Mark ở vũ trụ khác · Another Universe Genes and Rare Diseases #6 · Blood · Evolution

Sickle cell: the disease that was once a shield

!

Herrick 1910 · Pauling 1949 · Allison 1954 in Kenya · gene editing 2023

ConditionSickle cell disease, an inherited disorder of haemoglobin
GeneHBB (beta chain), the same gene as beta-thalassemia in #5
The storyone changed letter causes a severe disease, yet one copy of the gene protects against malaria
LabelSolid, with caution on the word "cure"

For reading and understanding, not a substitute for medical advice.

Source: NHGRI (genome.gov); Allison (1954), BMJ 1(4857): 290–294; FDA announcement of 8 Dec 2023 on Casgevy and Lyfgenia.

Recap in 1 minute

Essence

One letter changed, a bent cellOne amino acid is swapped in the beta chain. Short of oxygen, the haemoglobin forms fibres that bend the red cell.
Bent cells block vesselsCausing severe pain crises, anaemia and damage to many organs.
One copy is a shieldCarriers of one copy have about a tenth of the odds of severe malaria. So the gene is common where malaria was common.
Not rare worldwideAbout 7.74 million people were living with the disease in 2021, by GBD estimates.
Gene editing exists, but is newSince Dec 2023, most people in trials have had no severe pain crisis for at least a year. Price above 2 million USD.

Source: NHGRI (genome.gov); Taylor et al. (2012), Lancet Infectious Diseases 12(6): 457–468 (odds ratio 0.09); GBD 2021, Lancet Haematology 10(8): e585–e599; FDA 8 Dec 2023 (via summaries); prices from US press (secondary source).

Part 1

Question

If a gene causes such a severe disease, natural selection should have removed it long ago. Why is the sickle gene common in parts of Africa, and more than 100 years after it was first seen under a microscope, how far has treatment come?

The question is Mark's, based on Allison (1954) and GBD 2021.

1910SeenA doctor in Chicago sees red cells bent like sickles.
→
1949–1957UnderstoodThe haemoglobin differs by a single amino acid.
→
1954 and todayShield, then editingThe gene resists malaria; in 2023 gene editing arrives.
Part 2

The story

Who, where, when: from a clinic in Chicago to villages around Lake Victoria.

1910 · Chicago, USAWalter Clement Noel, a dental student from the island of Grenada, has severe anaemia. Intern Ernest Irons looks at his blood and sees red cells that are "elongated and sickle-shaped". Physician James Herrick publishes the case: the first description of the disease in Western medical literature.
1949 · A "molecular disease"Linus Pauling and colleagues run haemoglobin through an electric field: patients' haemoglobin moves differently from healthy people's. For the first time a disease is traced to a specific molecule.
1954 · Kenya and UgandaBritish doctor Anthony Allison finds the sickle trait common on the coast and around Lake Victoria, where malaria is heavy, but rare in the highlands. Children with the trait carry fewer malaria parasites. He publishes in the BMJ: the sickle gene shields against malaria.
1956–1957 · One letterVernon Ingram in Cambridge compares the haemoglobin piece by piece: just one amino acid differs. At position 6 of the beta chain, glutamic acid is replaced by valine.
Today · WorldwideIn 2021 about 515,000 babies were born with the disease, most in sub-Saharan Africa. In the US, about 1 in 365 African American babies is born with it, and about 1 in 13 carries the gene.

Source: NHGRI (genome.gov), education kit "1956: Cause of disease traced to alteration"; Herrick (1910), Archives of Internal Medicine 6: 517–521; Pauling et al. (1949), Science 110; Allison (1954), BMJ 1(4857): 290–294 and Trans R Soc Trop Med Hyg 48(4): 312–318; HHMI BioInteractive (geography, secondary source); GBD 2021, Lancet Haematology 10(8); HHS Office of Minority Health, citing CDC data (1/365, 1/13).

