Cyclosporine

What it is

Cyclosporine is a ring-shaped peptide of eleven amino acids, originally isolated as a product of the soil fungus Beauveria nivea. It is not a vitamin, a mineral or a herbal extract; it is a fungal metabolite developed into a prescription immunosuppressant in the early 1980s. It exists in two oral forms that behave differently in the body: the older non-modified form (Sandimmune) and the microemulsion, or modified, form (Neoral and generics), which is better absorbed. It is also made as an eye drop for dry eye disease.

What the research says

Cyclosporine suppresses the immune system, and the evidence base is different for each thing it is used for. In kidney transplantation it works, but a Cochrane review of 30 trials found tacrolimus better at preventing graft loss and acute rejection. In severe psoriasis a 2025 Cochrane network meta-analysis found ciclosporin beats apremilast for clearing skin but ranks well below the modern biologics, and Cochrane's confidence in the non-targeted systemic drugs is limited by trial conduct. In rheumatoid arthritis a 48-week placebo-controlled trial showed disease-modifying effect at the cost of rising creatinine. For dry eye the Cochrane review found trials too poorly reported to pool, with Schirmer test estimates pointing in opposite directions. The harms are the central fact about this drug: dose- and duration-dependent kidney damage, hypertension in roughly half of kidney transplant recipients, and a raised risk of infection and cancer, including skin cancer.

Evidence grade: Well established.

How it works

Drug class: Calcineurin inhibitor immunosuppressant; a cyclic polypeptide of 11 amino acids produced by the fungus Beauveria nivea

Cyclosporine damps down the immune system by shutting off T-lymphocytes, the white cells that drive transplant rejection and several inflammatory skin and joint diseases. The label says the exact mechanism is not known, but describes specific and reversible inhibition of immune cells early in their activation cycle, with T-helper cells the main target, and blocking of the signalling molecules those cells release, including interleukin-2. Unlike many immunosuppressants it does not suppress the bone marrow. It is absorbed incompletely and variably from the gut, which is why blood level monitoring matters. (Source 1)

Boxed warning

Only physicians experienced in management of systemic immunosuppressive therapy for the indicated disease should prescribe Neoral. At doses used in solid organ transplantation, only physicians experienced in immunosuppressive therapy and management of organ transplant recipients should prescribe Neoral. Patients receiving the drug should be managed in facilities equipped and staffed with adequate laboratory and supportive medical resources. The physician responsible for maintenance therapy should have complete information requisite for the follow-up of the patient.

(Source 2)

What it is used for

• This is the use cyclosporine was built on and it works, but a Cochrane review of 30 randomised trials in 4,102 kidney transplant recipients found tacrolimus reduced graft loss (RR 0.56, 95% CI 0.36 to 0.86 at six months) and acute rejection (RR 0.69, 95% CI 0.60 to 0.79 at one year) compared with cyclosporine, at the cost of more insulin-requiring diabetes. Evidence: established. (Source 3)
• A 48-week placebo-controlled trial in 122 patients found improvement in joint counts, pain and function and radiographic retardation of joint destruction, with creatinine rising about 26% by week 48. The label itself says long-term data are limited. Evidence: limited. (Source 4)
• It clears skin: a label dose-titration trial reported 75% or greater PASI improvement in 51% of patients at 8 weeks and 79% at 16 weeks, and the 2025 Cochrane network meta-analysis places ciclosporin above apremilast. But it sits well below the modern biologics, and Cochrane's confidence in the non-targeted systemic drugs is limited by trial conduct. Evidence: established. (Source 5)
• Two topical products are FDA-approved for dry eye, but the 2019 Cochrane review of 30 trials in 4,009 people found almost all trials too incompletely reported to pool, with Schirmer test results at six months ranging from clearly in favour to clearly against. The certainty was low throughout. Evidence: disputed. (Source 6)
• Off-label in the US. The 270-patient CONSTRUCT trial found ciclosporin produced the same quality-adjusted survival, the same colectomy rate (48% vs 41%, p=0.223) and the same time to colectomy as infliximab, so it is a reasonable alternative rather than a proven advance. Evidence: limited. (Source 7)
• Off-label in the US (licensed for this in some other countries). The 149-trial network meta-analysis behind the 2023 US atopic dermatitis guidelines rated cyclosporine's efficacy and safety as less certain than the JAK inhibitors and the anti-IL-13 biologics. Evidence: limited. (Source 8)
• Off-label. Cochrane found cyclosporin may increase complete remission compared with placebo, corticosteroid or no treatment (RR 3.50, 95% CI 1.09 to 11.20) on low-certainty evidence from only 74 children, with uncertain effects on kidney failure, hypertension and infection. Evidence: limited. (Source 9)

