu S, Acara C, Akman FC, et al. Tumor budding as a prognostic marker in laryngeal carcinoma. Pathol Res Pract 2010;206: Luo WR, Gao F, Li SY, Yao KT. Tumour budding and the expression of cancer stem cell marker aldehyde dehydrogenase 1 in nasopharyngeal carcinoma. Histopathology 2012;61: Kazama S, Watanabe T, Ajioka Y, Kanazawa T, Nagawa H. Tumour budding at the deepest invasive margin correlates with lymph node metastasis in submucosal colorectal cancer detected by anticytokeratin antibody CAM5.2. Br J Cancer 2006;94: Almangush A, Salo T, Hagstr om J, Leivo I. Tumour budding in head and neck squamous cell carcinoma a systematic review. Histopathology 2014;65: E1590 HEAD & NECK DOI /HED APRIL 2016