Staging

“Breast carcinoma is a heterogeneous disease. Classic prognostic and predictive factors for staging include patient age, tumor grade, histologic subtypes, hormone receptor [HR] and HER2 status, vascular invasion, and tumor margins. We also look at markers of proliferation and possibly tumor infiltrating lymphocytes,” explained Aysegul A. Sahin, MD, Director of Educational Operations, and Professor, Department of Pathology, The University of Texas MD Anderson Cancer Center. According to the 7th edition of the AJCC Cancer Staging Manual, breast cancers were classified according to T (tumor), N (node), and M (metastasis).

“There is major biological diversity in breast cancer. In general, small tumors have a significantly better prognosis compared with node-positive or larger tumors or those with multiple lymph node metastases. But the same stage of breast cancer can behave differently and have different responses to treatment based on tumor biology,” she noted. “Survival within each stage shows wide variation.”

Different subtypes of breast cancer are associated with different prognoses. For example, HER2-positive (HER2+) and basal-like tumors have a worse prognosis than luminal estrogen receptor–positive (ER+) tumors. Breast cancer can have the same TNM stage, but ER status, progesterone receptor (PR) status, and HER2 status will affect prognosis.

“The 7th edition of the AJCC staging system did not take biological factors into account. Yet treatment recommendations and subsequent responses to therapy are dictated by biological factors,” Dr. Sahin told the audience. The 8th edition of the AJCC staging system retains TNM staging as the basis but is considered an improvement because it incorporates biological markers into prognostic groups that recognize tumor biology.¹

Dr. Sahin and colleagues at MD Anderson Cancer Center developed a prognostic model that incorporates tumor biology, tumor grade, and ER status based on approximately 3,800 patients with invasive breast cancer treated at their institution between 1997 and 2006. The model was externally validated in a SEER dataset of 26,711 cases of invasive breast cancer.² The model had limitations, including use of retrospectively collected data with treatment not assigned, coding errors were possible in a population-based validation dataset, and the model was developed in the pre-trastuzumab era. To adjust for the missing data on trastuzumab, the dataset was updated to incorporate trastuzumab-treated HER2+ breast cancer.³

“The winners in the updated dataset were stage, grade, ER/PR status, and HER2 status” she said.

The 8th edition of the AJCC staging system incorporated these biomarkers and molecular subtyping for invasive breast cancer. When available, multigene panels are included as stage modifiers for patients with HR-positive (HR+), HER2-negative (HER2–), lymph node–negative breast cancer: the 21-gene Oncotype Dx recurrence score, 70-gene MammaPrint risk score, 12-gene EndoPredict molecular score, and the 50-gene PAM50 risk of recurrence score.¹

To validate the 8th edition of AJCC, the new staging system was compared with that of the 7th edition. Forty percent of all patients with stage I–III invasive breast cancer according to the 7th edition were restaged based on the staging system in the 8th edition. Prognostic staging was upstaged in 31% of patients and downstaged in 20.6%,⁴ she said.

Additional Clarifications in the 8th Edition

For clinical practice, additional clarifications are added to the 8th edition of the AJCC staging system to define pT and pN status more clearly.

“A major improvement in the AJCC staging 8th edition is clarification of the pathologic evaluation of both T stage and N status using a lot of illustrations and descriptive explanations,” Dr. Sahin said.

“Maximum invasive tumor size is a reasonable estimate of tumor volume for the primary tumor. Most of the time, clinical tumor size is similar to pathologic size, but there can be major discrepancies between clinical and pathologic size,” Dr. Sahin explained. Gross measurement of tumor size, confirmed through microscopic assessment, does not include small microscopic satellite foci around the primary tumor. The definition of tumor by size did not change from the 7th edition. However, a change to the staging system is that lobular carcinoma in situ is no longer considered a malignancy and is removed from staging as Tis.

