Place · Level 3
H. pylori · the one cause of cancer you can cure
它没碰到胃酸, 靠尿素酶造碱性云 · Correa 级联走几十年 · CagA 像针管扎进细胞 · 沿着饭桌传播, 该改的是公筷不是人 · 根除降风险但不归零
Story path
- 1How it survives the acidHow it survives the acid
- 2The Correa cascade · a decades-long roadThe Correa cascade · a decades-long road
- 3CagA · a syringe into your cellsCagA · a syringe into your cells
- 4Why China carries the loadWhy China carries the load
- 5How it's found · how it's treatedHow it's found · how it's treated
- 6After eradication · the honest ledgerAfter eradication · the honest ledger
Chapter 1
How it survives the acid
How it survives the acid
Stomach acid kills nearly every bacterium you swallow with food. One survives — not because it withstands the acid, but because it never touches the acid.
It's *Helicobacter pylori* (H. pylori). It carries one piece of equipment: urease, an enzyme that splits urea. Gastric juice already contains a little urea, and urease splits it into ammonia and carbon dioxide. Ammonia is alkaline, and it neutralizes the acid seeping in around the bacterium on the spot — a cloud of alkali wrapped around itself (Scott 2002).
The device even has a power-saving switch. A small gate on the bacterial surface lets only urea through. It normally stays shut and opens only as acid closes in — no effort wasted when there's no acid.
But it still has to leave the strong acid of the stomach cavity. A layer of mucus covers the stomach wall, thick as gel. It was long assumed the corkscrew body screwed its way in, but the real mechanism is neater: urease raises the pH nearby, the mucus turns from gel to fluid, and the bacterium swims through on its flagella (Celli 2009).
Under the mucus, it settles on the surface of the stomach's own cells. The pH there is near neutral — the only mild spot in the stomach. So it doesn't live *in* the acid; it lives underneath it. For how stomach acid is made and how the mucus barrier works, see the digestive system island.
It's *Helicobacter pylori* (H. pylori). It carries one piece of equipment: urease, an enzyme that splits urea. Gastric juice already contains a little urea, and urease splits it into ammonia and carbon dioxide. Ammonia is alkaline, and it neutralizes the acid seeping in around the bacterium on the spot — a cloud of alkali wrapped around itself (Scott 2002).
The device even has a power-saving switch. A small gate on the bacterial surface lets only urea through. It normally stays shut and opens only as acid closes in — no effort wasted when there's no acid.
But it still has to leave the strong acid of the stomach cavity. A layer of mucus covers the stomach wall, thick as gel. It was long assumed the corkscrew body screwed its way in, but the real mechanism is neater: urease raises the pH nearby, the mucus turns from gel to fluid, and the bacterium swims through on its flagella (Celli 2009).
Under the mucus, it settles on the surface of the stomach's own cells. The pH there is near neutral — the only mild spot in the stomach. So it doesn't live *in* the acid; it lives underneath it. For how stomach acid is made and how the mucus barrier works, see the digestive system island.
One number · the stomach is nearly sterile
Since strong acid kills bacteria, the stomach was never home to many. A gram of gastric contents holds roughly 10¹ to 10³ bacteria; a gram of colonic contents holds 10¹¹ to 10¹² (Liu 2024). About a billion-fold difference.Hold onto that contrast — the last scene uses it to dismantle a widespread claim.
liu-2024-gastric-microbiota
Chapter 2
The Correa cascade · a decades-long road
The Correa cascade · a decades-long road
H. pylori doesn't turn into gastric cancer overnight. It walks a decades-long road, one step at a time — and the first half of that road can be walked back.
The pathologist Correa mapped the road, and it now carries his name (Correa 1992; Correa & Piazuelo 2012):
Chronic gastritis — the bacterium moves in and the stomach wall inflames long-term. The vast majority of people stop here for life and feel nothing.Atrophic gastritis — inflammation burns too long, and some of the glands that make acid and digestive enzymes are worn away. The wall thins.Intestinal metaplasia — stomach lining cells are replaced by cells that look like intestine. The stomach grows intestine in the wrong place.Dysplasia — cells start growing out of order.Gastric cancer.
