The presence or absence of ENE is required to assign N category. ENE is defined as extension through the lymph node capsule into the surrounding connective tissue, with or without associated stromal reaction.²²-²⁵ Unambiguous evidence of gross ENE (i.e., defined as invasion of skin, infiltration of musculature/fixation to adjacent structures on clinical examination, cranial nerve, brachial plexus, sympathetic trunk or phrenic nerve invasion with dysfunction) is a sufficiently high threshold to classify these as clinical ENE(+).

Tumor diameter refers to the maximum clinical diameter of the CC lesion (preoperatively based on physical exam). Tumor diameter larger than 2 cm changes the T category to T2. Multiple studies corroborate a correlation between tumor diameter and more biologically aggressive disease, including local recurrence and metastasis in multivariate analysis.²⁰,²¹,²⁶-²⁹ Two of these studies point toward size of 2 cm as a threshold beyond which tumors are more likely to metastasize to lymph nodes.²⁰,²¹ A 2.1 fold risk of nodal metastasis for tumors larger than 2 cm was noted when prospectively reviewing risk factors for poor prognosis in 615 patients with CSCC.²¹ Another study of 985 CSCC patients found that tumors 2 cm or larger were associated with a 5.6-fold higher risk of local recurrence, a 7.0-fold higher risk of nodal metastasis, and a 15.9-fold higher risk of death from CSCC.²⁰ The 2 -cm threshold was decided upon for assigning a T2 category because of the existing published data that a clinical diameter larger than 2 cm is associated with a poor prognosis. In addition, this breakpoint allows continued congruence between CC and Head and Neck Staging. A further cut point of 4 cm was included for assigning a T3 category because one study showed this to be predictor of particularly poor outcomes, with a 4.5-fold increase in disease-specific death in tumors 4 cm or larger in diameter.²⁸

Although 2 cm is recognized by many to be an important size cutoff, the metastatic potential of tumors smaller than 2 cm cannot be ignored, as they too can metastasize. Multiple studies have identified several other factors independently associated with elevated risks of recurrence, metastasis, and/or death. These factors are weighted on an equal basis with size of greater than 2 cm because there are no clear means of differentiating the significance of prognostic factors.

Recent studies show that both tumor thickness (measured in millimeters) and the tissue level of invasion are important variables for the prognosis of CSCC. Prospective studies show that increasing tumor thickness²¹,³⁰ and anatomic DOI³⁰ correlate with an increased risk of metastases. In one study, no metastases were present with primary tumors less than 2 mm in depth (tumor thickness), but 16% of cases with tumors greater than 6 mm in depth had metastases.²¹ This study also found that tumors greater than 6 mm in depth had a 6.0-fold higher risk of local recurrence and nodal metastasis on multivariate analysis.²¹ Another study reported increasing metastatic rates as tumor invasion progressed from dermis (0.0% risk) to subcutaneous adipose tissue (4.1% risk), to muscle or bone (12.5% risk).³⁰ A 5-20% increase in nodal metastasis risk has been reported for each 1-mm increase in tumor thickness.²⁷,³¹

Traditional Breslow depth is measured from the granular layer to the base of the tumor. In CSCC, however, the granular layer is lost and simply measuring from the surface of the tumor to the base may overestimate prognostic impact because the dead keratotic surface of tumors may contribute little prognostically, and some exophytic CSCCs, such as keratoacanthomas, have a low risk of metastases. Thus, the authors recommend that millimeter depth be measured from the granular layer of adjacent normal skin to the base of the tumor to avoid these issues. Such measurement is assumed in the staging system herein.

Two studies employing multivariate analysis have shown DOI past the subcutaneous fat to be associated with poor outcomes.²⁰,²⁹ Invasion past subcutaneous fat was associated with a 9.3-fold increased risk of nodal metastasis and a 13.0-fold increased risk of death from CSCC. A smaller study of 256 patients with high-risk CSCC (defined as those with one of the following risk factors: perineural or lymphovascular invasion, poorly differentiated histology, depth beyond subcutaneous fat, diameter of at least 2 cm, location on the ear, or location on the vermilion lip) found that invasion past subcutaneous fat was associated with 7.2-fold higher risk of nodal metastasis and 4.1-fold higher risk of death from CSCC.²⁹

Based on the data above, the 8^th Edition CC staging system incorporates deep invasion, defined as either greater than 6 mm depth as measured above or invasion past subcutaneous fat (to fascia, muscle, perichondrium, periosteum, etc.), as one of the high-risk features in the T category. Differentiation between the prognostic contributions of millimeter thickness versus tissue level of invasion will depend on future studies.

