Odds of a nodule being a lung cancer

[Ray D. Ayshun]

I'm seeing someone later today who is not an operative candidate and has a solitary nodule in a place that biopsy is not worth attempting. The shorter question is, isn't there a nomogram somewhere that allows you to enter a few variable about the patient and characteristics of the nodule that will give you an idea of the likelihood that it's a cancer?

---

Members don't see this ad.

 

[OTN]

Is it PET-avid?

 

[RADONC4285]

I'm seeing someone later today who is not an operative candidate and has a solitary nodule in a place that biopsy is not worth attempting. The shorter question is, isn't there a nomogram somewhere that allows you to enter a few variable about the patient and characteristics of the nodule that will give you an idea of the likelihood that it's a cancer?

Not aware of an actual normogram, but there are some high risk features, which include: older age, female sex, family history of lung cancer, emphysema, larger nodule size, location of the nodule in the upper lobe, part-solid nodule type, and spiculation. I would also add demonstrated growth over sequential scans and PET uptake. see reference below. I think treating suspicious lesions with SBRT in patients at high risk of biopsy is reasonable. we do it not infrequently. Have a balanced discussion with the patient and let them decide.

Probability of Cancer in Pulmonary Nodules Detected on First Screening CT

Major issues in the implementation of screening for lung cancer by means of low-dose computed tomography (CT) are the definition of a positive result and the management of lung nodules detected on the scans. We conducted a population-based prospective ...

www.ncbi.nlm.nih.gov

 

[RADONC4285]

I'm seeing someone later today who is not an operative candidate and has a solitary nodule in a place that biopsy is not worth attempting. The shorter question is, isn't there a nomogram somewhere that allows you to enter a few variable about the patient and characteristics of the nodule that will give you an idea of the likelihood that it's a cancer?

actually here is a calculator i found on uptodate:

UpToDate

www.uptodate.com

 

[ROforbetterorworse]

I'm seeing someone later today who is not an operative candidate and has a solitary nodule in a place that biopsy is not worth attempting. The shorter question is, isn't there a nomogram somewhere that allows you to enter a few variable about the patient and characteristics of the nodule that will give you an idea of the likelihood that it's a cancer?

@Ray D. Ayshun - I like this paper on empiric SBRT linked below

Agree with @OTN at minimum get a PET/CT

Empiric Radiotherapy for Lung Cancer Collaborative Group multiinstitutional evidence-based guidelines for the use of empiric stereotactic body radiation therapy for non-small cell lung cancer without pathologic confirmation - Berman - Translational L

tlcr.amegroups.com

 

[TheWallnerus]

I'm seeing someone later today who is not an operative candidate and has a solitary nodule in a place that biopsy is not worth attempting. The shorter question is, isn't there a nomogram somewhere that allows you to enter a few variable about the patient and characteristics of the nodule that will give you an idea of the likelihood that it's a cancer?

A big (unmentioned) variable: chronology & growth of lesion over time. A growing lesion increases the pre-test/pre-biopsy probability of lung cancer. I believe in the recent lung CT screening trials there was some nomogram-ing involving time? IMHO higher SUVs increase this pre-test probability too. If a nomogram doesn't exist which allows SUV input, size & growth over time... it needs to exist! My intracranial nomogram says that lesions that double in size (volume!) in one year or less and have SUV>10 have a >95% chance of being malignant.

 

[Ray D. Ayshun]

@Ray D. Ayshun - I like this paper on empiric SBRT linked below

Agree with @OTN at minimum get a PET/CT

Empiric Radiotherapy for Lung Cancer Collaborative Group multiinstitutional evidence-based guidelines for the use of empiric stereotactic body radiation therapy for non-small cell lung cancer without pathologic confirmation - Berman - Translational L

tlcr.amegroups.com

Yes, of course to the PET. More meant, entire work-up minus biopsy. According to up-to-date is 68% in her case.

 

[Ray D. Ayshun]

Is it PET-avid?

No PET yet, and weird up-to-date doesnt incorporate PET. I thought I used one that did in residency. In any case, if it's an infectious consolidation, it increases my success rate.

 

[TheWallnerus]

No PET yet, and weird up-to-date doesnt incorporate PET. I thought I used one that did in residency. In any case, if it's an infectious consolidation, it increases my success rate.

