SQUIRE: Statistically significant, but is it clinically relevant?
| Jun 3, 2014
By Arnold DuBell and Stephanie Hawthorne
As treatment of lung cancer patients has become more driven by histology and biomarkers, one group of patients had been left out of the clinical advances made in recent years with targeted therapies: non-small cell lung cancer (NSCLC) patients whose tumors have squamous histology. A few different drugs are now being studied in this patient population, and the first of these trials was reported on Monday at the 2014 annual meeting of the American Society of Clinical Oncology (ASCO). Eli Lilly had stated via press release in August 2013 that necitumumab, their second-generation anti-EGFR monoclonal antibody (mAb), significantly improved overall survival (OS) when combined with gemcitabine and cisplatin as a first-line treatment in 1,093 patients with squamous NSCLC in the Phase III SQUIRE study. That press release did not provide quantitation for the OS benefit found in SQUIRE, but such data was finally made available at the ASCO 2014 session on metastatic NSCLC.
The data presented at the conference suggest that while the benefit may be statistically significant, it may not be clinically relevant. Necitumumab was associated with a six-day (0.2-month) improvement in median progression-free survival (mPFS; 5.7 months versus 5.5 months, HR 0.85, p=0.020), and a 1.6-month improvement in median overall survival (mOS; 11.5 months versus 9.9 months, HR 0.84, p=0.012). Interestingly, the only subgroup not to show a survival benefit was the group of patients who were 70 years or older. Finally, the response rate was also slightly but insignificantly improved in the necitumumab-containing arm (31% versus 29%, p=0.400).
One bright spot for necitumumab was its toxicity profile, which enabled the agent to be generally well-tolerated. Necitumumab was associated with increases in Grade 3 or higher hypomagnesemia (9.3% versus 1.1%), skin rash (7.1% versus 0.4%) and venous thrombolic events (5.0% versus 2.6%).
During this session a comparison was rightly made to Erbitux® (cetuximab, Bristol-Myers Squibb). As noted above, necitumumab is a second-generation anti-EGFR mAb; past history with the first-generation anti-EGFR mAb Erbitux provides a note of caution. Erbitux had been tested in the first-line setting in two Phase III trials in NSCLC patients of all histologies: FLEX and CA225-099. Erbitux was shown to provide a modest yet significant benefit in the FLEX study when added to cisplatin and vinorelbine (HR 0.871, p=0.044) but failed to show a benefit when added to carboplatin and paclitaxel in the CA225-099 trial (HR 0.89, p=0.17). In an attempt to find a biomarker-defined population that benefited from treatment with Erbitux, investigators developed an EGFR IHC score algorithm, termed the “H-score,” to identify “high EGFR-expressing” patients. While an H-score greater than 200 was predictive for patients enrolled in FLEX, it was not predictive for Erbitux benefit in the CA225-099 trial. Unfortunately, in an exploratory analysis presented for the SQUIRE study, the H-score was not predictive for necitumumab activity.
The SQUIRE data, following that of the REVEL study, highlighted the difficulty of finding new effective therapies for patients with NSCLC. Both trials showed minimal efficacy benefits, and both trials probably reached statistical significance thanks (in part) to the large patient populations enrolled. However, Julie Brahmer, the discussant for both the SQUIRE and REVEL presentations, noted that according to ASCO’s recently released recommended efficacy targets for new agents in both non-squamous and squamous NSCLC, SQUIRE does not come close to meeting the threshold of a 2.5- to 3-month improvement in overall survival and a Hazard Ratio of 0.77 to 0.8. Based on the SQUIRE data, Dr. Brahmer rightfully concluded that it is time to reconsider the notion that wild-type EGFR is a target in NSCLC. Although SQUIRE did present statistically significant survival data in an underserved patient population, questions arise about the clinical relevance of these data. Physicians, along with their patients, will need to have serious discussions to answer this question.
 Thatcher, Abstract 8008, ASCO 2014
 Pirker, Lancet, 373: 1525-1531, 2009
 Lynch, J Clin Oncol, 28: 911-917, 2010
 Pirker, Abstract O01.06, WCLC 2011
 Perol, Abstract LBA8006, ASCO 2014