JK07

JK07 is a lab-made protein designed to do something no approved cardiovascular drug has done: actually regrow damaged heart tissue. It fuses neuregulin-1, a natural growth signal for heart muscle, to an antibody that blocks ErbB3, engineered to trigger cardiac repair while defusing the cancer-risk problem that killed earlier attempts at this mechanism. It is dosed intravenously every four weeks. Salubris Biotherapeutics, the US arm of Shenzhen Salubris Pharmaceuticals, has enrolled 282 patients across both heart failure phenotypes (reduced and preserved ejection fraction) in the Phase 2 RENEU-HF trial (NCT06369298), with 215 of those in the HFrEF cohort [1]. The HFrEF cohort is fully enrolled and follow-up is underway. If JK07 shows even modest functional or biomarker improvement without the liver and oncology signals that derailed prior neuregulin drugs, it opens a genuinely new mechanistic category in heart failure. Current standard of care reduces workload and neurohormonal strain but does not rebuild lost cardiomyocytes. Those standards are SGLT2 inhibitors (drugs like empagliflozin that lower glucose reabsorption and, mechanistically less understood, help failing hearts), ARNI (angiotensin receptor-neprilysin inhibitor, i.e., sacubitril/valsartan / Entresto), beta-blockers, and MRAs (mineralocorticoid receptor antagonists such as spironolactone and eplerenone). None of them rebuild lost cardiomyocytes. That is the pitch, and it is a large one. The trade-off: two prior neuregulin programs (Acorda's cimaglermin alfa and Zensun's neucardin) generated encouraging early signals and then stalled, so investors should treat JK07 as a mechanism-restart bet, not a fresh idea with no baggage.

First-in-class biologic. No approved drug hits this target for this indication anywhere in the world. Phase 2 status: active, not recruiting, fully enrolled at 282 patients per ClinicalTrials.gov, with the HFrEF cohort at 215 subjects [1]. The prior Phase 1 (NCT04210375) was a single-ascending-dose study in HFrEF patients that read out safety, pharmacokinetics, and exploratory efficacy [2]; results were subsequently published in JACC: Basic to Translational Science in 2025 [8]. No FDA breakthrough therapy, fast track, orphan, or RMAT designations have been publicly disclosed. Timing has slipped. In a September 2025 update, Salubris guided that HFrEF-cohort topline results were expected in the first half of 2026 [9]. As of this writeup (mid-July 2026), no topline has been announced, and the trial page has not been updated with a new date. That is a mild negative signal on its own but not diagnostic; cardiology readouts often push by a quarter or two when adjudication of endpoints is being cleaned up. Any communication from Shenzhen Salubris in its next earnings cycle or from Salubris Biotherapeutics directly is the near-term catalyst. The corporate structure matters here. Salubris Biotherapeutics is the wholly-owned US subsidiary of Shenzhen Salubris Pharmaceuticals (SHE: 002294), a Chinese cardiovascular-focused company with real approved products at home: allisartan isoproxil (an angiotensin receptor blocker for hypertension), clopidogrel bisulfate, benazepril, and bivalirudin injections, among others [11]. That gives JK07 a parent with genuine cardiology commercial infrastructure in China, but no ex-China commercial presence and no US biologics manufacturing scale-up track record. There is no separate Nasdaq listing for the biotech arm and no publicly announced ex-China partnership or licensing deal to date, so the program's cash runway is a function of parent-company allocation, not public biotech markets. That structure is unusual for a first-in-class cardiovascular biologic and shapes how a Phase 2 readout will be received commercially.