Interactions

• Grapefruit and grapefruit juice (pharmacokinetic study): Grapefruit juice raises how much cyclosporine reaches the bloodstream. In a crossover pharmacokinetic study in ten kidney transplant recipients taking their usual dose, grapefruit juice increased total drug exposure by about a quarter and delayed absorption, without changing the peak level. The label tells patients to avoid it. Limit: This was a crossover study in only ten kidney transplant recipients and it reports no confidence intervals. The paper's sentence says grapefruit juice 'increased' both total exposure (P=0.029) and the 12-hour trough (P=0.09), but P=0.09 is not a statistically significant result by the paper's own standard, so only the change in total exposure is supported. The paper also prints concentration units as 'ng x ml' where it means ng/ml. (Source 10)
• Grapefruit and grapefruit juice (label position) (label): The FDA label states the interaction flatly and says grapefruit should be avoided. This is the regulator's position, dated by the SPL version; the measured numbers come from the pharmacokinetic study, not from the label. (Source 11)
• St John's wort (Hypericum perforatum) (case reports): St John's wort speeds up cyclosporine's breakdown and can drop blood levels far enough to lose a transplanted organ. The evidence is case reports, not trials: a 2002 systematic review found eleven case reports and two case series, with causality well established in most, and the mechanism traced to induction of both P-glycoprotein and CYP3A4. (Source 12)
• St John's wort (first published cases) (case reports): The signal entered the literature as a 2000 Lancet letter reporting acute rejection in two heart transplant patients attributed to a metabolic interaction with St John's wort. Limit: This is a two-patient case report published as a letter. The authors attribute the rejections to a metabolic interaction, but two cases cannot establish cause, and nothing here gives a rate or a risk. (Source 13)
• St John's wort (label position) (label): The label names St John's wort among the dietary supplements that decrease cyclosporine concentrations and states the consequence: subtherapeutic levels, rejection of transplanted organs and graft loss. (Source 14)
• Potassium supplements, potassium-containing drugs and a potassium-rich diet (label): Cyclosporine tends to raise blood potassium, so adding potassium from supplements, potassium-sparing drugs or a potassium-rich diet can push it too high. The label asks for potassium levels to be controlled in these situations; this rests on the drug's known effect on potassium handling rather than on an interaction trial. (Source 15)
• Food, particularly a fatty meal (modified / microemulsion form) (pharmacokinetic study): Taking the modified form with food lowers how much drug is absorbed. A high-fat meal eaten within half an hour before the dose cut total exposure by 13% and the peak level by 33%, and a low-fat meal did much the same. This is why the label asks for a consistent schedule in relation to meals. (Source 16)
• Fluconazole (pharmacokinetic study): Fluconazole roughly doubles cyclosporine exposure. In eight kidney transplant recipients on a stable cyclosporine dose, fourteen days of fluconazole 200 mg daily raised the area under the curve by 92% on average and the trough level by 157%, with apparent oral clearance falling 45%. (Source 17)
• Alcohol (as a component of the oral formulation) (label): We found no study of drinking alcohol while taking cyclosporine. What is on record is that the non-modified oral capsule itself contains dehydrated alcohol up to 12.7% by volume, and the intravenous form up to 32.9%, which matters for anyone avoiding alcohol for religious, medical or recovery reasons. (Source 18)

Stopping it

• Stopping cyclosporine for psoriasis usually means the psoriasis comes back. The label says true rebound is rare but relapse is the norm on cessation, as with other psoriasis therapies. (Source 5)
• In rheumatoid arthritis the label records that disease activity generally returns within about four weeks of stopping, and that long-term treatment data are limited. (Source 19)
• A meta-analysis of elective cyclosporine withdrawal after kidney transplantation found more acute rejection in those withdrawn, but no difference in graft loss or death; the authors cautioned that long-term consequences were still unknown at the time. (Source 20)
• The same meta-analysis gives the numbers behind that: a higher combined rate of acute rejection after withdrawal, with no difference in grafts lost or deaths per patient-year. (Source 20)
• In late survivors of heart transplantation with kidney impairment, a randomised trial of stopping cyclosporine and switching to rapamycin improved measured kidney function over six months, while patients kept on low-dose cyclosporine did not change. (Source 21)
• Some cyclosporine kidney injury reverses on stopping and some does not: of eleven psoriasis patients in whom the drug was discontinued after biopsy-proven nephropathy, creatinine returned to the normal range in seven. (Source 22)

What goes wrong

Cyclosporine-treated transplant recipients had significantly more constipation and cosmetic side effects than tacrolimus-treated recipients, and no difference in infection or cancer was demonstrated between the two. (Source 23)

• Systematic review, Moderate certainty.
• Size: 4,102 patients across 30 trials.
• Who: Kidney transplant recipients.
• How long: Up to three years.
• Result: More constipation and cosmetic side effects with cyclosporin; no difference shown in infection or malignancy between the two drugs.
• Funding: Cochrane review.

Cyclosporin-treated recipients experienced significantly more constipation and cosmetic side-effects.

Hypertrichosis and gum overgrowth were probably more common with cyclosporin than with tacrolimus in children treated for steroid-resistant nephrotic syndrome. (Source 9)

• Systematic review, Moderate certainty.
• Size: 29 studies, 1,248 children.
• Who: Children with steroid-resistant nephrotic syndrome.
• How long: Six to twelve months.
• Result: Hypertrichosis and gingival hyperplasia probably increased with cyclosporin; tacrolimus may reduce relapse during treatment (RR 0.22, 95% CI 0.06 to 0.90; 1 study, 34 children)
• Funding: Cochrane review.

Limit of this finding: The review grades almost everything here low or very low certainty and its own bottom line is that, apart from calcineurin inhibitors against placebo or cyclophosphamide, it remains unclear whether these treatments change outcomes. Kidney failure was not reported for this comparison at all.

Hypertrichosis and gingival hyperplasia probably increased with cyclosporin.

Repeat kidney biopsies in psoriasis patients given cyclosporine showed structural cyclosporine nephropathy in 21% after a mean of 23 months, rising to 30% on later biopsy. (Source 22)

• Official position, Low certainty.
• Size: 86 psoriasis patients biopsied; 13 of them re-biopsied.
• Who: Psoriasis patients treated with 1.2 to 7.6 mg/kg/day for a mean of 23 months.
• How long: Mean 23 months, with repeat biopsy after a mean of 2 further years.
• Result: Cyclosporine nephropathy in 18/86 (21%), rising to 26/86 (30%); renal tubular atrophy and interstitial fibrosis. Most affected patients (19/26) were on 5.0 mg/kg/day or more, above the highest recommended dose of 4 mg/kg/day. Creatinine returned to normal in 7 of 11 who stopped the drug. Reported inside the FDA label rather than as a separately retrievable paper.
• Funding: not applicable (regulatory document)

Kidney biopsies from 86 psoriasis patients treated for a mean duration of 23 months with 1.2 to 7.6 mg/kg/day of cyclosporine showed evidence of cyclosporine nephropathy in 18/86 (21%) of the patients. The pathology consisted of renal tubular atrophy and interstitial fibrosis. On repeat biopsy of 13 of these patients maintained on various dosages of cyclosporine for a mean of 2 additional years, the number with cyclosporine induced nephropathy rose to 26/86 (30%).