“If there are multiple foci of microinvasion, the size of only the largest focus is used to classify tumor size. The sum of the individual foci is not used,” she stated.

Inflammatory breast carcinoma is a clinical and pathological diagnosis. This subtype of breast cancer is characterized by diffuse erythema and edema involving approximately one-third or more of the skin around the breast. This type of breast cancer is aggressive and rapidly evolves from first symptom to diagnosis.

Multiple synchronous tumors are identified clinically and/or macroscopically. The presence of these are documented by the modifier of the T category. The maximum dimension of the largest tumor is used for clinical (cT) and pathologic (pT) classification. The size of multiple tumors is not added. Post-neoadjuvant therapy pathologic T category (ypT) is based on the largest focus of residual tumor. Adjacent fibrosis is not included in the ypT maximum dimension. When multiple foci of residual tumor are present, the (m) modifier is included.

The largest contiguous tumor deposit is used to measure pathologic nodes (pN). If a cancer is categorized as M1 (clinical or pathologic) before therapy, the cancer is still categorized as M1 regardless of the observed response to neoadjuvant therapy. Any residual invasive carcinoma detected by pathologic examination in the breast or lymph nodes precludes posttreatment classification as a complete pathologic response.

Challenges of New Staging

“The changes in the 8th edition of the AJCC Cancer Staging Manual are more helpful to pathologists than clinicians,” stated Timothy D. Gilligan, MD, Vice Chair for Education, and Associate Professor of Medicine, Cleveland Clinic Taussig Cancer Institute. Dr. Gilligan reviewed the changes and emphasized areas of confusion and uncertainty in the newer version of staging.

“We stage tumors to determine prognosis and the most appropriate treatment options, as well as to facilitate high-quality research. Staging has to be associated with outcomes and to help us make clinical decisions. So it may not be clinically relevant whether T stage is associated with N and/or M stage,” he said.

“What I would like to know in testicular cancer is which patients with stage I disease will experience relapse on surveillance. Will patients with stage II disease experience relapse if treated with retroperitoneal lymph node dissection only? Which patients with stage II and III disease will need to undergo resection of residual masses after chemotherapy?” he continued.

Dr. Gilligan explained that the 8th edition has made some changes, but that several of these changes are not helpful in clinical practice (Figure 1).

View Full Size

Changes in testis cancer staging.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 17, 5.5; 10.6004/jnccn.2019.5015

“The most bizarre change is related to placing discontinuous spermatic cord involvement in the pM1 category,” he told the audience. “We used to call invasion of the spermatic cord ‘T3.’ Now the new system says that noncontinuous invasion is considered a metastasis, but I don’t think in clinical practice people will treat it as if it is metastatic.”

Four Key Questions

“It is questionable whether these changes help us manage patients better,” Dr. Gilligan commented. Issues to be resolved include:

• Should patients with stage I seminoma with tumors <3 cm be treated differently from those with larger stage I tumors?

• Should patients with clinical stage I disease with invasion of the epididymis be treated differently?

• Where does hilar soft tissue come in?

• Should patients with spermatic cord metastasis be treated differently from those with only spermatic cord invasion?

“The rationale for staging seminomas based on size of the primary tumor is pretty clear,” Dr. Gilligan said. “Seminoma size is clearly linked to relapse rates: both size and lymphovascular invasion predict relapse. We think all of these patients should undergo surveillance unless there is compelling reason not to. However, patients with larger tumors may prefer radiation or chemotherapy if they know they have a higher risk of relapse. So tumor size can facilitate shared decision-making.”

The other area tumor size is informative is in patients with clinical stage I who are treated with a single cycle of carboplatin. Size predicts relapse after carboplatin, according to results of the SWENOTECA study, and large tumors were associated with unacceptably high relapse rates.⁵

“In light of these data, a single cycle of carboplatin does not make a lot of sense. In my opinion, based on the results of nonrandomized, uncontrolled studies, you need 2 cycles of carboplatin for high-risk disease. The data on 2 cycles are more reassuring, showing a relapse rate of about 2%,” Dr. Gilligan said. “An open question is whether rete testis invasion will be added to testis cancer staging in the future.”

