You are here

Comparing the eighth and the seventh editions of the American Joint Committee on Cancer staging system and the Brigham and Women's Hospital alternative staging system for cutaneous squamous cell carcinoma: Implications for clinical practice

J Am Acad Dermatol. 2019;80(1):106-113.e2

Background

The new eighth edition of the American Joint Committee on Cancer staging system (AJCC-8) incorporates changes regarding cutaneous squamous cell carcinoma (CSCC).

Objectives

We aimed to compare the AJCC-8 staging system with the previous seventh edition of the AJCC staging system (AJCC-7) and the Brigham and Women's Hospital (BWH) alternative staging system to identify their usefulness and the utility of their risk factors in defining prognostic groups in CSCC.

Methods

A series of 186 CSCCs of the head and neck were retrospectively collected. All 3 staging systems were compared from the standpoint of their ability to predict poor prognosis. Binary logistic regression models were built to determine which risk factors were most relevant.

Results

Poor prognosis was mainly associated with stage T2 of the AJCC-7, with stages T2b/T3 of the BWH system, and with stage T3 of the AJCC-8. The AJCC-8 and the BWH staging systems displayed overlap with each another in predicting poor prognosis, and both were superior to the AJCC-7. The new risk factors incorporated into the AJCC-8 and the poor degree of differentiation were independently associated with poor outcome.

Limitations

Retrospective study and few cases with bone invasion.

Conclusions

The AJCC-8 is more distinctive, monotonous, and homogeneous than the AJCC-7 and shows some overlap with the BWH system in stratification of tumors.

Key words: AJCC, American Joint Committee on Cancer, Brigham and Women's Hospital, cutaneous squamous cell carcinoma, poor degree of differentiation, prognosis, skin cancer staging.

Abbreviations used: AJCC-American Joint Committee on Cancer, AJCC-8-eighth edition of the American Joint Committee on Cancer staging system, AJCC-7-seventh edition of the American Joint Committee on Cancer staging system, BWH-Brigham and Women's Hospital, CSCC-cutaneous squamous cell carcinoma, SLNB-sentinel lymph node biopsy, ME-major events.

 

Capsule Summary

 

  • The eighth edition of the American Joint Committee on Cancer staging system incorporates changes regarding cutaneous squamous cell carcinoma.

  • Poor prognosis was mainly associated with stage T2 of the seventh edition of the American Joint Committee on Cancer staging system but was diverted to stage T3 of the eighth edition. The eighth edition is more distinctive, monotonous, and homogeneous than the former seventh edition and shows some overlap with the Brigham and Women's Hospital alternative staging system.

  • The subgroup of tumors staged as T3 in the eighth edition of the American Joint Committee on Cancer staging system, which also showed a poor degree of differentiation, is a subgroup with greater risk.

 

 

Forty years after publication of the first edition of the American Joint Committee on Cancer (AJCC) staging system, 1 the eighth and latest edition of the staging system (AJCC-8) was published in 2017 with relevant changes regarding skin cancer, including cutaneous squamous cell carcinoma (CSCC). 2 3 CSCC is the second most frequent cancer in humans, with 700,000 estimated cases per year in the United States, and it can be both locally invasive and metastatic. 4 Staging systems have been poorly implemented in practice for CSCC, and clinical guidelines have been based mainly on high-risk factors rather than on the T stage itself. 5 6 7 This practice has been due mainly to the heterogeneous outcomes that can occur within each T stage and to the fact that the majority of tumors are staged as T1 when the seventh edition of the AJCC staging system (AJCC-7) is used. In addition, several research groups have attempted to define alternative staging systems for CSCC during recent years. 8 9

In 2002, the AJCC stated that stratification of patients by using a cancer staging system should make it possible to sort patients into distinctive (meaning that outcomes should be different among staging categories), monotonous (meaning that outcomes should worsen as the stage increases), and homogeneous (indicating that outcomes should be similar within a staging category) groups. 10 Until 2010, very few changes had been included in the stratification of the T category of CSCC 11 despite the publication of relevant works in the 1990s 12 13 14 15 and 2000s. 16 17 18 19 20 21 The AJCC-7, which was published in 2010, included some high-risk factors that upgraded the T stage, even in small tumors. 11 Specifically, perineural invasion, poor degree of differentiation, tumor thickness greater than 2 mm, Clark levels IV to V, and location of the primary tumor (ear or lower lip), along with the tumor's horizontal size, were the relevant risk factors used for defining T stages. 11 The major criticism of the AJCC-7 was based on evaluation of the stage T2, as patients with stage T2 disease exhibited a wide range of heterogeneity in terms of risk and prognosis. 9 For example, a small 3-mm-thick CSCC in the ear with no other risk factors would be staged as T2 with use of the AJCC-7. A 7-mm-thick tumor with a diameter of 2.5 cm, poor degree of differentiation, and muscle and perineural invasion (even with the involvement of nerves greater than 0.1 mm) with a location within a high-risk area would also be staged as T2 with use of the AJCC-7 if the bone was not affected. In addition, the clustering of poor outcomes in low T stages (T1 and T2), with more than 80% of poor outcome events, implicated poor homogeneity and monotonicity and limited the prognostic usefulness of stages T3 and T4. 9

