H&N Unknown primary

Question 1

H&N cancers of an unknown primary represent what % of H&N cancers?

Answer 1

∼3%–5% of all H&N cancers are of an unknown primary.

Question 2

What is the most commonly presumed general site of origin for H&N cancers of an unknown primary?

Answer 2

The OPX is the presumed site of origin for most cases (less common are the NPX, hypopharynx [HPX], and larynx).

Question 3

What are the 2 most common originating sites/primary locations if the cancer is presumed to be of oropharyngeal origin?

Answer 3

Tonsils and base of tongue (BOT). Up to 80% of presumed oropharyngeal tumors are thought to originate from these 2 sites.

Question 4

Appx what % of pts with tonsillar primaries harbor Dz in both tonsils?

Answer 4

∼5%–10% of pts with tonsillar primaries harbor Dz in both tonsils.

Question 5

A primary can be identified in what % of H&N cancers of unknown primary?

Answer 5

A primary site of origin can be identified in ∼20%–40% of pts.

Question 6

What is the most common presentation for H&N cancers of an unknown primary?

Answer 6

Painless upper neck LAD (IB–III) is the most common presentation.

Question 7

What is the T staging if no primary H&N site is found after workup?

Answer 7

T0 (not TX) is the assigned T stage if no primary is found.

Question 8

On what are the overall stage groupings based if the primary is not known?

Answer 8

LN involvement and p16 status determine stage groupings, with p16(+) denoting more favorable prognostic stage groups than p16(-). (Please see
AJCC staging, 8th edition).

Question 9

What % of pts with an unknown primary present with bilat LAD (N2c)?

Answer 9

∼10% of pts present with bilat neck Dz.

Question 10

What does the workup include for pts with an unknown H&N primary (NCCN 2018)?

Answer 10

Unknown H&N primary workup: H&P including skin exam, EUA/panendoscopy + directed Bx with HPV testing, FNA of involved node, CT/MRI, PET/CT (before EUA), thyroglobulin, calcitonin, PAX8, and TTF
staining if adenoCa and/or anaplastic/undifferentiated, consider bilat tonsillectomy (bronchoscopy, esophagoscopy)

Question 11

If FNA is negative in H&N pts with an unknown primary, what other kind of nodal Bx can be attempted?

Answer 11

If FNA is negative, a core Bx can be attempted next. Avoid incisional/excisional Bx, b/c this would result in “neck violation.”

Question 12

What does cystic appearance of the involved LNs suggest in pts with H&N cancers?

Answer 12

Cystic appearance on imaging suggests HPV positivity/etiology.

Question 13

What is the significance of nodal location in terms of likely primary sites?

Answer 13

If upper neck nodes are involved, they are more likely to be d/t a H&N primary (e.g., if level I LNs, OC; if upper level V, NPX primary). If lower neck or SCV nodes are involved, they are more likely to be d/t a chest or
abdominal primary (below the clavicle). Bilat nodes suggest midline structures. Intraparotid LN involvement suggests a cutaneous primary.

Question 14

What is the significance of histology in terms of likely primary sites?

Answer 14

Squamous cell: more likely to be a H&N primary (upper aerodigestive mucosal axis)
Adenocarcinoma: more likely to be a chest or abdominal primary

Question 15

What sites are traditionally biopsied for level II nodal involvement?

Answer 15

The BOT, NPX, pyriform sinus, and tonsils are typically biopsied with level II LN involvement

Question 16

Why is it problematic to obtain the PET scan after endoscopy and Bx in pts with an unknown H&N primary?

Answer 16

Post-Bx inflammation may lead to false+ results. This is why some advocate that if used, PET scans should be done initially

Question 17

When is triple endoscopy indicated in pts with neck Dz and an unknown primary?

Answer 17

Triple endoscopy is generally done in pts with levels IV–V LAD (more likely to be lung/abdominal primary). Also, PET/CT C/A/P should be considered in such cases.

Question 18

Site-directed Bx will reveal the primary in roughly what % of unknown primary cases?

Answer 18

Site-directed Bx will reveal the H&N primary in ∼50% of cases.

Question 19

Unilat tonsillectomy will reveal the primary in appx what % of cases?

Answer 19

Unilat tonsillectomy will reveal the primary in ∼20%–25% of cases.

Question 20

PET/CT will reveal the primary in appx what % of unknown H&N primary cases?

Answer 20

PET/CT will reveal the primary in ∼15%–20% of cases.

Question 21

What do the data show in regard to bilat tonsillectomy for pts with an unknown H&N primary?

Answer 21

Data from the JHH (McQuone S et al., Laryngoscope 1998) showed improved diagnostic yields with bilat tonsillectomy. Additionally, it may render f/u with PET/CT easier.

Question 22

When should tonsillectomy be performed in pts being worked up for an unknown primary?

Answer 22

Tonsillectomy is generally performed at the time of direct laryngoscopy.

Question 23

What are the approximate predictive values of PET for pts with an unknown primary?

Answer 23

The PPV is ∼90% and the NPV is ∼75% for pts with an unknown primary.

Question 24

What % of pts with an unknown H&N primary have metastatic Dz on PET/CT?

Answer 24

∼10% have metastatic Dz on PET/CT—yet another reason to consider upfront PET.

Question 25

What is the general Tx paradigm for H&N cancers if a primary is found vs. if there is an unknown primary?

Answer 25

H&N cancer Tx paradigm:
If primary found: treat according to the primary location
If no primary found: Sg +/– RT, RT alone, or chemo/RT +/– neck dissection

Question 26

Which unknown primary pts can be treated with neck dissection alone?