Diagram of Allison's 1954 idea: in lowland areas with heavy malaria, such as the coast and around Lake Victoria, the sickle trait is common; in the highlands with little malaria, it is rare. Allison's idea (1954): the gene follows malaria highlands little malaria trait is rare coast around Lake Victoria heavy malariamany carriers heavy malariamany carriers
The gene is not spread at random. Each blue dot stands for carriers; the diagram is not to scale and only shows the direction of the finding. Allison did not give a single rate for the whole region.Source: Allison (1954), Trans R Soc Trop Med Hyg 48(4): 312–318 and BMJ 1(4857): 290–294; HHMI BioInteractive (secondary source). Self-drawn diagram.
Who gets the credit? A debate among historians

The familiar story puts Pauling at the centre. Some medical historians, such as Feldman and Tauber, call this account "romanticised": Pauling showed the haemoglobin was different, but Ingram showed where.

The disease also existed in Africa long before 1910, with its own names in many local languages. "First described" means first in Western medical literature.

Source: Feldman and Tauber, article on Project MUSE (muse.jhu.edu/article/3977); NHGRI (genome.gov).

Part 3

How the gene works

1. The HBB gene makes the beta chain of haemoglobin, the oxygen carrier in red cells. In patients, one letter of the gene is changed (codon GAG becomes GTG), so at position 6 of the beta chain glutamic acid is replaced by valine. This haemoglobin is called HbS.

2. When it releases oxygen, HbS sticks together into long fibres, pulling the red cell from a soft disc into a stiff sickle. Sickled cells jam in small blood vessels, causing severe pain, and break early, so patients are anaemic.

3. Everyone has two copies of HBB. One HbS copy (AS): a carrier, usually healthy, and less likely to get severe malaria. Two copies (SS): disease. Newborns still have fetal haemoglobin, which does not join HbS fibres, so signs usually start around 5 months of age.

Left: normal red cells are soft discs passing through a small vessel. Right: HbS forms fibres when short of oxygen, cells bend into sickles, jam in the small vessel and block blood flow. HbA: soft, round cells blood flows through a small vessel codon GAG · glutamic acid at position 6 HbS: low oxygen, fibres form HbS sticks into fibres block codon GTG · valine at position 6
One letter in the gene, one amino acid in the protein, a different shape of cell. Modern numbering (counting the initial amino acid) calls it position 7; older texts and NHGRI say position 6.Source: NHGRI (genome.gov); MedlinePlus Genetics "Sickle cell disease". Self-drawn diagram, not to scale.
2012 meta-analysis: compared with non-carriers, carriers of one sickle copy had an odds ratio of 0.09 for severe malaria, about one tenth; for uncomplicated malaria the ratio was 0.69. malaria risk for carriers of one copy (AS), vs non-carriers (= 1) non-carriers1 AS, uncomplicated0.69 AS, severe malaria0.09 severe malaria is what kills young children; that is where the shield is strongest
A strong shield against severe malaria, a weaker one against ordinary malaria. The cost: if two carriers have children, each pregnancy has a 25% chance of a child with the disease.Source: Taylor, Parobek and Fairhurst (2012), Lancet Infectious Diseases 12(6): 457–468 (0.69 is an incidence rate ratio, 0.09 an odds ratio); MedlinePlus Genetics (25%).
Part 4

Fact check: true stories and rumours

14 things people say about sickle cell, checked against medical sources. Tap a row for the reason.

GBD 2021 estimate: deaths coded directly to sickle cell disease about 34,400; total deaths related to the disease about 376,000, nearly 11 times as many. in 2021, how many deaths were linked to the disease? (GBD estimate) coded as the cause≈34,400 total related≈376,000 many children die of infection or malaria without the underlying disease being recorded the two numbers measure different things: read carefully before quoting
A disease "invisible" on death certificates. Where there is no newborn screening, children can die before being diagnosed.Source: GBD 2021 Sickle Cell Disease Collaborators (Thomson et al. 2023), Lancet Haematology 10(8): e585–e599; summary on healthdata.org.
SOLIDA child has the disease only if they inherit the sickle gene from both parents

Autosomal recessive. If both parents are carriers, each pregnancy has a 25% chance of an affected child, 50% of a carrier, 25% of neither (MedlinePlus Genetics).

