health · family: it worked in the lab
we traded theculture for blindness
Gonorrhea is becoming untreatable while the shift to molecular diagnostics limits our ability to detect drug resistance
Problem statement
Neisseria gonorrhoeae infects an estimated 87 million people per year and has progressively developed resistance to the majority of antibiotic classes used against it: sulfonamides, penicillins, tetracyclines, macrolides, and fluoroquinolones. Ceftriaxone resistance — the last-line treatment — rose from 0.8% to 5% in WHO surveillance data between 2022 and 2024. Extensively drug-resistant "super-gonorrhea" strains resistant to both ceftriaxone and azithromycin have been described since 2018 and have spread internationally. Yet in many countries molecular (NAAT) diagnostics have more or less replaced culture for detecting gonorrhea, and molecular tests still cannot accurately predict resistance to ceftriaxone. Molecular methods detect only the resistance mechanisms that are already known, so they cannot entirely replace culture-based susceptibility testing.
Why this matters
Without point-of-care resistance information, clinicians must treat empirically with the most powerful remaining antibiotics, accelerating resistance selection. The WHO/FIND/GARDP collaboration published two TPPs in 2020: one for a rapid diagnostic test (TPP1, <=30 minutes, <$3) and one for an AMR/susceptibility test (TPP2, <=60 minutes, <$25). The 2020 paper expected tests meeting them to reach the market within two years. The WHO's Enhanced Gonococcal Antimicrobial Surveillance Programme (EGASP) received data from 12 countries in 2024, and the TPP expert group said better resistance data are urgently needed in Africa, Eastern Europe, Central America, the Caribbean, and Asia. Surveillance remains thin for a pathogen that has only one remaining option for first-line treatment.
What’s been tried and why it hasn’t worked
Culture-based susceptibility testing remains necessary, but in many countries NAATs have more or less replaced culture for routine detection. Molecular prediction of resistance is complicated by N. gonorrhoeae's complex resistance mechanisms: cephalosporin resistance involves multiple interacting genes (penA, mtrR, porB, ponA), making genotype-to-phenotype prediction unreliable. Several molecular tests under development detect ciprofloxacin resistance, but the TPP paper expects them to be of limited use in most African and South East Asian regions because resistance to this antibiotic is already high. WHO's surveillance data put ciprofloxacin resistance at 95%. No molecular test reliably predicts ceftriaxone susceptibility — the drug that matters most.
What would unlock progress
The TPP paper points to two routes: (1) a lateral flow test for rapid diagnosis (TPP1), a format whose earlier versions often had sensitivities of 50% or lower; and (2) a test that detects genetic markers of resistance (TPP2), which the experts expected to need a molecular format and a district-hospital laboratory. The critical unmet need is a test that can predict ceftriaxone susceptibility from clinical specimens without culture — which likely requires either breakthrough genotype-phenotype mapping or a novel rapid phenotypic approach.
Entry points for student teams
A bioinformatics team could build a machine learning model that predicts ceftriaxone resistance from targeted gene panels and ask which minimal marker set reaches clinically useful accuracy, training on NCBI Pathogen Detection's Neisseria gonorrhoeae browser — a public, no-login FTP release whose metadata table carries paired `AST_phenotypes` and `AMR_genotypes` fields alongside the assemblies, refreshed weekly (https://ftp.ncbi.nlm.nih.gov/pathogen/Results/Neisseria_gonorrhoeae/). A microfluidics team could build the rapid phenotypic susceptibility device — capture organisms, expose them to ceftriaxone in microchannels, read growth/no-growth by fluorescence in 2–4 hours — and characterize it on spiked mock swab matrices: known-phenotype laboratory strains seeded into a cervical or urethral mucus simulant with commensal Neisseria as the interfering background, which is where the real engineering problem (viable capture from mixed flora without conventional culture) actually lives. That bench arm needs a BSL-2 microbiology lab and characterized strains — the 2024 WHO gonococcal reference panel of 15 strains spans every susceptible and resistant phenotype including new-penA ceftriaxone resistance (Unemo et al., J Antimicrob Chemother 2024) and is distributed through the WHO Collaborating Centre that curates it, so a faculty microbiologist has to request them. Running the device on actual clinical swabs takes an STI-clinic partnership and IRB approval that the clinic owns; the semester deliverable is the characterized device plus the clinical evaluation protocol. Relevant disciplines: microbiology, microfluidics, bioinformatics, antimicrobial resistance research.
Genes
Structural cousins: stuck the same way, in other fields
Sources
Ferreyra C et al., "Developing target product profiles for Neisseria gonorrhoeae diagnostics in the context of antimicrobial resistance: An expert consensus," PLOS ONE 15(9):e0237424, 2020. WHO, "More countries report rising levels of drug-resistant gonorrhoea," Nov 19, 2025. go to source 1 ↗ go to source 2 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
- The diagnostic displacement paradox (better detection technology destroys resistance monitoring capability) is structurally interesting — a case where technological progress in one dimension creates regression in another.
- Pharyngeal gonorrhea as a hidden AMR reservoir is a key research frontier: asymptomatic infections with poor antibiotic penetration and horizontal gene transfer from commensal Neisseria species.
- The two TPPs (TPP1 for diagnosis, TPP2 for AMR) mirror the TB diagnostic/DST split — related but distinct technical challenges requiring different approaches.