In US controlled psoriasis trials at recommended doses, 1.0% of patients stopped cyclosporine because of hypertension and 5.4% because of rising creatinine. (Source 24)

• Official position, Low certainty.
• Size: Psoriasis patients in US controlled clinical studies (182 on Neoral and 185 on Sandimmune in the incidence table)
• Who: Adults with severe recalcitrant plaque psoriasis.
• How long: Controlled trial durations.
• Result: Discontinuation for hypertension 1.0%; for increased creatinine 5.4%. Newly occurring hypertension 27.5% on Neoral vs 25.4% on Sandimmune; increased creatinine 19.8% vs 15.7%. One death reported in a 27-year-old man continued on cyclosporine despite renal deterioration.
• Funding: not applicable (regulatory document)

In psoriasis patients treated in US controlled clinical studies within the recommended dose range, cyclosporine therapy was discontinued in 1.0% of the patients because of hypertension and in 5.4% of the patients because of increased creatinine.

Hypertension occurs in roughly half of kidney transplant recipients taking cyclosporine and in most heart transplant recipients. (Source 25)

• Official position, Low certainty.
• Size: 892 patients in the kidney, heart and liver transplant trial programme.
• Who: Solid organ transplant recipients.
• How long: Not stated.
• Result: Hypertension, usually mild to moderate, in approximately 50% of patients after renal transplantation and in most cardiac transplant patients. The principal adverse reactions are renal dysfunction, tremor, hirsutism, hypertension and gum hyperplasia.
• Funding: not applicable (regulatory document)

Hypertension, which is usually mild to moderate, may occur in approximately 50% of patients following renal transplantation and in most cardiac transplant patients.

Solid organ transplant recipients had about nine times the skin cancer rate of matched controls, and the older agents cyclosporin and azathioprine carried more risk than the newer ones. (Source 26)

• Cohort study, Low certainty.
• Size: 2,852 Welsh transplant recipients matched to 13,527 population controls.
• Who: Solid organ transplant recipients in Wales, 2000-2018.
• How long: Up to 18 years.
• Result: Skin cancer incidence 1,203.2 per 100,000 person-years at risk in recipients vs 133.9 in matched controls. Highest adjusted incidence rate ratio in heart recipients (IRR 10.82, 95% CI 3.64-32.19), lowest in liver recipients (IRR 2.86, 95% CI 1.15-7.13). Tacrolimus and mycophenolate were associated with lower skin cancer risk than cyclosporin and azathioprine.
• Funding: not stated.

Contemporary immunomodulators such as tacrolimus and mycophenolate were associated with a reduction in skin cancer risk when compared to their predecessors, cyclosporin and azathioprine.

Glomerular capillary thrombosis resembling haemolytic-uraemic syndrome has been found in cyclosporine-treated patients and can progress to graft failure. (Source 25)

• Official position, Certainty not rated.
• Size: Not quantified in the label.
• Who: Transplant recipients.
• How long: Not stated.
• Result: Platelet-fibrin thrombi occluding glomerular capillaries and afferent arterioles, microangiopathic haemolytic anaemia, thrombocytopenia and falling renal function; similar findings seen with other post-transplant immunosuppressants.
• Funding: not applicable (regulatory document)

Glomerular capillary thrombosis has been found in patients treated with cyclosporine and may progress to graft failure.

The Neoral boxed warning is a position of the FDA label: it restricts prescribing to experienced physicians, warns of infection and cancer, and warns that Neoral and Sandimmune are not interchangeable. (Source 2)

• Official position, Certainty not rated.
• Size: Not applicable.
• Who: All patients prescribed Neoral.
• How long: Not applicable.
• Result: A boxed warning covering prescriber experience, increased susceptibility to infection and development of neoplasia including lymphoma, and the bioavailability difference from Sandimmune.
• Funding: not applicable (regulatory document)

Neoral, a systemic immunosuppressant, may increase the susceptibility to infection and the development of neoplasia.

The Sandimmune boxed warning adds a specific warning against mg-for-mg switching between Sandimmune and Neoral capsules, and records that Sandimmune absorption is erratic on chronic use. (Source 27)

• Official position, Certainty not rated.
• Size: Not applicable.
• Who: All patients prescribed Sandimmune.
• How long: Not applicable.
• Result: Inappropriate switching may raise exposure and adverse reaction risk, or lower exposure and efficacy. Blood concentration monitoring is advised because absorption of Sandimmune capsules on chronic administration was found to be erratic.
• Funding: not applicable (regulatory document)

Do not switch between Sandimmune capsules, 25 mg to Neoral capsules, MODIFIED, 25 mg (or between Sandimmune capsules, 100 mg to Neoral capsules, MODIFIED 100 mg) on a mg-to-mg basis to achieve the same total daily cyclosporine dosage. Inappropriate switching may lead to increased cyclosporine exposure which may increase the risk of cyclosporine-associated adverse reactions or decreased cyclosporine exposure which may decrease the efficacy of cyclosporine.

