The scenario presents itself with increasing frequency in obesity medicine practice: a patient who has completed 36 weeks of tirzepatide (Mounjaro/Zepbound) titration, achieved 10 mg or 15 mg/week, and lost between 18 and 24% of starting body weight. Metabolic markers have normalized. The question — whether from the patient managing insurance costs, or a clinician reassessing the long-term plan — is predictable: what happens if the drug is stopped? SURMOUNT-4 (NCT05556512) was constructed precisely to answer that question at scale, and the trial's published data in JAMA 2023 provide the most controlled evidence available on tirzepatide maintenance dosing, withdrawal outcomes, and cardiometabolic trajectory following discontinuation.
SURMOUNT-4: Study Architecture and the Clinical Question It Was Designed to Answer
SURMOUNT-4 was a Phase 3, multicenter, randomized, double-blind, placebo-controlled trial. Eligible participants were adults with BMI ≥30 kg/m², or ≥27 kg/m² with at least one weight-related comorbidity, excluding those with type 2 diabetes. The trial employed a sequential two-phase design that mirrors the clinical decision arc a prescriber faces after successful induction therapy.
During the 36-week open-label lead-in, all participants received subcutaneous tirzepatide titrated from 2.5 mg/week upward using the standard 4-week step schedule, targeting the maximum tolerated dose of 10 mg or 15 mg weekly. At the completion of the lead-in, 670 participants who had tolerated their dose were randomized 1:1 to one of two arms for the subsequent 52-week double-blind phase: continued tirzepatide at the established dose, or matched placebo injection. The primary endpoint was percent change in body weight from randomization to week 88. Secondary endpoints included the proportions achieving ≥5%, ≥10%, and ≥20% total weight loss from original baseline, alongside changes in waist circumference, systolic blood pressure, fasting plasma glucose, HbA1c, and fasting lipid parameters.
The design is not merely academic. By randomizing only participants who demonstrated tolerability and response during the lead-in, the trial specifically models the clinical situation of an established responder facing a maintenance or discontinuation decision — the highest-relevance population for the question the trial was built to answer.
Lead-In Phase Outcomes: Quantifying What Gets Put at Risk
Interpreting withdrawal-arm outcomes requires full appreciation of what was achieved before randomization. Across the 36-week open-label lead-in, participants reached a mean weight loss of approximately 20.9% from original baseline — a magnitude consistent with the 15 mg arm of SURMOUNT-1 (NCT04184622), where the primary endpoint at 72 weeks was 20.9% mean reduction in the tirzepatide 15 mg arm (n=630, p<0.001 vs. placebo; Jastreboff AM et al., NEJM 2022;387:205–216). Approximately 84% of SURMOUNT-4 lead-in participants titrated to the 15 mg/week dose; 16% maintained at 10 mg/week due primarily to gastrointestinal tolerability during titration.
Metabolic improvements across the lead-in were clinically meaningful beyond the weight endpoint alone. Reductions in waist circumference, systolic blood pressure, fasting plasma glucose, HbA1c (particularly in participants with pre-diabetes at enrollment), and fasting triglycerides were documented before randomization. This composite metabolic improvement established the reference state from which both arms diverged — and from which the withdrawal arm's regression must be interpreted. A 14% body weight regain after a 21% loss looks different when the underlying metabolic improvements are also being quantified in parallel.
The Withdrawal Arm: What 52 Weeks Without Tirzepatide Actually Looks Like
Among the 334 participants randomized to placebo, the aggregate trajectory was consistent and clinically significant. By week 88 — 52 weeks after the last tirzepatide dose — this arm had regained approximately 14.0% body weight from the randomization point. Net from original baseline, total weight loss had fallen from approximately −20.9% at randomization to approximately −9.9% at week 88. Approximately two-thirds of the weight lost during the lead-in was recovered within the post-discontinuation window.
The physiological mechanism driving this trajectory is well-characterized at the receptor level. Tirzepatide carries a half-life of approximately 5 days, placing plasma concentrations at negligible levels within approximately 25 days of the final weekly injection. As dual GIP receptor (GIPr) and GLP-1 receptor (GLP-1R) agonism ceases, the pharmacologically-mediated effects on gastric emptying rate, central satiety signaling, and adipocyte lipolysis dissipate. Pre-treatment appetite homeostasis — including the compensatory leptin resistance and upregulated orexigenic signaling patterns characteristic of obesity biology — reasserts itself, driving caloric intake back toward the pre-treatment equilibrium. The regain trajectory appeared steepest in the early post-discontinuation period, consistent with the rapid PK clearance, with some deceleration as a new lower-weight plateau was approached across the 52-week window.
Cardiometabolic markers tracked the weight regain trajectory in the withdrawal arm with clinical relevance independent of the scale endpoint:
- Waist circumference: Regained a significant proportion of the circumference reduction achieved during the lead-in, reflecting restoration of visceral and subcutaneous adipose tissue compartments.
