An anatomic checklist for accurate staging of grossly invasive thyroid cancer

Introduction

The American Joint Commission on Cancer (AJCC) cancer staging system stratifies patients according to disease-specific mortality risk. This static determination is based on initial tumor characteristics (size/extent of the primary tumor, regional lymph node status and distant metastasis) plus clinical information included up to 4 months after surgery. The current AJCC schema is not applicable in the setting of recurrent surgical disease. Risk stratification for disease recurrence is clinically more relevant for thyroid cancer patients than predicting disease-specific mortality, which is generally low. The American Thyroid Association (ATA) risk of recurrence (ROR) provides a framework for predicting a patient’s risk of recurrent disease (1, 2).

AJCC thyroid cancer staging acknowledges that gross extrathyroidal extension (ETE) identified during surgery is a key prognosticator and offers the following guidance: ‘The surgeon’s description of gross ETE must also be included’ (3). AJCC staging incorporates all pertinent information from preoperative evaluations, intraoperative findings, postoperative pathology, as well as pertinent information that becomes available up to 4 months after surgery (3). The Surgical Affairs Committee of the ATA highlights the importance of communicating intraoperative surgical findings. The operative report is the only current standard for conveying this information. However, it is time-consuming and challenging for pathologists, endocrinologists, nuclear medicine specialists and radiation oncologists to glean important anatomic details from these reports (4, 5). When one depends on scouring the surgeon’s operative report, there are numerous potential pitfalls that can lead to understaging and inaccurate AJCC and ATA-ROR determinations.

We address this information gap by presenting a newly developed checklist, which includes anatomic figures that advise pathologists on the anatomic details of intraoperative findings. This level of detail not only improves staging but also helps in planning for adjuvant radiation therapy and interpreting postoperative imaging. It can also improve our understanding of clinical scenarios currently lumped together as pT4a disease. For instance, involvement of the RLN, esophagus and laryngotracheal complex is all similarly staged as pT4a. It is unclear as to whether these different types of structural involvement could be associated with different outcomes. The additional anatomic detail in the checklist is intended to supplement, but not replace, the information included in the College of American Pathologists thyroid protocol and will facilitate outcomes research (6).

The absence of an AJCC staging system for recurrent disease makes formal reporting in this setting challenging and nonuniform. The rationale for including a checklist for recurrent disease is that anatomic organs are more frequently resected. Gross extranodal extension (ENE) involving anatomic structures should be accurately identified in the pathology report. Distinguishing recurrent disease that involves lymph nodes from structural recurrence involving nerves, bone, vessels, the laryngotracheal complex and the esophagus is important, but there is currently no mechanism for doing so.

In recognition of the importance of ease of checklist completion in order to support surgeon and pathologist adoption, both an online and a printed version of the form (https://tiro.expert/analysis-tools) are readily available. The purpose of this study is to validate the effectiveness of the anatomic checklist in supporting both AJCC and ATA-ROR staging.

Materials and methods

This retrospective study has obtained IRB approval from the Icahn School of Medicine at Mount Sinai, New York (STUDY-18-01236). Over a 24-month period, two senior head and neck surgeons (A and B) tracked the accuracy of final staging in a consecutive cohort of both primary and recurrent thyroid cancer patients that demonstrated clinically gross invasion of surrounding structures. Only cases with ETE or ENE were included in this study. Surgeon A used the checklist for the last six of 19 primary cases but not for the first 13. Surgeon B did not use the checklist for the first case, which served as the impetus for this study. The checklist was developed and utilized for surgeon B’s next nine cases (Fig. 1). Before sign-out, the pathologist searched the surgeon’s operative note in cases without a checklist submission. We compared the accuracy of final pathology staging on primary resections submitted for both surgeons. The checklist includes anatomic details regarding the laryngotracheal complex, nerves, muscles, bone, vasculature and pharyngoesophagus, which represent the sites usually involved by gross ETE or ENE.

View Full Size

A breakdown of reported cases with associated information regarding the inclusion or omission of the checklist and the results of accuracy of staging/classification.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

Preoperative clinicopathologic information related to the results of molecular testing and the function of the vocal cords is often known by the surgeon and may be included in a clinic record but is not traditionally included in an operative note. These vital pieces of information are incorporated in the anatomic checklist as line items to be completed.