In the same Cochrane review, people using topical cyclosporine may be more likely to have treatment-related side effects, chiefly burning and stinging, than people using the vehicle drops. (Source 6)

• Systematic review, Low certainty.
• Size: All but two of the 30 included RCTs reported adverse events (4,009 participants in the review)
• Who: People with dry eye of varying severity and cause.
• How long: Six weeks to twelve months.
• Result: Treatment-related adverse events RR 1.33, 95% CI 1.00 to 1.78; low-certainty evidence. The lower bound touches 1.00, so the review's own wording is 'may be more likely' rather than a demonstrated increase.
• Funding: Cochrane review.

Participants treated with CsA may be more likely to have treatment-related adverse events than those who treated with vehicle (RR 1.33, 95% CI 1.00 to 1.78; low-certainty evidence).

In the worldwide psoriasis trial programme behind the Neoral label, 2.2% of patients developed a tumour, most of them skin cancers in people who had already had PUVA, and two had a lymphoproliferative malignancy. (Source 22)

• Official position, Low certainty.
• Size: 1,439 psoriasis patients treated with cyclosporine in clinical trials worldwide, plus 7 postmarketing reports.
• Who: Psoriasis patients treated with cyclosporine.
• How long: Not stated in the label.
• Result: Tumours in 32/1439 (2.2%); skin malignancies in 16 (1.1%), of whom all but 2 had previously had PUVA; among those 16, 11 had 18 squamous cell carcinomas and 7 had 10 basal cell carcinomas; two lymphoproliferative malignancies (one non-Hodgkin's lymphoma needing chemotherapy, one mycosis fungoides that regressed on stopping); four cases of benign lymphocytic infiltration; the remaining 13 cases (0.9%) in various organs.
• Funding: not applicable (FDA-approved label)

Limit of this finding: Most of these skin cancers occurred in people who had already had PUVA light therapy, which itself causes skin cancer, so the label's figures cannot be read as the effect of cyclosporine alone. The two lymphoproliferative malignancies are two events in 1,439 people, far too few to estimate a rate from.

There were two lymphoproliferative malignancies; one case of non-Hodgkin’s lymphoma which required chemotherapy, and one case of mycosis fungoides which regressed spontaneously upon discontinuation of cyclosporine.

What the evidence supports

For clearing moderate-to-severe plaque psoriasis, ciclosporin beat apremilast but sat well below the biologics in the ranking. (Source 28)

• Meta-analysis, Low certainty.
• Size: 67,889 randomised participants across 204 trials, 26 treatments.
• Who: Adults with moderate-to-severe plaque psoriasis, mean age 44.4 years, mean baseline PASI 20.5.
• How long: Induction phase, 8 to 24 weeks after randomisation.
• Result: Ciclosporin and methotrexate superior to apremilast for reaching PASI 90; the highest-ranked drugs were infliximab, xeligekimab, bimekizumab, ixekizumab and risankizumab. Cochrane states its confidence in the non-targeted systemic treatments is limited by concerns about study conduct.
• Funding: Review funded by the French Society of Dermatology and the French Ministry of Health; 157 of 204 included studies declared pharmaceutical company funding and 27 reported no funding source.

Limit of this finding: The review prints '57% (94/169)' for studies at high risk of bias on serious adverse events, but 94 out of 169 is 55.6%, not 57%. The fraction is the reliable figure. The percentage is a rounding or typing slip in the published review and nothing else in that sentence depends on it. This is also a living review first published in 2025, so its conclusions may already have been updated.

Ciclosporin and methotrexate were superior to apremilast for reaching PASI 90.

A 48-week placebo-controlled trial found cyclosporine 5 mg/kg/day improved joint counts, pain and function in active rheumatoid arthritis and retarded joint destruction on X-ray, while serum creatinine rose by about a quarter. (Source 29)

• Randomized trial, Low certainty.
• Size: 122 patients with active rheumatoid arthritis.
• Who: Patients with active RA.
• How long: 48 weeks (46 weeks of treatment)
• Result: Significant improvement in tender joints, swollen joints, pain score, morning stiffness and Lee's functional index; radiographic retardation of joint destruction. Serum creatinine rose 17.5 micromol/L (23%) at week 24 and 21.8 micromol/L (26%) at week 48; five patients needed antihypertensives and two withdrew for rising creatinine.
• Funding: not stated (Norwegian Arthritis Study Group)

Limit of this finding: The paper contradicts itself on length: its title and methods describe a 48-week study and the creatinine data are reported at week 48, but the objective sentence says 46 weeks. Treat it as a 48-week trial, which is what the results are keyed to. Note also that the 5 mg/kg/day the trial called 'low-dose' is above the 4 mg/kg/day maximum in the current Neoral label for rheumatoid arthritis.

The study shows that cyclosporine seems to have disease-modifying effects in RA.

In children with steroid-resistant nephrotic syndrome, cyclosporin may increase complete remission compared with placebo, corticosteroid or no treatment, on low-certainty evidence. (Source 9)

• Systematic review, Low certainty.
• Size: 74 children across 4 studies (review total 29 studies, 1,248 children)
• Who: Children with idiopathic steroid-resistant nephrotic syndrome.
• How long: Two to six months.
• Result: Complete remission RR 3.50, 95% CI 1.09 to 11.20; complete or partial remission RR 3.15, 95% CI 1.04 to 9.57; low-certainty evidence. Effects on kidney failure, worsening hypertension and infection are very low certainty.
• Funding: Cochrane review.

Limit of this finding: The review grades almost everything here low or very low certainty and its own bottom line is that, apart from calcineurin inhibitors against placebo or cyclophosphamide, it remains unclear whether these treatments change outcomes. Kidney failure was not reported for this comparison at all.