What matters isn't the names of the stops but the time and the direction. The road takes decades; the further along, the harder to turn back.
The first two stops are clearly reversible: after eradication, both chronic gastritis and atrophy improve measurably (Liang 2024 meta-analysis, atrophy OR 2.96). Intestinal metaplasia has traditionally been called the point of no return, but that's now contested — the same meta-analysis found that early metaplasia improves after eradication too (OR 2.41), while other evidence suggests that for people already at metaplasia or dysplasia, eradication buys a smaller reduction in cancer risk.
So the honest statement isn't 'past metaplasia, it's hopeless.' It's: the earlier you eradicate, the more certain the benefit. Which is exactly why guidelines keep stressing that it should be dealt with before precancerous changes appear (Kyoto 2015).
The pathologist Correa mapped the road, and it now carries his name (Correa 1992; Correa & Piazuelo 2012):
Chronic gastritis — the bacterium moves in and the stomach wall inflames long-term. The vast majority of people stop here for life and feel nothing.Atrophic gastritis — inflammation burns too long, and some of the glands that make acid and digestive enzymes are worn away. The wall thins.Intestinal metaplasia — stomach lining cells are replaced by cells that look like intestine. The stomach grows intestine in the wrong place.Dysplasia — cells start growing out of order.Gastric cancer.
What matters isn't the names of the stops but the time and the direction. The road takes decades; the further along, the harder to turn back.
The first two stops are clearly reversible: after eradication, both chronic gastritis and atrophy improve measurably (Liang 2024 meta-analysis, atrophy OR 2.96). Intestinal metaplasia has traditionally been called the point of no return, but that's now contested — the same meta-analysis found that early metaplasia improves after eradication too (OR 2.41), while other evidence suggests that for people already at metaplasia or dysplasia, eradication buys a smaller reduction in cancer risk.
So the honest statement isn't 'past metaplasia, it's hopeless.' It's: the earlier you eradicate, the more certain the benefit. Which is exactly why guidelines keep stressing that it should be dealt with before precancerous changes appear (Kyoto 2015).
Don't read the road as a verdict
Being infected with H. pylori does not mean you'll get gastric cancer. A long-term Japanese follow-up puts numbers on it: of 1,246 infected people followed a mean of 7.8 years, 2.9% developed gastric cancer; among 280 uninfected people, not one did (Uemura 2001).Read both directions. The vast majority of infected people never reach cancer, so a positive test is not a death sentence. But the zero in the uninfected group also shows the risk genuinely isn't zero.
What makes this worth taking seriously is the next trait: of all known causes of cancer, this is very nearly the only one a single course of treatment can remove.
uemura-2001-hp-gastric-cancer
Chapter 3
CagA · a syringe into your cells
CagA · a syringe into your cells
Two people carry H. pylori; one is fine for life, the other reaches gastric cancer. Whether the strain carries a protein called CagA is among the most important forks in that road.
Once a virulent strain docks onto a stomach cell, it extends an extremely fine tube — formally, a type IV secretion system. What it does is act like a syringe: it punctures the cell and injects CagA straight in (Hatakeyama 2004).
Inside, CagA starts making trouble. The host's own enzymes tag it with phosphate, and the tagged CagA becomes an adaptor that grabs the host's signaling proteins. The captured circuits scramble, with two consequences: the cell loses track of up and down (polarity breaks), and it gets pushed to divide without stopping.
More to the point, there is more than one CagA. Its tail carries a repeated sequence, and by that sequence it splits into an East Asian type and a Western type. The East Asian segment grips host signaling proteins more tightly (Ji 2024). A tighter grip pushes the downstream signal harder, and the risk of malignant transformation rises with it.
