Medicines that block PI3K, AKT or mTOR commonly raise blood sugar, often in the first weeks. Blood sugar and HbA1c are checked before and during treatment. Metformin is usually the first treatment, and the cancer medicine may be paused or its dose lowered. Tell your team at once about great thirst, passing a lot of urine, vomiting, confusion or trouble breathing. Your own team's plan comes first.
Hyperglycaemia is an on-target effect of blocking insulin signalling through PI3K and AKT, and a labelled metabolic effect of mTOR inhibitors. The US labels report hyperglycaemia in 65% of patients on alpelisib (grade 3 in 33%, grade 4 in 3.9%), and raised fasting glucose in 85% on inavolisib, 37% on capivasertib with fulvestrant and 69% on capivasertib with abiraterone. The median time to onset was 15 days on alpelisib (events of grade 2 or higher), 15 days on capivasertib with fulvestrant, 71 days on capivasertib with abiraterone and 7 days on inavolisib. Ketoacidosis, including fatal cases, is in the alpelisib, capivasertib and inavolisib labels.
The alpelisib, capivasertib, inavolisib and gedatolisib labels ask for fasting glucose and HbA1c before starting, with glucose optimised first. The first three repeat HbA1c every 3 months; the gedatolisib label says at regular intervals and asks for glucose control before each infusion. Fasting glucose schedules differ by drug: alpelisib at least weekly for 2 weeks then at least every 4 weeks; inavolisib every 3 days in week 1, weekly to day 28, every 2 weeks for 8 weeks, then every 4 weeks; capivasertib on day 3 or 4 of the dosing week in weeks 1, 2, 4, 6 and 8, then monthly; everolimus before starting and annually in people without diabetes. The dose tables are written on fasting glucose: no change up to 160 mg/dL (8.9 mmol/L); above 160 to 250 mg/dL (8.9 to 13.9 mmol/L) capivasertib and inavolisib are withheld while alpelisib and gedatolisib continue; above 250 mg/dL (13.9 mmol/L) alpelisib, capivasertib and inavolisib are interrupted, and gedatolisib is withheld if a retest after hydration stays above 250 mg/dL; above 500 mg/dL (27.8 mmol/L) the drug is withheld and stopped for good if the value is confirmed on a recheck within 24 hours (alpelisib), if it persists at 24 hours or there are life-threatening effects (capivasertib), if it recurs within 30 days (inavolisib) or if it recurs (gedatolisib).
The labels do not set an order of glucose-lowering medicines, though the alpelisib label prints the metformin schedule used in SOLAR-1. Manufacturer-funded expert consensus (Gallagher 2024 for alpelisib, funded by Novartis; Iyengar 2025 for capivasertib, writing funded by AstraZeneca) puts metformin first and an SGLT2 inhibitor or a thiazolidinedione such as pioglitazone next. Iyengar advises ketone checks on an SGLT2 inhibitor; the Gallagher panel did not require them. Gallagher rates insulin and sulfonylureas as generally not appropriate first or second, noting a potential risk that insulin stimulates the PI3K pathway; the reviews by Tankova and by Goncalves and Farooki cite preclinical models in which insulin partly reactivated it; Iyengar advises avoiding insulin secretagogues because of hypoglycaemia. The UK JBDS and UK Chemotherapy Board guideline (2023), which is independent and covers all anticancer treatment, instead advises considering gliclazide 40 mg daily when plasma glucose is 12 mmol/L or more on two occasions and ruling out DKA or HHS at 20 mmol/L or more. No ESMO, ASCO or diabetes society guideline specific to these drugs was found.
Why: The labels ask for blood sugar to be optimised before the first dose. A raised fasting glucose or an HbA1c at the upper limit of normal or above is among the risk factors the alpelisib label names for high blood sugar on treatment.
“Before initiating treatment with PIQRAY, test fasting plasma glucose (FPG), HbA1c, and optimize blood glucose.”
Novartis, via DailyMed (US National Library of Medicine) 2026
Why: The alpelisib label names these as risk factors. In SOLAR-1, grade 3 to 4 hyperglycaemia was reported in 56% of people aged 75 or over against 36% of younger people.