Neuregulin-1 (NRG-1) is a signaling protein that heart muscle cells need to survive, adapt to load, and, in early life, divide. It binds two receptors on the cardiomyocyte surface: ErbB3 and ErbB4. When NRG-1 activates ErbB4 in particular, the cell turns on survival programs, maintains its energy machinery, and repairs sarcomere damage. In failing hearts, NRG-1 signaling is blunted. The therapeutic bet: replace it, and the heart heals. UniProt characterizes ErbB4 as required for normal cardiac muscle differentiation during embryonic development and for postnatal cardiomyocyte proliferation, which is the mechanistic anchor for the entire class [7]. The ErbB family biology is worth spelling out carefully, because JK07's design lives or dies on it. There are four receptors: ErbB1 (EGFR), ErbB2 (HER2), ErbB3, and ErbB4. They function as dimers. ErbB3 is essentially kinase-dead on its own, meaning its intracellular signaling domain has almost no catalytic activity; it only signals productively when it partners with ErbB2 as a heterodimer. That ErbB2/ErbB3 pair is one of the most potent proliferative signaling units in the body and is central to the oncogenic biology of breast, lung, and gastric cancers. ErbB4, in contrast, can signal as a homodimer or as an ErbB2/ErbB4 heterodimer, and its ErbB4-dominant activity in cardiomyocytes drives survival and repair rather than proliferation. This distinction is why blocking ErbB3 while feeding ErbB4 has a coherent biological rationale rather than just a marketing story: the mitogenic risk of NRG-1 runs mainly through ErbB2/ErbB3, and the cardiac benefit runs mainly through ErbB2/ErbB4 and ErbB4 homodimers. Two prior tries proved the concept is not free money. Acorda's cimaglermin alfa (GGF2) showed early ejection-fraction improvements in a Phase 1 HFrEF study but was hampered by transaminase elevations and a hepatotoxicity signal that halted development [3]. Zensun's neucardin (recombinant human NRG-1) reported a positive Phase 2 in China with functional and mortality signals, but ex-China development stalled and safety questions lingered [4]. The core problem: recombinant NRG-1 activates both ErbB3-containing and ErbB4-containing dimers indiscriminately, so chronic systemic dosing carries a legitimate oncology concern via ErbB2/ErbB3. JK07's design targets that specific weakness. Salubris describes JK07 as a recombinant fusion of an active NRG-1 polypeptide fragment with a fully human IgG1 antibody targeting ErbB3, such that the antibody arm blunts ErbB3 signaling while the NRG-1 payload preferentially engages the cardiac-protective ErbB4 pathway [5][8]. Elegant in theory. The Phase 1 single-ascending-dose data (published in JACC: Basic to Translational Science, 2025) supports the intended pharmacology at the doses tested [8]; whether the pharmacology holds up at chronic dosing across a larger, more heterogeneous population is the question RENEU-HF is designed to answer.

RENEU-HF (NCT06369298) is a Phase 2 randomized, double-blind, placebo-controlled study enrolling 282 patients across two cohorts: HFrEF (heart failure with reduced ejection fraction, LVEF ≤40%, n=215, randomized 1:1:1 to placebo, low-dose JK07 at 0.045 mg/kg IV, and high-dose JK07 at 0.09 mg/kg IV) and HFpEF (preserved ejection fraction, LVEF ≥50%, remaining ~67 subjects) [1][9]. Dosing is intravenous every four weeks. The primary endpoint for the HFrEF cohort (Cohort 1) is change in LVEF at 26 weeks. Secondary endpoints include change in NT-proBNP (N-terminal pro-B-type natriuretic peptide, a blood protein released when ventricles are stretched; falling NT-proBNP means the heart is under less strain), KCCQ (Kansas City Cardiomyopathy Questionnaire, a patient-reported quality-of-life score in heart failure), and 6-minute walk distance. Sponsor: Salubris Biotherapeutics. Two design choices deserve scrutiny. First, running HFrEF and HFpEF in the same Phase 2 protocol is ambitious. These are biologically different diseases. HFrEF is defined by systolic pump failure and has a decade of drug development wins (ARNI, SGLT2i, beta-blockers, MRAs). HFpEF is a stiff-ventricle disorder that has resisted almost every drug thrown at it. A drug that works on cardiomyocyte survival might plausibly help both, but the risk is that a modest effect in each cohort dilutes to nothing in a pooled analysis. Salubris's decision to run the HFrEF cohort as the key-looking arm with its own primary endpoint mitigates this concern. Second, the January 2025 interim analysis on the first 68 subjects reported the trial was 'safe and well tolerated with no meaningful differences in adverse event frequency and severity between active and placebo groups,' with 'trends towards improvement' on preliminary efficacy signals [10]. That is a real if guarded positive readthrough, and it justified opening the high-dose cohort. Enrollment complete at 282 (215 HFrEF) is a decent-sized Phase 2 for a small-biotech cardiology program and is meaningfully larger than the Acorda and early Zensun studies that set expectations for the class.