Compared with placebo, corticosteroid or no treatment, cyclosporin may increase the number who achieve complete remission (RR 3.50, 95% CI 1.09 to 11.20; 4 studies, 74 children) or complete or partial remission (RR 3.15, 95% CI 1.04 to 9.57; 4 studies, 74 children) by two to six months (low-certainty evidence).

What the evidence does not support

In kidney transplantation tacrolimus outperformed cyclosporine on graft loss and acute rejection, so cyclosporine is effective but no longer the best-performing calcineurin inhibitor. (Source 23)

• Systematic review, Moderate certainty.
• Size: 4,102 patients across 30 randomised trials (123 reports)
• Who: Kidney transplant recipients receiving primary immunosuppression.
• How long: Six months to three years.
• Result: Graft loss at six months RR 0.56, 95% CI 0.36 to 0.86 favouring tacrolimus, persisting to three years; acute rejection at one year RR 0.69, 95% CI 0.60 to 0.79; steroid-resistant rejection RR 0.49, 95% CI 0.37 to 0.64; insulin-requiring diabetes RR 1.86, 95% CI 1.11 to 3.09 against tacrolimus.
• Funding: Cochrane review; individual trial funding not stated in the abstract.

At six months graft loss was significantly reduced in tacrolimus-treated recipients (RR 0.56, 95% CI 0.36 to 0.86), and this effect was persistent up to three years.

Cochrane's confidence in the psoriasis results for non-targeted systemic drugs, the group ciclosporin belongs to, is limited by how the trials were conducted. (Source 28)

• Meta-analysis, Low certainty.
• Size: 204 trials, 67,889 participants.
• Who: Adults with moderate-to-severe plaque psoriasis.
• How long: 8 to 24 weeks.
• Result: For PASI 90, 31% of studies (51/165) were at high risk of bias; for serious adverse events 94 of 169 studies were at high risk of bias (the review prints this as 57%, although 94/169 is 55.6%)
• Funding: French Society of Dermatology and French Ministry of Health.

Limit of this finding: The review prints '57% (94/169)' for studies at high risk of bias on serious adverse events, but 94 out of 169 is 55.6%, not 57%. The fraction is the reliable figure. The percentage is a rounding or typing slip in the published review and nothing else in that sentence depends on it. This is also a living review first published in 2025, so its conclusions may already have been updated.

Our confidence in the results for non-targeted systemic treatments is limited due to concerns regarding study conduct.

Almost all topical cyclosporine dry eye trials reported their results so incompletely that between-group effects could not be calculated. (Source 6)

• Systematic review, Low certainty.
• Size: 30 RCTs, 4,009 participants.
• Who: People with dry eye.
• How long: Six weeks to twelve months.
• Result: Reporting deficiencies precluded calculation of between-group effect estimates or meta-analysis in almost all trials.
• Funding: Cochrane review.

Almost all trials had deficiencies in the reporting of results (e.g. reporting P values or direction only), precluding the calculation of between-group estimates of effect or meta-analysis.

In atopic dermatitis the comparative evidence for cyclosporine remains uncertain, while JAK inhibitors and the anti-IL-13 biologics have high-certainty evidence. (Source 8)

• Meta-analysis, Low certainty.
• Size: 28,686 patients with moderate-to-severe atopic dermatitis across 149 trials, 75 interventions.
• Who: Adults and children with moderate-to-severe atopic dermatitis.
• How long: Trial durations varied.
• Result: Efficacy and safety of cyclosporine (with azathioprine, oral corticosteroids, methotrexate, mycophenolate and phototherapy) rated less certain than the targeted agents.
• Funding: Guideline-commissioned review for the AAAAI/ACAAI Joint Task Force.

Efficacy and safety of azathioprine, oral corticosteroids, cyclosporine, methotrexate, mycophenolate, phototherapy, and many novel agents are less certain.

The dry eye review's authors concluded that the evidence for topical cyclosporine is inconsistent and at times no different from vehicle or artificial tears, and that every published trial was short term. (Source 30)

• Systematic review, Low certainty.
• Size: 30 RCTs, 4,009 participants.
• Who: People with dry eye.
• How long: Six weeks to twelve months; no trial examined longer-term disease modification.
• Result: No pooled estimate; the authors describe the body of evidence as inconsistent and sometimes not different from vehicle or artificial tears over the periods the trials reported.
• Funding: Cochrane review.

we found that evidence on the effect of CsA on ocular discomfort and ocular surface and tear film parameters such as corneal fluorescein staining, Schirmer's test, and TBUT is inconsistent and sometimes may not be different from vehicle or artificial tears for the time periods reported in the trials.

Where the evidence is mixed

In steroid-resistant acute severe ulcerative colitis, ciclosporin and infliximab produced the same quality-adjusted survival and the same colectomy rate over three years. (Source 7)

• Randomized trial, Moderate certainty.
• Size: 270 patients randomised at 52 UK hospitals; 121 per group analysed for the primary outcome.
• Who: Adults admitted unscheduled with severe ulcerative colitis who failed intravenous hydrocortisone within about 5 days.
• How long: Up to 3 years.
• Result: Mean area under the CUCQ curve 564.0 (SD 241.9) infliximab vs 587.0 (226.2) ciclosporin; mean adjusted difference 7.9, 95% CI -22.0 to 37.8, p=0.603. Colectomy in 55/135 (41%) vs 65/135 (48%), p=0.223.
• Funding: not stated in the abstract (NIHR-commissioned pragmatic trial)

Limit of this finding: This trial was open label and pragmatic: the patients and their local doctors knew which drug was given, and only the chief investigator and the analysts were masked. It should not be described as a blinded trial. The safety data are now inside the quoted passage: serious adverse reactions and serious adverse events were similar, and there were three deaths in the infliximab group and none in the ciclosporin group, which is three events in 270 patients and far too few to compare the two drugs on mortality.