Thus, the AJCC-8 emerged with the aim of overcoming these limitations. This new edition considers other relevant prognostic factors of CSCC to stratify the T stage, such as the influence of a tumor thickness greater than 6 mm, 16 perineural invasion (when the nerve diameter is at least 0.1 mm), 9 22 and invasion of the CSCC beyond the subcutaneous fat. 9 However, this new edition excludes some well-known risk factors 23 and stages only tumors of the head and neck (H&N), because most CSCC cases occur in the H&N area. For this reason, the AJCC-8 was developed for these tumors by the Head and Neck Expert Panel and thus was not dedicated to staging tumors located outside of the H&N area. 2 Alternatively, the Brigham and Women's Hospital (BWH) staging protocol includes some risk factors of both staging systems, but not all. 9 22

The aim of this study was to compare the T stages in a series of patients with CSCC by using both AJCC staging systems (AJCC-7 and AJCC-8) and the BWH alternative staging system, because the BWH classification system has been reported to be superior to AJCC-7 9 and is able to stage tumors beyond the H&N area. Moreover, we have attempted to explore the significance of risk factors to predict poor outcomes in CSCC through binary logistic regression.

Patients and methods

Patients and variables

We evaluated a retrospectively collected cohort of 186 CSCCs located in the H&N. The study included 103 male and 83 female patients with a median age of 85.5 years (range, 47-105) and a median follow-up of 79 months (range, 18.5-190.2). The data used were derived from a database of 2391 CSCCs registered from January 2008 to December 2016 at the University Hospital of Salamanca, Spain. Tumors other than those from the H&N (n = 430), recurrent tumors (n = 20), and tumors of the H&N with missing clinical data (n = 549) were excluded from the study. Of the remaining 1392 tumors, all that had evolved poor outcome were selected (n = 52) and 1 out of every 10 CSCCs with a good prognosis was randomly chosen (n = 134) to complete the cohort of 186 CSCCs. The local institutional ethics review board approved the study.

The following data were considered: (1) patient history (age and sex; previous history of actinic keratosis, nonmelanoma skin cancer [NMSC], and CSCC; and immunosuppression), (2) tumor traits (location of the primary tumor [lower lip, ear, or other locations in the H&N]; tumor size and tumor thickness, as measured using the OV100 software [Olympus, Center Valley, PA]; degree of differentiation [classified as good to moderate or poor]; perineural invasion [PNI] [both the presence and infiltration of nerves ≥0.1 mm in diameter]; and Clark level IV-V), (3) staging system (all tumors were classified according to the AJCC-7, AJCC-8, and the BWH alternative staging system), and (4) disease-specific poor outcome events (DSPOs) (local recurrence, metastasis [to the parotid gland or to lymph nodes], and disease-specific death). Specifically, instances of metastasis and disease-specific death were considered major events (MEs) of poor outcome.

Statistical analysis

The McNemar test was used on paired categoric data to analyze changes in T stages among classification systems. The binary logistic regression was applied to identify those variables used in the staging systems that were associated with DSPOs and MEs. The Wald backward approach was used by discarding the nonsignificant variables at each step via use of the IBM SPSS software (version 24, IBM Corporation, Armonk, NY). For all procedures, P values less than .05 were considered significant.

Results

T-stage classification using the AJCC-7, AJCC-8, and BWH staging systems

The characteristics of our cohort are described in Table I . First, the tumors were classified according to both the AJCC-7 and the AJCC-8. Some tumors were staged the same with use of both systems, but for other tumors the T stage was either upgraded or downgraded with use of the AJCC-8. Specifically, 71 of the 186 tumors (38.2%) remained in the same T stage with use of both systems, 95 tumors (51.1%) were up-staged with use of the AJCC-8, and 20 tumors (38.7%) were down-staged with use of the AJCC-8 ( Table II ). Next, we stratified our cohort by using the BWH alternative staging system. When the BWH staging system was compared with the AJCC-8, 77 tumors (41.39%) remained in the same stage, 92 tumors (49.47%) were down-staged, and 17 tumors were up-staged (9.14%) with use of the BWH system ( Table III ).

Variable Frequency %
Patient history  
Median age, y (range) 85.5 (47-105)
Sex 103 M:83 F  
History of AK 113 61.4
History of NMSC 81 43.5
History of CSCC 71 38.2
Immunosuppression 30 16.13
Tumor traits    
Median horizontal size, mm (range) 15 (4-73)
Median thickness, mm (range) 5.5 (1-28)
Degree of differentiation    
Good to moderate 139 74.7
Poor 47 25.3
Perineural invasion 43 26.1
Perineural invasion of nerves >0.1 mm 31 17.7
Invasion beyond the subcutaneous fat 36 19.4
Clark level IV or beyond 148 79.5
Muscle invasion 36 19.4
Bone invasion 3 1.6
Tumor location    
Lower lip 20 10.8
Ear 23 12.4
Other location 140 76.8
Growth pattern    
No infiltrative 99 53.5
Infiltrative 87 46.8
Desmoplasia 37 19.9
Poor prognosis    
Disease-specific poor outcome 52 28
Local recurrence 37 19.9
Major events 23 12.4
Metastasis to the parotid gland 6 3.2
Nodal metastasis 23 12.4
Disease-specific death 6 3.2

Chronic lymphocytic leukemia, HIV, poorly controlled diabetes, and long-term use of immunosuppressants.

