The Curious Pictsweet Case: Only Pre-Compute Biology Was Admitted

Imagine Manhattan built five hundred stories high. The ten billion people racing between those floors are proteins, never sleeping, building and breaking and repairing. The DNA is the same in every room. What makes an eye cell an eye cell and not a heart cell is which switches sit on that DNA, open or shut. [5] Age, radiation, diet, chemical toxins nudge those switches out of place. Tissue fails and skin wrinkles when enough switches drift. [6]

​

The modern basics we did not know in 1996:

​

37 trillion

cells make a human body [4]

10 billion

moving parts and a set of switches is what makes a living cell [3]

The Yamanaka Factor

​In 2006, Shinya Yamanaka showed that four proteins could walk into that city and throw the switches all the way back to a stem cell: a blank that can become any tissue again. He won the Nobel Prize. [7] Later work showed a lighter dose can slide an old cell toward youth while leaving its identity intact. Eye cells, skin cells, heart cells. [8] The human trials, aimed at maintenance of youth at the scale of the whole body. [9] The capital forming around that research already runs to billions, because every added year of healthy life is worth more than any drug ever sold. [10]​​

Administrative Dossier Assembled Before Biology Was Computational

​In 1996, California’s Developmental and Reproductive Toxicant Identification Committee voted to list cadmium as a Propositon 65 chemical known to the state to cause reproductive toxicity. [1] The administrative record was a product of its decade: paper studies, early epidemiology, observational mammalian assays, a world before affordable sequencing, before genome-wide maps, before any machine that could watch ten thousand genes flip at once had come within a mile of a toxicology committee. [2] The record was assembled before biology became a computational science. Nobody could count the residents, because nobody had the tools.​​

Pictsweet Riling's Two-Dimentional Science Map

​The 1996 record is a flat street map: a few thousand named kinds of thing, drawn in two dimensions, with the streets labeled and the buildings blank. What we have now is the census. Every resident, every floor, every switch, and, for the first time, the knowledge that the switches can be thrown.

​

That matters for cadmium because cadmium throws switches. It is one of the standard examples in the literature on chemicals that alter the marks laid over DNA during development: it inhibits the enzymes that lay down DNA methylation. [11] It crosses into the placenta and accumulates there. [12] Maternal exposure has been associated with altered methylation in cord blood and placental tissue in human cohorts, [13] and with sex-specific effects on birth weight. [14]

Whether a given injury needs months of accumulation or a single exposure inside a window of days is a question the census can ask. The map cannot. In 1996 the map had two kinds of harm to a fetus: the sudden structural kind, the missing limb, and the slow quantitative kind, the low birth weight, assumed to accumulate the way rust does. The switches were a third kind, and nobody had drawn them.

The regulation does not say to dig out an old map and lay it over the future Manhattan. It says the opposite.

The regulation California applies to cadmium did not say to dig out an old paper map and lay it over the future Manhattan. It says the opposite. The rate of exposure “shall be based on the pattern and duration of exposure that is relevant to the reproductive effect which provided the basis for the determination.” [15] Is relevant. Present tense. The words “administrative record” appear nowhere in it. And the regulation’s own example of how to tell one pattern from the other is whether the chemical is a teratogen: “an exposure of short duration is appropriate for a teratogenic chemical, whereas a chronic or protracted exposure is appropriate for one that retards fetal growth.” [15] The question is not optional. A fact-finder applying that sentence must decide whether cadmium is a teratogen before it can decide anything else. The Pictsweet ruling threw that entire question out of the courtroom.

Appellate Court said that scientific findings  “are subject to perpetual revision”

The agency that wrote the regulation said how to find the effect: “a person needs only to consult the list to determine which effects provide the basis for the listing.” [16] The list says “developmental, male.” [17] A reader who wants more may, “if more specific information is desired,” look at the committee transcript; the same statement of reasons says the no-effect level rests on “whatever proof is available,” and that “nothing in this article shall preclude a person from using evidence, standards, assessment methodologies, principles, assumptions or levels not described in this article.” [18] The statute above the regulation asks for evidence “of comparable scientific validity” to the listing evidence: a floor of quality, not a date. [19] For forty years every safe-harbor trial in California was tried on the science of the day. The lead case used blood-lead models developed decades after the listing; the toluene case heard what “current studies” showed; the mercury case said that scientific findings “can easily become dated and outmoded as science develops” and “are subject to perpetual revision.” [20] No court had ever confined a safe-harbor trial to the year of the listing.

The Court: the law does not permit teratogenicity evidence - where statute mandates it

​The Pictsweet's court did. The court itself framed the choice in one sentence: if the law “requires or permits the court to evaluate plaintiff’s post-listing scientific evidence and if the court concludes that, yes, cadmium is a teratogen, then, yes, the shorter duration factor leads the arithmetic to show that the LOE is greater than the MADL, and, yes, defendants needed to include a Proposition 65 warning.” [21] Then it held that the law did not permit it. Pattern and duration, it ruled, must be found in the 1996 record, “not post-listing science,” thus excluding every expert who would have testified to what the census shows. [22]

​

"An absence of evidence is not the equivalent of substantial evidence. If an absence of evidence could satisfy the burden of proof, the concept of burden of proof would have no meaning." Roddenberry v. Roddenberry (1996) 44 Cal.App.4th 634.

