Robert O. Edwards, Jr. v. Douglas A. Collins
CourtUnited States Court of Appeals for Veterans Claims
Date FiledAugust 4, 2026
Docket24-6641
StatusPublished
📰 News Coverage: Read the LAWS.com news report on this case
Full Opinion
Case: 24-6641 Page: 1 of 43 Filed: 08/04/2026
UNITED STATES COURT OF APPEALS FOR VETERANS CLAIMS
No. 24-6641
ROBERT O. EDWARDS, JR., APPELLANT,
V.
DOUGLAS A. COLLINS,
SECRETARY OF VETERANS AFFAIRS, APPELLEE.
On Appeal from the Board of Veterans' Appeals
(Argued April 15, 2026 Decided August 4, 2026)
Stephanie M. Leacock, with whom Glenn R. Bergmann was on the brief, both of Rockville,
Maryland, for the appellant.
Jonathan G. Scruggs with whom Danielle A. Runyan, Acting General Counsel; David L.
Quinn, Assistant Chief Counsel; Mark D. Vichich, Deputy Chief Counsel, Emily K. Cincinnati,
and Jack A. Jorgest were on the brief, all of Washington, D.C., for the appellee.
Before ALLEN, Chief Judge, and FALVEY and JAQUITH, Judges.
FALVEY, Judge, filed the opinion of the Court. JAQUITH, Judge, filed a concurring
opinion.
FALVEY, Judge: Scientific reports created under the mandate of the Agent Orange Act of
1991 remain an important source of evidence in the veterans benefits adjudication system. This
case presents the issue of whether these reports could be constructively of record given varied
statutory and regulatory changes in the modernized VA appeals structure. The Court must also
decide whether the agency of original jurisdiction (AOJ) or the Board of Veterans' Appeals can
consider independent evidence submitted through past Board decisions that are unrelated to the
given appeal.
Our answer to both questions is yes. Nothing prohibits the possibility that the Agency was
in constructive possession of a report while the AOJ was receiving evidence. And information
from an independent and relevant source is not diluted because it happens to be communicated
through another veteran's Board decision. With this established, we find that not all the pertinent
evidence was considered in the present appeal. Further discussion by the Board is consequently
necessary, so we will remand the matter for additional adjudication.
Case: 24-6641 Page: 2 of 43 Filed: 08/04/2026
I. RESPECTED REPORTS
Retellings of the Agent Orange Act and its residuum are plentiful. Euzebio v. McDonough,
989 F.3d 1305, 1310-15 (Fed. Cir. 2021); LeFevre v. Secretary of Veterans Affairs, 66 F.3d 1191,
1193-96 (Fed. Cir. 1995); Aviles-Rivera v. McDonough, 35 Vet.App. 268, 271-72 (2022). But it is
a narrative worth repeating before reciting the facts and evidentiary concerns that brought us here.
This thread commences with the Veterans' Dioxin and Radiation Exposure Compensation
Standards Act, which required VA to resolve Agent Orange claims for benefits based on "exposure
during service" in Vietnam. Pub. L. No. 98-542 at § 5(a)(1)(A), 98 Stat. 2725 (1984). To meet this
requirement, the Act explained that a panel of agency experts called the Dioxin Council would
communicate to the Agency various scientific findings regarding possible health risks attributed
to herbicide exposure. Id. at § 5(b)(1)(A)-(B). Based on the Council's evaluation of the evidence,
VA would then promulgate regulations providing presumptive service connection for disabilities
with a high statistical connection to exposure. Id. at § 5(b)(2).
The Dioxin Act's tenure was ultimately fleeting. Over its lifespan, the Council found that
just one disease, chloracne, had a "cause-and-effect relationship" with herbicide exposure such
that VA found service connection presumptively appropriate. See 50 Fed. Reg. 34,458 (Aug. 26,
1985); Nehmer v. U.S. Veterans' Admin., 712 F. Supp. 1404, 1408 (N.D. Cal. 1989). And,
following a class action lawsuit, the Agency's cause-and-effect test was jettisoned for an inquiry
into whether "a significant statistical association" existed between a given disease and exposure.
Nehmer, 712 F. Supp. at 1409; see 54 Fed. Reg. 40,391 (Oct. 2, 1989).
Congress dissipated remaining concerns—and the Dioxin Council—through the Agent
Orange Act. Rather than rely on a collateral cadre of VA experts to classify connections between
conditions and exposure, the Act directed the Agency to outsource this responsibility to an
unaligned organization: the National Academies of Sciences, Engineering & Medicine (NAS).