Major events are metastasis and death due to CSCC.

Percentages refer to the total number of patients (N = 186). The median follow-up period was 79 months (range, 18.5-190.2).
AK , Actinic keratosis; CSCC , cutaneous squamous cell carcinoma; F, female; M , male; NMSC , nonmelanoma skin cancer.

Table IDescriptive data on the cohort of 186 patients with CSCCs of the head and neck

Systems compared No changes in T stage n (%) Up-staging/down-staging   n (%)
AJCC-7 (N = 186) vs AJCC-8 (N = 186) T1-T1 58 (31.2% [ ]) Up-staged T1-T2 1 (0.5%)
T2-T2 10 (5.4%) T1-T3 14 (7.5%)
T3-T3 3 (1.6%) T2-T3 80 (43.0%)
Total tumors with no change = 71 (38.2%) Down-staged T2-T1 20 (10.8%)

Percentages refer to the initial number of patients (N = 186).

AJCC-7 , American Joint Committee on Cancer staging system, seventh edition; AJCC-8 , American Joint Committee on Cancer staging system, eighth edition; CSCC , cutaneous squamous cell carcinoma.

Table IIChanges in T-stage assignment of the CSCC tumors of our cohort of 186 patients after evaluation of the tumors with use of the AJCC-7 versus the AJCC-8

Systems compared No changes in T stage n (%) Up-staging/down-staging   n (%)
AJCC-8 (N = 186) vs BWH (N = 186)     Down-staged T3-T1 11 (5.9%)
T1-T1 64 (34.4% ) T3-T2a 37 (19.9%)
T2-T2a 8 (4.3%) T3-T2b 44 (25.3%)
T3-T3 5 (2.7%) Up-staged T1-T2a 14 (7.5%)
Total tumors with no change = 77 (41.4%) T2-T2b 3 (1.6%)

Percentages refer to the initial number of patients (N = 186).

AJCC-8 , American Joint Committee on Cancer staging system, eighth edition; BWH , Brigham and Women's Hospital; CSCC , cutaneous squamous cell carcinoma.

Table IIIChanges in T-stage assignment of the CSCC tumors of our cohort of 186 patients after evaluation of the tumors with use of the AJCC-8 versus the BWH staging system

 

When we compared each T stage of the AJCC-8 with its counterpart in the other 2 staging systems by using the McNemar test, we observed significant differences for T2 and T3 ( P < .0001) of the AJCC-7 and all 3 T stages of the BWH staging system ( P < .05) ( Supplemental Table I ; available at http://www.jaad.org ).

Distinctiveness, homogeneity, and monotonicity among staging systems

The highest proportion of tumors that evolved into poor prognosis appeared in T3 with use of the AJCC-8 (84.6% of DSPOs and 86.96% of MEs). Also, as the T stage increased, the proportion of tumors with poor outcome also increased. In contrast, when the AJCC-7 was used, the highest proportion of tumors that evolved poor prognosis belonged to the stage T2 (76.9% of DSPOs and 91.3% of MEs). Thus, with the AJCC-8, the higher the T stage, the poorer the outcome; however, this was not the case when the AJCC-7 was used to evaluate these tumors. With the BWH classification system, the majority of tumors in our cohort that evolved poor prognosis were staged as T2b (42.30% of DSPOs and 39.13% of MEs). Homogeneity was assessed by comparing poor outcome between low T stages (T1 and T2 in both AJCC staging systems and T1 and T2a in the BWH staging system), and monotonicity was assessed by comparing poor prognosis between high T stages (T3 in both AJCC staging systems and T2b and T3 in the BWH system), as previously described 9 ( Table IV ). Venn diagrams show the distribution of T stages among classification systems, as well as the homogeneity and the monotonicity ( Supplemental Fig 1 ; available at http://www.jaad.org ).

