Most studied
A knee worn, but not bone-on-bone.
The most trials behind it, led by a 146-patient phase 3 trial that found better pain and function at 6 and 12 months than placebo.
Do stem cell injections work? It depends on the joint, and yours may be a good fit.
For many people with a worn knee, injections can reduce pain and help delay a replacement. How much depends on the joint and how far it has worn, and this page sorts that out one joint at a time, straight from the published trials.
Ten questions, no email, no call. The result names a sensible next step for your joint.
Does stem cell therapy work for knees, shoulders, and spines?
For many worn knees, yes: injections can reduce pain and help delay a replacement. The shoulder and the spine have smaller, earlier trials, some of them encouraging. How worn the joint is matters as much as which joint: a knee at grade II to III (joint space still there) tends to respond better than one at grade IV (bone-on-bone).
Looking for information on IV treatment, or on neurologic or autoimmune conditions? See more here: IV treatment and neurologic and autoimmune conditions.
Most studied
Early evidence
The spine and the disc.
Smaller trials, some encouraging: in one, donor cells placed in the disc improved pain and function at one year.
Early evidence
A partial rotator cuff tear.
Early and small: one study saw pain fall and partial tears shrink on MRI over two years. Controlled trials are still catching up.
A surgeon's case
A tendon torn through, a joint that needs hardware.
Here a surgeon is the right call, and we will say so and name one we trust. A joint that turns hot, red and swollen, often with a fever, or back pain with sudden leg weakness or loss of bladder or bowel control: go to an emergency room today.
The studies behind these cards 12 studies on our reading list, by what each found
Knee
- Gupta 2023: helped against its comparison
- Lee 2025: helped against its comparison
- Vega 2015: helped against its comparison
- Matas 2019: a small gain, or no control group
- Cochrane 2025: a small gain, or no control group
- Mautner 2023: no better, or missed its main goal
- Sadeghirad 2024: no better, or missed its main goal
Spine and disc
- Noriega 2017: helped against its comparison
- Amirdelfan 2021: helped against its comparison
- Beall 2025: no better, or missed its main goal
Shoulder
- Jo 2020: a small gain, or no control group
- Chun 2022: no better, or missed its main goal
- Helped against its comparison
- A small gain, or no control group
- No better, or missed its main goal
Curious which card is yours? Ten questions, about 2 minutes, point you to a next step. The medical reasons we decline anyone are on who this is not for.
Find your next step →Stem cell therapy success rate: why there is no number here.
One percentage cannot tell you how your joint will do: response varies by joint, by how carefully patients were screened, and by what counted as a response. We won't quote a rate we can't stand behind. You get something more useful: a straight read of your own imaging.
Why one number can't be right for you
-
Each joint has its own evidence.
One figure blends the knee, the shoulder, and the spine into a single number.
-
Careful screening changes the result.
A clinic that treats everyone and one that reads imaging first will report different numbers.
-
Success can be defined many ways.
Unless a clinic says how it measured a response, its number cannot be checked.
The honest part: many patients improve, and no one can yet promise which ones. That is why your imaging is read before anything is booked.
Want a read on your own joint instead of a percentage? Start with ten questions.
Find your next step →What “worked” should mean to you, not to a scoring sheet.
On the first call your coordinator asks what you want back. Every later call is measured against it.
What you want back
Walk the dog two miles without stopping.
- 1 month
- 3 months
- 6 months
“Be able to do the activities that I want to do and live the life that I want to live.”
Scott M., 68 · Spine pain. A real patient, not a script. Individual results vary.
-
Less pain, and the relief holds.
At night, on the stairs, first thing in the morning.
-
What you can do again.
The walk, the round of golf, the floor with a grandchild.
-
A surgery or replacement pushed back.
For early-to-moderate wear, some patients delay one. Surgery stays open either way.
Start with what you want back. Ten questions, no email.
Find your next step →How long stem cells take to work, and how you will know.
The first 4 weeks are usually quiet while the cells work. A response usually shows between weeks 4 and 12 and can keep building through the first year; the spine is slower, 3 to 6 months. The week-by-week timeline has the detail.