“Monitor fasting glucose more frequently for the first few weeks during treatment with PIQRAY in patients with risk factors for hyperglycemia, such as obesity (BMI ≥ 30), elevated FPG, HbA1c at the upper limit of normal or above, use of concomitant systemic corticosteroids, or age ≥ 75”
Novartis, via DailyMed (US National Library of Medicine) 2026
Why: The alpelisib panel (funded by Novartis) rated starting without this consultation as inappropriate, and the label says people with a history of diabetes may need more intensive glucose-lowering treatment.
“For all patients with type 2 diabetes mellitus and/or HbA1c 6.5% to <8.0%, it is inappropriate to consider alpelisib treatment without a pre-treatment endocrinology consultation.”
npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024
Why: The capivasertib expert group (writing funded by AstraZeneca) says dietary change plays an important role in preventing and managing high blood sugar, and that the advice should be tailored to the person, including whether they are overweight or losing weight from cancer. The alpelisib panel adds a dietitian as needed.
“Patients should be encouraged to follow a low-carbohydrate, high-fiber diet, engage in regular physical activity, avoid substantial weight gain, and limit alcohol consumption.”
Why: In one study, metformin started a week before alpelisib appeared to lower the number and severity of high blood sugar events, but nausea, vomiting and diarrhoea were more frequent and more severe. The capivasertib expert group (writing funded by AstraZeneca) also suggests considering metformin beforehand for people at raised risk.
“Consider premedication with metformin prior to the initiation of PIQRAY in combination with fulvestrant based on patient risk factors for hyperglycemia, gastrointestinal tolerability, and clinical situation”
Novartis, via DailyMed (US National Library of Medicine) 2026
Why: In the trials the median time to the first rise was 15 days on alpelisib (grade 2 or higher) and 7 days on inavolisib.
“After initiating treatment with PIQRAY, monitor fasting glucose (FPG or fasting blood glucose) at least once every week for the first 2 weeks, then at least once every 4 weeks, and as clinically indicated.”
Novartis, via DailyMed (US National Library of Medicine) 2026
Why: These three labels all set this interval, alongside the more frequent fasting glucose checks. The gedatolisib label says only at regular intervals.
“Monitor HbA1C every 3 months and as clinically indicated.”
Genentech, via DailyMed (US National Library of Medicine) 2026
Why: The label says to advise patients of the signs and to consider home monitoring. Its patient leaflet lists thirst, a dry mouth, passing more urine, blurred vision, confusion, nausea, vomiting, fruity-smelling breath and difficulty breathing among the symptoms to tell the team about right away.
“Consider consultation with a healthcare practitioner with expertise in the treatment of hyperglycemia and initiation of fasting glucose monitoring at home for patients who have risk factors for hyperglycemia or who experience hyperglycemia.”
AstraZeneca, via DailyMed (US National Library of Medicine) 2026
Why: The panel (funded by Novartis) preferred metformin, up to 2,000 mg a day provided the GFR is above 45 mL/minute/1.73 m². The alpelisib label prints the SOLAR-1 schedule: 500 mg once daily, raised as tolerated to 500 mg twice daily, then 500 mg with breakfast and 1,000 mg with dinner, then 1,000 mg twice daily if needed.
“For patients who developed hyperglycemia while on alpelisib, metformin (short-acting or extended-release) is the preferred first-line anti-hyperglycemic agent.”
npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024
Why: The alpelisib panel (funded by Novartis) rated either as an appropriate second or third medicine. The expert papers prefer medicines that do not act on the PI3K pathway.
“Either an SGLT2i or a TZD is an appropriate second- or third-line agent, or first-line therapy in metformin-intolerant patients.”
npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024
Why: The Goncalves and Farooki review (supported by Novartis) says the symptoms include nausea, vomiting and fatigue, can be mistaken for side effects of alpelisib, and can come with a normal blood sugar. The experts differ: the alpelisib panel did not require ketone monitoring and left it to the clinician.