Classification system Disease-specific poor outcome Major events
Stage Yes No Stage Yes No
Distinctiveness among the staging systems
AJCC-7 T1 (n = 73) 11/73 (15.06%) 62/73 T1 (n = 73) 3/73 (4.11%) 70/73
T2 (n = 110) 40/110 (36.36%) 70/110 T2 (n = 110) 20/110 (18.18%) 90/110
T3 (n = 3) 1/3 (33.33%) 2/3 T3 (n = 3) 0/3 (0%) 3/3
AJCC-8 T1 (n = 78) 6/78 (7.69%) 72/78 T1 (n = 78) 2/78 (2.56%) 76/78
T2 (n = 11) 2/11 (18.18%) 9/11 T2 (n = 11) 1/11 (9.09%) 10/11
T3 (n = 97) 44/97 (45.36%) 53/97 T3 (n = 97) 20/97 (20.62%) 77/97
BWH T1 (n = 75) 8/75 (10.66%) 67/75 T1 (n = 75) 3/75 (4%) 72/75
T2a (n = 59) 17/59 (28.81%) 42/59 T2a (n = 59) 6/59 (10.17%) 53/59
T2b (n = 47) 22/47 (46.8%) 25/47 T2b (n = 47) 9/47 (19.14%) 38/47
T3 (n = 5) 5/5 (100%) 0/5 T3 (n = 5) 5/5 (100%) 0/5
Homogeneity among the staging systems
AJCC-7 T1 and T2 51 of 52 (98.07%) T1 and T2 23 of 23 (100%)
AJCC-8 T1 and T2 8 of 52 (15.38%) T1 and T2 3 of 23 (13.04%)
BWH T1 and T2a 25 of 52 (48.07%) T1 and T2a 9 of 23 (39.13%)
Monotonicity among the staging systems
AJCC-7 T3 1 of 52 (1.93%) T3 0 of 23 (0%)
AJCC-8 T3 44 of 52 (84.61%) T3 20 of 23 (86.96%)
BWH T2b and T3 27 of 52 (51.93%) T2b and T3 14 of 23 (60.86%)

For disease-specific poor outcome, n = 52; for major events, n = 23.

AJCC-7 , American Joint Committee on Cancer staging system, seventh edition; AJCC-8 , American Joint Committee on Cancer staging system, eighth edition; BWH , Brigham and Women's Hospital.

Table IVDistinctiveness, homogeneity, and monotonicity among the different staging systems

 

Poor prognosis depending on T-stage changes among classification systems

Taking into account the changes in T stage with use of the AJCC-8 versus with use of the AJCC-7 and the associated poor outcomes, it was observed that events developed during follow-up in tumors in which the T stage increased from T1 to T3 (5 of 14 [35.7 %]) and in almost half of those in which the T stage increased from T2 to T3 (38 of 80 [47.5 %]). More importantly, DPSOs did not develop in any of the tumors in which the T stage decreased with use of the AJCC-8. Thus, the increase in T stage with use of the AJCC-8 system was associated with a higher risk of poor outcome, and its decrease was associated with good prognosis. When tumors were compared by using the BWH and AJCC-8 systems, it was observed that in a significant proportion of the tumors that were down-staged with the BWH staging system (39 of 93), DSPOs developed during follow-up, whereas only 1 of the 17 tumors that was up-staged behaved aggressively ( Table V ).

Systems being compared DSPOs MEs MPAR LYMPHM LR DSD
AJCC-8 vs AJCC-7
Up-staged (n = 95) T1-T2 1 0/52 (0) 0/23 (0) 0/6 (0) 0/23 (0) 0/37 (0) 0/6 (0)
T1-T3 14 5/52 (9.6%) 1/23 (4.3%) 1/6 (16.6%) 1/23 (4.35%) 4/37 (10.8%) 0/6 (0)
T2-T3 80 38/52 (73.1%) 20/23 (86.7%) 4/6 (66.6%) 20/23 (86.7%) 26/37 (70.3%) 6/6 (100%)
Down-staged (n = 20) T2-T1 20 0/52 (0) 0/23 (0) 0/6 (0) 0/23 (0) 0/37 (0) 0/6 (0)
AJCC-8 vs BWH
Up-staged (n = 17) T1-T2a 14 1/52 (1.92%) 0/23 (0) 0/6 (0) 0/23 (0) 1/37 (2.7%) 0/6 (0)
T2-T2b 3 0/52 (0) 0/23 (0) 0/6 (0) 0/23 (0) 0/37 (0) 0/6 (0)
Down-staged (n = 95) T3-T2a 37 14/52 (26.92%) 5/23 (21.2%) 1/6 (16.6%) 5/23 (21.7%) 9/37 (24.32.9%) 1/6 (16.67%)
T3-T2b 44 22/52 (42.3%) 9/23 (39.13%) 2/6 (33.3%) 10/23 (43.48%) 17/37 (45.94%) 3/6 (50%)
T3-T1 11 3/52 (5.76%) 1/23 (4.3%) 1/6 (16.6%) 1/23 (4.3%) 2/37 (5.4%) 0/6 (0)
AJCC-7 , American Joint Committee on Cancer staging system, seventh edition; AJCC-8 , American Joint Committee on Cancer staging system, eighth edition; BWH , Brigham and Women's Hospital; DPSO , disease-specific poor outcome; DSD , disease-specific death; LR , local recurrence; LYMPHM , lymphatic metastasis; ME , major event; MPAR , metastasis to the parotid gland.