-
1 week
Soreness, then a call.
24 to 72 hours of soreness after a knee, hip or shoulder injection; about a week after a spine injection.
-
1 month
Early days, and that is normal.
The cells work slowly, through signaling to nearby tissue.
-
3 months
Where responders start to notice.
This read is the one month 6 is compared against.
-
6 months
The read that counts.
Measured against where you started. If nothing has changed, you hear it plainly, with the options: a second treatment, a surgeon's opinion, or both, never a second treatment by default. After that, a call once a year.
Every call is read against where you started.
Find your next step →Four steps, and the physician team decides at the third one.
Nobody is quoted, booked, or treated before the physician team has read your imaging.
-
Call
A patient coordinator, about 15 minutes.
Your history, your goals, and what to send.
-
Collect records
Gather and share what you already have.
Your X-ray or MRI report, and what has been tried.
-
Review
The physician team evaluates it.
They review your imaging and history and decide whether cells are a good fit for your joint.
-
Plan
A written answer, no commitment required.
A plan and an itemized quote, if the answer is yes.
Candidacy is confirmed from your imaging and the call; on arrival the physicians examine you to plan how the treatment is done. Either way you get a clear answer in writing, and if cells are not the right tool, the name of a surgeon we trust.
Steps 1 and 2 are the only ones you do.
Book a call →Stem cell injections in Tijuana, inside a hospital.
Celva operates inside Hospital Angeles Tijuana, a 116-bed private hospital in the Angeles Health network, 25 to 30 minutes from San Diego. The physician team, the anesthesiologist and hospital imaging are on site for your session.
Medical Director
Dr. Cesar Amescua, MD
Anesthesiology & Sedation
Dr. Rocio Ambrosio Nuño, MD
Interventional Radiology
Dr. Alejandro Castillo, MD
Anesthesiology & Pain Medicine
Dr. Melina Mariscal, MD
The physicians on every chart. Full bios →
Most patients are home the same day.
Find your next step →Good questions to ask first.
How does a stem cell injection compare with a cortisone shot?
They do different jobs. Cortisone calms inflammation for a while; cells are designed to calm inflammation and support the joint's own repair. Trials against placebo and hyaluronic acid have been encouraging (Gupta 2023, Vega 2015, Lee 2025), and in the largest head-to-head with a steroid (Mautner, 2023, 480 knee patients) the two came out about even at one year.
How do you know if I am likely to respond?
Your imaging is the best guide, and it is what the physician team reads for. Grade 2 to 3 knees tend to respond better than grade 4, and in the spine the patients who do well have imaging findings that match where it hurts. No one can promise a result, but you will know where you stand before anything is booked.
How should I read stem cell therapy reviews?
As real experiences, and a good place to start. A patient's story tells you what treatment was like for one person; a controlled trial tells you how often it helps. Read both: the patient stories above, and the published trials on our reading list.
How safe is it for the joint?
Serious side effects were rare across the 25 trials in the Cochrane knee review, and in the 146-patient phase 3 knee trial, cartilage did not worsen on MRI. The most common effect is soreness and some swelling at the injection site, usually for 24 to 72 hours.
Are your cells like the ones the trials tested?
The same kinds of cells, yes. We use donor bone-marrow and umbilical-cord MSCs (mesenchymal stem cells), grown in the on-site cGMP-compliant lab and released after 6 lab tests. The Gupta knee trial used donor bone-marrow MSCs, and the closest published match to our umbilical-cord cells (Matas, 2019) found two doses six months apart beat hyaluronic acid on pain at 12 months.
Your joint. Your imaging. A straight answer.
Start with ten questions, no email asked. When you want a person, the first call is with a Celva patient coordinator, about 15 minutes, and the physician team answers in writing which answer your joint is in.
These therapies are not FDA-approved. Treatments are performed in Mexico by licensed physicians in a hospital setting. Individual results may vary. Not medical advice. Spine candidacy undergoes additional screening including neurologic assessment. Seek emergency care for sudden weakness, numbness, or loss of bladder or bowel control.