“Monitor for euglycemic DKA (checking urine or blood ketone levels) in patients receiving SGLT2i, and advise urgent care if DKA symptoms appear”
Why: These raise insulin in the blood. In laboratory studies insulin partly switched the PI3K pathway back on in tumour cells, and the panel warns that insulin given while alpelisib is held may cause low blood sugar. This is a caution from laboratory work and expert opinion, not from a trial in patients. The alpelisib label allows insulin for 1 to 2 days until high blood sugar resolves, the Tankova review says insulin is required for ketoacidosis, and the UK guideline for all cancer treatments uses gliclazide first (see variants).
“Insulin, sulfonylureas, and dipeptidyl peptidase-4 inhibitors (DPP4i) are generally not appropriate first- or second-line agents.”
npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024
Why: The alpelisib label says so in these words, and the capivasertib, inavolisib and gedatolisib tables are also written on fasting glucose.
“Dose modifications and management should only be based on fasting glucose values (FPG or fasting blood glucose).”
Novartis, via DailyMed (US National Library of Medicine) 2026
Why: The label and the panel (funded by Novartis) agree on holding alpelisib at this level. Capivasertib and inavolisib are also withheld; gedatolisib is withheld only if a retest stays above 250 mg/dL (see variants). In the alpelisib and inavolisib trials the median time to an improvement of at least one grade was 8 days.
“For patients with FBG > 250 mg/dL (first episode), the panel recommended holding alpelisib, adding another anti-hyperglycemic agent to metformin, and an endocrinology consult.”
npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024
Why: This is grade 4, the top row of the label's table. The other labels differ on what follows (see variants); the inavolisib label adds an assessment for volume depletion and ketosis.
“If fasting glucose is confirmed at > 500 mg/dL or 27.8 mmol/L, permanently discontinue PIQRAY treatment.”
Novartis, via DailyMed (US National Library of Medicine) 2026
Why: The capivasertib patient leaflet says ketoacidosis needs treatment in a hospital and can lead to death, and fatal ketoacidosis is also in the alpelisib and inavolisib labels. The capivasertib expert group (writing funded by AstraZeneca) says confusion, difficulty breathing, nausea, vomiting or fruity-smelling breath need immediate medical evaluation. The UK guideline says to rule out DKA or HHS when plasma glucose is 20 mmol/L or more or there are suggestive symptoms.
“Withhold TRUQAP immediately when ketoacidosis is suspected. If ketoacidosis is confirmed, permanently discontinue TRUQAP.”
AstraZeneca, via DailyMed (US National Library of Medicine) 2026
Why: The label names these as situations known to raise the risk of severe high blood sugar or ketoacidosis, and its patient leaflet gives the same instruction to patients.
“Withhold TRUQAP in clinical situations known to increase the risk of severe hyperglycemia or ketoacidosis (e.g., suspected serious infection or acute illness).”
AstraZeneca, via DailyMed (US National Library of Medicine) 2026
Why: This is the alpelisib label's schedule. The capivasertib label is similar (twice weekly until back to baseline, weekly for 2 months, then every 2 weeks), and the inavolisib label restarts its starting schedule.
“If a patient experiences hyperglycemia after initiating treatment with PIQRAY, monitor fasting glucose as clinically indicated, and at least twice weekly until fasting glucose decreases to normal levels.”
Novartis, via DailyMed (US National Library of Medicine) 2026
How care is usually carried out, read from the guidelines and standards cited. Your own team's protocol and their judgement come first; this page is for understanding it, not for doing it yourself.