Pub. L. No. 102-4, 105 Stat. 11, at § 3(a)-(b) (1991) (codified in part at 38 U.S.C. § 1116). Under
this partnership, the NAS produces reports that routinely review and summarize scientific evidence
before gauging the strength of any association between various disabilities and herbicide exposure.
Id. at §§ 3(c), 3(g). The Secretary takes it from there and determines whether presumptive service
connection is warranted for diseases discussed in the NAS reports. Id. at § 3; see 38 U.S.C.
§ 1116(b)-(c).
2
Case: 24-6641 Page: 3 of 43 Filed: 08/04/2026
As the Federal Circuit once explained, "[t]he importance and relevance of the NAS reports
to Agent Orange claims are well-known and well-established—they are the result of decades of
veteran engagement." Euzebio, 989 F.3d at 1320. We have likewise emphasized the important role
that NAS reports play in cases regarding Agent Orange exposure. See, e.g., Davis v. McDonough,
36 Vet.App. 142, 153 (2023).
Although the reports are famously used as a cornerstone of presumptive service-connection
determinations, they might also be used as evidence in direct service-connection arguments where
Agent Orange exposure is alleged. And their publication in the Federal Register apprises the
Agency as to the findings published therein. Euzebio, 989 F.3d at 1314 (citing Fed. Crop Ins. Corp.
v. Merrill, 332 U.S. 380, 385 (1947) ("Congress has provided that the appearance of rules and
regulations in the Federal Register gives legal notice of their contents.")). Those findings may then
implicate the Secretary's duty to assist. Id. (citing 38 U.S.C. § 5103A(d)(2)(b) (requiring the
Agency to provide an exam when evidence "indicates that the disability or symptoms may be
associated" with service)).1
II. A TALE OF TWO UPDATES
This case concerns two NAS reports—Updates 2010 and 2012—and the extent to which
evidentiary restrictions prevent their consideration. The facts of this case begin as one might
expect. Robert O. Edwards, a Navy veteran, was exposed to herbicides during his service in the
Vietnam War. Record (R.) at 2603, 5211. He was later diagnosed with squamous cell carcinoma
of the right tonsil.2 R. at 2941. Mr. Edwards filed a claim for that condition in November 2019,
alleging that the cancer was caused by his exposure to Agent Orange. R. at 2972. Because VA's
regulations do not permit presumptive service connection for squamous cell carcinoma, see 38
C.F.R. § 3.309 (e) (2026), Mr. Edwards needed to prove direct service connection, see Washington
v. Nicholson, 19 Vet.App. 362, 367 (2005) (listing the elements required to prove direct service
connection).
1
The NAS reports relay their conclusions about diseases and herbicide exposure in terms of an association
between the two. And they generally place diseases into one of four categories: (1) sufficient evidence of an
association; (2) limited or suggestive evidence of an association; (3) inadequate or insufficient evidence to determine
an association; or (4) limited or suggestive evidence of no association. See UPDATE 2018, at 67-68.
2
Squamous cell carcinoma of the tonsil is a "common malignant tumor of the oropharynx." Christopher J.
Skilbeck, et al., Squamous Cell Carcinoma of the Tonsillar Remnant—Clinical Presentation and Oncological
Outcome, 2011 HEAD & NECK ONCOLOGY 3:4 (2011).
3
Case: 24-6641 Page: 4 of 43 Filed: 08/04/2026
In response to a request for medical evidence, Mr. Edwards submitted two Board decisions
from unrelated cases to the AOJ.3 R. at 2610, 2778. Both granted service connection for squamous
cell carcinoma, both featured medical exams furnished by VA for the claimants, and both relied
heavily on opinions from private medical examiners to make their determinations. R. at 2610-11,
2613-15, 2620-23.
In the first decision, dated April 2015, the Board granted service connection to another
veteran for carcinoma of the tongue. R. at 2612. It considered three medical opinions, two provided
by the Agency and the third by a private examiner. The Board found that one VA exam was
inconclusive and unhelpful to deciding the matter. R. at 2615. The Board explained that the other
VA exam's explanation that Agent Orange exposure's complications did not include tongue
carcinoma was probative. Id. It then turned to the private medical examiner's opinion, which it
found adequate and compelling enough to put the evidence into equipoise. Id.