PETting is huge. For so many cancers. Even now I think the trial looking at 30 Gy for HPV OPSCC is using PET data to guide therapy. And of course we have that new PET BgRT Reflexion machine on horizon. Personally I put PET in with IMRT and IGRT as being biggest rad onc advancements of 21st century.

 

[evilbooyaa]

I treat with empiric SBRT without hesitation if it is a suspicious nodule that is new or enlarging over time, PET-avid, not amenable to biopsy (either location or patient's medical condition) and (generally) discussed in multidisciplinary tumor board as being good candidate for empiric SBRT. Usually it's more of a spiculated mass as well rather than a nice spherical nodule, but not mandatory.

I will offer it to patients without MDT, I've had one follow-up patient say he doesn't want to bother with MDT but otherwise most do get discussed. I always discuss risk that what I am treating is not actually cancer and document accordingly.

 

[TheWallnerus]

Some of the ABR staff informs that there was an old thread on this topic way back when. May wanna ask pain med guy how patient is doing!

Stereotactic radiation lung mass without biopsy​

 

[RadOncMegatron]

I'm seeing someone later today who is not an operative candidate and has a solitary nodule in a place that biopsy is not worth attempting. The shorter question is, isn't there a nomogram somewhere that allows you to enter a few variable about the patient and characteristics of the nodule that will give you an idea of the likelihood that it's a cancer?

Here is a good one:

Clinical prediction model to characterize pulmonary nodules: validation and added value of 18F-fluorodeoxyglucose positron emission tomography​

Gerarda J Herder 1, Harm van Tinteren, Richard P Golding, Piet J Kostense, Emile F Comans, Egbert F Smit, Otto S Hoekstra
Affiliations expand

• PMID: 16236914
• DOI: 10.1378/chest.128.4.2490

A little older study, but it was published in CHEST for what it's worth. The VUMC folks are also very solid. Not sure if there is better more updated version.

 

[OTN]

PETting is huge. For so many cancers. Even now I think the trial looking at 30 Gy for HPV OPSCC is using PET data to guide therapy. And of course we have that new PET BgRT Reflexion machine on horizon. Personally I put PET in with IMRT and IGRT as being biggest rad onc advancements of 21st century.

Someone, somewhere, at some point, is going to have to explain to me how Reflexion differs in any way other than throughput from: Diagnostic PET/CT ---> CT sim ---> fuse PET/CT to CT sim images --> plan ---> treat.

So far not a single person from the company has been able to do so.

 

[evilbooyaa]

Someone, somewhere, at some point, is going to have to explain to me how Reflexion differs in any way other than throughput from: Diagnostic PET/CT ---> CT sim ---> fuse PET/CT to CT sim images --> plan ---> treat.

So far not a single person from the company has been able to do so.

Reflexion I believe has an option where you can basically adaptively plan to FDG avid areas (where it auto-contours 'targets' based on threshold SUV values) and treat on the fly, without having to do a separate CT sim.

One initial consideration I heard being bounced around was giving something like 4 to 8Gy x 1 to sites of FDG-avid disease (after verification by physician that it was safe) on a q every few months basis.

Whether this is actually feasible or not, I don't know.

Some of the ABR staff informs that there was an old thread on this topic way back when. May wanna ask pain med guy how patient is doing!

Stereotactic radiation lung mass without biopsy​

Re-read through that whole thread, cool stuff. Glad I mostly agree with myself from 3-4 years ago.

 

[Palex80]

This is basically a WIN-WIN. Hit it.


Let's be serious now... I have treated a few such nodules. Sometimes in patients with prior metastatic disease to the lung (mainly CRC), who didn't want to undergo further thoracic surgery. Sometimes in patients that were quite old or had a relevant contraindication to attempt "risky" biopsy (lots of emphysema, post-contralateral-pneumonectomy, pulmonary hypertonia). It will always leave you with the question mark "What did I actually do there?", but bearing in mind the very low rate of complications post-SBRT, I feel fine doing it.

You may hear about the argument at the tumor board "But what about the loss of pulmonary function? SBRT will lead to focal fibrosis, when you are perhaps treating a benign leasion?". I always say more or less the same "Cutting out that segment with VATS to prove if the nodule is malignant or not will not improve pulmonary function either".

There are some of these "too risky to perform biopsy" nodules where you will find LOTS of emphysema and a PET-positive spot somewhere in that bullous emphysema. These patients may benefit from cutting out some of that emphysema (and the nodule) in terms of an LVRS (lung volume reduction surgery), so these patients should undergo VATS, if possible. Done wisely, they can achieve 3 goals with one procedure: better lung function, diagnosis and treratment. A SPECT-CT is generally indicated in these patients to quantify the function of the lung parenchyma around the lesion.