Two layers of guidance exist and they do not fully agree. The labels, which are the right source for monitoring and dose rules, agree on fasting glucose and HbA1c before starting with glucose optimised first, and on dose tables written on fasting glucose. The alpelisib, capivasertib and inavolisib labels set fasting glucose checks that are most frequent in the first weeks, HbA1c every 3 months, and interruption above 250 mg/dL (13.9 mmol/L); each has its own rule for stopping the medicine for good above 500 mg/dL (27.8 mmol/L), and the capivasertib label also stops it for confirmed ketoacidosis. They differ on the schedule and on grade 2: capivasertib and inavolisib are withheld above 160 mg/dL (8.9 mmol/L), alpelisib and gedatolisib are not. The gedatolisib label is lighter: HbA1c at regular intervals, glucose control before each infusion, and withholding above 250 mg/dL only if a retest stays above it. On which glucose-lowering medicine to use there is no independent, graded guideline. Expert consensus funded by the manufacturers (Gallagher 2024, funded by Novartis; Iyengar 2025, writing funded by AstraZeneca; the reviews by Tankova 2022 and Goncalves and Farooki 2022, supported by Novartis) puts metformin first, then an SGLT2 inhibitor or a thiazolidinedione such as pioglitazone, and keeps insulin and sulfonylureas out of the first lines: the alpelisib papers because raised insulin may blunt the cancer medicine, a concern that rests on preclinical models, and the capivasertib paper because insulin secretagogues can cause hypoglycaemia. The independent UK guideline for all anticancer treatment (JBDS and UK Chemotherapy Board, 2023) checks HbA1c and random glucose at baseline and random glucose at each visit, advises considering the sulfonylurea gliclazide 40 mg daily at 12 mmol/L or more on two occasions, and rules out DKA or HHS at 20 mmol/L or more; it does not address the insulin concern. Both are shown here; neither is graded.
Novartis, via DailyMed (US National Library of Medicine) 2026 · AstraZeneca, via DailyMed (US National Library of Medicine) 2026 · Genentech, via DailyMed (US National Library of Medicine) 2026 · Celcuity, via DailyMed (US National Library of Medicine) 2026 · Novartis, via DailyMed (US National Library of Medicine) 2026 · npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024 · npj Breast Cancer (Iyengar et al.); medical writing and editorial assistance funded by AstraZeneca 2025 · Cancers (Tankova et al.); medical writing and publication charge funded by Novartis 2022 · Integrative Cancer Therapies (Goncalves and Farooki); supported by Novartis Pharmaceuticals Corporation 2022 · UK Chemotherapy Board and Joint British Diabetes Societies for Inpatient Care 2023
Where it differs: Fasting glucose on day 3 or 4 of the dosing week in weeks 1, 2, 4, 6 and 8, then monthly; HbA1c every 3 months. Withhold at a fasting glucose of 161 to 250 mg/dL (9 to 13.9 mmol/L) until it is 160 mg/dL or less: restart at the same dose if that takes 28 days or fewer, one dose lower if longer. At 251 to 500 mg/dL (14 to 27.8 mmol/L): withhold until it is 160 mg/dL or less; restart one dose lower if recovery takes 28 days or fewer, stop for good if longer. Above 500 mg/dL, or life-threatening effects of hyperglycaemia at any level: withhold; stop for good for life-threatening effects or if it persists at 500 mg/dL or more after 24 hours. Withhold at once when ketoacidosis is suspected and stop for good if it is confirmed.
Evidence: US prescribing information. In CAPItello-291 fasting glucose rose in 37% (grade 3 in 2%, grade 4 in 1.1%), with ketoacidosis in 0.3%; in CAPItello-281 with abiraterone it rose in 69% (grade 3 in 12%, grade 4 in 1.2%) and ketoacidosis occurred in 1.2%.
AstraZeneca, via DailyMed (US National Library of Medicine) 2026
Where it differs: Fasting glucose every 3 days in the first week, weekly for the next 3 weeks, every 2 weeks for the next 8 weeks, then every 4 weeks; HbA1c every 3 months. Above 160 to 250 mg/dL (8.9 to 13.9 mmol/L): withhold until it is 160 mg/dL or less, then restart at the same dose. Above 250 to 500 mg/dL: withhold; same dose if it falls to 160 mg/dL or less within 7 days, one dose lower if that takes 8 days or more or if it recurs within 30 days. Above 500 mg/dL: withhold, assess for volume depletion and ketosis, restart one dose lower once it is 160 mg/dL or less, and stop for good if it recurs within 30 days.