- Systolic blood pressure: Increased in the placebo arm relative to the tirzepatide continuation arm at week 88.
- Fasting glucose and HbA1c: Deteriorated toward pre-treatment values, with the signal amplified in participants with baseline pre-diabetes — the sub-population at highest risk of clinical T2D conversion following metabolic regression.
- Fasting triglycerides: Worsened relative to the continuation arm, consistent with the well-documented dyslipidemia pattern that accompanies adiposity re-accumulation and the loss of GLP-1R-mediated hepatic lipid effects.
These parallel regressions across four cardiometabolic dimensions are not secondary findings to the weight endpoint — they represent the clinical burden of discontinuation for patients whose cardiovascular risk was meaningfully reduced by the induction-phase response. The SURMOUNT-4 withdrawal data are most accurately characterized as documenting a multi-dimensional metabolic rebound, not simply a number on a scale.
Continuation Arm: Durability Profile and Tolerability at Week 88
The 336 participants who continued tirzepatide through week 88 demonstrated ongoing — if attenuated — weight reduction beyond the lead-in gains. From the randomization point, the continuation arm lost an additional mean of approximately 5.5% body weight, yielding a cumulative total mean loss from original baseline of approximately 25.8% at week 88. The between-arm difference at week 88 was approximately 21.8 percentage points of body weight — a clinically substantial effect-size estimate that makes the maintenance-versus-discontinuation decision quantifiable in a way that prior GLP-1 receptor agonist data had not fully captured for a dual agonist.
The incremental loss beyond an already-large induction-phase reduction is mechanistically informative. Tirzepatide's GIPr engagement contributes to peripheral effects on adipocyte lipolysis, energy expenditure, and lipid partitioning that are partially independent of the GLP-1R-mediated satiety pathway. This dual-receptor profile may partly explain the resistance to the complete blunting of effect that typically characterizes single-agonist weight-loss plateaus at 36–52 weeks. The data do not suggest unlimited continued weight loss on maintenance dosing, but they are consistent with a sustained, moderate incremental effect beyond the standard titration window.
Tolerability in the continuation arm at week 88 reflected the expected accommodation pattern. Gastrointestinal adverse events — nausea, vomiting, diarrhea, constipation — were concentrated during the titration phase, with rates declining substantially at steady-state maintenance dosing. Serious adverse events at week 88 were reported in approximately 5.8% of the continuation group versus approximately 7.2% of the placebo group. This directional difference may partially reflect the cardiometabolic deterioration in the withdrawal arm, including blood pressure changes with cardiovascular implications, but the trial was not powered to isolate individual serious adverse event categories.
Tirzepatide Maintenance Dosing: What PK/PD Data Imply About Dose Reduction
SURMOUNT-4 randomized participants to continue at their established induction dose — 10 mg or 15 mg/week — through the entire 52-week blinded phase. The trial was explicitly not designed to evaluate dose de-escalation as a maintenance strategy. This is a clinically material gap: many prescribers managing long-term obesity pharmacotherapy will face a dose-reduction question driven by patient tolerability burden, insurance formulary constraints, or out-of-pocket cost, for which there is currently no dedicated RCT evidence.
Available PK/PD data provide partial mechanistic framing. Tirzepatide achieves steady-state plasma concentrations within 4–8 weeks of dose initiation or change, given the ~5-day half-life. Phase 2 dose-ranging data from NCT03131687 and NCT03131700 demonstrated dose-dependent weight loss across the 5 mg, 10 mg, and 15 mg weekly arms, with statistically significant effect-size differences between dose levels at all measured time points. This dose-response relationship implies that de-escalation from a 15 mg induction dose to 10 mg or 5 mg will produce a reduction in receptor occupancy and a predictable — though not yet RCT-quantified — attenuation of the weight-maintenance effect. The magnitude of that attenuation, and whether it is clinically acceptable relative to the SURMOUNT-4 full-withdrawal regain data, is unresolved.
In clinical practice, any de-escalation decision should incorporate systematic post-change surveillance. A minimum monitoring framework following a dose reduction would include:
- Weight and waist circumference at 4 and 8 weeks post-change
- Fasting glucose and blood pressure at 8 weeks post-change
- Reassessment of GI tolerability burden at both intervals to determine whether the dose reduction achieved its intended tolerability benefit
- Explicit documentation of the absence of RCT maintenance data for the new dose in the clinical record
Safety Monitoring Requirements During Long-Term Tirzepatide Use
The FDA approval of tirzepatide (Zepbound) for chronic weight management in November 2023 carries prescribing guidance that reflects the cumulative SURMOUNT safety dataset. Long-term maintenance therapy introduces monitoring considerations that extend beyond those relevant during the active titration window, several of which are either under-discussed in routine practice or inadequately addressed by standard annual metabolic panels.