SOLIDThe disease comes from one amino acid swapped in the beta chain

Valine replaces glutamic acid at position 6 (NHGRI; Ingram 1956–1957). Same HBB gene as beta-thalassemia, but a different error.

SOLIDCarriers of one copy are strongly shielded against severe malaria

2012 meta-analysis: odds ratio 0.09, about one tenth of the non-carrier risk (Taylor et al. 2012).

SOLIDBabies with the disease usually show no signs in the first months

Signs usually start around 5 months of age (MedlinePlus), as fetal haemoglobin fades. That is why many countries test at birth.

SOLIDHydroxyurea cuts pain crises by nearly half

MSH trial (299 adults): median 2.5 crises per year vs 4.5 on placebo; stopped early for benefit (Charache et al. 1995, NEJM 332(20): 1317–1322).

SOLIDGene editing freed most trial participants from severe crises for at least a year

Casgevy: 29 of 31 evaluable people per FDA (93.5%); 29 of 30 in the 2024 NEJM paper (97%). The denominators differ because of different data cut-offs.

SOLIDAbout 515,000 babies were born with the disease in 2021, most in sub-Saharan Africa

GBD 2021 estimate (Lancet Haematology 2023). A modelled estimate, but the best global source available.

CAUTION"Gene therapy has cured sickle cell disease"

Results are very good, but median follow-up was about 19 months (NEJM 2024). Recipients can still pass the gene to their children, need high-dose chemotherapy beforehand, and Lyfgenia carries a warning about blood cancer. "Potentially life-changing" is more accurate than "cured".

CAUTION"The sickle gene is a good gene"

Only for people with one copy, where malaria exists. Two copies cause severe disease. It is an evolutionary trade-off, not a gift.

CAUTION"376,000 people die of sickle cell every year"

That is total related mortality; deaths coded directly are about 34,400. Both are modelled estimates with wide uncertainty (GBD 2021).

CAUTION"Carrying the gene is completely harmless"

Usually benign. Very rarely there are complications under extreme exertion or dehydration. In 2010 the US college sports body (NCAA) required athletes to be tested; the American Society of Hematology (ASH) opposed it and urged safety measures for everyone instead of screening (via news, secondary source).

FALSE"Sickle cell disease is contagious"

It is inherited, present from conception. It does not spread through contact, blood or mosquitoes (MedlinePlus).

FALSE"Carriers will eventually develop the disease"

Carrier (AS) and disease (SS) are different genotypes and stay the same for life (MedlinePlus).

FALSE"Once the FDA approves a drug, it is surely safe for the long term"

Voxelotor got accelerated approval in 2019. On 25 Sep 2024 the company withdrew it from all markets because later studies showed more vaso-occlusive crises and deaths in the treated group. The European Medicines Agency said there was no clear evidence the drug caused the deaths, but the benefits no longer outweighed the risks.

Source: as noted in each row; full list in Part 9.

Part 5

Where things stand

From an old drug to gene editing. Each milestone has a year and a source.