Reconciliation 2026-08-21: The C37 entry-point triage flagged the lead suggestion for having students capture N. gonorrhoeae directly from clinical swab specimens, which takes a clinic partnership, IRB approval, and handling of a fastidious BSL-2 pathogen in mixed flora. Confirmed on re-reading — the phrase "directly from clinical swab specimens" was doing the work of an unstated access chain. The genomics suggestion was already facility-free but named no data, so it now leads and carries a verified public source; the microfluidics suggestion is retargeted onto spiked mock swab matrices with characterized laboratory strains, with an explicit access line for the BSL-2 lab and strain request, and the clinical-specimen work becomes an evaluation protocol handed to the clinic that owns the access. Resources verified by fetch: NCBI Pathogen Detection's N. gonorrhoeae release directory, public and unauthenticated, current snapshot PDG000000032.684 dated 2026-08-21, whose metadata TSV header includes `AST_phenotypes`, `AMR_genotypes` and `number_drugs_tested` columns — https://ftp.ncbi.nlm.nih.gov/pathogen/Results/Neisseria_gonorrhoeae/latest_snps/Metadata/ ; and the 2024 WHO gonococcal reference strain panel (n=15, plus 14 superseded strains), Unemo M et al., J Antimicrob Chemother 2024, doi:10.1093/jac/dkae176, https://pmc.ncbi.nlm.nih.gov/articles/PMC11290888/ — the panel is a physical strain collection obtained through its curating WHO Collaborating Centre, not a download, and is labelled as such in the entry points.
Reconciliation 2026-09-29: The whole write-up was read against its sources, all fetched on 2026-09-29. The TPP paper was read in full from https://pmc.ncbi.nlm.nih.gov/articles/PMC7462286/, including Figures 2 and 3 and the two supplementary files that hold the full TPP tables. The WHO news release was read in full from the address in the Source line. The 2024 reference-strain paper named in the note above (Unemo et al.) was fetched from https://pmc.ncbi.nlm.nih.gov/articles/PMC11290888/; its introduction and discussion were read and the rest of its text was searched. (1) The Source line gave the first author as "Toskin I". The record names Cecilia Ferreyra as first author and carries the copyright line "2020 Ferreyra et al". Toskin appears only in the reference list. The Source line now says "Ferreyra C et al." (2) The text said gonorrhea infects 82 million people per year. The TPP paper gives an estimated 87 million cases per year, and no source read here gives 82 million. The text now says 87 million. (3) The text said the bacterium developed resistance to "every antibiotic class used against it over 80 years." The TPP paper says "the majority of available classes" and gives no span of years. The text now says the majority of classes. (4) The text said ceftriaxone resistance rose from 0.8% to 5% "globally." The WHO release reports those figures from its surveillance programme, which had data from 4 countries in 2022 and 12 in 2024. The text now says "in WHO surveillance data." (5) The text said strains resistant to both ceftriaxone and azithromycin "have been reported across multiple continents." The 2024 paper says such strains have been described since 2018 and refers to their international spread. The text now says that. (6) The text said the shift to molecular tests "destroyed our ability" to test for resistance and that laboratories worldwide are losing the capacity to culture gonorrhea. No source read here says this. The 2024 paper says that in many countries molecular tests have more or less replaced culture for detection, that molecular methods detect only known resistance mechanisms and cannot entirely replace culture-based testing, and that predicting ceftriaxone resistance is still a challenge. The WHO release reports the culture-based surveillance programme growing from 4 to 12 countries. The Problem Statement now reports what the 2024 paper says, and the title on line 1 now reads "While the Shift to Molecular Diagnostics Limits" in place of "Because the Shift to Molecular Diagnostics Eliminated." The first sentence of "What's Been Tried" was changed in the same way; its details about special media, a CO2 atmosphere and 24 to 48 hours of incubation are in none of the sources and were removed. (7) The text said the two TPPs were developed in 2020 and "Neither exists." The TPPs were finalized in 2019 and the paper was published in 2020. The paper does not say that no such test exists; it says tests meeting the TPPs were expected to reach the market within two years. The text now says that. (8) The text said the WHO programme has limited or no data from Eastern Europe, Central Asia, Latin America, the Eastern Mediterranean and most of Africa, "the regions with the majority of cases." Neither source gives that list or that claim. The WHO release says 12 countries reported in 2024, and the TPP paper says better resistance data are urgently needed in Africa, Eastern Europe, Central America, the Caribbean and Asia. The text now reports those two statements. The closing sentence of that section said we are losing the ability to track "the pathogen most likely to become completely untreatable"; it now says surveillance remains thin for a pathogen with one remaining first-line treatment option, which is how the 2024 paper describes ceftriaxone. (9) The text said ResistancePlus GC (SpeeDx) is FDA-approved with 94.8% sensitivity and 100% specificity. None of the sources names this test or gives these figures, so the sentence was replaced with what the TPP paper says about ciprofloxacin tests under development and with the WHO figure of 95% ciprofloxacin resistance. (10) The text described a lateral flow test called NG-LFA and a multiplex HRM PCR test with 98.6% sensitivity, 99.2% specificity and a cost under $1 per sample. None of the sources mentions either test or these figures. The sentence now describes the two routes set out in the TPP paper. Claims (9) and (10) may be true, but they need their own cited sources before they return. Confirmed against the sources and left as written: the title, journal, volume, article number and year of the TPP paper; the title and date of the WHO release; the five named antibiotic classes; ceftriaxone resistance of 0.8% and 5% for 2022 and 2024; TPP1 at 30 minutes or less and under $3; TPP2 at 60 minutes or less and under $25; the WHO, FIND and GARDP collaboration; the statement that no molecular test reliably predicts ceftriaxone susceptibility; the role of penA; gene transfer from other Neisseria species in the throat. Not confirmed and left as written: that mtrR, porB and ponA act together in cephalosporin resistance (the 2024 paper lists them as resistance genes, but it was not read closely enough to settle this); the phrase "poor antibiotic penetration" in the second Source Note, which was not found in any source; the short title, which carries no number. The entry points were checked in the note above and were not rechecked. Genes untouched.