Issue No. 12
11 August 2026
Issue No. 12 · 11 August 2026
Lung · Biomarkers

One co-mutation identified who needed more than osimertinib.

Osimertinib plus chemotherapy. In treatment-naive advanced EGFR-mutated non-small cell lung cancer selected for a concurrent TP53 mutation, adding pemetrexed and carboplatin to osimertinib more than doubled median progression-free survival, from 15.6 to 34.0 months. Grade 3 or higher treatment-related events rose from 14.9 to 62.4 percent.

34.0
months median progression-free survival, versus 15.6 on osimertinib alone
0.44
hazard ratio for progression or death (95% CI 0.32 to 0.60)
62%
grade 3 or higher treatment-related events, versus 14.9
Source: JAMA · phase 3, open-label, 17 sites in China, EGFR and TP53 co-mutated advanced NSCLC, n=294
Also in this issue
01Radioligand therapy moves into hormone-sensitive prostate cancer
02High-dose vitamin D in colorectal cancer, and a subgroup that will not die
03Dr. Signal's hot take: the question was never whether
04100 Words on the mutations we sequence and ignore
This week’s calls
Four readouts · three to act on · one to watch
9 min read · or skim the calls in 60 seconds. Tap any row to jump.
Dr. Signal's Hot Take

The question was never whether.

Three of this week's four trials added something. SOLARIS added vitamin D to first-line chemotherapy in unselected colorectal cancer and got 11.8 months against 10.3. SKYSCRAPER-03 added an anti-TIGIT antibody to consolidation in unselected stage III lung cancer and got a hazard ratio of 0.96. The osimertinib trial added the same chemotherapy FLAURA2 had already tested, but only in patients carrying a TP53 co-mutation, and more than doubled progression-free survival.

We are still running trials on everyone and calling the result a drug failure.

The difference was not the drug. It was the denominator. TP53 status sits on most sequencing reports we already order, and most of us glance past it. Vitamin D may yet work in left-sided tumours; that subgroup has now appeared three times. Selection is the cheapest intervention available, and it is the one we keep skipping.

Tell me where I got it wrong · @DrRishabhOnco
This week · four readouts

The Verdicts

Every practice-changing trial of the week, distilled to a single call. Act on it, or wait.

How we call it

Every verdict weighs four things: the trial phase, whether the endpoint was mature, the size of the effect, and what it was measured against. Signal is reserved for randomised phase 3 evidence strong enough to act on now. Watch means the result is real but not yet enough to change practice, and most readouts land here.

Signal · act nowWatch · wait
WatchGU · Prostate

PSMAddition

The first phase 3 of radioligand therapy in hormone-sensitive prostate cancer met its primary endpoint, cutting the risk of radiographic progression or death by 28 percent. But medians were not reached in either arm, overall survival was not significantly different, and quality-of-life measures trended the wrong way.
rPFS HR 0.72 · OS not significant
Free of radiographic progression or death at analysis
76% 177Lu-PSMA-617 plus ADT and ARPI
70% ADT plus ARPI alone
Overall survival HR 0.84 (95% CI 0.63 to 1.13, p=0.13); grade 3 or worse adverse events 51 versus 43 percent
Read the detail

PSMAddition. [177Lu]Lu-PSMA-617 at 7.4 GBq every 6 weeks for up to six cycles, added to androgen deprivation therapy plus an androgen receptor pathway inhibitor, versus ADT plus ARPI alone, in PSMA-positive metastatic androgen pathway modulator-naive or sensitive prostate cancer, randomised phase 3, 169 sites in 20 countries, n=1144, second interim analysis at 311 events. Radiographic progression or death occurred in 139 of 572 (24 percent) versus 172 of 572 (30 percent), hazard ratio 0.72 (95% CI 0.58 to 0.90, p=0.0021), meeting the primary endpoint. Time to PSA progression favoured the radioligand strongly (hazard ratio 0.42). Median radiographic progression-free survival was not reached in either arm, and overall survival did not differ (hazard ratio 0.84, 95% CI 0.63 to 1.13, p=0.13) at 184 deaths. Dry mouth affected 46 versus 4 percent, all grade 1 or 2. Composite time-to-worsening hazard ratios for FACT-P, EQ-5D-5L and pain all sat above 1.0, though every confidence interval crossed it. The FDA approved this regimen on 31 July on the strength of these data, so Watch is a deliberate disagreement with the regulator rather than a wait for one. A genuine positive phase 3 that has so far bought delayed imaging progression, not longer or better life. In a disease measured in years, six cycles of radioligand on top of standard doublet therapy needs a survival curve before it becomes routine, and the regulatory filing cites a later cutoff with overall survival still crossing 1 (hazard ratio 0.80, 95% CI 0.63 to 1.01).