The anatomic checklist

This anatomic checklist (Figs 2, 3, 4, 5, 6, 7) includes the currently accepted details integrated into the AJCC staging system (8th edition) for thyroid cancer (3) and the information needed for ATA-ROR classification (2), reported in the ‘minimum data set’ section of the checklist (Fig. 2). The information included in the checklist evolved out of critical analysis, feedback, modification and consensus-building among highly experienced senior thyroid surgeons and pathologists of the authorship. The checklist facilitates immediate, efficient and accurate AJCC and ATA-ROR staging rather than requiring the pathologist to obtain critical information by reading the operative note or calling the surgeon directly.

View Full Size

Intraoperative gross extrathyroidal/extranodal disease checklist. The checklist is divided into sections to capture key data points necessary for accurate AJCC staging and ATA ROR classification. It includes fields for documenting preoperative clinicopathological findings, surgical outcomes, disease status and presence of distant metastatic disease.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

View Full Size

The second part of the anatomical checklist shows the laryngotracheal complex (LT-), indicating the thyroid (Th), cricoid (Cr) and trachea (T) subsites with delineation of the layers of resection, either subperichondrial or full thickness. The surgeon can specify the number of tracheal rings involved (T_x). In addition, it includes bony structures (B-), such as the clavicular heads (Clav) and sternum (St), indicating whether the involvement is superficial (within the periosteal plane) or necessitates partial or complete resection of the bone.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

View Full Size

The third part of the checklist illustrates cervical nerve involvement (N-), highlighting the recurrent laryngeal (RLN), superior laryngeal (SLN) and vagus (Vag) nerves. Each nerve subsite is further classified to indicate the layer of resection. The classification system includes RLN⁰, SLN⁰ and Vag⁰ for cases where the nerve can be preserved through subepineural dissection, and RLN¹, SLN¹ and Vag¹ for more substantial invasion requiring nerve sacrifice.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

View Full Size

The next part of the checklist allows reporting of involvement of the pharyngoesophagus (PE-), including the pharynx (Ph) and esophagus (E) subsites, detailing classifications for depth of invasion: muscularis only (Ph⁰ and E⁰) and full thickness including mucosa (Ph¹ and E¹).

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Involvement of vascular structures (VAS-), including the major arteries and veins, is highlighted in the next part of the checklist: the common carotid artery (A_CC), innominate artery (A_INN), internal jugular vein (V_IJ) and innominate vein (V_INN). The classification system delineates the degree of vessel involvement – from dissection in the subadventitial plane (superscript⁰) to complete transection or encasement (superscript¹).

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

View Full Size

The next part of the anatomic checklist shows involvement of the muscles of the neck (M-), including the sternocleidomastoid (SCM), strap (SM) and cricothyroid (CT) muscles. Muscle classification is binary, with no subclassification as to the degree of muscle involvement.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

Here, we describe the definitions and formatting of this classification system. The base initials used in the classification system designate the anatomic structure (e.g. T signifies trachea), and the superscript designates the degree of invasion. The superscript of either (⁰) or (¹) indicates the extent of surgery performed on that structure. Generally, the designation (⁰) indicates that the structure was involved by the cancer but dissected in a plane not requiring sacrifice (i.e. epineural layer, perichondrial layer and adventitial layer). There are two exceptions to this convention. For the pharyngoesophagus, if the muscularis is resected and the lumen is not entered, then the designation is also superscript (⁰) (e.g. Ph⁰ and E⁰). Alternatively, the superscript (¹) reflects that a portion or all of that structure was resected for tumor clearance and resulted in a transmural visceral defect. The involvement of various cervical muscles is a binary determination and does not require a superscript as to the degree of involvement.

Pharyngeal/esophageal defects (PE-)

Pharyngoesophageal (PE-) invasion is classified by site and degree of invasion (muscularis only versus muscularis extending to mucosa) (Fig. 5). Pharyngeal invasion limited to the muscularis is classified as Ph⁰. Full-thickness invasion extending into the mucosa is classified as Ph¹. Esophageal defects are similarly classified: muscularis only (PE-E⁰) versus extension into the mucosa (PE-E¹). Examples of anatomic staging are provided (Figs 8 and 9).

View Full Size

(A) Recurrent thyroid cancer in a central compartment lymph node (black arrow). (B) Resection required removal of the RLN as well as a portion of the cervical esophageal wall, leading to the creation of a transmural defect amenable to primary repair performed over a bougie seen in the lumen of the esophagus (white arrow). Defect classified as Rec N-RLN ¹ PE-E ¹ .

