[Ruth Li]

Update: the following pairs all start at the same position, so they may be interchangeable fields that are located in the same place in the NAACCR layout file, and may not be an issue.
State at DX Geocode 2000 & GeoLocationID – 2000
State at DX Geocode 2010 & GeoLocationID – 2010
State at DX Geocode 2020 & GeoLocationID – 2020
Subsq RX 2nd Course Surg & Subsq RX 2nd Course Codes
Subsq RX 3rd Course Surg & Subsq RX 3rd Course Codes

So it seems there is only 1 set of fields that are flipped, which is what you found:
Number of Examined Para-Aortic Nodes
Number of Positive Pelvic Nodes

I checked our v18 test software from CDC (Prep Plus) and the layout in there is the same as in chapter 7, which is # Examined Para-Aortic Nodes in position 1807 and # Positive Pelvic Nodes in position 1809.

There are 16 EDITS that include EITHER one of these 2 data fields.

[Ruth Li]

Hi Monica,

Thank you so much for finding this! I did a comparison of Chapter 7 (record layout) that is on the NAACCR website with an export of the Layout that is in v18C metafile and found the following:

v18C layout has 9 fields that are NOT on the Chapter 7 layout:
(start position 1647, same as NPCR specific field) EDP MDE Link
(start position 1648) EDP MDE Link Date
(start position 1656) Height
(start position 1658) Weight
(start position 1661) Tobacco Use Cigarettes
(start position 1662) Tobacco Use Other Smoke
(start position 1663) Tobacco Use Smokeless
(start position 1664) Tobacco Use NOS
(start position 1665) Source Comorbidity

Not accounting for the above, there are 6 “pairs” of swapped fields:
State at DX Geocode 2000
GeoLocationID – 2000
State at DX Geocode 2010
GeoLocationID – 2010
State at DX Geocode 2020
GeoLocationID – 2020
Number of Examined Para-Aortic Nodes
Number of Positive Pelvic Nodes
Subsq RX 2nd Course Surg
Subsq RX 2nd Course Codes
Subsq RX 3rd Course Surg
Subsq RX 3rd Course Codes

See attached screenshot and Excel spreadsheet – green columns are from chapter 7, I did a check each for position, length, item#, and name.

Attachments:

You must be logged in to view attached files.

[Ruth Li]

Clarification – the edit was updated for v18.

[Ruth Li]

Response from NAACCR:

It looks like TNM Path N, Reg Nodes Ex – Ed 7 (CoC) needs to be updated to allow cN1 and cN2 when regional nodes examined are 00 or 99. Personally, I would leave the edit in the metafile. I think the circumstance you described is pretty rare.

When it does arise, I, I would have the registrar leave the pN data item blank, but calculate the stage as if a cN1 or cN2 is in the data item. This is a valid coding option according to AJCC and all of the other standard setters and it will pass all edits.

At this point we do not plan on releasing a v16f. However, if we do, this edit will be corrected. It will also be correct for pre-2018 cases when we release the v18 metafile.

[Ruth Li]

Responses from NAACCR:

To the question “this edit is not listed in any of the edit sets. Is there a reason it was left out?”:

That was an oversight. We skipped the process of polling the standard setters to see if they wanted the edit in their edit set.

CoC, NPCR, and SEER are all on the Edits WG and they approved the edit. NPCR modified their .dll to accommodate these changes as well.

To the question ““TNM Path cN0 2016 (CoC)” edit, when I tested… it passed when it shouldn’t for cutaneous melanoma where pT is NOT pT1A”:

It does look like we have a problem with the logic. It is letting any pT value pass with a cN0. We will look into and should have a fix for v18. If you like we can send you the logic to correct the edit on your end. [Update] Error can be corrected by removing 31 from the following logic statement in the edit:

if (INLIST(#S”TNM Path N”, “0”, “cxbb”, 2,3))
if (INLIST(#S”TNM Path T”, “A”, “pxbb”, 2,3) or
INLIST(#S”TNM Path T”, “IS”, “pxxb”, 2,3) or
INLIST(#S”TNM Path T”, “ISU, ISD”, “pxxx”, 2,3) or
INLIST(Sitegrp, “16A,16B,027,028,31,36A”))
return PASS;

[Ruth Li]

Response from NAACCR:

It looks like this is an edit issue. Edit updated, both SSF 8 and SSF 10 set to 000 if = to 998, 988, or 999

[Ruth Li]

Per the CAnswer Forum discussion, I believe the edit in question is actually “Primary Site, TNM CLIN Stage Valid B-Ed 7 (COC)”.

