17 Nov: Exploring exploratory analyses

For Lung Cancer Awareness Month #LCAM I’m going to summarize 30 important lung cancer trials over 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial. #lcsm 1/14 Image
We have previously looked at adjuvant chemotherapy (Nov 2) and osimertinib (14 November). Today we’ll look at another influential adjuvant study, and ask whether it ought to be as influential as it is. 2/14
This trial enrolled people exclusively who had had stage Ib cancers removed surgically. In the staging system in use at the time, stage Ib was any tumour larger than 3 cm, with no lymph nodes involved (today many of these would be given a higher stage). 3/14
This trial randomized people to either chemotherapy with carboplatin/paclitaxel or to observation. The primary outcome was overall survival, and the trial was designed with a power of 80% to detect an improvement in 5 year OS from 50% to 63% (see 12 Nov) 4/14
The trial ran into a series of troubles. Accrual was very slow, requiring reduction in the sample size and corresponding adjustments in the statistical plan. Then the trial was stopped at an interim analysis due to a survival signal that disappeared with longer follow up. 5/14 Image
In the end, the trial was negative. There was no evidence that adjuvant carboplatin/paclitaxel increased survival for people with this stage of disease. This was in contrast to other adjuvant trials reporting results around the same time. 6/14 Image
Possible reasons for this trial being negative when others are positive include:

1. This trial was too small (underpowered)
2. Less (or no) benefit for node-negative tumours
3. This trial used carboplatin while others used cisplatin
7/14
Though the authors are frank in their assessment that the trial is negative, they go on to present an “Exploratory Analysis” looking at the subgroup whose tumours were larger than 4 cm. This is different from subgroup analysis, in that it was not pre-specified during study design
A major problem with such analyses is that statistics can’t tell us the risk of an outcome being due to chance. If I presented 40 subgroup analyses and one was positive, you’d be skeptical. If I did 40 exploratory analyses and published only the positive one, how would you know?
The exploratory analysis suggested benefit to chemo in this >4 cm subgroup, albeit only significant due to the statistical expedient of a one-sided test (more on this another day). 10/14 Image
This is a very shaky basis on which to suggest chemotherapy to a group of people, many of whom are already cured by surgery. Some further strength was given to this position, however, when the Canadian adjuvant trial (see 2 Nov) did a similar analysis and got a similar result. Image
I feel no hesitation recommending adjuvant chemo to those with resected node-positive lung cancers. While most guidelines also recommend chemo for node-negative tumours over 4 cm, we should recognize that this recommendation is based on weaker evidence. 12/14
The strength of the recommendation comes out in the discussion of treatment options with the patient. A treatment decision reflects a balance of patient preference and judgement, along with a doctor’s advice. When the advice is less strong, patient preference plays a bigger role
Tomorrow we’ll talk about another chemo trial, and make some comments about placebos and blinding.
14/14 Image

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Garth Nicholas

Garth Nicholas Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @Garth_Nicholas1

24 Nov
24 November: And now a word from our sponsor

For Lung Cancer Awareness Month #LCAM I’m going to review 30 important lung cancer trials over 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial. #lcsm 1/16 Image
To date we have reviewed at a couple of trials looking at the role of surgery in multidisciplinary management (Nov 6 & 8). Today we’ll look at a proper randomized trial of two surgical procedures for staging the mediastinum (the middle of the chest, between the lungs). 2/16
Knowing whether cancer has spread to mediastinal nodes is essential for staging a tumour. As we have seen, staging is required for any treatment decisions. Mediastinal nodes have numbers corresponding to the locations in the diagram below. 3/16 Image
Read 16 tweets
23 Nov
23 November: ALEX and clarity

This year for Lung Cancer Awareness Month #LCAM I’m going to summarize 30 important lung cancer trials over 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial. 1/11 #lcsm
Today we’re returning to ALK-positive lung cancer. Way back on 5 November we looked at the PROFILE study that established crizotinib rather than chemotherapy as the second-line standard of care.
Today’s study compares crizotinib to a newer generation of ALK drug, alectinib. 2/11
This trial enrolled 303 previously untreated people between 2014 and 2016. Primary outcome was progression-free survival, with an 80% power to detect an increase in median PFS form 10.9 to 16.8 months. Particular attention was paid to brain metastases. 3/11
Read 11 tweets
22 Nov
22 November: PACIFIC, intention to treat

For Lung Cancer Awareness Month #LCAM I’m going to summarize 30 important lung cancer trials over 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial. #lcsm 1/13 Image
We previously looked at locally-advanced lung cancer on 4, 6, and 9 November. We have established standard treatment as ~60 Gy radiotherapy with concurrent chemotherapy for those that are not resectable by lobectomy. Today’s trial looked at adding immunotherapy. 2/13 Image
The antibody in this trial is durvalumab. Like the previously mentioned pembrolizumab (Nov 13, 16) and nivolumab (Nov 19), durvalumab inhibits the interaction of PD-1 and PD-L1. Unlike the other two, durvalumab binds to PD-L1. Clinically, the difference seems negligible. 3/13
Read 13 tweets
21 Nov
21 November: Screening part 2

This year for Lung Cancer Awareness Month #LCAM I’m going to summarize 30 important lung cancer trials over 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial. #lcsm 1/17 Image
Yesterday we went over the major results of the NLST. We discussed screening in general and the concept of overdiagnosis in particular. Today we’ll look at the Dutch-Belgian NELSON study, the next largest randomized study in this field. 2/17
NELSON enrolled 13 195 people between 2000 and 2004. They were randomly assigned to no screening, or to CT scans at baseline, 1 year, 3 years, and 5.5 years later. The trial was powered to detect a 25% reduction in lung cancer mortality over the 10 years from enrollment. 3/17
Read 17 tweets
20 Nov
20 November: Screening part 1

This year for Lung Cancer Awareness Month #LCAM I’m going to summarize 30 important lung cancer trials over 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial. #lcsm 1/17 Image
I’m going to spend the next two days on screening studies. Honestly, I have some trepidation in posting on this topic. There are advocates of screening and sceptics about screening, both vocal, and I’m probably going to disagree with most of them. Let’s look at some evidence.2/17
I’m going to review two large screening studies: NLST and NELSON.

Screening is the testing of asymptomatic individuals at risk for a disease. An underlying assumption is that earlier detection results in better outcomes. For many cancers this seems to be the case. 3/17
Read 17 tweets
19 Nov
19 Nov: Ipi-nivo-meso, stratified randomization

For Lung Cancer Awareness Month #LCAM I’m going to review 30 important lung cancer trials in 30 days. These posts are directed at non-medical professionals, with descriptions of the results and of what makes a good trial.#lcsm 1/18 Image
We previously discussed a mesothelioma trial on 8 November. On 15 November I also briefly alluded to a study that established platinum/pemetrexed as the standard chemotherapy, in 2004. Today’s study expands the use of immunotherapy into this disease. 2/18
All of our previous immunotherapy trials (13, 16 Nov) have been about pembrolizumab, an antibody to PD-1. This trial uses the similar antibody, nivolumab, along with a second antibody, ipilimumab, which targets a different part of the immune response, called CTLA-4. 3/18
Read 18 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Thank you for your support!

Follow Us on Twitter!

:(