Strains circulating in East Asia more often carry the more virulent type — that's part of the molecular explanation for East Asia's gastric cancer rate. Only part: diet, salt, genetics, and screening intensity all sit in there too. And this layer's evidence comes mostly from cell and structural experiments, not population trials.
Once a virulent strain docks onto a stomach cell, it extends an extremely fine tube — formally, a type IV secretion system. What it does is act like a syringe: it punctures the cell and injects CagA straight in (Hatakeyama 2004).
Inside, CagA starts making trouble. The host's own enzymes tag it with phosphate, and the tagged CagA becomes an adaptor that grabs the host's signaling proteins. The captured circuits scramble, with two consequences: the cell loses track of up and down (polarity breaks), and it gets pushed to divide without stopping.
More to the point, there is more than one CagA. Its tail carries a repeated sequence, and by that sequence it splits into an East Asian type and a Western type. The East Asian segment grips host signaling proteins more tightly (Ji 2024). A tighter grip pushes the downstream signal harder, and the risk of malignant transformation rises with it.
Strains circulating in East Asia more often carry the more virulent type — that's part of the molecular explanation for East Asia's gastric cancer rate. Only part: diet, salt, genetics, and screening intensity all sit in there too. And this layer's evidence comes mostly from cell and structural experiments, not population trials.
This layer doesn't change what you do
Now that you know about CagA, should you ask for strain typing? No need, for now.The guideline line is blunt: if infection is found, eradication is recommended — regardless of strain — because for an individual, the action is the same either way (Maastricht VI 2022). Typing is currently a research and epidemiology tool, not a fork in clinical decisions.
So this scene isn't here to add a test. It's here so you understand why East Asia cares so much about this, and why the same bacterium adds up differently in different places.
malfertheiner-2022-maastricht-vi
Chapter 4
Why China carries the load
Why China carries the load
About one in two people in China carries H. pylori. And its main site of transmission is the dining table.
It spreads by mouth-to-mouth and fecal-oral routes, and transmission within the household is among the chief sources (Chinese family-based consensus 2022). Chinese eating habits happen to pave that road flat:
Communal dishes — everyone's chopsticks take turns in the same plateNo serving chopsticks — your chopsticks, wet with saliva, go back into shared foodElders chewing food before feeding a child — the most direct mouth-to-mouth channel
The result is infection by the household. A national survey of 10,735 families across 29 provinces found that on average 71% of households had at least one infected member (Chinese family-based consensus 2022).
So what needs to change is how people eat, not who they are. Serving chopsticks and spoons, individual portions, and no pre-chewed feeding — those few things are enough. There's no need to quarantine an infected person's bowl and chopsticks, and certainly no reason to move anyone away from the table.
A counter-example helps calibrate the fear. Hepatitis B was misunderstood in China for decades, and countless people were shut out because of it — yet hepatitis B is precisely not spread by eating together or sharing utensils (CDC); it travels by blood, sexual contact, and mother-to-child. The two are entirely different: H. pylori really does move along the dining table, so serving chopsticks help; hepatitis B doesn't, so segregating someone's dishes was only ever discrimination. For that thread, see the hepatitis B story.
Put the conclusion the other way round: use serving chopsticks against H. pylori, and common sense about hepatitis B. Neither one requires isolating a person.
It spreads by mouth-to-mouth and fecal-oral routes, and transmission within the household is among the chief sources (Chinese family-based consensus 2022). Chinese eating habits happen to pave that road flat:
Communal dishes — everyone's chopsticks take turns in the same plateNo serving chopsticks — your chopsticks, wet with saliva, go back into shared foodElders chewing food before feeding a child — the most direct mouth-to-mouth channel
The result is infection by the household. A national survey of 10,735 families across 29 provinces found that on average 71% of households had at least one infected member (Chinese family-based consensus 2022).
So what needs to change is how people eat, not who they are. Serving chopsticks and spoons, individual portions, and no pre-chewed feeding — those few things are enough. There's no need to quarantine an infected person's bowl and chopsticks, and certainly no reason to move anyone away from the table.
