The Skeptical Oncologist

Why conduct post approval studies at all?

Atezolizumab previously received accelerated approval in second-line metastatic or advanced urothelial cancer based on response rates from a single arm trial. The results of post approval confirmatory phase 3 are now published and demonstrate that atezolizumab did not improve survival versus chemotherapy (11.1 v 10.6 months, HR 0.87, p = 0.41). The concept of accelerated approval is to grant early and conditional approval and access to drugs in diseases of unmet need, and that the decision to fully approve or revoke be made based on results of confirmatory phase 3 trials. That means, if the confirmatory larger phase 3 trial shows that the assumed benefit with the drug didn’t exist, the approval be revoked. Naturally, the FDA has decided to revoke the approval. No, I am joking. The post approval studies don’t seem to matter at all. You can improve response rate in a single arm trial, gain an expedited approval and your approval will always remain. Previously, we also saw that a drug that was given accelerated approval based on response rates in a phase 2 received full approval after it failed to improve survival in a subsequent confirmatory phase 3.  Why bother conducting confirmatory studies at all?

(Im)precision in precision oncology

It goes without question that precision medicine requires precise measurement and validation of biomarkers. So it was very pleasing to see this study which showed that for 3 genes (BRAF, EGFR and KRAS) studied, both laboratory developed tests and FDA-approved companion diagnostics had more than 97% accuracy for all 3 genes combined. However, in another study among 34 samples from 40 prostate cancer patients, the concordance in the results from two commercial liquid biopsy tests, both of which were CAP-accredited for cell-free DNA next-generation sequencing, was very disappointing. There was complete concordance in only 35% samples. I have previously argued that imprecision in the measurement of biomarkers could fail precision oncology. My conclusion from that commentary still remains: ” Until the biomarkers have been proven with analytical and external validity, exercises in biomarker-based precision medicine are likely to be futile.”

Biomarkers for PD-1 inhibitors

One of the most important breakthroughs in oncology would be to find a good biomarker for PD-1 inhibitors because these drugs are effective for a subset of patients across many cancers but come with expensive price and varied side effects. Knowing which patients will and will not respond to these agents is important. Although we don’t know that yet for individual patients, a study has shown that the response with these drugs are fairly linearly correlated with the tumor mutational burden across cancer types. Given that these drugs have now received tissue-agnostic approval, this study can provide useful insights as to whether these drugs would provide response in a certain tumor type. However, unless we have predictive biomarkers that can be applied to an individual patient, there are miles to go before we sleep.

New questions with Neoadjuvant therapy

The breast cancer community is indebted to EBCTCG for various practice changing meta-analyses that form the current evidence-base to treat breast cancers. A new meta-analysis by the same group has compared neoadjuvant versus adjuvant chemotherapy for early breast cancer using individual patient data from 4756 women. First,  good news. There was no difference between neoadjuvant and adjuvant chemotherapy in terms of distant recurrence rates, breast cancer mortality rates or overall mortality rates. The bad news? Neoadjuavnt chemotherapy was associated with a significant increase in local recurrence rates versus adjuvant (21.4% v 15.9% at 15 years, HR 1.37, p = 0.0001). Thus, although neoadjuvant chemotherapy allows better chances for breast conserving surgery, there is a trade off of increased local recurrence to consider.

Avastin is not for adjuvant

Bevacizumab has been approved for a variety of cancers in the metastatic setting, although the absolute margin of benefit is debatable in some. However, one thing we have learned through many trials: bevacizumab is not a drug for the adjuvant setting. It has failed 3 trials in the adjuvant setting of colorectal cancer and now a new RCT shows that it doesn’t improve survival in the adjuvant setting of non small cell lung cancer either, but it did increase toxicities.

No GOLD in the Olympiad

Here is another example of why positive phase 2 doesn’t necessarily translate to positive phase 3. Olaparib plus paclitaxel was shown to improve survival versus palcitaxel alone in a previous phase 2 in gastric cancer. The results from the phase 3 GOLD trial are out but it seems olaparib marginally missed the golden success. However, come to think of it, even the results from the Olympiad trial don’t look very convincing as I have discussed in a recent editorial which brings me to my selfie.

Let me take a selfie

Olaparib was tested in Olympiad trial among women with metastatic HER2 negative breast cancer with BRCA mutation and met the primary endpoint of PFS benefit. However, in this editorial, I conclude that “although olaparib seems to have won the Gold with OlympiAD, the patients probably have not. We need to stop celebrating a gold-plated bronze as a true gold so that one day our patients can finally get the gold they deserve.”

Bishal Gyawali, MD is a consultant with the Anticancer Fund. He lives in Nepal.  A version of this post first appeared at

Leave a Reply

Your email address will not be published. Required fields are marked *