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

View Full Size

(A) Primary resection of invasive thyroid cancer with involvement of the internal jugular vein (black arrow, white trim) and the cervical esophageal musculature (solid black arrow), as well as the left RLN. (B) The specimen included a total thyroidectomy and central compartment lymph nodes that were removed in a subperichondrial plane involving two tracheal rings. Lateral compartment resection required resection of the internal jugular vein (white arrow). AJCC staging combined with anatomic information: pT4a N1b Mx PE-E⁰ Vas-V_IJ ¹ LT-T₂ ⁰ N-RLN¹.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

Data entry

In an effort to enhance the speed and efficiency of postoperative data entry by surgeons, the authors developed two options for accessing and completing the checklist before submitting it to the pathologist. Surgeons can utilize a printed version or an online version of the checklist. Both options are available at https://tiro.expert/analysis-tools. The website allows the surgeon to complete a de-identified electronic version of the checklist, wherein by checking the relevant items, the website automatically produces a final summary of the anatomic designations that can be forwarded to the pathologist and/or entered into the patient’s electronic health record (Fig. 10). In addition to the final anatomic classification, the output form includes both the AJCC and ATA-ROR classifications at the end to make these designations as easy as possible to incorporate into the final pathology report.

View Full Size

Electronically generated completed checklist output by surgeon B for patient 25. The patient is staged as pT4a N1b Mx N-RLN¹ M-SCM (sacrifice RLN, resection of the sternocleidomastoid muscle). The checklist includes a minimum data set, basic intraoperative findings per AJCC staging (8th edition) and ATA ROR classifications, clinicopathologic information and a visual representation of nerve and muscle invasion. A table of AJCC staging and ATA ROR is included at the end of each report to help facilitate staging and reporting.

Citation: European Thyroid Journal 14, 2; 10.1530/ETJ-24-0289

• Download Figure
• Download figure as PowerPoint slide

The surgeons informally kept track of the time required to complete the anatomic form and submit it to the pathologist.

Results

During the 24-month interval, 29 primary resections requiring AJCC staging and six recurrent cases were identified. For surgeon A, 12/13 final reports were understaged when an anatomic checklist was not submitted. In one case (#5), the sacrifice of the RLN was clearly stated in the operative report and appropriately noted in the final pathology report. In the latter part of the study period, surgeon A adopted the use of an anatomic checklist, where the final six cases were accurately staged (Fig. 1). For surgeon B, the index resection was understaged (Table 1, patient 20). This led to the development and implementation of this checklist initiative. The remaining nine patients were correctly staged.

In total, there were 14 final pathology reports for both surgeon A and B in which a checklist was not submitted. In all but one case, the disease process was understaged. A total of 15 cases (A and B combined) were accurately staged when accompanied by a surgeon-completed anatomic checklist (Fig. 1). There was a statistically significant improvement in the accuracy of staging as reported in the final pathology reports when an anatomic checklist was submitted as compared to when it was not (P < 0.01, Fisher exact test, two-tailed).

In the primary setting, the most commonly underreported structure was gross invasion of the pharyngoesophageal musculature (n = 8), followed by strap muscles (n = 7) and RLN sacrifice (n = 4). The cricothyroid muscle (n = 4) and tracheal perichondrium (n = 1) were the other structures missed in the primary cases.

In the six recurrent resections, gross nerve invasion requiring nerve sacrifice was not reported in four cases (RLN = 3 and vagus = 1). One recurrent case involved resection of the epineurium of the RLN (Table 2). All but one of the recurrences were located in the thyroid bed, leading to resection of all or a portion of the RLN in four cases, esophageal musculature in two, and laryngotracheal perichondrium in two. The pathology report for the one lateral compartment recurrence failed to report the resected carotid artery and vagus nerve. In all recurrent cases, a checklist was not incorporated, as it is not an obligatory reporting element to perform AJCC staging in recurrent cancer.

Both surgeon A and B reported that the time required to complete and send the checklist was as short as 30 s, and in no case did it require more than 90 s.

Discussion

Most thyroid cancers are low risk, and accurate staging usually does not require details of the preoperative and intraoperative findings. Surgeons and pathologists are often lulled into routines that require a jolt when dealing with grossly invasive thyroid cancer. For surgeon B, patient 20 (Table 1) was the impetus for implementing the anatomic checklist. For surgeon A, the adoption of the checklist was delayed until patient 14, after which all invasive cases were accompanied by a checklist. We conclude from this investigation that the use of a checklist is the most time-efficient process for surgeon-pathologist informational handoff. We envision its adoption as an addition to conventional tumor node metastasis (TNM) classification (e.g. Table 1 and Fig. 9). Preoperative information that is often known by the surgeon and entered into a clinic note, such as the results of molecular testing and the status of vocal cord function, has significant clinical and prognostic implications and is included in the anatomic checklist.