Table VPoor prognosis depending on changes in T stage when assessment is done using the AJCC-7 versus the AJCC-8 and the AJCC-8 versus the BWH systems

 

Risk factors independently associated with poor outcome

In a final step to determine which variables were the most important in the stratification of tumors according to their risk of evolving poor outcomes, several binary logistic regression models were built. All risk factors included in all of the 3 staging systems analyzed were used as starting variables, and DSPOs and MEs were considered response variables. The significant risk factors of each staging system after a backward stepwise selection are shown in Table VI .

Staging system Response variables Selected predictive variables OR 95% CI P value
AJCC-7 Disease-specific poor outcome Location to the ear or lower lip 2.75 1.02-7.42 .045
PNI 6.72 3.15-14.31 .0001
Major events Degree of differentiation 3.31 1.23-8.91 .018
PNI 2.93 1.08-7.93 .034
AJCC-8 Disease-specific poor outcome PNI >0.1 mm 7.99 3.42-18.64 .0001
Thickness >6 mm 2.6 1.27-5.30 .009
Major events PNI >0.1 mm 5.47 2.12-14.10 .0001
Thickness >6 mm 2.32 0.91-5.87 .077
BWH Disease-specific poor outcome PNI >0.1 mm 5.53 2.16-14.13 .0001
Beyond subcutaneous fat 2.212 0.87-5.62 .095
Major events Degree of differentiation 4.18 1.63-10.75 .003
Beyond subcutaneous fat 4.45 1.69-11.68 .002
AJCC-7 + AJCC-8 + BWH § (&) Disease-specific poor outcome PNI >0.1 mm 7.99 3.42-18.65 .0001
Thickness >6 mm 2.59 1.27-5.30 .009
Major events Degree of differentiation 4.19 1.63-10.75 .003
Beyond subcutaneous fat 4.45 1.69-11.68 .002

Starting variables entered in the logistic regression for AJCC-7 were tumor size (<2 cm vs >2 cm), tumor thickness (<2 mm vs >2 mm), PNI (no vs yes), location of the primary tumor (other locations vs ear and lower lip), degree of differentiation (good and moderate vs poor), Clark level (I-II vs IV-V), bone invasion (no vs yes).

Starting variables entered in the logistic regression for AJCC-8 were tumor size (<2 cm vs >2 cm), tumor size (<4 cm vs >4 cm), tumor thickness (<6 mm vs > 6 mm), invasion beyond the subcutaneous fat (no vs yes), PNI of nerves >0.1 mm (no vs yes), bone invasion (no vs yes).

Starting variables entered in the logistic regression for the BWH staging system were tumor size (<2 cm vs >2 cm), degree of differentiation (good and moderate vs poor), tumor thickness (<2 mm vs >2 mm), PNI of nerves >0.1 mm (no vs yes), bone invasion (no vs yes).

§ Variables entered in the logistic regression of all 3 staging systems were included.

Boldface indicates statistical significance.
AJCC , American Joint Committee on Cancer; AJCC-7 , American Joint Committee on Cancer staging system, seventh edition; AJCC-8 , American Joint Committee on Cancer staging system, eighth edition; BWH , Brigham and Women's Hospital; CI , confidence interval; CSCC , cutaneous squamous cell carcinoma; OR , odds ratio; PNI , perineural invasion.

Table VIVariables used in the stratification of AJCC T stages that are significantly associated with prognosis in CSCC after building of binary logistic regression models

 

Finally, all risk variables included in the AJCC-7, AJCC-8, and BWH staging systems were combined. It was observed that PNI of 0.1 mm or more and a tumor thickness greater than 6 mm were significantly associated with DSPOs and the invasion of the tumor beyond the subcutaneous fat and the poor degree of differentiation were both independently associated with MEs ( Table VI ). Furthermore, of the tumors staged as T3 with the AJCC-8, 38 exhibited a poor degree of differentiation. Of those tumors, 12 tumors developed MEs during follow-up and 20 developed DSPOs.

Discussion

In 2010, the AJCC-7 incorporated some high-risk factors and built a completely new T stage, 11 but it was found to exhibit limitations. 8 9 Recently, the AJCC-8 emerged as an upgraded classification system. 2 Here, it has been shown that the AJCC-8 is better than the AJCC-7 in terms of prognosis stratification.

The vast majority of CSCC tumors associated with DSPO in the AJCC-7 were staged as T2 in our cohort, which included tumors with a heterogeneous prognosis (which is a weak point that others have already highlighted). 9 Conversely, when the AJCC-8 was used on this same cohort, poor prognosis was diverted toward the T3 stage. Additionally, classification with the AJCC-8 showed that T1 tumors displayed the best outcome and T2 tumors had an intermediate prognosis. Thus, the AJCC-8 is more homogeneous and monotonous than its predecessor, the AJCC-7. Furthermore, the proportion of poor outcomes in each T stage was more distinctive with the AJCC-8 than with the AJCC-7.

