In a San Diego neighborhood, a research team from UC San Diego has spent years building trust among unhoused people who use drugs, on both sides of the border with Tijuana, offering snacks, small payments, and survey questions. Since spring, they have found that roughly 10% of their research participants have active syphilis, infections that would mostly have gone undiagnosed and untreated otherwise. The team is now caught between three problems at once: a national shortage of the penicillin that cures the disease, federal restrictions on foreign research collaborations that complicate their cross-border monitoring, and a population that is hard to reach and easy to lose track of.
It is tempting to read this as simply a sad story about vulnerable people and bad luck. It is more usefully read as something sharper. Syphilis is a disease medicine solved long ago. We have known how to cure it cheaply for roughly eighty years, we have a reliable test, and we understand exactly how it spreads. So its resurgence, and the resurgence of its most tragic form, congenital syphilis in newborns, is not a medical mystery. Every reason the disease is spreading is a failure in a system built around the cure, not a gap in the cure itself. That distinction is the whole story, and it makes syphilis a uniquely clear diagnostic of how well or badly the machinery of public health is working.
A solved problem coming undone
Start with what is not in doubt. Benzathine penicillin G, the drug that cures syphilis, has been available since the 1940s. It is not expensive, not experimental, and not scientifically contested. A single injection, or a short series, reliably cures the infection at most stages, and timely treatment of a pregnant patient reliably protects the baby. On the science, syphilis is as close to a settled problem as medicine has.
That is precisely what makes its return alarming in a way a genuinely hard disease would not be. When a condition we cannot yet treat kills people, the failure is medicine's, and the remedy is research. When a condition we have cured for eight decades starts killing babies again, the failure is not in the medicine. It is in every non-medical link between the cure and the patient, and the fact that the disease is medically trivial to stop means its spread specifically indicts those links. You cannot blame the science, because the science works. You are left looking at manufacturing, supply, funding, and access, which is exactly where the failures are.
Failure one: a cure that cannot be reliably made
The first broken link is the one the San Diego team is living with directly. Benzathine penicillin G is old, off-patent, and cheap, which sounds like a virtue and is, economically, a vulnerability. Cheap, off-patent drugs carry thin margins and little incentive for companies to build redundant, resilient production, so they tend to be made by very few manufacturers, sometimes only one. Benzathine penicillin G in the US essentially comes down to a single manufacturer, Pfizer's King Pharmaceuticals, selling it as Bicillin L-A, and that concentration is exactly what turns a single disruption into a national crisis.
That is what happened. A shortage that began around surging demand in 2023 worsened after a July 2025 voluntary recall of certain lots for particulate contamination, and the projected recovery has since been pushed out repeatedly, with deliveries and full recovery now stretching well into 2026 and, by some estimates, 2027. The FDA has resorted to temporarily authorizing imported alternatives to fill the gap. This is the general pathology of essential-generic shortages: the drugs that are cheapest and most fundamental are the most shortage-prone, precisely because being cheap and fundamental attracts no investment in resilient supply. A cure that cannot be manufactured dependably is, for the patient who cannot get it this month, not fully a cure at all. The knowledge is intact; the supply is not.
Failure two: no substitute for the people who need it most
The second broken link is what makes this shortage far worse than an ordinary one. For most bacterial infections, a drug shortage is an inconvenience, because you can switch to another antibiotic. Syphilis in pregnancy does not work that way. Benzathine penicillin G is the only treatment that reliably crosses the placenta and protects the fetus, and the standard alternative for non-pregnant adults, doxycycline, is not a safe substitute in pregnancy. For a pregnant patient with syphilis, there is no equivalent. It is that specific drug or a failed defense of the baby.
So the harm of the shortage does not fall evenly. It concentrates on exactly the population with no alternative, pregnant patients and their infants, and the consequence is congenital syphilis, which can cause stillbirth, newborn death, and lifelong disability. The numbers are moving in the wrong direction fast: congenital syphilis rates have risen more than 200% over five years in the US. Nearly every one of those cases is preventable with a test most pregnant patients already receive and a cheap injection given on time. When the injection is not available, or the patient is not reached in time, the disease does something it has no medical reason to do anymore: it damages or kills a newborn. A shortage that would be survivable for most infections becomes, for this one population, a direct pathway to preventable infant harm, because the one group that cannot switch drugs is the group the disease punishes most severely.
Failure three: throttling detection as the disease surges
The third broken link is about finding the disease, and it is where the San Diego team's federal-funding problem enters. The reason their work matters is that they locate infections that would otherwise stay hidden, the roughly 10% of their participants carrying active syphilis who were not going to be diagnosed through ordinary channels. Hidden, untreated cases are the reservoir that keeps an epidemic spreading, so reaching marginalized, hard-to-track populations is not charity work at the margins; it is core epidemic control.