There was no significant difference between groups in quality-adjusted survival (mean AUC 564·0 [SD 241·9] in the infliximab group vs 587·0 [226·2] in the ciclosporin group; mean adjusted difference 7·9 [95% CI -22·0 to 37·8]; p=0·603).

For dry eye disease the Cochrane review of 30 trials could not pool most results, and the Schirmer tear test estimates at six months pointed in opposite directions. (Source 6)

• Systematic review, Low certainty.
• Size: 4,009 participants across 30 RCTs.
• Who: People with dry eye of varying severity and cause.
• How long: Six weeks to twelve months.
• Result: Schirmer test mean differences at six months ranged from -4.05 (95% CI -6.67 to -1.73) to 3.26 (95% CI -1.52 to 5.00); ocular surface staining inconsistent between two trials (-0.35, 95% CI -0.69 to -0.01 in one; 0.58, 95% CI 0.06 to 1.10 in the other); all low-certainty.
• Funding: Cochrane review (NIH-supported authors)

Four trials reported MD in Schirmer test scores at 6 months and the estimates ranged from -4.05 (95% CI -6.67 to -1.73) to 3.26 (95% CI -1.52 to 5.00) (low-certainty evidence).

Head-to-head data on whether the modified microemulsion formulation is actually better than the older one are not clear-cut once blinding is taken into account. (Source 31)

• Meta-analysis, Low certainty.
• Size: Published Neoral versus Sandimmune comparisons (number of trials not given in the abstract)
• Who: Adult and paediatric kidney, liver and heart transplant recipients, de novo and stable.
• How long: Varied.
• Result: Graft loss similar in all analyses; rejection lower with Neoral in de novo renal, liver and cardiac transplants (P<0.05); but in blinded studies there were more adverse events with Neoral (P<0.05), and in open-label studies no difference.
• Funding: not stated.

However, when limiting the analysis to only randomized prospective trials, and specifically assessing blinded studies, the data become less clear.

Where the research disagrees

Whether the modified microemulsion formulation (Neoral and generics) is genuinely better than the older non-modified formulation (Sandimmune), and how dangerous switching between them is

• FDA label for Neoral (boxed warning), Regulatory position, SPL version 30, effective 10 August 2026, resting on pharmacokinetic comparisons: Neoral and Sandimmune are not bioequivalent and cannot be used interchangeably without physician supervision. (Source 2)
• FDA label for Sandimmune (boxed warning), Regulatory position, SPL version 37, effective 10 July 2026: Do not switch between Sandimmune capsules, 25 mg to Neoral capsules, MODIFIED, 25 mg (or between Sandimmune capsules, 100 mg to Neoral capsules, MODIFIED 100 mg) on a mg-to-mg basis to achieve the same total daily cyclosporine dosage. Inappropriate switching may lead to increased cyclosporine exposure which may increase the risk of cyclosporine-associated adverse reactions or decreased cyclosporine exposure which may decrease the efficacy of cyclosporine. (Source 27)
• Dunn and colleagues, Transplantation 1999 (meta-analysis of published comparisons), Meta-analysis of published Neoral versus Sandimmune comparisons; found lower rejection with Neoral overall but more adverse events with Neoral in blinded studies: However, when limiting the analysis to only randomized prospective trials, and specifically assessing blinded studies, the data become less clear. (Source 31)

Whether cyclosporine still belongs in first-line transplant immunosuppression

• FDA label for Neoral, Regulatory position; the indication remains in force: Neoral is indicated for the prophylaxis of organ rejection in kidney, liver, and heart allogeneic transplants. (Source 3)
• Webster and colleagues, Cochrane 2005, Systematic review and meta-analysis of 30 randomised trials, 4,102 kidney transplant recipients: At six months graft loss was significantly reduced in tacrolimus-treated recipients (RR 0.56, 95% CI 0.36 to 0.86), and this effect was persistent up to three years. (Source 23)

How much

• Reference intake: There is no reference intake for cyclosporine; it is a prescription immunosuppressant whose dose is set and adjusted by the prescriber against blood levels and kidney function. As a position, the label records starting doses from a 1994 survey of US transplant centres: about 9 mg/kg/day for kidney, 8 mg/kg/day for liver and 7 mg/kg/day for heart recipients, in two divided doses, then adjusted to a target blood concentration. For psoriasis and rheumatoid arthritis the labelled starting dose is 2.5 mg/kg/day in two divided doses. (Source 32)
• Upper limit: No upper intake level exists in the nutrition sense. The highest daily dose on the label for the non-transplant indications is 4 mg/kg/day: in psoriasis the dose may be raised in steps of about 0.5 mg/kg/day to a maximum of 4.0 mg/kg/day, and in rheumatoid arthritis to a maximum of 4 mg/kg/day. Reported as the label's position only. (Source 33)
• Studied: The 48-week Norwegian placebo-controlled rheumatoid arthritis trial used cyclosporine 5 mg/kg/day. (Source 29)
• Studied: The CONSTRUCT trial in acute severe ulcerative colitis used ciclosporin 2 mg/kg/day by continuous infusion for up to seven days, then twice-daily tablets delivering 5.5 mg/kg/day for twelve weeks. (Source 7)
• Studied: The psoriasis biopsy series reported in the label covered patients treated with 1.2 to 7.6 mg/kg/day for a mean of 23 months, above the labelled maximum in most of those who developed nephropathy. (Source 22)
• Studied: The grapefruit juice pharmacokinetic study gave ten kidney transplant recipients their usual steady-state dose with either 8 ounces of grapefruit juice or 8 ounces of water. (Source 10)

A common belief, and what the research shows

The belief: All cyclosporine capsules are the same drug at the same dose, so one brand or generic can be swapped mg-for-mg for another.