In the second decision, from September 2015, the Board granted service connection to
another veteran for carcinoma of the right forearm. R. at 2622. There, the veteran's private
examiner opined that, "although there was no definitive proof that herbicides cause skin cancers,
it was his experience that exposure to herbicides was definitely a factor in the development of skin
cancers," including the examined condition. R. at 2623. The Board conceded that negative nexus
opinions from VA exams were valueless, but noted that sun exposure is a known cause of the
condition and that the claimant had reported exposure to significant sunlight while in Vietnam. R.
at 2623. It ultimately found that service connection was appropriate because the evidence was in
equipoise as to whether the carcinoma was connected to exposure to herbicides and sun during
service. Id.
Here we introduce Update 2010, the first NAS report at issue today. In his formulation of
a positive nexus opinion, the private examiner reviewed Update 2010 and various other studies.
Id. After reading the report, the examiner said that there was a "biological mechanism by which
the Veteran’s herbicide exposure might have caused his cancer." 4 Id. Specifically, immune
3
The Court recognizes that Mr. Edwards also submitted a third unrelated Board decision, from March 2018,
to the AOJ. R. at 2617-19. Both parties seem to agree that this decision was correctly dismissed by the Board in the
decision on appeal. We too agree. After all, the May 2018 decision concerned entitlement to an earlier effective date
for a compensable rating for squamous cell carcinoma, while this matter concerns the preliminary issue of entitlement
to service connection. R. at 2617.
4
The Update 2010 report found that there was inadequate or insufficient evidence to determine an association
between squamous cell carcinoma of the tongue (including the tonsils) and herbicide exposure. NAT'L ACAD. OF SCI.,
4
Case: 24-6641 Page: 5 of 43 Filed: 08/04/2026
alterations associated with herbicide exposure may increase susceptibility to an infection in the
oral cavity and precipitate squamous cell carcinomas in that area. Id. Based on this theory and
other unenumerated studies, the examiner opined that "scientifically, it was more likely than not
that the Veteran's squamous cell cancer of the tongue was caused by or related to his in-service
exposure to Agent Orange." Id.
The AOJ was unconvinced by Mr. Edwards's submission of the unrelated Board decisions.
It denied service connection in a March 26, 2020, decision. R. at 2583. That decision did not note
any consideration of NAS reports in its summary of the evidence. R. at 2598-99. The AOJ
continued its denial in a higher-level review decision that reasoned that the submitted Board
decisions could not prove a connection because they were based "on the medical evidence and
service histories of other veterans." R. at 2267-75. This decision too noted no consideration of the
NAS reports in its summary of the evidence. Id. Apparent at this stage in the proceedings is that
the AOJ perceived no indication of a connection between Mr. Edwards's service and disability, for
it might have otherwise discussed the Secretary's duty to assist. See 38 U.S.C. § 5103A.
Mr. Edwards appealed to the Board in May 2021. R. at 2235. He and his representative
chose to pursue an appeal decision through the direct review docket and submitted an informal
presentation arguing his case. R. at 20-21, 2235. These two features about Mr. Edwards's appeal
are significant.
First, consider the process selected. R. at 2235. Under the Veterans Appeals Improvement
and Modernization Act of 2017 (AMA), a claimant picks one of three appellate avenues: the
hearing docket, additional evidence docket, or direct review docket. 38 U.S.C. § 7105(b)(3)(A)-
(C); see Andrews v. McDonough, 34 Vet.App. 151, 157 (2021). These paths each present "different
restrictions on the evidentiary record" before the Board and the window during which claimants
are permitted to submit unconsidered evidence thereto. Davis, 36 Vet.App. at 147. By electing the
direct review route, Mr. Edwards voluntarily restricted his evidentiary record to "the evidence of
record at the time of the decision of the [AOJ] on appeal." 38 U.S.C. § 7113(a). As a result, the
Board was limited to reviewing the evidence of record on March 26, 2020, the date of the AOJ's
denial of service connection.
VETERANS AND AGENT ORANGE: UPDATE 2010, at 17-18 (2010).
5
Case: 24-6641 Page: 6 of 43 Filed: 08/04/2026
Second, consider the informal presentation given to the Board. Its pages reference Update
2012, the other NAS report at issue here. Arguing that there is a positive nexus between herbicide
exposure and squamous cell carcinoma, Mr. Edwards cited Update 2012, which repeats the
"biological mechanism" theory also featured in Update 2010.5 R. at 20. He indeed says that the
evidence of record here "reflects an . . . infection leading to squamous cell carcinoma of the tonsil."
Id. The informal presentation then asserted that the Secretary should have furnished Mr. Edwards
with a medical exam, for the theory articulated in the NAS report was sufficient to indicate a
connection between the cancer and herbicide exposure. Id.; see 38 U.S.C. § 5103A(d)(2)(b).