Evidence: US prescribing information. In INAVO120 fasting glucose rose in 85% (grade 3 in 12%, grade 4 in 0.6%); 46% were treated with oral glucose-lowering medicines and 7% with insulin; median onset 7 days. The trial required an HbA1c below 6% and a fasting glucose below 126 mg/dL. In the phase 1 study (funded by Genentech) hyperglycaemia remained frequent in people with risk factors despite early metformin treatment.
Genentech, via DailyMed (US National Library of Medicine) 2026 · ESMO Open (Gambardella et al.); study funded by Genentech 2025
Where it differs: No dose change up to 250 mg/dL (13.9 mmol/L); reduce by one level if a value above 160 mg/dL has not come down within 21 days of treatment. Above 250 to 500 mg/dL: interrupt; restart one dose lower if it falls to 160 mg/dL or less within 3 to 5 days; stop for good if it has not within 21 days. Above 500 mg/dL: interrupt, recheck within 24 hours, and stop for good if confirmed. The label says insulin may be used for 1 to 2 days until the hyperglycaemia resolves.
Evidence: US prescribing information, based on SOLAR-1: hyperglycaemia in 65%, grade 3 in 33%, grade 4 in 3.9%, ketoacidosis in 0.7%; 87% of those affected were given glucose-lowering medicines, 76% metformin. In METALLICA, metformin started 7 days before alpelisib gave grade 3 to 4 hyperglycaemia in 2.1% of people with normal blood sugar and 15% of those with impaired blood sugar, with more nausea, vomiting and diarrhoea.
Novartis, via DailyMed (US National Library of Medicine) 2026
Where it differs: Given by infusion; glucose control is to be achieved before each infusion, and home monitoring started for an HbA1c above 6.4%. No dose change up to 250 mg/dL (13.9 mmol/L). At 251 to 500 mg/dL: hydrate and retest; withhold if it remains above 250 mg/dL. Above 500 mg/dL: withhold, restart one dose lower once below 250 mg/dL, and stop for good if it recurs.
Evidence: US prescribing information. In VIKTORIA-1 fasting glucose rose in 46% with fulvestrant and palbociclib and 57% with fulvestrant, grade 3 in 0.9% and 1.8%. The trial excluded an HbA1c above 6.4%.
Celcuity, via DailyMed (US National Library of Medicine) 2026
Where it differs: Lighter monitoring and no glucose-specific dose table. Everolimus: fasting glucose before starting and annually in people without diabetes, more often in people with diabetes; withhold for a grade 3 metabolic event until it improves to grade 2 or less and restart at half the dose, and stop for good for grade 4; the label's table gives no step for grade 1 or 2. Temsirolimus: test serum glucose before and during treatment; an increase in, or start of, insulin or oral glucose-lowering treatment may be needed.
Evidence: US prescribing information. Everolimus: hyperglycaemia as a laboratory finding in up to 75%; the label gives grade 3 and grade 4 rates of up to 15% and up to 0.4% for hyperglycaemia, raised cholesterol and raised triglycerides together. Temsirolimus: 89% had at least one raised glucose and 26% reported hyperglycaemia as an adverse event. The consensus papers on choice of glucose-lowering medicine were written for alpelisib and capivasertib, not for mTOR inhibitors.
Novartis, via DailyMed (US National Library of Medicine) 2026 · Wyeth (Pfizer), via DailyMed (US National Library of Medicine) 2025
Where it differs: Baseline HbA1c and random plasma glucose for everyone starting anticancer treatment, random glucose at each visit, a blood glucose meter for those at high risk, gliclazide 40 mg in the morning when glucose is 12 mmol/L or more on two occasions (its summary says 'consider commencing'; the dose may be increased in 40 mg steps), and exclusion of DKA or HHS at 20 mmol/L or more or with suggestive symptoms. It lists everolimus, temsirolimus and alpelisib among the medicines that raise blood sugar but has no pathway specific to them, and does not discuss insulin-raising medicines and PI3K inhibitors.
Evidence: Consensus expert opinion of a UK Chemotherapy Board and JBDS working party (January 2023), independent of manufacturers and ungraded. Its gliclazide recommendation was developed from steroid-induced hyperglycaemia guidance.