Discuss on X
SignalGI · Colorectal

SOLARIS

High-dose vitamin D3 added to first-line chemotherapy and bevacizumab did not improve progression-free survival in metastatic colorectal cancer. The phase 2 SUNSHINE result that prompted this trial did not replicate. The supplementation worked, in that it corrected deficiency; it just did not treat the cancer.
Median PFS 11.8 versus 10.3 months · HR 0.92
Median progression-free survival
11.8 mo high-dose vitamin D3 (4000 IU daily)
10.3 mo standard-dose vitamin D3 (400 IU daily)
Overall survival 25.6 versus 27.0 months (HR 1.05); objective response 51 versus 44 percent (P=.12)
Read the detail

SOLARIS (Alliance A021703). High-dose vitamin D3 (8000 IU daily for 14 days, then 4000 IU daily) versus standard-dose (400 IU daily), added to mFOLFOX6 or FOLFIRI plus bevacizumab, in previously untreated metastatic colorectal cancer, double-blind randomised phase 3, 151 US centres, n=455, median follow-up 20 months. Median progression-free survival was 11.8 versus 10.3 months (hazard ratio 0.92, 95% CI 0.73 to 1.16, one-sided log-rank P=.25). Overall survival was 25.6 versus 27.0 months (hazard ratio 1.05) and objective response 51 versus 44 percent (P=.12). Grade 3 or higher toxicity was not meaningfully different, and vitamin D-associated toxicities were rare. Adherence was 96 percent in both arms and the high-dose regimen reliably corrected insufficiency. A prespecified subgroup showed benefit in left-sided primaries (13.8 versus 10.2 months, hazard ratio 0.74, interaction P=.02) and harm in right-sided (hazard ratio 1.35). The authors call this exploratory and it should be read that way. The question is closed for unselected patients. The left-sided subgroup is the third time this signal has appeared, which makes it worth a dedicated trial and not worth a prescription.

Discuss on X
SignalLung · EGFR
Cover story

Osimertinib plus chemotherapy

In advanced EGFR-mutated lung cancer selected for a concurrent TP53 mutation, adding pemetrexed and carboplatin to first-line osimertinib more than doubled median progression-free survival, 34.0 against 15.6 months. This is the first phase 3 to prospectively select an EGFR population by co-mutation rather than treat everyone the same.
Median PFS 34.0 versus 15.6 months · HR 0.44
Median progression-free survival
34.0 mo osimertinib plus pemetrexed and carboplatin
15.6 mo osimertinib alone
Grade 3 or higher treatment-related events 62.4 versus 14.9 percent; overall survival immature at 30.6 percent
Read the detail

Osimertinib plus chemotherapy in TP53 co-mutated EGFR-mutated NSCLC (NCT04695925). Osimertinib 80 mg daily with pemetrexed 500 mg/m2 and carboplatin AUC 5 every 3 weeks for four cycles then maintenance osimertinib plus pemetrexed, versus osimertinib alone, in treatment-naive stage IV or recurrent nonsquamous NSCLC harbouring both an EGFR-sensitising mutation and a concurrent TP53 mutation, randomised, open-label, phase 3, 17 sites in China, n=294, median follow-up 25.1 and 26.1 months. Median progression-free survival was 34.0 versus 15.6 months (difference 18.4 months, 95% CI 9.9 to 22.3; hazard ratio 0.44, 95% CI 0.32 to 0.60, P<.001), with 24-month rates of 60.8 versus 26.7 percent and median duration of response 32.7 versus 15.3 months. Benefit held across prespecified subgroups including brain metastases and L858R. Grade 3 or higher treatment-related events were 62.4 versus 14.9 percent, driven by cytopenias; 26.2 versus 1.4 percent discontinued for toxicity, and one patient died of treatment-related thrombocytopenia and pulmonary haemorrhage. Overall survival is immature at 30.6 percent (48.4 versus 36.5 months, hazard ratio 0.57, nominal P=.01). Global health status declined faster in the combination arm during induction. Single country, open-label, investigator-assessed endpoint, AstraZeneca funded. An effect this large on a prospectively selected population is hard to explain away, and it answers the question FLAURA2 left open: not whether to intensify, but in whom. TP53 status is already on most next-generation sequencing reports and most of us have been ignoring it.