The Board then issued the August 5, 2024, decision on appeal here. It first reasoned that
the unrelated Board decisions did not obligate VA to furnish an exam because their references to
separate medical opinions were "too general and conclusory to indicate that there may be a nexus"
linking the cancer to service. R. at 6-7. The decision then turned from the duty-to-assist argument
and towards the merits of Mr. Edwards's theory of direct service connection.
There, too, the Board began its discussion by addressing the other Board decisions provided
by Mr. Edwards. It emphasized that such decisions could be persuasive authority only, for "each
case before the Board is decided on the basis of the individual facts" presented. R. at 8 (citing 38
C.F.R. § 20.1303). And the Board did not share the view that these past decisions were analogous
to this case. As to the April 2015 decision, the Board found that Mr. Edwards failed to submit a
private opinion, whereas that veteran submitted such an opinion. R. at 8. Though it plainly read
the decision, the Board did not mention Update 2010. As to the September 2015 decision, the
Board distinguished it because Mr. Edwards, unlike the claimant there, did not provide a private
medical opinion. R. at 9. The Board also explained that sunlight exposure was not alleged as a
cause of the disability here while it was deemed a possible cause in that decision. Id.
The informal presentation's reference to the Update 2012 NAS report was mentioned in
the Board's decision, just not in the manner Mr. Edwards wanted. Because the appeal arrived via
the direct review pathway, the Board found that the report's contents could not be considered either
as argument or evidence. R. at 9. To consider that report—or any such report—would be to rely
on materials received by the Agency after the AOJ's decision. Id.; see 38 U.S.C. § 7113(a).
5
Like its predecessor, the Update 2012 report found that there was inadequate or insufficient evidence to
show an association between squamous cell carcinoma of the tongue (including the tonsils) and herbicide exposure.
NAT'L ACAD. OF SCI., VETERANS AND AGENT ORANGE: UPDATE 2012, at 8-9 (2012).
6
Case: 24-6641 Page: 7 of 43 Filed: 08/04/2026
Mr. Edwards appealed the decision to this Court. In essence, he would have us remand the
matter and require the Secretary to consider the NAS reports and furnish an exam. See Appellant's
Brief (Br.) at 5-6. In his telling, the Board was required to address Update 2012 because the
Agency constructively possessed that report long before it denied the claim at the AOJ level. Id.
at 13-14. He further disputes the Board’s reasons or bases for its decision that a medical exam was
not required, especially given that the conclusion found within Update 2010 was raised to the AOJ
through the unrelated April 2015 decision. Id. at 7-13.
III. CONSERVING CONSTRUCTIVE POSSESSION
Mr. Edwards first argues that the Board failed to consider the Update 2012 NAS report that
should have been treated as evidence of record under our judicially wrought doctrine of
constructive possession. Because one could reasonably expect an Agency-controlled report like
Update 2012 to be a relevant document in his case, see Euzebio, 989 F.3d at 1319, Mr. Edwards
would have the Board discuss whether that evidence was included in the record before he decided
to walk down the direct review pathway, Appellant's Br. at 13-14. We agree; had Congress or VA
intended to deliver a knock-out blow to constructive possession via the AMA, it would have clearly
stated that intent in the statute or regulations.
A. The Doctrine
This Court sometimes consults a doctrine called constructive possession when determining
which documents reside in the appellate record. See Varad v. McDonough, 37 Vet.App. 198, 204
(2024); Monzingo v. Shinseki, 26 Vet.App. 97, 100-02 (2012) (per curiam), overruled on other
grounds by Euzebio, 989 F.3d at 1305; Bowey v. West, 11 Vet.App. 106, 109 (1998). The genesis
of this doctrine in our Court was Bell v. Derwinski, 2 Vet.App. 611, 613 (1992), in which we held
that some documents "generated within the VA" were evidence that should have been considered
by the Board. As the Court explained, because some pieces of evidence were created and held by
VA, "the Secretary had constructive, if not actual, knowledge of those items." Id.
The doctrine has since evolved, but its emphasis on dominion, control, and knowledge of
a document has remained the same. Its current iteration is best articulated in the Federal Circuit's
decision in Euzebio. There, the court repeated the holding in Bell that "evidence that is 'within the
Secretary's control' and 'could reasonably be expected to be a part of the record before the Secretary
and the Board,'" is constructively part of the record. Euzebio, 989 F.3d at 1319 (quoting Bell, 2
7
Case: 24-6641 Page: 8 of 43 Filed: 08/04/2026
Vet.App. at 613). The "correct standard" by which to test expectations, the court continued, "is
relevance and reasonableness." Id. That version of the test has since been applied by this Court on
multiple occasions. See, e.g., Davis, 36 Vet.App. at 150-55.