Science makes no difference under this ruling, Pictsweet admits

The defense had said what that meant in open court. Asked whether “five blue ribbon human epi studies published in the Lancet,” every one “showing by high statistical significance that cadmium is a human teratogen,” would “make no difference,” defense counsel answered: “That’s correct.” Asked whether the court was therefore “looking to see what OEHHA did in 1996, not what they could have done in 2024,” he answered: “Yes.” [23]

​

"Until a trier of fact is exposed to evidence and concludes that the evidence supports a position, it cannot be said that anything has been proved." Stull v. Sparrow, 92 Cal. App. 4th 860 (2001).

The court took the map. What the map showed was not what the ruling claimed. The committee that listed cadmium never said which effect it listed for; the defense toxicologist conceded it “just tallied votes,” “didn’t state, in so many words, what the basis would be,” and “did not, specifically, identify the pattern and duration that should be used.” [24]

​

"Exclusion of relevant evidence... was reversible error notwithstanding the fact that findings against [the party] were supported by substantial evidence." Loucks v. Luckel, 107 Cal. App. 2d 217, 236 P.2d 905 (1951).

​

The committee’s own hearing transcript, read into the record at trial, reports that “major and minor malformation have also been observed.” [25] The 1996 staff review the court relied on lists ten teratogenic effects among the seventeen developmental effects it discusses, and hedges the malformation data as “not a consistent finding,” which is a judgment of weight, not a finding of absence. [26]

The Court found that cadmium is not a teratogen by excluding evidence of teratogenicity

So the court found, as fact, that cadmium “was not listed as a reproductive toxicant based on evidence of malformations,” from a committee that recorded no reasoning, over a transcript that recorded malformations, on the word of a retained witness who admitted the committee identified nothing. [27] After excluding of al evidence of teratogenicity, it found that cadmium is not a teratogen, therefore protracted exposure, therefore average the servings over a lifetime, therefore the product is under the safe harbor, therefore judgment for the defendant. [28]

The the Court did the opposite for Pictsweet: it allowed post-1996 science​

And it did the opposite in the same breath, twice. The sixty-day averaging window on which the defendant’s exposure number rests comes from “a 2006 study by Misell,” a decade after the listing. [29] The consumption figures come from a federal survey and a fourteen-day commercial diary that the court accepted as “the successor” to the 1980s report the regulation names, because the old report was old. [30] When the State itself set cadmium’s daily limit in 2001, it searched the literature published since the listing and reviewed fourteen new studies “along with the studies in OEHHA (1996).” [31] Current science for the denominator, and for the defendant’s own numerator, and for the State’s own number. 1996 for the plaintiff. The map is admitted only where it helps.

​

Put the ruling in the allegory. A tenant is injured in the five-hundred-story building. The question is how she was hurt. The court admits the 1996 street map, which shows the block and the address, and excludes the census, which shows the floor, the room, and the switch. Then it rules that because the map shows no floors, there was no fall.

​

There is an argument for construing an agency’s listing by its record. It is the argument for stability, for not relitigating in every courtroom what the State decided once. But this regulation does not fix a number. It fixes a method, which is a scientifically valid study, and it directs the fact-finder to apply that method to a biological question. A method that must be applied with the biology of 1996 is not what the regulator could have imagined in a nightmare. It is an instruction to be wrong on purpose.

​

The State did not list cadmium because it believed the science was finished. It listed cadmium because the science was sufficient to warn. A rule that turns “sufficient to warn in 1996” into “the only science admissible in 2026” does not honor the listing. It embalms it. Somewhere in the city, on a floor the map does not show, the switches are still being thrown. For pregnant women and the generations to come.

References

1. Statement of Decision, Chemical Toxin Working Group, Inc. v. The Pictsweet Co., No. 24STCV08902 (L.A. Super. Ct. July 30, 2026), p. 17 (DARTIC “voted to list cadmium as a reproductive toxicant at its December 1996 meeting”); Office of Environmental Health Hazard Assessment, The Proposition 65 List (cadmium: developmental and male reproductive toxicity, listed May 1, 1997).

2. Schena M, Shalon D, Davis RW, Brown PO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270 (1995) 467–470.

3. Milo R. What is the total number of protein molecules per cell volume? A call to rethink some published values. BioEssays 35 (2013) 1050–1055; Milo R, Phillips R. Cell Biology by the Numbers. Garland Science, 2015.

4. Bianconi E, et al. An estimation of the number of cells in the human body. Annals of Human Biology 40 (2013) 463–471; Hatton IA, et al. The human cell count and size distribution. PNAS 120 (2023) e2303077120.

