More than 2,000 plaintiffs sued Merck & Co. alleging its live-attenuated shingles vaccine, Zostavax, caused shingles or related injuries. Centralized in the Eastern District of Pennsylvania in August 2018, the litigation became a causation test case for vaccine-related mass torts.
No case reached a jury. This analysis covers the factual record, MDL structure, expert rulings, Vaccine Court routing, settlement posture, and practice consequences for 2026.
The Injury Theory and the Wild-Type Causation Problem
FDA prescribing materials describe Zostavax as a live-attenuated varicella-zoster virus vaccine, built on the Oka/Merck strain and indicated to prevent shingles in older adults. The FDA licensed it on May 25, 2006, for adults 60 and older, and the indication later expanded to adults 50 and up. Merck discontinued sales effective November 18, 2020. The federal proceeding was centralized as MDL No. 2848, In re: Zostavax (Zoster Vaccine Live) Products Liability Litigation, No. 2:18-md-02848 (E.D. Pa. 2018).
Plaintiffs alleged that the vaccine's live virus caused them to develop shingles or related injuries, and that Merck failed to adequately warn of that risk. The theory ran into a biological problem the court treated as central. Dormant wild-type VZV from childhood chickenpox is carried by virtually every person over age 30 in the United States, and that latent virus reactivates as shingles on its own. The Third Circuit framed the obstacle directly: plaintiffs "would have to account for and exclude the 'obvious alternative cause' of shingles for the Group A cases: the wild-type chickenpox strain of the [VZV] latent in almost every person over the age of 30."
Only one method separates the two causes. The district court cited "compelling medical authority" that PCR testing "is the only way to tell" whether shingles came from latent wild-type virus or Zostavax's attenuated strain. Both sources produce identical shingles rashes, so laboratory strain-identification was the reliable proof of specific causation.
How the MDL Was Sorted Into Groups A, B, and C
The Judicial Panel on Multidistrict Litigation centralized the proceeding on August 2, 2018, transferring 57 actions then pending in nine districts, with 41 additional related actions identified as potential tag-alongs, and the inventory grew past 2,000 cases. The MDL proceeded before Judge Bartle in the Eastern District of Pennsylvania, using a case-management structure familiar from other large MDL proceedings. Merck & Co., Inc. and Merck Sharp & Dohme Corp. were the manufacturer defendants, with distributor McKesson Corporation also named as a co-defendant. No other corporation, including Bayer, was a defendant or a Merck parent in this litigation.
Judge Bartle sorted the cases by alleged injury, and that sorting decided which cases lived and which died:
- Group A: shingles or shingles-related injuries, more than 1,700 actions as of March 2022; of these, roughly 500 alleged shingles plus non-shingles injuries and 1,189 alleged shingles only, with the remainder unclassified.
- Group B: non-shingles illnesses, with approximately 500 cases.
- Group C: hearing loss injuries.
Only Group A faced the wild-type-versus-vaccine-strain differentiation problem, because a shingles plaintiff had to rule out the alternative cause carried by nearly every adult. That difference explains why Group A collapsed while settlement talks covered Groups B and C, and the 1,189 shingles-only cases became the dismissed subset.
The Exclusion of the Plaintiffs' Causation Expert
The Group A collapse began with a specific-causation ruling. On December 1, 2021, the court excluded the specific-causation opinions of Mark Poznansky, M.D., a Harvard Medical School professor retained to testify in five Group A bellwether cases, under Rule 702 and Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993).
Dr. Poznansky's differential diagnosis "required him to explain why the wild-type virus from chickenpox was not responsible for plaintiffs' shingles". He did not do so. At oral argument, plaintiffs' counsel conceded, with what the court called "appreciated candor," that nothing in Dr. Poznansky's reports ruled out wild-type reactivation for any plaintiff. The court found that "no straining or stretching of the words of Dr. Poznansky can satisfy Daubert or Rule 702."
Two fallback arguments failed with it:
- The "rash looks the same" argument. The court rejected it because a shingles rash "looks the same for both" causes.
- The timing argument. The court held that timing alone did not show causation. Temporal proximity relies on post hoc reasoning and fails as a differential-diagnosis method.
The distinction between general and specific causation drove the analysis. The court noted that even "15% of the recipients of Zostavax may suffer from shingles as a result of being vaccinated is of no help to the issue of causation in a specific case". For that analysis, the court accepted that Zostavax can cause shingles in a population. The excluded opinion failed to show the vaccine caused shingles in each named plaintiff. Summary judgment for Merck followed in all five bellwether cases through Pretrial Orders 411, 413, 415, 417, and 419.
The Lone Pine Order and Rule 41(b) Dismissal
With the five bellwethers lost and more than 1,700 Group A cases remaining, Judge Bartle issued Pretrial Order No. 426 on March 30, 2022. This Lone Pine order required every remaining Group A plaintiff to "serve laboratory reports or other records documenting that strain-identification testing detected vaccine-strain varicella zoster virus ('VZV') in a rash sample". Plaintiffs had 90 days to comply. The order's purpose was to "weed out non-meritorious from meritorious claims and move along these 1,700 or more cases toward a final resolution".