The missing anatomic detail on the recurrent thyroid cancer cases again highlights the gap between what the surgeon identifies intraoperatively and what the pathologist can confirm histologically. We have since adopted the checklist for both primary and recurrent cancers that demonstrate involvement of adjacent structures. We believe that the inclusion of anatomic structures resected and/or the designation of macroscopic (R2) or microscopic (R1) disease left in the tumor bed is important information for all clinicians to understand, as well as radiologists who will interpret post-treatment surveillance scans. An example of a completed electronic output form is shown in Fig. 10. At the end of that online report, a table of AJCC staging and ATA ROR is included to help facilitate staging understanding and inclusion in the final pathology report.

Currently, the responsibility for identifying that a particular case requires intraoperative anatomic detail usually falls upon the pathologist. However, there is no standard regarding when pathologists must seek preoperative and intraoperative details. The limitations of non-surgeons, not present in the operating room, to accurately interpret the operative report for data required for accurate staging have been previously documented (5). It is related to both the failure to include that information in the operative note and the failure of the non-surgeons to interpret what is dictated. There is cognitive dissonance regarding the absence of an established, efficient mechanism for information transfer from surgeon to pathologist, in the face of requiring accurate AJCC staging and ATA ROR determination so vital for contemporary clinical management.

Optimal multidisciplinary therapeutic planning for patients with grossly invasive thyroid cancer requires consideration of many factors (2). Parsing out the degree of invasion into surrounding structures is important (4). Documenting the extent of resected disease, as well as disease left behind (R2), is also important and is usually not evident to a pathologist. Cancer can be dissected away from multiple structures without leaving any histological evidence and therefore may not be included in the final pathology report, reflecting the need for this anatomic checklist. Unlike the surgeon’s operative report, it is the final pathology report that follows patients throughout their thyroid cancer journeys. This important document should unambiguously reflect the true extent of the disease. The best opportunity to document the intraoperative extent of the disease is immediately after surgery when the details of surgery are fresh in the surgeon’s mind. The speed of checklist completion was viewed very favorably by both surgeons A and B. When compared to the time required by pathologists to obtain intraoperative data by either analyzing the surgeon’s operative report or the need to call the surgeon to obtain that information, the overall efficiency was very evident. The ability to automatically generate an electronic classification form that is devoid of patient identifiers and can be emailed to the pathologists efficiently completes the information transfer. In addition, the relative rarity of grossly invasive thyroid cancer creates the opportunity to establish a registry to promote multi-institutional collaboration and case accrual with outcome data, which would further our understanding of these advanced thyroid malignancies.

Conclusion

The inherent complexity of invasive thyroid cancer mandates a multidisciplinary approach and requires a detailed understanding by all involved clinicians of the intraoperative findings and extent of surgery performed. Here, we demonstrate the benefit of an anatomic checklist to enhance surgeon-pathologist communication and improve final staging accuracy. The decision to include that checklist in the patient record will be made at an institutional level. This granular information regarding the precise location of residual disease and the location/extent of structural invasion will be invaluable for tracking surveillance scans and decision-making regarding adjuvant radiation therapy. This added value justifies its inclusion in the patient record and should accompany the permanent pathology report. We introduce a novel anatomic checklist with a convenient shorthand notation to provide uniformity. This classification system lends itself to data analysis, which could never be achieved by a pathologist gleaning information from the operative note.

The extent of anatomic detail will require further prospective multi-institutional studies to validate and/or modify the details included herein. Outcomes data are needed to determine the significance of including anatomic detail to the current TNM staging system. We hope that the collection of this data will promote and facilitate this type of research.

Declaration of interest

M U, M B-W, M Z, A Shaha, Z B, R M T serve as consultants to TIRO. The authors declare that there are no conflicts of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This research idid not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.

Author contribution statement

M U, M B-W contributed to conceptualization, writing original draft, data curation and formal analysis; R L C was responsible for data curation, formal analysis and reviewing and editing the manuscript; M Z, M S, J S, contributed to writing, reviewing and editing the manuscript; J F, A Silberzweig, J K J, contributed to reviewing and editing the manuscript; R G, A Shaha, Z B, R M T, contributed to conceptualization, reviewing and editing the manuscript.