That surveillance is being made harder at exactly the wrong moment. Federal restrictions on foreign research collaborations have complicated the team's cross-border monitoring in Tijuana, part of a broader tightening of the research-funding environment. There are stated reasons for scrutinizing foreign research partnerships, concerns about oversight and accountability of overseas collaborators that are not frivolous, and this analysis takes no position on the merits of those policies in general. But the specific cost here is real and worth naming plainly: infectious diseases do not respect borders, and a syphilis epidemic that moves between San Diego and Tijuana cannot be monitored on only one side of the line. Constraining the cross-border work degrades the detection of a disease that is actively crossing the border, and it does so at the same moment the supply of the cure is failing. Detection and treatment, the two things that stop an epidemic, are being throttled together.
Why congenital syphilis is the clearest possible indictment
Put the three failures next to each other and congenital syphilis emerges as something like a diagnostic instrument for the health system itself. Because the disease is so cheaply and reliably curable when everything works, its spread, and especially its spread to newborns, is almost a pure measurement of how badly the surrounding systems are functioning. A rising congenital syphilis rate is not evidence of a tough new pathogen. It is evidence that prenatal care is not reaching people, or that the drug is not on the shelf, or that surveillance is missing cases, or all three.
This is what makes the disease's return genuinely ominous beyond its own toll. Many public-health failures are ambiguous, tangled up with genuine scientific difficulty, so you cannot cleanly separate "the problem is hard" from "the system is failing." Congenital syphilis offers no such ambiguity. The problem is not hard. A functioning system had nearly eliminated it. Its resurgence therefore reads as a direct signal that the delivery machinery, manufacturing, supply chains, the prenatal-care safety net, disease surveillance, has decayed, because nothing else can explain a curable disease killing babies at rising rates. You can, in effect, read the health of the whole system off the syphilis numbers, and right now they are saying something is broken.
Who ends up paying
It is worth being honest about where the failures land, because it is not random. The people caught in this convergence are among the most marginalized, unhoused, using drugs, moving across a border, with the least access to consistent care and the least political voice to demand better supply or funding. The woman whose story opens the STAT account came to the research site for gummies and a small payment and left having learned she had a serious, treatable infection she had no idea she carried. That is the system working at the very last moment it can, catching a case that everything upstream had missed.
The pattern is that system failures concentrate on those least equipped to absorb them. A well-resourced pregnant patient with good prenatal care and a pharmacy that has the drug is unlikely to become a congenital syphilis statistic. The failures fall on the people already carrying the most disadvantage, and on their children, which is both the human tragedy of the story and a reason the numbers stay partly invisible: the harm accrues to populations the rest of the system is not watching closely.
How to read it
The clarifying way to understand the San Diego situation, and the national one it reflects, is that syphilis has become a solved medical problem in the process of becoming an unsolved logistical and political one. Nothing about the disease got harder. What decayed is everything around the cure: the resilience of manufacturing a cheap essential drug, the protection of the one population that has no alternative to it, and the funding and reach of the surveillance that finds hidden cases. None of the fixes are glamorous or scientific. They are the unheroic work of making sure an eighty-year-old drug is actually on the shelf, that pregnant patients can get it in time, and that the people tracking the epidemic can keep doing so across the border it crosses.
The disease's return is best read as a warning indicator rather than a medical event. Because syphilis is so easy to cure when the systems work, its spread is a uniquely honest report on whether they are working, and the report is not good. The tragedy is that every congenital case represents a failure that was entirely preventable with tools we have had for generations, which means the suffering is not the price of a hard scientific problem but the consequence of letting solved problems come undone. For anyone who can act on it, the practical core is unchanged from what it has been for decades: syphilis is testable and curable, and the people most at risk are reachable if the will and the supply exist. The medicine did its part a long time ago. Everything failing now is downstream of that, which is exactly why it did not have to happen.
Primary sources
- STAT for the reporting on the UC San Diego team's syphilis monitoring among unhoused people who use drugs in San Diego and Tijuana, the finding that roughly 10% of participants had active syphilis, the effect of NIH restrictions on foreign research collaborations on their cross-border work, and the patient narrative.
- Medical Daily for the shortage's origins predating 2026 and worsening after Pfizer's July 2025 voluntary recall of certain penicillin G benzathine lots for particulate contamination, King Pharmaceuticals as the Pfizer subsidiary and effectively sole US manufacturer, the extended delivery and recovery timelines, and the point that syphilis in pregnancy has no substitute treatment because benzathine penicillin G must reach the fetus on schedule.
- The CDC's Bicillin L-A updates for benzathine penicillin G being the only recommended treatment for syphilis in pregnancy, Bicillin L-A being the only FDA-approved BPG formulation in the US, and the temporary authorized importation of Lentocilin.
- The California Department of Public Health advisory for prioritization guidance directing scarce long-acting penicillin to pregnant patients and exposed infants and the use of doxycycline only for non-pregnant adults.
- The New Mexico Department of Health alert for the more-than-200% rise in congenital syphilis over five years and the temporary FDA importation of Extencilline and Lentocilin.
- MedFinder and provider briefings for the 2023 shortage onset amid surging syphilis rates and manufacturing constraints, the single-domestic-manufacturer concentration, and the Q3-2026-and-beyond recovery projections.