Discuss on X
SignalLung · Stage III

SKYSCRAPER-03

Tiragolumab plus atezolizumab as consolidation after chemoradiation did not beat durvalumab in unresectable stage III lung cancer. The primary endpoint was missed, overall survival was flat, and immune-mediated toxicity was higher. The tiragolumab programme was discontinued in July 2025.
PFS 19.4 versus 16.6 months · HR 0.96, not significant
Median progression-free survival, PD-L1 positive (primary endpoint)
19.4 mo tiragolumab plus atezolizumab
16.6 mo durvalumab
Hazard ratio 0.96 (95% CI 0.75 to 1.23, P=0.76): not significant. PD-L1 all-comers 14.2 versus 13.8 months, HR 1.00
Read the detail

SKYSCRAPER-03. Tiragolumab 840 mg plus atezolizumab 1680 mg every 4 weeks versus durvalumab, for 13 cycles as consolidation in locally advanced unresectable stage III NSCLC without progression after platinum-based concurrent chemoradiation, open-label randomised phase 3, 177 sites, n=829 (419 PD-L1 positive), median follow-up 33.0 months. The primary endpoint of independently reviewed progression-free survival in PD-L1-positive disease was not met: 19.4 versus 16.6 months, hazard ratio 0.96 (95% CI 0.75 to 1.23, P=0.76). In PD-L1 all-comers, 14.2 versus 13.8 months (hazard ratio 1.00). Overall survival was flat in both populations (PD-L1 positive hazard ratio 0.99; all-comers 45.6 versus 45.8 months, hazard ratio 0.98). Grade 3 or 4 events were similar (26.5 versus 25.7 percent), but immune-mediated events were higher with the doublet (76.4 versus 62.2 percent), as was the need for systemic corticosteroids (33.7 versus 23.5 percent) and withdrawal for toxicity (15.5 versus 9.0 percent). Grade 5 treatment-related events were lower (0.5 versus 1.7 percent). Durvalumab remains the standard after chemoradiation. Together with KEYVIBE-006 this closes anti-TIGIT in stage III, and the tiragolumab programme has already been discontinued. Worth reading as the fourth or fifth consecutive failure to improve on PACIFIC, which is itself the finding.

Discuss on X
Sent here by a colleague?
Get the verdict on every practice-changing trial, every Tuesday.
Subscribe free →
100 Words
One idea, one hundred words.
We order next-generation sequencing on almost every advanced lung cancer, then read one line of it. EGFR, ALK, ROS1, and on to treatment. The co-mutations scroll past. TP53 is altered in roughly half of EGFR-mutated tumours, it has been known for years to shorten time on a tyrosine kinase inhibitor, and until this week nobody had run a phase 3 that used it to choose. The result was a doubling of progression-free survival in the half who needed more, and a spared course of carboplatin for the half who did not. The report was always there. We were reading the first line.

Dr. Rishabh Jain
Consultant Medical Oncologist · Editor, OncoSignals Weekly
@DrRishabhOnco
Want this slot? Email oncosignals@gmail.com
Recent · regulatory

On the label

6 Aug
Vusolimogene oderparepvec (Tudriqev) plus nivolumab
Accelerated approval for unresectable advanced cutaneous melanoma that has progressed on anti-PD-1 therapy. The oncolytic viral therapy cleared on its third review cycle, after two complete response letters and a 10 to 3 advisory committee vote on 30 July.
Melanoma
31 Jul
Lutetium Lu 177 vipivotide tetraxetan plus an ARPI
Approved for PSMA-positive metastatic androgen pathway modulation-naive or sensitive prostate cancer, on the strength of PSMAddition. Eligibility is set by PSMA PET. The regulatory counterpart to this issue's first verdict, which we have called Watch.
Prostate
One question
In EGFR-mutated advanced lung cancer with a concurrent TP53 mutation, adding pemetrexed and carboplatin to osimertinib raised median progression-free survival from 15.6 to 34.0 months, and grade 3 or higher events from 14.9 to 62.4 percent. Will you use TP53 status to decide?
Tap to vote, then see how peers are reading it.

Stop scrolling for the data. Get the verdict instead.

One issue a week. Five minutes. Every number sourced. Read by oncologists who would rather treat patients than chase PubMed.

Subscribe free →
Next week · 18 August. We track the readouts between congresses, and call only the ones that change Monday morning in clinic.