The BWH stratification has proved to be superior to the AJCC-7 9 and may have some advantages over the AJCC-8 as well, being a good choice for staging tumors beyond the H&N area. However, according to our data, this system did not appear to have great advantages over the AJCC-8 in staging CSCCs of the H&N. Whereas the BWH system shows clear differences with respect to the AJCC-7, 9 many overlaps are observed with the AJCC-8 in terms of homogeneity and monotonicity. More series, including series with T4 tumors, are needed to further evaluate these similarities and differences.

The T-stage categories are based on the presence of some high-risk factors that are combined to sort tumors in terms of prognosis. Our data showed that all 3 of the new risk factors incorporated into the AJCC-8, namely, PNI of 0.1 mm or more, thickness greater than 6 mm, and invasion of the tumor beyond the subcutaneous fat, are significantly associated with poor prognosis in CSCCs. Moreover, the poor degree of differentiation, which is considered a risk factor in the BWH staging system 9 and in the AJCC-7, 11 is a relevant risk factor according to our data but was not considered in the AJCC-8, 2 even though several works have highlighted its prognostic relevance. 20 23 24 25 26 27 In our cohort, MEs developed in almost 1 in 3 tumors staged T3 (according to the AJCC-8) and having a poor degree of differentiation. Hence, this apparent subset could be more aggressively managed, and adjuvant radiotherapy and sentinel lymph node biopsy (SLNB) could be considered in some cases. SLNB, although well established in the management of melanoma, 28 is not routinely recommended in CSCC. The risk of SLNB positivity in intermediate-thickness melanoma is around 20%, 28 and the risk of nodal metastasis in the AJCC-8 T3 subgroup with poor degree of differentiation in our series is greater than 30%, which means that some of these tumors may benefit from SLNB; however, clinical trials are needed to address this issue. The problem with the degree of differentiation may lie with the subjectivity that could exist upon tumor evaluation. 2 29 According to the AJCC Expert Panel, specific associations of the degree of differentiation with prognosis of CSCC independently of other risk factors are not definitive, 2 and thus, degree of differentiation was not included in the AJCC-8. However, this is an issue that should be assessed in future studies.

Another weak point of the AJCC-8 may be that the newly incorporated prognostic factors are given similar importance even though not all are equally as important. Also, it has been shown that the presence of 2 or more risk factors is more highly predictive of poor outcomes than is presence of only1 factor. 9 22 Considering the odds ratio in the logistic regression models in our study, PNI of 0.1 mm or more seems to be the most important risk factor in predicting poor outcome, whereas the relevance of a thickness greater than 6 mm seems to be less relevant. Hence, establishing a hierarchy of risk factors would be useful in the management of CSCC.

The retrospective design of our study implies that the patients and their records were not conceived of before this particular study. However, the archived slides did provide the possibility of a detailed re-examination of all histopathologic tumor traits. Another limitation of this study could be the absence of T4 tumors in our series, but given their rarity, the present study may still prove useful to physicians in the stratification and management of most CSCCs.

According to our data, the AJCC-8 is more distinctive, monotonous, and homogeneous than the AJCC-7, and it is better at predicting poor outcomes in CSCC. However, the AJCC-8 is not able to adequately stratify all patients with CSCC because it is specific to the H&N area. 2 29 In addition to this limitation, some high-risk factors, such as the degree of differentiation, are not considered in the AJCC-8 but may be useful in identifying subgroups of tumors with a greater risk of poor outcomes. The BWH alternative staging system shows overlap with the AJCC-8 both in high-stage and low-stage tumor assignment and allows classification of CSCCs beyond the H&N area. This system also considers the cumulative risk of risk factors when they are combined, which may also improve the accuracy of the AJCC-8. Finally, according to our data, the new AJCC-8 seems to be considerably better than its predecessor, the AJCC-7. But future studies are required to better refine T3 tumors in AJCC-8 and to build a staging system for CSCC that can be used to classify tumors located outside the H&N area.

We thank Emma Keck for the English editing.

Appendix

 

fx1
 

Supplemental Fig 1
Venn diagrams representing the distribution of T stages among the different classification systems. Each circle represents 1 staging system; sectors correspond with the overlap of 2 or 3 staging systems. Numeric data in each sector: disease-specific poor outcome (DSPO) in the corresponding sector, major events (MEs) in the corresponding sector, and total number of patients in the corresponding sector (total). A , B , and C , How tumors are classified in the different staging systems. In these schemes, subcategories T2a and T2b were grouped within stage T2 for the Brigham and Women's Hospital (BWH) alternative staging system. The greatest overlap among all 3 systems occurs for T1 tumors (50 tumors are staged as T1 among all 3 systems [see the central sector]) ( A ); there is a medium overlap for T2 tumors ( B ); and the smallest overlap among all 3 systems occurs for T3 tumors (none of the 186 tumors are staged as T3 in all 3 systems at the same time) ( C ). D , Classification of low-stage tumors among all 3 staging systems (ie, homogeneity). There is no equivalence between the seventh edition of the American Joint Committee on Cancer (AJCC-7) staging system and the other 2 systems (ie, eighth edition of the American Joint Committee on Cancer [AJCC-8] and BWH alternative staging system). A total of 86 tumors were classified within low T stages with use of all 3 classification systems ( central sector ), and it is in these tumors in which 8 DSPOs and 3 MEs occurred. All but 1 tumor that evolved DPSOs (n = 51) were classified as T1 or T2 with the AJCC-7, which means that the AJCC-7 is the worst staging system in terms of homogeneity. The majority of tumors classified as T1 or T2 with the AJCC-8 are classified as T1 or T2a with the BWH alternative staging system. Thus, both the AJCC-8 and the BWH systems display great overlap (64.2 % for the BWH system and 96.6 % for the AJCC-8 system). E , Classification of high-stage tumors among all 3 staging systems (ie, monotonicity). The scheme shows no equivalence between the AJCC-7 staging system and the other 2 classification systems (ie, AJCC-8 and BWH). The majority of tumors classified as T3 with the AJCC-8 system are classified within the high T stages in the BWH system (T2b and T3); hence, these systems have an overlap of 94% and 51%, respectively. However, use of the AJCC-8 resulted in 48 high-stage cases that were not classified the same way with use of the BWH system, according to which 17 DSPOs and 6 MEs occurred. According to our data, the AJCC-8 seems to be the best staging systems in terms of monotonicity.