A counter-example helps calibrate the fear. Hepatitis B was misunderstood in China for decades, and countless people were shut out because of it — yet hepatitis B is precisely not spread by eating together or sharing utensils (CDC); it travels by blood, sexual contact, and mother-to-child. The two are entirely different: H. pylori really does move along the dining table, so serving chopsticks help; hepatitis B doesn't, so segregating someone's dishes was only ever discrimination. For that thread, see the hepatitis B story.
Put the conclusion the other way round: use serving chopsticks against H. pylori, and common sense about hepatitis B. Neither one requires isolating a person.
The numbers · prevalence and cancer burden
Prevalence: a meta-analysis pooling 412 studies and 1.37 million people put mainland China's combined prevalence at 44.2% (95% CI 43.0-45.5), implying about 589 million people infected (Ren 2022). The past decade shows a slow decline, but the order of magnitude is unchanged.Cancer burden: China had roughly 479,000 new gastric cancers and 373,000 deaths in 2020 — over 40% of the global gastric cancer burden (Yin 2025). China, Japan, and Korea together account for about three quarters of global cases.
High prevalence isn't the only reason gastric cancer runs high, but it is the largest — and the only removable — one. Another accomplice is high-salt preserved food; that correlational debate is unpacked honestly in the kimchi story, so it isn't repeated here.
ren-2022-china-hp-prevalenceyin-2025-gastric-cancer-burden-china
Chapter 5
How it's found · how it's treated
How it's found · how it's treated
The usual way to find it is to breathe out. The test works precisely because the survival kit from scene one gives the bacterium away.
The breath test's logic is clean:
1. You drink a small cup of water in which the urea has been swapped for a labeled version (carbon-13 or carbon-14)
2. If H. pylori is in your stomach, its urease splits that urea as usual, and the carbon dioxide it releases carries the label
3. The labeled carbon dioxide enters the blood, reaches the lungs, and you breathe it out
4. The machine reads your breath: label present means bacteria present; no label means none
In other words, it doesn't measure the bacterium — it measures urease activity. The very kit that keeps it alive in acid is the signature it can't hide. Accuracy is good enough: meta-analysis puts sensitivity around 96% and specificity around 93% (Ferwana 2015).
For treatment, China's first-line regimen is bismuth quadruple therapy for 14 days (2022 Chinese guideline): a proton-pump inhibitor (PPI) plus bismuth, plus two antibiotics, taken for 14 days.
Why not the simpler triple therapy? Because antibiotic resistance broke it. Resistance rates in China: clarithromycin 30.7%, metronidazole 70.1%, levofloxacin 33.0% (Zeng 2024). At those levels, clarithromycin-based standard triple therapy is no longer reliable. Meanwhile amoxicillin (2.4%) and tetracycline (2.5%) have stayed low and stable — bismuth quadruple therapy stands on those two intact pillars plus bismuth's own antibacterial action. This isn't 'more drugs is better'; it's a choice resistance forced.
The two-edged nature of long-term PPI use is covered in more detail on the GERD island, so it isn't expanded here.
The breath test's logic is clean:
1. You drink a small cup of water in which the urea has been swapped for a labeled version (carbon-13 or carbon-14)
2. If H. pylori is in your stomach, its urease splits that urea as usual, and the carbon dioxide it releases carries the label
3. The labeled carbon dioxide enters the blood, reaches the lungs, and you breathe it out
4. The machine reads your breath: label present means bacteria present; no label means none
In other words, it doesn't measure the bacterium — it measures urease activity. The very kit that keeps it alive in acid is the signature it can't hide. Accuracy is good enough: meta-analysis puts sensitivity around 96% and specificity around 93% (Ferwana 2015).
For treatment, China's first-line regimen is bismuth quadruple therapy for 14 days (2022 Chinese guideline): a proton-pump inhibitor (PPI) plus bismuth, plus two antibiotics, taken for 14 days.