What the research shows: They are not interchangeable. The Neoral boxed warning states that "Neoral and Sandimmune are not bioequivalent and cannot be used interchangeably without physician supervision." and the Sandimmune boxed warning spells out the consequence: "Do not switch between Sandimmune capsules, 25 mg to Neoral capsules, MODIFIED, 25 mg (or between Sandimmune capsules, 100 mg to Neoral capsules, MODIFIED 100 mg) on a mg-to-mg basis to achieve the same total daily cyclosporine dosage. Inappropriate switching may lead to increased cyclosporine exposure which may increase the risk of cyclosporine-associated adverse reactions or decreased cyclosporine exposure which may decrease the efficacy of cyclosporine." The underlying pharmacokinetics show why: "In studies of renal transplant, rheumatoid arthritis and psoriasis patients, the mean cyclosporine AUC was approximately 20% to 50% greater and the peak blood cyclosporine concentration (Cmax) was approximately 40% to 106% greater following administration of Neoral compared to following administration of Sandimmune."

Questions and answers

What is it?

Cyclosporine is a ring-shaped peptide made of eleven amino acids, originally found as a product of the soil fungus Beauveria nivea. It is a prescription immunosuppressant, not a supplement or a nutrient. It comes as an older non-modified oral form, a better-absorbed modified microemulsion form, an injection, and eye drops for dry eye. (Source 34)

What does it do in the body?

It switches off the white blood cells that drive transplant rejection and several inflammatory diseases. The label says the exact mechanism is unknown but describes specific and reversible inhibition of immune cells early in their activation cycle, with T-helper cells the main target, and suppression of the signalling molecules they release, including interleukin-2. Unusually for an immunosuppressant, it does not suppress the bone marrow. (Source 1)

Is it good or bad for you?

Both, and which one dominates depends on dose and duration. It saves transplanted organs and clears severe psoriasis. It also causes dose- and duration-dependent kidney damage: repeat biopsies in psoriasis patients found cyclosporine nephropathy in 21% after a mean of 23 months, rising to 30% on later biopsy, with tubular atrophy and interstitial fibrosis. Hypertension affects around half of kidney transplant recipients, and transplant recipients overall had about nine times the skin cancer rate of matched controls. (Source 22)

How do you get more of it?

This is not something to get more of. Blood levels are deliberately steered by the prescriber, and both the dose and the target concentration differ by transplanted organ and by co-medication. What is worth knowing is what unintentionally raises levels: grapefruit juice increased total drug exposure in kidney transplant recipients, and fluconazole nearly doubled it. The label asks for a consistent schedule in relation to meals and tells patients to avoid grapefruit. (Source 10)

If it is harmful, what reduces it?

When levels are too high, the dose is lowered or the drug stopped under supervision, and some of the damage reverses: creatinine returned to the normal range in seven of eleven psoriasis patients in whom the drug was discontinued after biopsy-proven nephropathy. Enzyme-inducing drugs and supplements lower levels fast, which is a hazard rather than a treatment: St John's wort has driven levels low enough to cause graft loss. (Source 22)

Why might someone be low in it or missing it?

Nobody is naturally low in cyclosporine, but people do end up with levels lower than intended. The two documented routes are poor absorption, which the label calls incomplete and variable from the gut, and interactions that speed up its breakdown, above all St John's wort, which has produced subtherapeutic levels, rejection and graft loss. (Source 1)

Which whole foods contain it or feed it?

No whole food contains cyclosporine. It is a fungal metabolite isolated from Beauveria nivea and manufactured as a drug; there is no dietary source and no food that feeds it. Food does affect the drug, but in the opposite direction: a high-fat meal before a dose of the modified form cut exposure by 13% and the peak level by 33%. (Source 16)

We searched: We read the DESCRIPTION, CLINICAL PHARMACOLOGY and Absorption sections of both the Sandimmune and Neoral labels and searched PubMed for dietary sources of cyclosporine; the only food-related findings concern how meals and grapefruit change absorption of the drug, not any dietary source of it.

What happens if you do not have it?

For someone who does not need it, nothing: it is not a nutrient. For someone who does, stopping it has measurable consequences. After elective withdrawal in kidney transplant recipients there was a higher rate of acute rejection, though no difference in graft loss or death in that meta-analysis. In psoriasis, relapse on stopping is the norm, and in rheumatoid arthritis disease activity generally returns within about four weeks. (Source 20)

How can you test for it?

Yes, and the test matters more here than for most drugs. Transplant centres treat blood concentration monitoring as an essential part of care, usually on trough samples, alongside kidney function and sometimes biopsy. The important limitation is that assays are not interchangeable: older non-specific assays reported concentrations roughly twice those of the modern parent-compound-specific assays, so a number only means something alongside the method that produced it. (Source 35)