Version of T1 T1 (AJCC-8) P value Version of T2 T2 (AJCC-8) P value Version of T3 T3 (AJCC-8) P value
T1 (AJCC-7)   Yes No .5 (NS) T2 (AJCC-7)   Yes No .0001 T3 (AJCC-7)   Yes No .0001
Yes 58 15 Yes 10 100 Yes 3 0
No 20 93 No 1 75 No 94 89
T1 (BWH)   Yes No .0001 T2 (BWH)   Yes No .003 T3 (BWH)   Yes No .030
Yes 64 11 Yes 11 95 Yes 5 0
No 14 97 No 0 80 No 92 89

The stage T2 in the BWH's staging system includes the subcategories T2a and T2b; as these 2 subcategories exhibit differences in prognosis, the grouping of both of them in only 1 T2 group in this table for methodologic purposes might be somewhat imprecise. Boldface indicates statistical significance.

AJCC-7 , American Joint Committee on Cancer, seventh edition; AJCC-8 , American Joint Committee on Cancer, eighth edition; BWH , Brigham and Women's Hospital; NS , not significant.

Supplemental Table IPairwise comparison of each T stage between the AJCC-7 and AJCC-8 staging systems and between the AJCC-8 and BWH staging categories (McNemar test)

 

Dr Moyano-Bueno and Dr Viñolas-Cuadros contributed equally to this article.

Funding sources: Dr Cañueto is partially supported by a grant (GRS 1342/A/16) from the Regional Health Office of Castile and Leon and by the program INT/M/16/17 from the Regional Health Office of Castile and Leon. Dr Pérez-Losada was partially supported by FEDER and the MICINN (grants SAF2014-56989-R and SAF2017-88854-R), the Instituto de Salud Carlos III (PIE14/00066), and the We Can Be Heroes Foundation.

Conflicts of interest: None disclosed.