Our model estimates a 5% chance this drug is eventually approved. It starts from the historical base rate for Phase 2 drugs in this area (about 27%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by larger-than-typical enrollment for this phase; it is held back by the sponsor's thin or weak approval record, its few secondary endpoints, and weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Efficacy risk sits at the top. HFpEF is a graveyard: candesartan and spironolactone missed or split their HFpEF Phase 3s. Sacubitril/valsartan (Entresto) missed its primary endpoint in PARAGON-HF but was subsequently approved by the FDA in February 2021 for chronic heart failure based on subgroup analyses showing benefit in patients with LVEF below normal, making it a partial rather than clean HFpEF win [12]. SGLT2 inhibitors are the cleaner Phase 3 result in true HFpEF. If JK07's mechanism turns out to be more relevant to systolic dysfunction than diastolic dysfunction, the HFpEF cohort could show a diluted signal even if the HFrEF cohort works. Endpoint selection matters: LVEF change is meaningful in HFrEF but nearly useless in HFpEF where ejection fraction is already normal. Safety risk is class-specific and real. Cimaglermin alfa's hepatotoxicity halted its development [3]. Recombinant NRG-1 signaling also drives proliferation via ErbB2/ErbB3 heterodimers, which is why chronic systemic NRG-1 dosing has always carried an implicit oncology concern. The anti-ErbB3 fusion design is the theoretical answer, and the Phase 1 SAD and Phase 2 interim (68 subjects) have not surfaced red flags [8][10], but that follow-up window is short relative to the timescale on which mitogenic signals become detectable. Any transaminase pattern that looks like the Acorda experience would tank the program. Execution risk. Salubris Biotherapeutics has no US commercial infrastructure and no prior FDA biologics filing. Manufacturing scale-up for a fusion protein at chronic-therapy volume is nontrivial. A partnership or licensing deal with a major cardiovascular player would derisk this substantially and has not been publicly announced. Comparable cardiac-biologic licensing precedents are thin: MyoKardia's mavacamten went to Bristol Myers Squibb via full acquisition rather than a straight license, and Novartis's serelaxin failed key, denting the reference price for cardiac-biologic deals broadly. There is no clean recent benchmark, which cuts both ways. IP and commercial risk. The JK07 fusion construct is presumably covered by composition-of-matter patents in the Salubris family, but the specific patent set and expiry horizon have not been surfaced in this pass and should be treated as a pending datapoint. A chronic IV biologic dosed every four weeks is a real access hurdle against oral SGLT2 inhibitors that cost pennies. Payer coverage for a novel biologic in a category with a cheap generic backbone will require outcomes data, not just biomarker change.

Worth watching, with a clear trigger. The signal-versus-noise question resolves on one thing: does the RENEU-HF HFrEF cohort read out a clean safety profile with an objective LVEF gain that echoes any meaningful fraction of the Phase 1 dose-dependent signal. If yes, JK07 becomes the first credible cardiac-regenerative program with real plausibility, and the whole neuregulin class comes back on the map after a decade in the wilderness. If safety wobbles or efficacy is flat, the mechanism moves closer to dead. Specific data points to look for. LVEF change at 26 weeks in the HFrEF cohort - the primary endpoint and the direct read against the Phase 1 signal (2 vs 12 vs 7 vs 2 points across the dose escalation). NT-proBNP change (biological confirmation). KCCQ score change (patient-reported, historically hard to move). Transaminase elevation frequency and pattern. Any oncology adverse events, even ambiguous ones. Timeline. Salubris guided HFrEF topline for H1 2026 in its September 2025 update [9]. That guidance has slipped past mid-July 2026 without a topline announcement, which is worth flagging without over-reading. Watch for the next Shenzhen Salubris parent-company earnings disclosure, an American Heart Association 2026 late-breaker, or a European Society of Cardiology late-breaker abstract for the first meaningful data. Company context. Salubris Biotherapeutics is small, privately funded by a Chinese parent, and unlikely to run a US Phase 3 alone. A partnership announcement with a large cardiovascular player before or immediately after the Phase 2 readout would be a strong positive signal. Silence following a readout would suggest the internal data did not clear the bar for outside partnership interest.