Even though constructive possession is a mainstay of our evidentiary toolkit, the doctrine
was not articulated in the operative statutes or regulations at the time of its announcement. Its birth
was instead Athenian, seemingly bursting fully formed from the mind of this Court. See Bell, 2
Vet.App. at 613. Federal courts, however, are not all powerful and cannot create rules from
nothing. See Com. of Pennsylvania v. Wheeling & Belmont Bridge Co., 54 U.S. (13 How.) 518,
580-81 (1851) (Taney, J., dissenting); Bilharz v. Collins, 38 Vet.App. 366, 378-79 (2024). There
must be some reasoning behind the doctrine's design.
Yet it is challenging to pinpoint how Bell initially concluded that constructive possession
was a doctrine applicable in this Court. The opinion meanders through authority, including our
jurisdictional statute and caselaw about duty-to-assist errors, to contextualize the legal topography.
See Bell, 2 Vet.App. at 612-13 (first citing 38 U.S.C. § 7252(b), then citing Murincsak v.
Derwinski, 2 Vet.App. 363, 372 (1992)). The notion that evidence that the Secretary has
"constructive" knowledge and possession of must be included in the record, however, is not
explicitly supported by any binding authority. Id. As best we can tell, the Bell Court created a new
rule of decision using federal common law powers and informed by settled common law
understandings about possession.6 It is indeed difficult to view the decision differently, given that
its citations focused on peripherally related concepts and did not possess the hallmarks of an
interpretive conclusion.7
6
Such rules of decision are generally disfavored here. Cf. Rodriguez v. FDIC, 589 U.S. 132, 133 (2020).
Aside from constructive possession, we struggle to name another surviving doctrine created in the interstices of a
statute.
7
But not impossible. One might, for instance, view Bell as a poor interpretive exercise, focusing on section
7252(b). See id. Under this view, to the extent that common law played a role in that decision, it was either to
contextualize section 7252(b) or to suggest that "before the Secretary and the Board" had a common-law meaning.
See, e.g., D'Oench, Duhme & Co. v. FDIC, 315 U.S. 447, 468 (1942) (Jackson, J., concurring). Yet, even under this
view, our decision that constructive possession survives the imposition of the AMA would not change. Our focus
would become whether the best reading of the new statutory scheme displaces the doctrine. See Lackey v. Stinnie, 604
U.S. 192, 670 (2025) (citing Henson v. Santander Consumer USA Inc., 582 U.S. 79, 89 (2017); SEC v TheStreet.Com,
273 F.3d 222, 233 n.11 (2d Cir. 2001). And, for the same reasons as we discuss in our interpretive endeavor below,
VA has failed to provide a reading of the AMA that would plausibly, clearly, or most likely displace constructive
possession. That same failure dooms a view of constructive possession tied only to the Federal Circuit's precedential
endorsement or acceptance of the Bell doctrine. See Euzebio, 989 F.3d at 1319.
8
Case: 24-6641 Page: 9 of 43 Filed: 08/04/2026
Courts can import background principles into their view of what a given authority requires.
See D'Oench, 315 U.S. at 468. And they may fill the interstices of federal legislation concerning
federal interests. Clearfield Trust Co. v. United States, 318 U.S. 363, 367 (1943); see Henry J.
Friendly, In Praise of Erie—and of the New Federal Common Law, 39 N.Y.U. L. REV. 383, 421-
22 (1964). Applying these general principles and a generous definition of federal common law
best accounts for the Court's decision in Bell to look beyond the written law when crafting the rule
of decision that constructively possessed evidence was properly included in the record. See
Thomas W. Merrill, The Common Law Power of Federal Courts, 52 U. CHI. L. REV. 1, 5 (1985)
(defining "federal common law" as a "federal rule of decision that is not mandated on the face of
some authoritative federal text").8
Reading between the lines of our appellate process, we found that documents that were
effectively before VA were part of the record, although nothing in the written law demanded that
conclusion. Bell, 2 Vet.App. at 613. But like all good common law reasoning, our conclusion in
Bell was not formed in a void, but through historic and settled legal precepts. See id.; D'Oench,
315 U.S. at 468; Edmonds v. Compagnie Generale Transatlantique, 443 U.S. 256, 259 (1979)
(employing long-recognized legal understandings when operating in a federally controlled area of
law). The precept at issue here, an idea that possession "occurs when one has knowledge of the
thing possessed 'coupled with the ability to maintain control over it'," has a lengthy pedigree. Aqua
Log, Inc. v. Georgia, 594 F.3d 1330, 1336 (11th Cir. 2010) (quoting United States v. Wynn, 544
F.2d 786, 788 (5th Cir. 1977)).