1995 · HydroxyureaThe MSH trial in 299 adults was stopped early because the drug cut pain crises by nearly half (median 2.5 vs 4.5 per year). It makes the body produce more fetal haemoglobin. The FDA approved it for adults in 1998.
2019 → 2024 · A lesson about new drugsVoxelotor received FDA accelerated approval in 2019, then was withdrawn from all markets by its own maker on 25 Sep 2024 after new safety data.
16 Nov 2023 · The UK goes firstThe UK medicines regulator (MHRA) approves Casgevy, the world's first approved therapy using CRISPR.
8 Dec 2023 · FDA approves two gene therapiesCasgevy cuts a "switch" so the patient's own stem cells make fetal haemoglobin again: 29 of 31 evaluable people had no severe crisis for at least 12 months. Lyfgenia adds a corrected beta gene: 28 of 32 had complete resolution of crises 6 to 18 months after treatment, with a warning about blood cancer.
2024 · Results in the NEJM44 people received Casgevy; 29 of 30 evaluable (97%) had no severe crisis for at least 12 months, with median follow-up of 19.3 months.

Source: Charache et al. (1995), NEJM 332(20): 1317–1322; ClinicalTrials.gov NCT00000586; Pfizer 25 Sep 2024 and EMA (via Pharmaceutical Journal, Sickle Cell Anemia News); MHRA 16 Nov 2023 (via Pharmaceutical Technology); FDA 8 Dec 2023 (via summaries and company releases); Frangoul et al. (2024), NEJM 390(18): 1649–1662. Drug names are for reference only, not treatment suggestions.

The price, and who gets it

In the US, Casgevy is priced at about 2.2 million USD and Lyfgenia about 3.1 million USD for a one-time treatment (press, secondary source). Patients need high-dose chemotherapy to "clear space" in the bone marrow and weeks in hospital at a major centre.

Meanwhile most babies with the disease are born in sub-Saharan Africa (GBD 2021), where many are not yet tested at birth. For them, newborn screening, infection prevention and low-cost hydroxyurea may save more lives than gene therapy for years to come. This is Mark's comment from the sources above.

Part 6

Verdict

Scores are Mark's judgement based on the sources in Part 9, not data from a study.

Understanding of the gene
5/5
Evidence for the malaria shield
4/5
Gene therapy: long-term data
3/5
Access where most patients live
1/5
Strengths
  • The first disease traced to a molecule; very clear mechanism
  • A classic example of natural selection in humans
  • Hydroxyurea is cheap with a randomised trial; gene therapy highly effective in trials
Weaknesses
  • Gene therapy has about 2 years of follow-up and costs millions
  • An approved drug was withdrawn after 5 years
  • Global figures are modelled; the places with most disease have the least data

Label: SOLID. One letter in a gene is both a shield and a burden, and the promise of a "cure" should be read alongside follow-up time and price.

More: why was this disease neglected for so long?

The disease was described in 1910 and its mechanism understood in the 1950s, but adults had no FDA-approved drug until 1998: 88 years after Herrick's case. Most patients live in places that receive little research investment.

2023 changed that, but a question remains: what does a treatment mean if most people who need it cannot reach it?

The "More" section is Mark's comment, based on sourced milestones in Parts 2 and 5 (Herrick 1910, Ingram 1956–1957, FDA approval of hydroxyurea 1998) and GBD 2021.

Part 7

Lessons: five questions when you read "cured"

There will be more news about gene therapy. Five questions help you read it more calmly.

For reading and understanding, not a substitute for medical advice. This is a way to read news, not a guide to choosing treatment.

Five questions when reading cure news, applied to Casgevy: how long, median 19.3 months follow-up; how many people, 44 treated; compared with what, single-arm trial versus before treatment; what risks, high-dose chemotherapy; who can get it, about 2.2 million USD in the US. 1 · how long?19.3 monthsmedian follow-up 2 · how many?44 peopletreated 3 · compared to?single armvs before treatment 4 · what risks?high-dose chemobefore infusion 5 · who gets it?≈2.2 million USDin the US (press) These questions do not make a good result any less good. They only tell you how sure the news can be.
Tried on Casgevy, the first gene-editing therapy. The five questions are a news-reading habit suggested by Mark.Source: Frangoul et al. (2024), NEJM 390(18): 1649–1662 (44 people, 19.3 months, single-arm design); price from US press (secondary source).
Ask "followed for how long?""No crises for 12 months" is a real result, but not yet "for life".
Ask "how many, compared with whom?"A few dozen people and no comparison group: a notable result that needs more data.
Ask "what are the risks?"Voxelotor is the example: approved, then withdrawn when more data came in.
Ask "who can get it?"A great treatment that few people can use is only half good news.
Self-check: read a headline with five questions
  1. Does the news say "cured" or "free of symptoms for X months"?
  2. How many people were treated, and how many were followed long enough?
  3. Compared with whom: placebo, an older drug, or themselves before?
  4. Are side effects and long-term risks mentioned?
  5. What does it cost, where is it done, and is it available in my country?