Answer 26

Generally, N1 (<3 cm) without ECE. B/c of this, some advocate upfront neck dissection at the time of direct laryngoscopy. (Coster JR et al., IJROBP
1992)

Question 27

For which H&N pts is upfront neck dissection a reasonable approach?

Answer 27

Upfront neck dissection is reasonable if better staging is desired (e.g., if the path is unclear), if the neck has been “violated” (i.e., after incisional Bx), and with a small, unilat, single +node (N1).

Question 28

What % of pts with N1 Dz fail at the primary site after neck dissection alone?

Answer 28

∼25% of N1 pts ultimately fail at the primary site after neck dissection alone. However, this can vary from 10%–50%.

Question 29

What is the approximate overall neck failure rate after neck dissection alone?

Answer 29

The overall neck failure rate is ∼15% after neck dissection alone. (Coster JR et al., IJROBP 1992)

Question 30

What pathologic factor is associated with the highest risk of Tx failure for pts with an unknown primary?

Answer 30

ECE is associated with the highest risk of Tx failure in these pts.

Question 31

What is the approximate neck failure rate after neck dissection if there is evidence of ECE?

Answer 31

The approximate neck failure rate after neck dissection alone is ∼60% with ECE. (Coster JR et al., IJROBP 1992)

Question 32

What are the indications for PORT in pts with an unknown H&N primary?

Answer 32

≥N2 Dz, ECE/+margin, or neck violation (e.g., after open/excisional Bx)

Question 33

What do the standard RT fields include in pts with an unknown H&N primary?

Answer 33

The fields generally include both neck and the mucosal sites at risk (NPX, OPX, HPX, larynx). Some advocate omission of the HPX/larynx from the RT fields, especially if HPV+.

Question 34

What are the historical 5-yr LC and OS rates after definitive RT for pts with an unknown H&N primary?

Answer 34

University of Florida data (Erkal HS et al., IJROBP 2001): LC 78% and OS 47%
Danish data (Grau C et al., Radiother Oncol 2000): OS 37%

Question 35

What factors have been traditionally associated with inf OS after definitive RT for H&N tumors of an unknown primary?

Answer 35

More advanced N stage, ECE, and lower RT doses have been associated with inf outcomes. (Erkal HS et al., IJROBP 2001)

Question 36

What standard RT fields have been traditionally used for H&N tumors of an unknown primary?

Answer 36

Opposed lats matched with an ant low neck/SCV field (with post neck electron fields after 40–44 Gy). However, conformal techniques such as IMRT should be now used as standard of care.

Question 37

What were the anatomical borders of the traditional setup for the 2D lat fields used for H&N tumors of an unknown primary?

Answer 37

Anterior: behind OC/hard palate
Superior: to base of skull to include NPX
Posterior: below tragus to post edge of spinous processes
Inferior: sup edge of thyroid cartilage; if level III or IV, inf edge of cricoid to cover larynx

Question 38

What definitive RT doses are generally employed?

Answer 38

1. CTV1: 66 (2.2. Gy/fx) or 70 Gy (2 Gy/fx) to gross Dz with margin (5–8
   mm)
2. CTV2: 50–66 Gy (2 Gy/fx) to intermediate-risk areas 1 nodal station above
   and below in ipsi side and putative primary mucosal sites (NPX, OPX,
   HPX)
3. CTV3: 50–54 Gy to low-risk areas and uninvolved neck

Question 39

What evidence supports the omission of the larynx/HPX from the standard RT fields?

Answer 39

University of Florida data (Baker CA et al., Am J Clin Oncol 2005): larynx sparing RT is just as effective with less toxicity.

Question 40

What is the evidence in favor of bilat neck irradiation for H&N tumors of an unknown primary?

Answer 40

Loyola data (Reddy SP et al., IJROBP 1997): contralat nodal failure is higher (44%) in pts receiving unilat nodal RT (vs. 14% for bilat nodal RT).
Also, there is a higher primary emergence rate with unilat RT (44% vs. 8%). No difference in 5-yr OS.

Question 41

What are a few of the advantages of IMRT for H&N tumors of an unknown primary?

Answer 41

Greater parotid sparing, can consider concurrent chemo (Klem ML et al., IJROBP 2008), dose painting to avoid sequential CDs, can use SIB dosing (e.g., 212 × 33 = 69.96 Gy, 180 × 33 = 59.4 Gy, and 170 × 33 = 56.1 Gy).

Question 42

When is neck dissection entertained after definitive RT for H&N tumors of an unknown primary?

Answer 42

Post-RT neck dissection is considered with persistence of Dz (e.g., on PET or clinically; LN >1 cm and/or PET+). Some still consider it standard for all pts with ≥N2 Dz, although more commonly, elective neck dissection after RT is not performed if there is no clinical evidence of Dz on clinical exam and radiographic restaging.

Question 43

Within what timeframe after RT should neck dissection be performed if decided upon upfront (i.e., regardless of response to RT)?

Answer 43

Neck dissection should occur ∼3–4 mos (and no later than 6 mos) after RT.

Question 44

What are common acute side effects from RT to the H&N region?

Answer 44

Pain, mucositis, hoarseness, and malnutrition (weight loss).

Question 45

What are common long-term complications from RT to the H&N region?

Answer 45

Xerostomia, dysphagia, neck scarring and edema (especially if combined with neck dissection), hypothyroidism, and laryngeal dysfunction (aspiration,
hoarseness, etc.).

Question 46

After RT, when should PET be performed to assess for nodal response?

Answer 46

PET should be performed no sooner than 3 mos after RT. (NCCN 2018)