This aspect of ownership traces its roots to the same British soil that provided other
common law doctrines that still shape our law. Edward R. Cohen, The Finders Cases Revisited,
48 TEX. L. REV. 1001, 1015-16 (1970). Dominion and control over an object, the common law
said, constitute legal possession, even without evidence of physical possession. See OLIVER
8
We take this opportunity to note two things about federal common law. First, we understand that some
definitions are broad, see generally Louise L. Weinberg, Federal Common Law, 83 NW. U. L. REV. 805 (1989), while
others would cabin federal common law powers, see MARTIN L. REDISH, THE FEDERAL COURTS IN THE POLITICAL
ORDER 29-46 (1991). Indeed, it is difficult to pin down this elusive concept. See Merrill, supra, at 5; RICHARD H.
FALLON, JR., ET. AL., HART AND WECHSLER'S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 643-656 (7th ed.
2015). We presented a broader view to account for the possible ways the Court in Bell could have viewed their
lawmaking powers, but do not adopt that view as a general matter. Second, federal common law has an interpretive
side, which encourages courts to import a well-settled common law understanding of a subject into a statutory text
invoking that subject. See, e.g., Kousisis v. United States, 605 U.S. 114, 124 (2025); Rios v. Nicholson, 490 F.3d. 928,
930-32 (Fed. Cir. 2007) (citing Rosenthal v. Walker, 111 U.S. 185, 193 (1884)). Considering common law definitions
as part of an interpretive exercise gives us less pause than does creating a broader rule of decision.
9
Case: 24-6641 Page: 10 of 43 Filed: 08/04/2026
WENDELL HOLMES, JR., THE COMMON LAW, 216 (1881). Indeed, former law students will
doubtless remember that a "wounded, circumvented, and ensnared" fox is the rightful property of
his pursuer. Pierson v. Post, 3 Caines 175, 179 (N.Y. 1805). Our revisitation was not uncommon;
even today, use of the doctrine is ramulose, emerging in a variety of contexts. See, e.g., Henderson
v. United States, 575 U.S. 622, 626 (2015) (outlining constructive possession in the criminal law);
Cline v. Kaplan, 323 U.S. 97, 100 (1944) (bankruptcy); Smith v. Gale, 144 U.S. 509, 526 (1892)
(real estate).
Yet even well-established common law doctrines must yield to legislative directives. See
Am. Elec. Power Co. v. Connecticut, 564 U.S. 410, 423-24 (2011). Discerning when a doctrine is
displaced, however, is no easy task. After all, when omnipresent, federal common law commands
respect from the legislature and the judiciary alike. "Congress is understood to legislate against a
background of common-law adjudicatory principles," Astoria Fed. Sav. & Loan Assn. v. Solimino,
501 U.S. 104, 108 (1991), and our cases employ "a presumption favoring the retention of long-
established and familiar principles," Isbrandtsen Co. v. Johnson, 343 U.S. 779, 783 (1952).
Still, those background principles are preempted when a statute "'speak[s] directly' to the
question addressed by the common law." United States v. Texas, 507 U.S. 529, 534 (1993) (quoting
Mobil Oil Corp. v. Higginbotham, 436 U.S. 618, 625 (1978)). As applied here, we will presume
that our doctrine of constructive possession remains in inertia unless a clear statement disposes of
it. See Smith v. Brown, 35 F.3d 1516, 1523 (Fed. Cir. 1994) (explaining that canons of statutory
interpretation apply also to regulations). After all, the legislature would not likely banish the
longtime legal understanding of possession silently, especially since it was well-established in our
caselaw by the time Congress passed the AMA.9
The Secretary declares that our strain of constructive possession has been rendered mostly
defunct, if not by our precedent, then by commands in the AMA and its implementing regulations.
Secretary's Br. at 14-23. The AMA's neoteric docket selection system, he asserts, is incompatible
with a doctrine that could require VA to consider new evidence at the Board level. Secretary's
9
There is, to be sure, a strong argument that clear statement rules, like the one preserving federal common
law, exist uncomfortably in a textualist interpretation. See ANTONIN SCALIA, A MATTER OF INTERPRETATION 27–29
(1997) (arguing that clear statement rules, as substantive canons, "load the dice for or against a particular result").