The way the edit is in v16E, according to the table it references (CLST7VAL), it actually allows for a stage grouping of 1A instead of 99 when blood involvement is unknown (BX) (attached screenshot #1). I tested it with the TNM staging info in the CAnswer forum thread and it passed the edit with stage group as 1A (attached screenshot #2). In the administrative notes, the reference table was last updated for 57B (primary cutaneous lymphoma) in v16D. Not sure why, for the edit, stage grouping is allowed instead of 99 when peripheral blood involvement is unknown?

Ruth

Attachments:

You must be logged in to view attached files.

[Ruth Li]

For regional nodes positive: per CS for breast http://web2.facs.org/cstage0205/breast/Breast_fab.html , Note 1 states “Record this field even if there has been preoperative treatment.” So that would indicate the regional nodes positive would, in this case, include the ALND that happened POST-chemotherapy.

Donna Gress at the CAForum confirmed that the pathologic N category for this case is pN1a: http://cancerbulletin.facs.org/forums/node/73735 She points out this is because staging rules are different from 1st crs of tx rules.

Suggestion for the edit logic: in this case scenario, I believe the systemic/surgery sequence is coded 7 [surgery both before and after systemic therapy – systemic therapy was administered between 2 separate surgical procedures to the primary site; regional LN; surgery to other regional site(s), distant site(s), or distant LN(s)] and scope of regional lymph node surgery would also be a 7 [sentinel node bx & code 3,4,6 at different times]? Could one or both of these fields be included in consideration in the edit logic so an exception can be made when LN surgery happens before and after systemic tx such that staging of LN and examination of LN may not correspond directly?

[Ruth Li]

Mystery solved! The latest version of the Registry Plus suite of software (POST Feb 2017) is needed to run with 16E edits.

[Ruth Li]

I did a query for 2016 GIST cases we have in pending (we also use CRS Plus) and consolidated one. I think I was able to replicate the error you are seeing. Is the attached screenshot the invalid Path N error that you see? If so, that is a CRS Plus error message, and not an edit related error.

We don’t have the latest CRS Plus release yet (we are still on the 2/27/17 version), but according to an email from CDC on 6/12, since the Feb 2017 release, the latest TNM DLL has additional clinical N values added to TNM Path N for all GIST schemas. If you have the latest version of the Registry Plus suite of software, but are still seeing this error, then it seems something’s wrong with the latest TNM DLL…

Attachments:

You must be logged in to view attached files.

[Ruth Li]

In our software system the valid values does not contain the “T”, “N”, “M”, so it would be just “c0” in the pN field instead of “cN0″. Does just entering “c0” resolve the error?

[Ruth Li]

There is an AJCC schema for GIST tumors of the stomach – chapter 16 in the 7th edition manual. It covers histology codes 8935 and 8936 for ICD-O-3 topography codes for esophagus, stomach, small intestine, colon, recto-sigmoid junction, rectum, and retro-peritoneum & peritoneum. So the TNM fields should not be 88 for these cases unless they are in situ.

You mentioned that cN0 was not an option, is that within the registry software? We use Web Plus and we do currently see a cN0 option (attached screenshot), although not a cN1 option. That may be because we don’t have the latest TNM DLL yet, but 16E edits should allow for cN1 also in the pN category for these cancers.

Attachments:

You must be logged in to view attached files.

[Ruth Li]

We have not ran into an issue with this edit (yet?). What is the site/histology combination and behavior code of the case(s) where you are seeing the error?

[Author]

Posts