Why not the simpler triple therapy? Because antibiotic resistance broke it. Resistance rates in China: clarithromycin 30.7%, metronidazole 70.1%, levofloxacin 33.0% (Zeng 2024). At those levels, clarithromycin-based standard triple therapy is no longer reliable. Meanwhile amoxicillin (2.4%) and tetracycline (2.5%) have stayed low and stable — bismuth quadruple therapy stands on those two intact pillars plus bismuth's own antibacterial action. This isn't 'more drugs is better'; it's a choice resistance forced.
The two-edged nature of long-term PPI use is covered in more detail on the GERD island, so it isn't expanded here.
Before testing and after treatment · two traps
Before testing: the breath test reads urease activity, so any drug that suppresses bacterial activity pushes the result toward a false negative — proton-pump inhibitors (PPIs), antibiotics, and bismuth all count. How long to stop them beforehand is your doctor's call; don't take a PPI and go blow into a tube.Retesting: finishing the course doesn't mean eradication succeeded — you must retest. And not immediately: right after stopping the drugs the bacteria may merely be suppressed rather than cleared, which reads as a false negative too. The consensus definition: still undetectable 4 to 6 weeks after treatment ends (Maastricht VI 2022).
Both traps are the same trap. The breath test measures activity, not presence. Suppress the activity and the test goes blind.
malfertheiner-2022-maastricht-vi
Positive test = endoscopy right away?
It doesn't. Whether to scope, and how often, depends on your age, family history, symptoms, and the extent and stage of changes already in your stomach — a doctor's judgment, not something a positive result triggers automatically.The European guideline tiers it: people with extensive atrophy or intestinal metaplasia need scheduled high-quality endoscopic surveillance, while people with mild changes confined to the antrum and no other risk factors are not advised to have routine surveillance (MAPS III 2025).
One situation doesn't wait for anyone's schedule, though. Alarm symptoms — trouble swallowing, vomiting blood or black stools, unexplained weight loss, persistent vomiting, anemia — mean see a doctor now, not search online.
Chapter 6
After eradication · the honest ledger
After eradication · the honest ledger
Eradicating H. pylori lowers gastric cancer risk but doesn't zero it. Both halves have to be said together.
The lowering first. In healthy infected people without precancerous lesions, eradication cuts gastric cancer incidence to about half (RR 0.54), and gastric cancer mortality falls with it (RR 0.61). In more concrete terms: treat about 72 people to prevent one gastric cancer (Ford 2020 meta-analysis).
The higher the risk, the clearer the benefit. A Korean randomized controlled trial recruited people whose first-degree relatives had gastric cancer (parents, siblings) and followed them about 9.2 years: 1.2% of the eradication group developed gastric cancer versus 2.7% on placebo, and among those actually cleared of the infection, risk fell by more than 70% (Choi 2020).
Now the not-zero. If your stomach has already reached atrophy or intestinal metaplasia, those structural changes don't vanish because the bacterium was cleared — residual risk remains. So endoscopic surveillance still applies to this group, at intervals your doctor sets from the actual extent and stage in your stomach (MAPS III 2025). Removing the cause is not the same as erasing the road already walked.
The ledger reads like this. H. pylori is a Group 1 carcinogen in the International Agency for Research on Cancer's classification (1994). Group 1 means the evidence that it causes cancer is conclusive — not that it is as dangerous as tobacco; headlines routinely eat that distinction. What makes it genuinely special is this: of all known causes of cancer, it is very nearly the only one a single course of treatment can remove. You can't take back the cigarettes you smoked, but you can show this bacterium the door.
IARC now recommends treating screen-and-treat as a public health priority in regions and populations where gastric cancer runs high (IARC 2025). China is exactly such a place.
The lowering first. In healthy infected people without precancerous lesions, eradication cuts gastric cancer incidence to about half (RR 0.54), and gastric cancer mortality falls with it (RR 0.61). In more concrete terms: treat about 72 people to prevent one gastric cancer (Ford 2020 meta-analysis).