Thus, "academics and the Supreme Court have rightfully plumbed the analytical foundations of clear-statement rules."
N.C. Coastal Fisheries Reform Grp. v. Capt. Gaston LLC, 76 F.4th 291, 296 n.5 (4th Cir. 2023) (citing Stephen E.
Sachs, The Law of Interpretation, 130 HARV. L. REV. 1079, 1122-28 (2017)). But an inferior court like ours can only
complain of discomfort before applying the clear statement rule prescribed to us.
10
Case: 24-6641 Page: 11 of 43 Filed: 08/04/2026
Supp. Br. at 4-7. He would have us determine that the Agency relegated post-AMA constructive
possession only to certain VA hospital records mentioned in a regulation that we will consult later.
Id. at 1-3.
B. The Caselaw
Despite the contentions of both parties, we have yet to describe constructive possession's
shape in post-AMA litigation. In many instances, this Court or the Federal Circuit employed the
doctrine, but none concern an entirely AMA veteran's attempts to prove constructive possession
of evidence after voluntarily closing the record. While the Secretary's citations address only
circumstances auxiliary to the doctrine's heartland, Mr. Edwards's offerings do little more than
restate the legacy conception of constructive possession.
Begin with Mr. Edwards's frequent references to the Federal Circuit's decision in Euzebio.
Appellant's Br. at 13-14; Appellant's Supp. Br. at 3-6. To be sure, its facts are initially reminiscent
of the background here: Following an adverse regional office decision in September 2011, the
veteran argued that the Board needed to consider the Update 2014 NAS report, published during
his appeal, before denying him a VA exam. Euzebio, 989 F.3d at 1315-16. The Board and this
Court disagreed because the update offered no "direct relationship to the claim on appeal." Id. at
1317 (quoting Euzebio v. Wilkie, 31 Vet.App. 394, 397 (2019)). The Federal Circuit reversed,
determining that we had narrowed our rules for constructive possession such that they conflicted
with the perceived purposes of the Secretary's statutory duty to assist and the Agent Orange Act.
Id. (first citing 38 U.S.C. § 5103A(a)(1), then citing Agent Orange Act, 105 Stat. at 11). It therefore
sent back the claim so that the Board could consider Update 2014 in the first instance. Id. at 1326.
Although instructive in framing the functions and fundamentals of constructive possession,
Euzebio is inapposite to the issues presented here. The procedure in Euzebio, after all, flew
naturally above the tangle of AMA commands that matter in this appeal. Because the decision
there was issued in July 2017, during the legacy era, the court had no reason to doubt whether the
constructive possession doctrine could apply. Id. at 1315; see Godsey v. Wilkie, 31 Vet.App. 207,
214 n.2 (2019). This matter, however, rests only within the AMA's encincture and questions the
doctrine's survival given the new evidentiary and duty-to-assist limitations.10 Thus, Euzebio is
10
The Court recognizes that, even in the legacy appeals system, the Secretary was vexed by the doctrine and
sought to dismiss it as colliding with 38 U.S.C. § 7252(b) because we construed the record of proceedings to include
evidence not before VA adjudicators at the regional level. Id. at 1322. But the Federal Circuit disagreed, id. at 1322-
23, and the Secretary does not attempt renewal of that argument. Even if he did, the query here depends on what we
11
Case: 24-6641 Page: 12 of 43 Filed: 08/04/2026
helpful in the pedagogic sense of outlining constructive possession, but it does not control our
disposition here.
The same is true for our decision in Davis, which Mr. Edwards would now have us employ
as a blueprint for AMA constructive possession cases. Appellant's Br. at 14 (asserting that Davis
implicitly found all Agent Orange Act reports to be constructively before the Board); Appellant's
Supp. Br. at 9. Our review, however, reflects that the Davis blueprint was not written to scale for
AMA cases. Because VA issued the regional office decision during the legacy period, the rules of
that time applied, including constructive possession. Davis, 36 Vet.App. at 145-46. Only later did
the AMA amend the appellate procedure. Mr. Edwards only traversed the modernized system.
Thus, as with Euzebio, our decision in Davis does not meaningfully inform this case because we
had no reason to square constructive possession with the limitations found within the AMA.