Thyroid C-cell surveillance. FDA labeling carries a boxed warning based on rodent carcinogenicity data showing dose-dependent C-cell adenomas and carcinomas at supratherapeutic exposures. Extrapolation to human risk remains unestablished — GIP receptors are not expressed on human thyroid C cells at functionally relevant levels, and no cases of tirzepatide-attributable medullary thyroid carcinoma (MTC) have been confirmed in the SURMOUNT clinical program. Despite this, prescribing information requires contraindication screening for personal or family history of MTC or MEN2 syndrome at every prescribing encounter. Serum calcitonin is not mandated but warrants consideration in patients with thyroid nodules, neck symptoms, or first-degree MTC family history.
Gallbladder disease. Cholelithiasis and cholecystitis were elevated across the SURMOUNT program relative to placebo arms. SURMOUNT-1 reported gallbladder-related adverse events in approximately 2.5% of participants in the tirzepatide 15 mg arm. The mechanism — accelerated biliary cholesterol supersaturation during rapid weight loss combined with reduced gallbladder contractility from GLP-1R agonism — persists as long as active weight loss continues. Extended maintenance therapy carries ongoing gallstone risk, particularly if incremental weight loss continues beyond the induction plateau. Right upper quadrant pain or nausea independent of the expected GI adverse-event profile warrants clinical evaluation and abdominal ultrasound.
Renal function. GLP-1R agonism reduces glomerular hyperfiltration in patients with obesity-associated hyperfiltration; rare acute kidney injury cases have been documented in the context of significant dehydration during GI adverse events. Basic metabolic panel assessment at 3–6 month intervals is appropriate for patients with stage ≥2 CKD, hypertension, or concurrent RAS inhibitor therapy. Any clinical presentation of acute nausea, vomiting, or diarrhea severe enough to reduce oral intake warrants hydration assessment and consideration of temporary dose interruption per FDA guidance.
Lean mass and body composition. SURMOUNT-4 did not report dual-energy X-ray absorptiometry (DEXA)-confirmed body composition outcomes. Available data from tirzepatide and semaglutide pharmacological weight-loss studies suggest lean mass reduction of approximately 20–40% of total weight lost — broadly consistent with the lean mass loss fraction observed in caloric-restriction models generally. For patients on long-term maintenance therapy in whom cumulative weight loss approaches or exceeds 25% of starting body weight, body composition monitoring and structured resistance exercise protocols are clinically warranted, though evidence-based optimal frequency for DEXA surveillance in this context has not been established by RCT.
Clinical Decision Framework: What SURMOUNT-4 Actually Supports
The SURMOUNT-4 data support one central inference: tirzepatide produces its metabolic effects through active receptor engagement, and discontinuation predictably reverses those effects on a timeline governed by the drug's PK profile. The withdrawal arm's ~14% weight regain, combined with parallel cardiometabolic regression, is inconsistent with a "treat until goal weight, then stop" model for patients with obesity-associated comorbidities. This parallels the evidence framework for other chronic metabolic condition pharmacotherapy — antihypertensives, statins, thyroid hormone replacement — where the underlying disease biology persists regardless of prior treatment response, and cessation reliably reverses the treated state.
Three clinical trajectories emerge from the data as distinct decision points:
- Continuation at induction dose: Best-supported by SURMOUNT-4 evidence. Requires structured safety monitoring at 3–6 month intervals and periodic individualized reassessment of benefit-risk, particularly as cumulative weight loss changes the patient's cardiovascular and metabolic risk profile.
- Dose de-escalation for maintenance: Mechanistically plausible, not yet RCT-confirmed at any specific de-escalation target. Requires post-change monitoring and documented clinical rationale. The absence of efficacy data for sub-induction maintenance doses is a material limitation on evidence-based guidance.
- Planned or unplanned discontinuation: The SURMOUNT-4 withdrawal trajectory should directly inform pre-cessation counseling. Approximately two-thirds of induction-phase losses are expected to return within 12 months. Intensified dietary behavioral support during the high-risk early post-discontinuation window — particularly weeks 1 through 24, where regain rate appears steepest — is a rational clinical adjunct to the absence of pharmacological maintenance, though no RCT has specifically evaluated behavioral intensification as a substitution strategy.
A meaningful gap persists in the evidence base as of mid-2026: no head-to-head RCT has compared structured tirzepatide dose de-escalation versus full-dose continuation versus planned discontinuation with concurrent lifestyle intensification. The SURPASS-CVOT trial (NCT04255433) and ongoing SURMOUNT extension data may partially address cardiometabolic endpoint durability, but the dosing-strategy question will likely require a purpose-built maintenance optimization trial to resolve with RCT-grade certainty. Until that evidence matures, clinical decisions about long-term tirzepatide maintenance dosing should be individualized, explicitly documented, and revisited at scheduled intervals rather than set as a fixed protocol at the point of reaching a weight target.
This article summarizes research and does not constitute medical advice. Consult a licensed clinician for diagnosis, treatment, or any decisions about medications or supplements.