References

  • 1 American Joint Committee for Cancer Staging and End Results Reporting. Manual for Staging of Cancer of 1977. (American Joint Committee, Chicago, IL, 1977)
  • 2 M.B. Amin, S. Edge, F. Greene. AJCC Cancer Staging Manual. 8th ed. 2017 Springer: New York, NY
  • 3 J. Canueto, C. Roman-Curto. Novel sdditions to the AJCC's new staging systems for skin cancer. Actas Dermosifiliogr. 2017;108:818-826
  • 4 P.S. Karia, J. Han, C.D. Schmults. Cutaneous squamous cell carcinoma: estimated incidence of disease, nodal metastasis, and deaths from disease in the United States, 2012. J Am Acad Dermatol. 2013;68(6):957-966
  • 5 A. Stratigos, C. Garbe, C. Lebbe, et al. Diagnosis and treatment of invasive squamous cell carcinoma of the skin: European consensus-based interdisciplinary guideline. Eur J Cancer. 2015;51(14):1989-2007
  • 6 Oncology NNCCNCPi. Squamous cell skin cancer. In. NCCN Guidelines 2016. Available at: <http://oncolife.com.ua/doc/nccn/Squamous_Cell_Skin_Cancer.pdf>(nccn CSCC). Accessed July 30, 2018.
  • 7 Network CCAAC. Basal cell carcinoma, squamous cell carcinoma (and related lesions): a guide to clinical management in Australia. In: 2008. Available at: <https://www.cancer.org.au/content/pdf/HealthProfessionals/ClinicalGuidelines/Basal_cell_carcinoma_Squamous_cell_carcinoma_Guide_Nov_2008-Final_with_Corrigendums.pdf>. Accessed July 30, 2018.
  • 8 S.J. Miller. Staging cutaneous squamous cell carcinoma. JAMA Dermatol. 2013;149(4):472-474
  • 9 P.S. Karia, A. Jambusaria-Pahlajani, D.P. Harrington, G.F. Murphy, A.A. Qureshi, C.D. Schmults. Evaluation of American Joint Committee on Cancer, International Union Against Cancer, and Brigham and Women's Hospital tumor staging for cutaneous squamous cell carcinoma. J Clin Oncol. 2014;32(4):327-334
  • 10 F.L. Greene, C.M. Balch, I.D. Fleming, F. April, D.J. Haller. AJCC Cancer Staging Manual. 6th ed. 2002 Springer: Chicago, IL
  • 11 S.B. Edge, D.R. Byrd, C.A. Compton, A.G. Fritz, F.L. Greene, A. Trotti. AJCC Cancer Staging Manual. 7th ed. 2010 Springer: New York, NY
  • 12 D.E. Rowe, R.J. Carroll, C.L. Day. Prognostic factors for local recurrence, metastasis, and survival rates in squamous cell carcinoma of the skin, ear, and lip. Implications for treatment modality selection. J Am Acad Dermatol. 1992;26(6):976-990
  • 13 H. Breuninger, B. Black, G. Rassner. Microstaging of squamous cell carcinomas. Am J Clin Pathol. 1990;94(5):624-627
  • 14 R. Doll. The Pierre Denoix Memorial Lecture: nature and nurture in the control of cancer. Eur J Cancer. 1999;35(1):16-23
  • 15 A. Eroglu, U. Berberoglu, S. Berreroglu. Risk factors related to locoregional recurrence in squamous cell carcinoma of the skin. J Surg Oncol. 1996;61(2):124-130
  • 16 K.D. Brantsch, C. Meisner, B. Schonfisch, et al. Analysis of risk factors determining prognosis of cutaneous squamous-cell carcinoma: a prospective study. Lancet Oncol. 2008;9(8):713-720
  • 17 C. Mourouzis, A. Boynton, J. Grant, et al. Cutaneous head and neck SCCs and risk of nodal metastasis - UK experience. J Craniomaxillofac Surg. 2009;37(8):443-447
  • 18 G.L. Clayman, J.J. Lee, F.C. Holsinger, et al. Mortality risk from squamous cell skin cancer. J Clin Oncol. 2005;23(4):759-765
  • 19 B.A. Moore, R.S. Weber, V. Prieto, et al. Lymph node metastases from cutaneous squamous cell carcinoma of the head and neck. Laryngoscope. 2005;115(9):1561-1567
  • 20 B.S. Cherpelis, C. Marcusen, P.G. Lang. Prognostic factors for metastasis in squamous cell carcinoma of the skin. Dermatol Surg. 2002;28(3):268-273
  • 21 A.S. Ross, F.M. Whalen, R. Elenitsas, X. Xu, A.B. Troxel, C.D. Schmults. Diameter of involved nerves predicts outcomes in cutaneous squamous cell carcinoma with perineural invasion: an investigator-blinded retrospective cohort study. Dermatol Surg. 2009;35(12):1859-1866
  • 22 A. Jambusaria-Pahlajani, P.A. Kanetsky, P.S. Karia, et al. Evaluation of AJCC tumor staging for cutaneous squamous cell carcinoma and a proposed alternative tumor staging system. JAMA Dermatol. 2013;149(4):402-410
  • 23 A.K. Thompson, B.F. Kelley, L.J. Prokop, M.H. Murad, C.L. Baum. Risk factors for cutaneous squamous cell carcinoma recurrence, metastasis, and disease-specific death: a systematic review and meta-analysis. JAMA Dermatol. 2016;152(4):419-428
  • 24 J.T. Krediet, M. Beyer, K. Lenz, et al. Sentinel lymph node biopsy and risk factors for predicting metastasis in cutaneous squamous cell carcinoma. Br J Dermatol. 2015;172(4):1029-1036
  • 25 K. Wermker, J. Kluwig, S. Schipmann, M. Klein, H.J. Schulze, C. Hallermann. Prediction score for lymph node metastasis from cutaneous squamous cell carcinoma of the external ear. Eur J Surg Oncol. 2015;41(1):128-135
  • 26 N.D. Brougham, E.R. Dennett, R. Cameron, S.T. Tan. The incidence of metastasis from cutaneous squamous cell carcinoma and the impact of its risk factors. J Surg Oncol. 2012;106(7):811-815
  • 27 A. Kyrgidis, T.G. Tzellos, N. Kechagias, et al. Cutaneous squamous cell carcinoma (SCC) of the head and neck: risk factors of overall and recurrence-free survival. Eur J Cancer. 2010;46(9):1563-1572
  • 28 D.L. Morton, J.F. Thompson, A.J. Cochran, et al. Final trial report of sentinel-node biopsy versus nodal observation in melanoma. N Engl J Med. 2014;370(7):599-609
  • 29 P.S. Karia, F.C. Morgan, J.A. Califano, C.D. Schmults. Comparison of tumor classifications for cutaneous squamous cell carcinoma of the head and neck in the 7th vs 8th edition of the AJCC Cancer Staging Manual. JAMA Dermatol. 2018;154(2):175-181