The higher the risk, the clearer the benefit. A Korean randomized controlled trial recruited people whose first-degree relatives had gastric cancer (parents, siblings) and followed them about 9.2 years: 1.2% of the eradication group developed gastric cancer versus 2.7% on placebo, and among those actually cleared of the infection, risk fell by more than 70% (Choi 2020).
Now the not-zero. If your stomach has already reached atrophy or intestinal metaplasia, those structural changes don't vanish because the bacterium was cleared — residual risk remains. So endoscopic surveillance still applies to this group, at intervals your doctor sets from the actual extent and stage in your stomach (MAPS III 2025). Removing the cause is not the same as erasing the road already walked.
The ledger reads like this. H. pylori is a Group 1 carcinogen in the International Agency for Research on Cancer's classification (1994). Group 1 means the evidence that it causes cancer is conclusive — not that it is as dangerous as tobacco; headlines routinely eat that distinction. What makes it genuinely special is this: of all known causes of cancer, it is very nearly the only one a single course of treatment can remove. You can't take back the cigarettes you smoked, but you can show this bacterium the door.
IARC now recommends treating screen-and-treat as a public health priority in regions and populations where gastric cancer runs high (IARC 2025). China is exactly such a place.
Myths dismantled · part one
Bad breath means H. pylori? No. Eighty to ninety percent of halitosis originates in the mouth itself — tongue coating, periodontal disease, oral hygiene (Memon 2023). Seeing a dentist first makes far more sense than a breath test first.Garlic or honey can kill it? No. Garlic inhibits it beautifully in a test tube, but tested in actual people it had no effect at all (Graham 1999). This is the classic lesson that in vitro isn't in vivo: your stomach never reaches that test-tube concentration, and the bacteria are hiding under the mucus anyway.
Probiotics can kill it? Also no. Their real role is adjunctive: added to a proper regimen, the gain in eradication rate is quite limited (RR 1.12), and their main benefit is fewer side effects during treatment (Lv 2015). Useful, but no substitute for antibiotics.
memon-2023-halitosis-aetiologygraham-1999-garlic-jalapeno-hplv-2015-probiotics-adjuvant
Myths dismantled · part two
Eradication wrecks the stomach's good bacteria? This treats the stomach as if it were the colon. That number from scene one earns its keep here: the stomach holds 10¹ to 10³ bacteria per gram, the colon 10¹¹ to 10¹² (Liu 2024) — about a billion-fold difference. Acid is too strong; the stomach was never a thriving microbial community, so there is no good flora there waiting to be wrecked. The disturbance antibiotics cause to the gut flora during the course is real, but it is short-term and recovers afterward. And the organism being shown the door is a Group 1 carcinogen.No symptoms, no problem? This is the one to watch — because the first half of the Correa cascade has no symptoms by design. Chronic gastritis, atrophy, metaplasia: most of the time you feel nothing, and by the time you do, the road is often long behind you. Hence the guideline line: a positive test warrants eradication, symptoms or not (Maastricht VI 2022).
But don't read that as panic. Look back at the number in the last scene: the vast majority of infected people never get gastric cancer. This is worth doing not because it's terrifying, but because it's cheap, well-defined, and done once you're done.
liu-2024-gastric-microbiotamalfertheiner-2022-maastricht-vi
The honest close
This page is education about mechanism, not medical advice. Whether to test, whether to treat, which regimen, whether to scope — let a doctor decide from your actual situation, especially the regimen, since the resistance picture varies by region and shifts over time.If alarm symptoms appear — trouble swallowing, vomiting blood or black stools, unexplained weight loss, persistent vomiting, anemia — don't search online; see a doctor.
To keep reading: how stomach acid and the mucus barrier work is on the digestive system island; the two-edged nature of long-term PPI use is on the GERD island; the correlational debate about high-salt preserved food and gastric cancer is in the kimchi story; and how food overall nudges the odds of cancer is in the cancer and nutrition story.