UK Chemotherapy Board and Joint British Diabetes Societies for Inpatient Care 2023
Where it differs: Carbohydrate under 50 g a day, or more than 12 hours without food before each dose, in place of a low-carbohydrate diet of 60 to 130 g a day.
Evidence: The alpelisib Delphi panel (funded by Novartis) rated a low-carbohydrate diet as appropriate for all patients and said a ketogenic diet or pre-dose fasting 'may also be appropriate'; the capivasertib expert opinion suggests a ketogenic diet 'if needed'. The Tankova review (writing funded by Novartis) advises against it until there are safety data, noting that urine ketones on such a diet may be mistaken for ketoacidosis. No completed randomised trial was read.
npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024 · npj Breast Cancer (Iyengar et al.); medical writing and editorial assistance funded by AstraZeneca 2025 · Cancers (Tankova et al.); medical writing and publication charge funded by Novartis 2022
Where it differs: The antibody-drug conjugate enfortumab vedotin also causes hyperglycaemia and ketoacidosis, in people with and without diabetes. Its label has one rule: withhold when blood glucose is above 250 mg/dL and restart at the same dose once it is 250 mg/dL or less, with close glucose monitoring in people with or at risk of diabetes. The advice above on metformin first and on avoiding insulin was not written for it; in its trials 5% of patients needed insulin.
Evidence: US prescribing information: hyperglycaemia in 17% of 720 patients on single-agent treatment, grade 3 to 4 in 7%, with one death from hyperglycaemia and one from ketoacidosis.
| Recommendation | Grade, as printed | Source |
|---|---|---|
| Fasting glucose and HbA1c before starting, with blood sugar optimised (alpelisib, capivasertib, inavolisib, gedatolisib labels) “Before initiating treatment with ITOVEBI, test fasting glucose levels (FPG or FBG), HbA1C levels, and optimize fasting glucose.” | Regulatory label; not graded | Genentech, via DailyMed (US National Library of Medicine) 2026 |
| Capivasertib monitoring schedule “After initiating treatment with TRUQAP, monitor or self-monitor fasting glucose levels on Day 3 or 4 of the dosing week during weeks 1, 2, 4, 6 and 8; then monthly while on treatment with TRUQAP; and as clinically indicated.” | Regulatory label; not graded | AstraZeneca, via DailyMed (US National Library of Medicine) 2026 |
| Inavolisib monitoring schedule “monitor or self-monitor fasting glucose levels once every 3 days for the first week (Day 1 to 7), then once every week for the next 3 weeks (Day 8 to 28), then once every 2 weeks for the next 8 weeks, then once every 4 weeks thereafter, and as clinically indicated.” | Regulatory label; not graded | Genentech, via DailyMed (US National Library of Medicine) 2026 |
| Everolimus monitoring: fasting glucose before starting and annually in people without diabetes “In non-diabetic patients, monitor fasting serum glucose prior to starting AFINITOR/AFINITOR DISPERZ and annually thereafter.” | Regulatory label; not graded | Novartis, via DailyMed (US National Library of Medicine) 2026 |
| Temsirolimus: glucose before and during treatment “Serum glucose should be tested before and during treatment with TORISEL.” | Regulatory label; not graded | Wyeth (Pfizer), via DailyMed (US National Library of Medicine) 2025 |
| Gedatolisib: glucose controlled before each infusion “Achieve optimal glucose control before starting each REVTORPYK infusion” | Regulatory label; not graded | Celcuity, via DailyMed (US National Library of Medicine) 2026 |
| Metformin beforehand appeared to lower hyperglycaemia and raised gut side effects (alpelisib) “In the METALLICA study, use of metformin starting 7 days prior to the initiation of PIQRAY appeared to decrease the incidence and severity of hyperglycemia events, but increased the incidence and severity of nausea, vomiting, and diarrhea adverse reactions” | Regulatory label; not graded | Novartis, via DailyMed (US National Library of Medicine) 2026 |