The list of questions is Mark's suggestion, based on sourced examples in this article (Frangoul 2024; FDA 8 Dec 2023; Pfizer 25 Sep 2024).

Part 8

Read next

NHGRI, "1956: Cause of disease traced to alteration"A short kit on Herrick, Pauling and Ingram.
Allison (1954), BMJ, "Protection afforded by sickle-cell trait against subtertian malarial infection"The original paper on the malaria shield, free on PubMed Central (PMC2093356).
Genes and Rare Diseases #5: ThalassemiaThe same HBB gene, a different kind of error, and two islands that prevented the disease by testing. Read
Genes and Rare Diseases #7: ProgeriaA gene error that makes children age fast, and the first drug for it. Read
Part 9

Sources

Main sources are original papers, NHGRI, GBD estimates and drug regulator decisions. This environment blocks PubMed, NIH and FDA, so figures were checked against excerpts in search results (16 Oct 2026), without opening full texts. Secondary sources are marked.

  1. NHGRI (National Human Genome Research Institute), "1956: Cause of disease traced to alteration", education kit. genome.gov
  2. Herrick JB (1910), "Peculiar elongated and sickle-shaped red blood corpuscles in a case of severe anemia", Archives of Internal Medicine 6: 517–521 (cited from NHGRI's reference list).
  3. Pauling L, Itano HA, Singer SJ, Wells IC (1949), "Sickle cell anemia, a molecular disease", Science 110. Archive: Oregon State University
  4. Allison AC (1954), "Protection afforded by sickle-cell trait against subtertian malarial infection", BMJ 1(4857): 290–294. PMC2093356; and Allison AC (1954), Trans R Soc Trop Med Hyg 48(4): 312–318.
  5. Taylor SM, Parobek CM, Fairhurst RM (2012), "Haemoglobinopathies and the clinical epidemiology of malaria: a systematic review and meta-analysis", Lancet Infectious Diseases 12(6): 457–468. PMC
  6. GBD 2021 Sickle Cell Disease Collaborators (Thomson AM et al., 2023), "Global, regional, and national prevalence and mortality burden of sickle cell disease, 2000–2021", Lancet Haematology 10(8): e585–e599. healthdata.org
  7. MedlinePlus Genetics, "Sickle cell disease". medlineplus.gov
  8. HHS Office of Minority Health, "Sickle cell disease topic guide" (citing CDC data). minorityhealth.hhs.gov
  9. Charache S et al. (1995), "Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia", NEJM 332(20): 1317–1322. NCT00000586
  10. Frangoul H et al. (2024), "Exagamglogene autotemcel for severe sickle cell disease", NEJM 390(18): 1649–1662.
  11. FDA, announcement of 8 Dec 2023 approving Casgevy and Lyfgenia; MHRA (UK), Casgevy 16 Nov 2023. Cited via summaries, Pharmaceutical Technology and company releases.
  12. Voxelotor: Pfizer statement 25 Sep 2024; EMA 26 Sep 2024. Cited via Pharmaceutical Journal and Sickle Cell Anemia News.
  13. Secondary sources: HHMI BioInteractive (geography after Allison); Feldman and Tauber on Project MUSE (historians' debate); US press (drug prices); news on NCAA 2010 and the ASH position.