Mr. Edwards pushes back, asseverating that we did not delineate a distinction between the
regimes and had instead indicated that constructive possession should apply no matter the posture
of the appeal. Appellant's Reply Br. at 10. But we did, in fact, separate the application of legacy
rules from the AMA's rules when discussing the evidence of record. Davis, 36 Vet.App. at 154-55
(considering a regulation that applies during legacy appeal periods). And, as a general principle, a
precedent's failure to comment on potential applications does not negate the consequences of any
factual differences between the initial case and a later case. Rather, "an expression in an opinion
yields later to the impact of facts unforeseen." Jaybird Mining Co. v. Weir, 271 U.S. 609, 619
(1926) (Brandeis, J., dissenting). Davis, then, is helpful in applying the doctrine, especially when
NAS reports are involved, see 36 Vet.App. at 156-57 (Falvey, J., concurring), but gets us no closer
to figuring out the doctrine's post-AMA durability.
The Secretary's position is likewise more grounded in hope than in precedential analogy.
His strongest champion is Hyatt v. Shinseki, 566 F.3d 1364 (Fed. Cir. 2009). Secretary's Br. at 21.
At issue in that case was whether certain court-martial records were included in the appellate
record for a claim for accrued benefits. Hyatt, 566 F.3d at 1370-71. The evidence supporting such
a claim must be limited to that of record at the time of a veteran's death. 38 U.S.C. § 5121(a).
Because the court-martial records were not part of the record at the time of the death in question,
the Federal Circuit found that section 5121(a) prohibited their inclusion. Hyatt, 566 F.3d at 1370.
consider to be before VA adjudicators at the AOJ level, so that past objection would be tangential to this controversy.
12
Case: 24-6641 Page: 13 of 43 Filed: 08/04/2026
It explained that the ability to "enlarge the universe of evidence upon which accrued benefits
claimants may rely" is controlled by Congress alone. Id. at 1371.
We understand the point the Secretary tries to make, but his comparison misses the mark.
Our decision in this case would not enlarge the universe of evidence that the Board needs to
consider. See Secretary's Br. at 21. To the contrary, the entire point of constructive possession is
to determine which evidence was already of record when the evidentiary window closed at the
AOJ level. Until otherwise decided, therefore, our doctrine is considered harmonious with the
command that "the evidentiary record before the Board shall be limited to the evidence of record"
at the time of the AOJ decision. See 38 U.S.C. § 7113(a) (indicating that the Board must consider
everything that is properly in the record before it); see Texas, 507 U.S. at 534.
The Hyatt decision indeed reasoned that including the court-martial records would expand
the universe of permissible evidence because those records violently mismatched with the pillars
of constructive possession. Hyatt, 566 F.3d at 1371 (finding that, unlike other records, the court-
martial records were "not generated by, submitted to, or otherwise within the VA's possession or
control" at the time of the veteran's death). This case, however, concerns the inclusion of NAS
reports and these documents not only live at the nucleus of possession, control, and relevance, but
they also exist outside the restrictions placed by Hyatt in the accrued benefits context.
The Federal Circuit, when discussing constructive possession in Hyatt and its later cases,
expressed no concern that the doctrine could violate the evidentiary restrictions placed on the
Board. See Lang v. Wilkie, 971 F.3d 1348, 1353-55 (Fed. Cir. 2020); Hyatt, 566 F.3d at 1370-72.
If anything, the court seemed to rely on the doctrine's staying power in our Court and its twin
prongs of relevance and reasonableness. See Euzebio, 989 F.3d at 1322-26; Hyatt, 566 F.3d at
1371. The Secretary's discussion of Hyatt, therefore, does little to aid his arguments.
His parenthetical references to other cases are similarly fruitless. Secretary's Br. at 21 (first
citing Mil. Veterans Advoc. v. Sec. of Veterans Aff. (MVA), 7 F.4th 1110, 1140 (Fed. Cir. 2021),
then citing Aviles-Rivera, 35 Vet.App. at 278). The reason that the AMA's restrictions would have
been "rendered meaningless" by including the NAS report in Aviles-Rivera is because that report
was published after the date of the AOJ decision that marked the end of the evidentiary window.
35 Vet.App. at 275. To include that report would indeed run afoul of the directive to bar evidence
following the AOJ decision. Id. at 178.
13
Case: 24-6641 Page: 14 of 43 Filed: 08/04/2026
But here, the Update 2012 NAS report was published many years before Mr. Edwards's
claim was denied by the AOJ. See R. at 2583. If the doctrine constructively places the report before
the AOJ by the date of its decision, therefore, the report could be a proper part of the record without
disturbing evidentiary deadlines. See 38 U.S.C. § 7113(a). In other words, the report woul