“Invasion of the epididymis is a more confusing area,” he said. Although older data show that epididymal invasion is associated with higher stage at the time of diagnosis, a multivariate analysis found that epididymal invasion was not consistently associated with relapse.

More recently, a small study of 56 patients showed that the presence of epididymal invasion was associated with higher risk of relapse.⁶ But a separate study of 94 patients undergoing surveillance found that epididymal invasion was present in only 3.2% and was not associated with relapse.⁷ Risk factors for relapse in this study were tumor size, lymphovascular invasion, and rete testis invasion.

“Epididymal invasion is rare and rarely reported. Why add this to staging? The rationale was that the cancer can’t invade the epididymis without first invading the hilar soft tissue, so if hilar soft tissue invasion constituted T2 disease, then epididymal invasion should, too,” he stated. “Some of these changes were made to try to reduce the huge variation in how testicular tumors are staged by practicing pathologists. Epididymal invasion was made pT2 largely to reduce the risk that patients would be understaged.”

Invasion of hilar tissue appears to be a marker of more aggressive cancer. A multivariate analysis of a study of 148 consecutive patients with nonseminomatous germ cell testis cancer identified 3 significant risk factors for stage II or III disease at the time of diagnosis: vascular invasion (P=.011), invasion into the rete testis (P=.007), and invasion into the hilar soft tissue (P=.017).⁸ However, it is not clear whether invasion of hilar soft tissue is associated with risk of relapse for patients with clinical stage I disease.

There are unresolved issues around spermatic cord invasion. In the 8th edition, lymphatic vascular invasion of the spermatic cord is pT2, whereas invasion of the spermatic cord parenchyma is pT3. “It is unclear whether the distinction between invading vessels and parenchyma have different prognostic implications,” Dr. Gilligan continued. “We don’t see this often, and small studies suggest no difference in outcome between these 2 features.”

Another unclear change is the distinction between continuous versus discontinuous involvement in vascular spaces and designating discontinuous involvement as stage pM1, and thus stage III. “We have no persuasive data illuminating whether discontinuous spermatic cord invasion should be managed as stage III in the absence of elevated markers or radiologically detectable distant metastases,” Dr. Gilligan said. “This change is a mystery to clinicians. We don’t know why it was changed.”

A major problem that has clinical implications is a large variation in staging by pathologists. According to a recent study, practicing pathologists, including genitourinary pathologists, display a huge variation in the staging of testicular cancer, especially when hilar soft tissue, epididymal invasion, and tunica vaginalis are reported.⁹ This study showed that for a tumor with hilar soft tissue invasion, 48% of pathologists described it as stage T1, 25% as T2, and 27% as T3. For epididymal invasion, the rates for T1, T2, and T3 were 83%, 14%, and 3%, respectively.

“This is alarming. I was surprised by the variation found in this large, well-done study,” Dr. Gilligan commented. “Clinicians could be helped if there was less variation in how similar cancers are staged.”

Take-Away Messages

Dr. Gilligan’s main take-away messages regarding the 8th edition of the AJCC Cancer Staging Manual are as follows:

• We don’t know whether invasion of the epididymis or hilar soft tissue predicts relapse of patients with stage I on surveillance;

• We don’t know whether lymphovascular invasion in the spermatic cord carries a worse prognosis than lymphovascular invasion in the testis; and

• We don’t know whether discontinuous invasion of the spermatic cord affects the prognosis for patients with stage I, stage II, and stage III disease more than simple invasion of the spermatic cord.

“It is questionable whether we should alter our management of these patients based on these histopathologic findings,” Dr. Gilligan stated. “We need better studies to answer these and other questions.”