| Prophylactic metformin for a baseline HbA1c of 5.7 to 6.4% (alpelisib) “Prophylactic metformin (short-acting or extended-release) is recommended for all patients with baseline HbA1c of 5.7–6.4%, and it may be appropriate for patients with HbA1c < 5.7%.” | Expert consensus (modified Delphi, funded by Novartis): recommended; not otherwise graded | npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024 |
| Metformin first line for alpelisib hyperglycaemia | Expert consensus (modified Delphi, funded by Novartis): preferred; not otherwise graded | npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024 |
| Insulin and sulfonylureas not first or second line with alpelisib | Expert consensus (modified Delphi, funded by Novartis): generally not appropriate | npj Breast Cancer (Gallagher et al.); funded by Novartis Pharmaceuticals Corporation 2024 |
| Insulin last line with PI3K inhibitors, because of pathway reactivation in cell lines “In PI3Ki-associated hyperglycemia, insulin should only be recommended as a last-line antihyperglycemic agent.” | Narrative review (supported by Novartis); not graded | Integrative Cancer Therapies (Goncalves and Farooki); supported by Novartis Pharmaceuticals Corporation 2022 |
| Sulfonylureas avoided as primary treatment; insulin or sulfonylureas acceptable as rescue “Insulin therapy, and/or subsequent sulfonylureas, are acceptable as rescue therapy, but only after treatment with more suitable hypoglycemic agents has proven insufficient.” | Narrative review (writing funded by Novartis); not graded | Cancers (Tankova et al.); medical writing and publication charge funded by Novartis 2022 |
| Avoid insulin secretagogues with capivasertib; consider metformin beforehand for those at raised risk “start metformin or escalate existing dose and add other agents to treat hyperglycemia, as needed, while avoiding insulin secretagogues, which can cause hypoglycemia.” | Expert opinion (writing funded by AstraZeneca); not graded | npj Breast Cancer (Iyengar et al.); medical writing and editorial assistance funded by AstraZeneca 2025 |
| Gliclazide 40 mg when glucose is 12 mmol/L or more on two occasions, for people starting any anticancer treatment “Consider commencing gliclazide 40 mg daily if plasma glucose ≥12 mmol/L on 2 occasions.” | Consensus expert opinion (JBDS and UK Chemotherapy Board); not graded | UK Chemotherapy Board and Joint British Diabetes Societies for Inpatient Care 2023 |
| Baseline HbA1c and random glucose for all people starting anticancer treatment “All oncology patients should have a baseline HbA1c and random plasma glucose checked prior to commencing anti-cancer therapy or steroids.” | Consensus expert opinion (JBDS and UK Chemotherapy Board); not graded | UK Chemotherapy Board and Joint British Diabetes Societies for Inpatient Care 2023 |
| Ketoacidosis can occur with a near-normal glucose “It is important to be aware that DKA can occur with relatively low glucose concentrations so-called euglycemic DKA (euDKA)” | Consensus expert opinion (JBDS and UK Chemotherapy Board); not graded | UK Chemotherapy Board and Joint British Diabetes Societies for Inpatient Care 2023 |
| A ketogenic diet should not be recommended yet “Therefore, it should not be recommended for people before more data on the safety of such an approach are available.” | Narrative review (writing funded by Novartis); not graded | Cancers (Tankova et al.); medical writing and publication charge funded by Novartis 2022 |
| Enfortumab vedotin: withhold above 250 mg/dL “Withhold until elevated blood glucose has improved to ≤250 mg/dL, then resume treatment at the same dose level.” | Regulatory label; not graded | Seagen, via DailyMed (US National Library of Medicine) 2026 |
Shares Inavolisib, Alpelisib, AKT, Capivasertib.
Shares Inavolisib, Alpelisib, AKT, Capivasertib.
Shares Temsirolimus, mTOR, Everolimus.
Shares Hyperglycaemia (PI3K/AKT inhibitor class effect), Inavolisib, Alpelisib, Capivasertib.
Shares mTOR, Alpelisib, Capivasertib, Everolimus.
Shares Hyperglycaemia (PI3K/AKT inhibitor class effect), AKT, Capivasertib.
Shares Alpelisib, Capivasertib, Everolimus.
Shares Gedatolisib, Alpelisib.