The requirement exposed a chokepoint that could never be reopened. PCR testing works only on active rash tissue; after healing, the sample is gone, and testing is permanently impossible.
Plaintiffs argued that none of the Group A plaintiffs had been PCR tested and that most rashes had already healed. The court found the argument unavailing: across nearly three years of discovery, over 6,000,000 pages produced, and nearly 40 witnesses deposed, plaintiffs cited no medical literature or expert opinion offering any other method to prove specific causation.
When the 90 days expired without laboratory reports, Merck moved to dismiss. Judge Bartle issued Pretrial Order No. 458 on December 6, 2022, dismissing 1,189 Group A shingles cases with prejudice under Rule 41(b) for failure to comply with PTO 426. Merck also sought summary judgment under Rule 56, and the Rule 41(b) route mattered on appeal because it draws deferential abuse-of-discretion review rather than de novo review.
The court applied the six Poulis factors from Poulis v. State Farm Fire & Cas. Co., 747 F.2d 863 (3d Cir. 1984), and found all six supported dismissal. These procedural mechanics distinguish this outcome from an ordinary product liability merits ruling.
Why the Claims Bypassed the Vaccine Court
The litigation proceeded in ordinary civil court because the Vaccine Injury Table omits herpes zoster vaccines. The National Vaccine Injury Compensation Program requires a petitioner to show receipt of "a vaccine set forth in the Vaccine Injury Table" under 42 U.S.C. § 300aa-11, and the operative Table at 42 CFR § 100.3 lists no herpes zoster vaccine.
For VICP coverage, CDC must recommend the vaccine for routine administration to children or pregnant women. The vaccine also needs federal excise-tax status and HHS addition to the program. Zostavax, indicated for adults 50 and older, fails the first. The civil-action bar in § 300aa-11(a)(2)(A) applies only to covered vaccines, so Zostavax claimants could sue manufacturers directly rather than file a VICP petition first. The district court exercised ordinary diversity and MDL jurisdiction under 28 U.S.C. §§ 1332 and 1407. Court of Federal Claims jurisdiction did not govern these claims.
The routing decided the causation standard. VICP's Table-injury presumptions offer a causation pathway unavailable in civil court. In civil litigation, Zostavax plaintiffs faced full Daubert scrutiny and Lone Pine exposure, the tools that dissolved Group A.
Where the Litigation Stands in 2026
The Third Circuit affirmed the dismissals on July 16, 2024, in In re: Zostavax (Zoster Vaccine Live) Prods. Liab. Litig., No. 23-1032 (3d Cir. July 16, 2024) (non-precedential under Third Circuit I.O.P. 5.7). The panel found no abuse of discretion and held that uncontradicted record evidence established PCR testing as the only way to prove specific causation. The court acknowledged that plaintiffs' inability to produce nonexistent PCR tests was neither willful nor in bad faith and affirmed because the cases had been "at a standstill" for over a year "with nothing to indicate they could ever succeed on the merits."
No case reached a jury. No compensatory or punitive damages were awarded. The remaining Group B and C inventory is resolving through case-management settlement proceedings. Any financial terms are not publicly disclosed; no dollar figure appears in any court order, and Merck's 2024 and 2025 Form 10-K filings do not name Zostavax or MDL 2848.
The wind-down continued into 2026. The New Jersey Superior Court issued dismissal orders dated March 25, 2026, in MCL No. 629 for failure to meet the production deadline.
Rule 16.1, adopted April 23, 2025 and effective December 1, 2025, formalizes early MDL management; Committee Notes discuss fact sheets, census methods, early discovery, and early motion practice, while Lone Pine orders remain a related MDL tool.
What FRCP 16.1 Changes for the Next Vaccine MDL
Rule 16.1 makes the Zostavax sequence harder to repeat on the same timeline. Early case-management conferences now invite courts to test threshold evidence before inventories mature for years.
- Plaintiff counsel: Validate methodology before filing large inventories, especially where an obvious alternative cause exists.
- Defense counsel: Build the merits record in bellwethers, request mechanism-specific Lone Pine relief, and preserve Rule 41(b) dismissal arguments.
Future vaccine MDLs will likely compress causation testing, discovery sequencing, and settlement pressure into earlier proceedings.
Practice Lessons After the Zostavax Dismissals
The Zostavax record leaves plaintiff counsel with a preservation mandate: identify the biological mechanism early, secure time-sensitive specimens, and test whether expert methodology can exclude obvious alternative causes before inventory growth. Defense counsel drew the complementary lesson: develop the general-versus-specific causation record, use bellwethers to test methodology, and seek mechanism-specific Lone Pine relief when threshold proof is missing.
Rule 16.1 will likely move those disputes earlier, making early documentation and expert screening central to settlement position. The through-line for both sides is that causation proof must be preserved at intake, before the evidentiary window closes. That is why intake preservation protocols aligning client screening with causation proof are where the next vaccine MDL will be won or lost.