Specific anatomic locations including the lip (vermilion and hair-bearing), ear, temple, and cheek have an increased risk of local recurrence and metastatic potential in multivariate studies and thus have been categorized as high risk in the this staging system.²⁰,²¹,²⁶ Location is not part of T categorization because studies have varied in how location was classified. In a large retrospective study of about 9,000 CSCC patients, tumors located on the ear/cheek and lip were 3.0 and 4.8 times more likely, respectively, to result in nodal metastasis than tumors located on other anatomic sites.²⁶ A prospective study found similar results with a 3.6-fold increased risk of nodal metastasis for tumors located on the ear.²¹ Another study found that location on ear or temple was independently associated with an increased risk of local recurrence, nodal metastasis, and death from CSCC.²⁰

Four studies have shown perineural invasion (PNI, defined as tumor cells within the nerve sheath) is an independent factor associated with poor outcomes.²⁰,²⁶,²⁸,³² Two additional studies showed small-caliber PNI (involving nerves less than 0.1 mm in caliber) to have no independent association with poor outcomes, indicating that invasion of small dermal nerve fibers alone may not significantly affect prognosis.³³,³⁴ In a study of 114 CSCC cases with PNI, the risk of nodal metastasis CSCC was significantly higher in patients with large-caliber (risk 17%) versus small-caliber (risk 4%) PNI.³³ Another study found that the risk of nodal metastasis and CSCC death in cases with large-caliber PNI was 18% and 22%, respectively.³⁴ Thus, larger caliber (greater than or equal to 0.1 mm) nerve invasion is a risk factor in the 8th Edition staging system. Because most nerves deep to the dermis are greater than 0.1 mm in caliber, nerve invasion beyond the dermis also is a risk factor.

Early studies recognized that the histological grade or degree of differentiation of a CSCC affects prognosis: the more well-differentiated, the less aggressive the clinical course.³⁵,³⁶ In 1978, Mohs, in his review of “microscopically controlled surgery,” reported significant differences in cure rates for well-differentiated tumors (99.4%) compared with poorly differentiated tumors (42.1%).³⁷ More recently, three studies have confirmed poor differentiation to be independently associated with recurrence, metastasis, and death.²⁰,²⁶,²⁹ Patients with poorly differentiated CSCCs have a 2.5-fold to 3.0-fold²⁰,²⁹ higher risk of local recurrence and a 3.3-fold to 6.1-fold²⁰,²⁶,²⁹ higher risk of nodal metastasis than patients with well-differentiated CSCCs. Death due to CSCC is also higher in poorly differentiated CSCCs with a 4.1-6.7 times higher risk reported.²⁰,²⁹

Other studies have found desmoplasia to be associated with poor outcomes.²¹,³⁸,³⁹ Desmoplasia, single-cell spread, and poor or sarcomatoid differentiation can often occur together and are all suggestive of an aggressive tumor phenotype. Thus, CC staging in the 8^th Edition continues to include aggressive histologic features (poorly differentiated tumors) as one of the several high-risk features and expands that definition to include desmoplasia and sarcomatoid presentations. Specific associations of these histologic subtypes independent of other risk factors are not definitive and, therefore, they have not been included as determinants of T categorization.