5. Allis CD, Jenuwein T. The molecular hallmarks of epigenetic control. Nature Reviews Genetics 17 (2016) 487–500; Waddington CH. The Strategy of the Genes. Allen & Unwin, 1957.

6. López-Otín C, et al. The hallmarks of aging. Cell 153 (2013) 1194–1217; Horvath S. DNA methylation age of human tissues and cell types. Genome Biology 14 (2013) R115; Waterland RA, Jirtle RL. Molecular and Cellular Biology 23 (2003) 5293–5300.

7. Takahashi K, Yamanaka S. Cell 126 (2006) 663–676; Takahashi K, et al. Cell 131 (2007) 861–872; Nobel Prize in Physiology or Medicine 2012 (Gurdon and Yamanaka).

8. Ocampo A, et al. Cell 167 (2016) 1719–1733 (mice); Lu Y, et al. Nature 588 (2020) 124–129 (retinal ganglion cells, mice); Life Biosciences, nonhuman primate optic neuropathy results presented at ARVO 2024 (company announcement; not peer-reviewed).

9. Life Biosciences, announced first-in-human trial of partial epigenetic reprogramming for optic neuropathies (company announcement; confirm status at publication).

10. Altos Labs launch, approximately $3 billion (company announcement, January 2022; MIT Technology Review, Jan. 19, 2022); Retro Biosciences, $180 million (reported 2022); NewLimit, $130 million Series B (company announcement, 2025).

11. Takiguchi M, et al. Experimental Cell Research 286 (2003) 355–365.

12. Kippler M, et al. Toxicology Letters 192 (2010) 162–168.

13. Kippler M, et al. Epigenetics 8 (2013) 494–503; Sanders AP, et al. Epigenetics 9 (2014) 212–221; Everson TM, et al. Environmental Health Perspectives 126 (2018) 017010.

14. Kippler M, et al. Environmental Health Perspectives 120 (2012) 284–289.

15. Cal. Code Regs., tit. 27, § 25821(b).

16. OEHHA, Final Statement of Reasons, Title 22, California Code of Regulations, Division 2, § 12801 (June 1989) (trial Exhibit 991), p. 74, as quoted in the Statement of Decision, p. 19.

17. Cal. Code Regs., tit. 27, § 27001(c) (cadmium: “developmental, male”); Statement of Decision, p. 24.

18. Final Statement of Reasons (1989), pp. 74 (“if more specific information is desired”), 81 (“whatever proof is available”), 66–67 (“nothing in this article shall preclude”); Plaintiff’s Phase I Trial Brief (Apr. 2026), pp. 14–15.

19. Health & Saf. Code, § 25249.10, subd. (c).

20. Environmental Law Foundation v. Beech-Nut Nutrition Corp. (2015) 235 Cal.App.4th 307, 327–328; DiPirro v. Bondo Corp. (2007) 153 Cal.App.4th 150, 172; People ex rel. Brown v. Tri-Union Seafoods, LLC (2009) 171 Cal.App.4th 1549, 1575–1576; see also Mateel Environmental Justice Foundation v. Office of Environmental Health Hazard Assessment (2018) 24 Cal.App.5th 220, 237 (“the best generally-accepted methodologies should be used in each situation”); Lee v. Amazon.com, Inc. (2022) 76 Cal.App.5th 200, 211.

21. Statement of Decision, p. 3.

22. Statement of Decision, pp. 17–18; March 13, 2026 Minute Order, p. 2.

23. Reporter’s Transcript, March 13, 2026, 116:1–19.

24. Reporter’s Transcript, March 10, 2026, 111:22–28, 114:4–18 (testimony of Dr. Boomhower), as quoted in Plaintiff’s Objections to the Proposed Statement of Decision, pp. 158–159.

25. DARTIC transcript (trial Exhibit 048), read into the record at Reporter’s Transcript, March 10, 2026, 108:1–9.

26. OEHHA, Evidence on the Developmental and Reproductive Toxicity of Cadmium (Oct. 1996) (trial Exhibit 42), pp. 31–33 (ten teratogenic effects among seventeen developmental effects, per Plaintiff’s Phase I Trial Brief, p. 19); § 2.5, p. 32 (“not a consistent finding”), as quoted in the Statement of Decision, p. 29.

27. Statement of Decision, p. 29 (finding (3)).

28. Statement of Decision, pp. 29–30; judgment entered August 13, 2026.

29. Statement of Decision, p. 33 (“he settled on a 2006 study by Misell, to fix the period of spermatogenesis at 60 days”).

30. Statement of Decision, p. 12 (NHANES as “the successor to the USDA report identified in the regulations”; “the 14-day National Eating Trends survey, or NET, from Circana, Inc.”); Cal. Code Regs., tit. 27, § 25821(c)(2).

31. OEHHA, Proposition 65 Maximum Allowable Daily Level (MADL) for Reproductive Toxicity for Cadmium (Oral Route) (2001) (trial Exhibit 41), p. 1, as quoted in Plaintiff’s Phase I Trial Brief, p. 18.