Staging

Over the past 3 decades, human papillomavirus (HPV)–positive status has been recognized as a new etiologic factor for cancer of the oropharynx. The number of HPV-positive cancers now surpasses HPV-negative cancers, and the incidence of HPV-associated oropharyngeal cancer is predicted to increase considerably over the next 10 years.

This is not all bad news, explained Jimmy J. Caudell, MD, PhD, Associate Member, Head and Neck Program, Moffitt Cancer Center. ”Patients who are HPV-positive have significantly better progression-free survival and overall survival than those who are HPV-negative. P16 positivity is a prognostic factor,” he said.

“The 7th edition of AJCC worked well for HPV-negative oropharyngeal cancers, but lost prognostic significance for HPV-positive cancers,” Dr. Caudell said. “We needed to improve staging for patients who are HPV-positive, who constitute the majority of patients with head and neck cancer.”

For the 8th edition, a model for a new staging system of HPV-positive head and neck cancers was developed based on ICON-S data.¹⁰ Because survival was similar for stages T4a and T4b, those categories were collapsed into T4. Nodal categories did not perform well, so they were also collapsed, and N1, N2a, and N2b all became N1.

“The model performed quite well in a training set and in a 1,900-patient validation group. We also had external validation from several institutions,” Dr. Caudell said.

Challenges of New Staging for HPV-Positive Cancers

“The new staging system changed the conversation with patients,” he said. “We used to explain that stage 4a/4b cancers were incurable. Now I have the pleasure of telling these patients that they have a better prognosis with current treatments. We also have to discuss with patients how they acquired HPV, and the potential need for multimodality treatments for many patients with stage I disease,” Dr. Caudell continued.

Recent data suggest that deintensification strategies could be considered for patients with stage I disease who have an 85% to 94% 5-year overall survival rate. Patients may be selected according to age, performance status, and extranodal extension.¹¹ The type of deintensification matters, because large trials have shown the superiority of cisplatin over cetuximab, Dr. Caudell commented.

Pathologic staging has changed along with clinical staging for HPV-positive oropharyngeal cancer. In the development model, pathologic staging based on the 7th edition was not prognostic but T staging was, and the number of positive lymph nodes was prognostic. Caveats to pathologic staging include the fact that most cancers in the model were early-stage and amenable to robotic or laser surgery. Unaccounted for risk factors include lymph node size, bilateral lymph nodes, perineural invasion, extranodal extension, and vascular invasion. The new model currently does not provide guidance for the selection of adjuvant radiation ± chemotherapy.

“Although we have a new staging system for HPV-positive head and neck cancers in the 8th edition of AJCC, current treatment guidelines have not changed. We await the results of de-escalation trials,” Dr. Caudell stated.

A footnote in the 2019 NCCN Guidelines for Head and Neck Cancers states that research is rapidly evolving regarding the impact of adverse features such as extranodal extension in patients with HPV-positive oropharyngeal cancer who have undergone resection.¹²

HPV-Negative Oropharyngeal Cancer

The 8th edition of the AJCC Cancer Staging Manual includes some changes for HPV-negative oropharyngeal cancer as well. Extranodal extension is recognized as a known high-risk postoperative prognostic factor, and is considered an indication for intensive adjuvant therapy.

Radiographic predictive power for extranodal extension has uncertain reliability. Unambiguous clinical evidence of extranodal invasion includes invasion of the skin, fixation to adjacent structures on clinical examination, and cranial nerve or brachial plexus invasion with dysfunction.

For HPV-negative cancers, the T category is unchanged; the T4a/T4b categories are retained.¹³ Separate pathologic criteria are now presented for involvement of regional lymph nodes.

“Extranodal extension is dependent on pathology,” Dr. Caudell suggested. Caveats for HPV-negative staging include the fact that patients included in training and validation datasets were primarily those with oral cavity tumors. The impact on current treatment paradigms is not clear, and the role of intensification strategies remains to be established.