References

1. DailyMed (FDA Structured Product Labeling), SPL version 37, effective 2026-07-10. SANDIMMUNE — CLINICAL PHARMACOLOGY, Mechanism of Action and Pharmacokinetics. 2026. Read the source
2. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL (cyclosporine capsules/oral solution, MODIFIED) — BOXED WARNING. 2026. Read the source
3. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — INDICATIONS AND USAGE, Kidney, Liver, and Heart Transplantation. 2026. Read the source
4. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — INDICATIONS AND USAGE, Rheumatoid Arthritis. 2026. Read the source
5. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — INDICATIONS AND USAGE, Psoriasis. 2026. Read the source
6. Cochrane Database Syst Rev. Topical cyclosporine A therapy for dry eye syndrome.. 2019. PMID 31517988, DOI 10.1002/14651858.CD010051.pub2. Read the source
7. Lancet Gastroenterol Hepatol. Infliximab versus ciclosporin for steroid-resistant acute severe ulcerative colitis (CONSTRUCT): a mixed methods, open-label, pragmatic randomised trial.. 2016. PMID 27595142, DOI 10.1016/S2468-1253(16)30003-6. Read the source
8. J Allergy Clin Immunol. Systemic treatments for atopic dermatitis (eczema): Systematic review and network meta-analysis of randomized trials.. 2023. PMID 37678577, DOI 10.1016/j.jaci.2023.08.029. Read the source
9. Cochrane Database Syst Rev. Interventions for idiopathic steroid-resistant nephrotic syndrome in children.. 2025. PMID 40337980, DOI 10.1002/14651858.CD003594.pub7. Read the source
10. Transplantation. Effect of grapefruit juice on cyclosporine pharmacokinetics in renal transplant patients.. 1996. PMID 8693526, DOI 10.1097/00007890-199607150-00023. Read the source
11. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — PRECAUTIONS, Drug Interactions: Grapefruit Juice. 2026. Read the source
12. Arch Surg. St John's Wort supplements endanger the success of organ transplantation.. 2002. PMID 11888457, DOI 10.1001/archsurg.137.3.316. Read the source
13. Lancet. Acute heart transplant rejection due to Saint John's wort.. 2000. PMID 10683008, DOI 10.1016/S0140-6736(99)05467-7. Read the source
14. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — PRECAUTIONS, Drug Interactions: St. John’s Wort. 2026. Read the source
15. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — PRECAUTIONS, Drug Interactions: Potassium-Sparing Diuretics. 2026. Read the source
16. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — CLINICAL PHARMACOLOGY, Pharmacokinetics: Absorption. 2026. Read the source
17. DailyMed (FDA Structured Product Labeling), SPL version 57, effective 2026-04-02. DIFLUCAN — CLINICAL PHARMACOLOGY, Drug Interaction Studies: Cyclosporine. 2026. Read the source
18. DailyMed (FDA Structured Product Labeling), SPL version 37, effective 2026-07-10. SANDIMMUNE — DESCRIPTION, capsule composition. 2026. Read the source
19. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — DOSAGE AND ADMINISTRATION, Rheumatoid Arthritis. 2026. Read the source
20. JAMA. Elective cyclosporine withdrawal after renal transplantation. A meta-analysis.. 1993. PMID 8418349. Read the source
21. Am J Transplant. Cyclosporine withdrawal improves renal function in heart transplant patients on reduced-dose cyclosporine therapy.. 2006. PMID 16939514, DOI 10.1111/j.1600-6143.2006.01527.x. Read the source
22. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — WARNINGS, Psoriasis. 2026. Read the source
23. Cochrane Database Syst Rev. Tacrolimus versus cyclosporin as primary immunosuppression for kidney transplant recipients.. 2005. PMID 16235347, DOI 10.1002/14651858.CD003961.pub2. Read the source
24. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — ADVERSE REACTIONS, Psoriasis. 2026. Read the source
25. DailyMed (FDA Structured Product Labeling), SPL version 37, effective 2026-07-10. SANDIMMUNE — ADVERSE REACTIONS. 2026. Read the source
26. Eur J Dermatol. The association between immunosuppression and skin cancer in solid organ transplant recipients: a control-matched cohort study of 2,852 patients.. 2021. PMID 34427560, DOI 10.1684/ejd.2021.4108. Read the source
27. DailyMed (FDA Structured Product Labeling), SPL version 37, effective 2026-07-10. SANDIMMUNE (cyclosporine capsules / cyclosporine injection) — WARNING: RECOMMENDATIONS FOR USE, USE WITH CORTICOSTERIODS, RISKS WITH INAPPROPRIATE SWITCHING, and MONITORING CYCLOSPORINE BLOOD LEVELS. 2026. Read the source
28. Cochrane Database Syst Rev. Systemic pharmacological treatments for chronic plaque psoriasis: a network meta-analysis.. 2025. PMID 40767824, DOI 10.1002/14651858.CD011535.pub7. Read the source
29. Arthritis Rheum. Radiologic evidence of disease modification in rheumatoid arthritis patients treated with cyclosporine. Results of a 48-week multicenter study comparing low-dose cyclosporine with placebo. Norwegian Arthritis Study Group.. 1994. PMID 7945477, DOI 10.1002/art.1780371015. Read the source
30. Cochrane Database Syst Rev. Topical cyclosporine A therapy for dry eye syndrome.. 2019. PMID 31517988, DOI 10.1002/14651858.CD010051.pub2. Read the source
31. Transplantation. The evaluation of the safety and tolerability of two formulations of cyclosporine: neoral and sandimmune. A meta-analysis.. 1999. PMID 10385078, DOI 10.1097/00007890-199906150-00004. Read the source
32. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — DOSAGE AND ADMINISTRATION, Newly Transplanted Patients. 2026. Read the source
33. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — DOSAGE AND ADMINISTRATION, Psoriasis. 2026. Read the source
34. DailyMed (FDA Structured Product Labeling), SPL version 37, effective 2026-07-10. SANDIMMUNE — DESCRIPTION. 2026. Read the source
35. DailyMed (FDA Structured Product Labeling), SPL version 30, effective 2026-08-10. NEORAL — DOSAGE AND ADMINISTRATION, Blood Concentration Monitoring in Transplant Patients. 2026. Read the source