In the AJCC Cancer Staging Manual, 6^th Edition (6^th Edition), the T4 designation was used for tumors that “invaded extradermal structures.” The most common and important instances of deep anatomic extension for CSCC involve extension to bone and perineural extension to skull base. Based on these considerations, in the 7^th Edition, T3 and T4 were reserved for these presentations of locally advanced disease consistent with data from several head and neck studies suggesting that CSCC extending to skull base is associated with poor prognosis, similar to advanced lymph node disease.⁹-¹²,⁴⁰-⁴² Subsequent cohort studies, however, have shown that although these presentations do connote a poor prognosis, they are very rare for primary CSCC and thus few tumors are in the T3 and T4 categories as designated by the 7^th Edition staging system. This resulted in most cases of poor outcomes occurring in what the 7^th Edition staged as T2 cases.²⁹,³⁴ To improve upon this, the 8^th Edition CC staging groups all bone and skull base invasion in T4 because they are likely similar in their poor prognosis.

Since the 6^th Edition, multiple studies have examined the outcomes in patients with CSCC and regional lymph node metastasis.¹⁰,¹²,⁴⁰,⁴¹ These studies show that the number of nodes involved and the size of lymph node metastasis correlate with poor prognosis.

Based on data from O'Brien, et al.,⁹ the NMSC Expert Panel decided that sufficient evidence exists to stage patients according to increasing nodal disease. Although preliminary data exist to suggest that cervical nodal disease may portend a worse prognosis than similar disease in the parotid, the data are insufficient to support this separation at this time. Separating facial nerve involvement or involvement of the skull base (now T4) from extensive parotid disease will further clarify the prognosis of these patients.

It is well known that immunosuppressed patients are at risk for developing malignancies, especially CSCCs. Organ transplant recipients develop CSCC 65 times more frequently than age-matched controls.⁴³,⁴⁴ The CSCCs in immunocompromised patients are more numerous and tend to recur and metastasize at a higher rate.¹⁵,¹⁶,⁴⁵-⁵¹ It has been reported that immunocompromised patients have a 7.2 times increased risk of local recurrence and a 5.3 times increased risk of any recurrence of disease.⁵² Mortality also is increased with skin cancer, the fourth most common cause of death in a renal transplant cohort.⁵³ In transplant recipients, CSCC develops 10-30 years earlier than in immunocompetent hosts.³,⁴ Strong consideration was given therefore toward including immunosuppression as a risk factor. However, in studies employing multivariate analysis, only a single study showed immunosuppression independently associated with poor outcomes,²¹ possibly because immunosuppression is a broad category with varying degrees of associated immune dysfunction and variable prognostic effects. It is therefore not part of the staging system. This factor (including type or cause of immunosuppression) should be collected by cancer registries and investigators as a potentially important prognostic factor. Centers collecting such data and performing studies may designate immunosuppressed status with an “I” after the staging designation.

In addition to the importance of the TNM factors outlined previously, the overall health of these patients clearly influences outcome. An ongoing effort to better assess prognosis using both tumor and nontumor-related factors is underway. Chart abstraction will continue to be performed by cancer registrars to obtain important information regarding specific factors related to prognosis. These data will then be used to further hone the predictive power of the staging system in future revisions.

Comorbidity can be classified by specific measures of additional medical illnesses.⁵⁴ Accurate reporting of all illnesses in the patient's medical record is essential to assessment of these parameters. General performance measures are helpful in predicting survival. The AJCC strongly recommends the clinician report performance status using the Eastern Cooperative Oncology Group (ECOG), Zubrod, or Karnofsky performance measures, along with standard staging information. An interrelationship between each of the major performance tools exists.

Lifestyle factors, such as tobacco and alcohol abuse, negatively influence survival. Accurate recording of smoking in pack-years and alcohol in number of days drinking per week and number of drinks per day will provide important data for future analysis. Nutrition is important to prognosis and will be indirectly measured by weight loss of greater than 5% of body weight in the previous 6 months.⁵⁵ Depression adversely affects quality of life and survival. Notation of a previous or current diagnosis of depression should be recorded in the medical record.⁵⁶

The role of tobacco as a negative prognostic factor is well established. However, exactly how this could be codified in the staging system is less clear. At this time, smoking is known to have a deleterious effect on prognosis but that effect is hard to accurately apply to the staging system.

Smoking history should be collected as an important element of the demographics and may be included in Prognostic Groups in the future. For practicality, the minimum standard should classify smoking history as never, less than or equal to 10 pack-years, greater than 10 but less than or equal to 20 pack-years, or greater than 20 pack-years.