 

[Palex80]

Someone, somewhere, at some point, is going to have to explain to me how Reflexion differs in any way other than throughput from: Diagnostic PET/CT ---> CT sim ---> fuse PET/CT to CT sim images --> plan ---> treat.

So far not a single person from the company has been able to do so.

Have another look at the fancy video, it looks so cool!

 

[RadOncDoc21]

View attachment 334757

This is basically a WIN-WIN. Hit it.


Let's be serious now... I have treated a few such nodules. Sometimes in patients with prior metastatic disease to the lung (mainly CRC), who didn't want to undergo further thoracic surgery. Sometimes in patients that were quite old or had a relevant contraindication to attempt "risky" biopsy (lots of emphysema, post-contralateral-pneumonectomy, pulmonary hypertonia). It will always leave you with the question mark "What did I actually do there?", but bearing in mind the very low rate of complications post-SBRT, I feel fine doing it.

You may hear about the argument at the tumor board "But what about the loss of pulmonary function? SBRT will lead to focal fibrosis, when you are perhaps treating a benign leasion?". I always say more or less the same "Cutting out that segment with VATS to prove if the nodule is malignant or not will not improve pulmonary function either".

There are some of these "too risky to perform biopsy" nodules where you will find LOTS of emphysema and a PET-positive spot somewhere in that bullous emphysema. These patients may benefit from cutting out some of that emphysema (and the nodule) in terms of an LVRS (lung volume reduction surgery), so these patients should undergo VATS, if possible. Done wisely, they can achieve 3 goals with one procedure: better lung function, diagnosis and treratment. A SPECT-CT is generally indicated in these patients to quantify the function of the lung parenchyma around the lesion.

I have no shame in doing it. My patients all understand and all say they rather get RT then another attempted or high risk associated biopsy. No complaints regarding the side effects either. I’ve stopped with the shaming game. Rad oncs need to start taking charge!

 

[Radonc90]

- In the absence of primary cancer elsewhere...
- If Bx is "too risky" (Yes, I had a pt who died from a Bx but that was very rare)...
- I'd repeat the CT Chest w/o contrast in about 2-3 months, I use growth as a sign of malignancy and treat at that time.
But make sure the pt stops smoking now to avoid further confusion.

 

[BobbyHeenan]

Aside: Sorry if I missed it if mentioned elsewhere in this thread.... but in the ASTRO guidelines for lung SBRT there is a little section about treating lesions not biopsied.

For me personally, I always present these cases at our lung tumor board, just to make sure I'm not missing anything and (hope) to get "consensus."

 

[medgator]

I have no shame in doing it. My patients all understand and all say they rather get RT then another attempted or high risk associated biopsy. No complaints regarding the side effects either. I’ve stopped with the shaming game. Rad oncs need to start taking charge!

I usually get the Pulmonary's blessing, otherwise it is no meat, no treat

 

[Lamount]

If it's a growing FDG-avid nodule that is >9 mm in a smoker in a medically inoperable patient, and pulm/IR won't touch it (for whatever reason) OR patient reasonably refuses biopsy, I will offer SBRT.

For me, this usually comes up with 1) patients with severe emphysema, 2) nodules that are right on the diaphragm, or 3) nodules near vascular/central structures --with 3) I am probably using 8-15 fractions

 

[Ray D. Ayshun]

I usually get the Pulmonary's blessing, otherwise it is no meat, no treat

Pulm sent it to me. 1.4 cm avid lesion just next to aortic arch. Emphysema on 2l oxygen.

 

[RadOncDoc21]

Pulm sent it to me. 1.4 cm avid lesion just next to aortic arch. Emphysema on 2l oxygen.

Send the patient over to me for a “second opinion.” Let’s just say the matter will be solved. It’s hard out here for a pimp!

 

[Ray D. Ayshun]

Send the patient over to me for a “second opinion.” Let’s just say the matter will be solved. It’s hard out here for a pimp!

She's being simmed Monday. Outcome here was never in doubt, just curious about the nomogram. Rounding up in the case, it was 69%, so I trust it.

 

[medgator]

If it's a growing FDG-avid nodule that is >9 mm in a smoker in a medically inoperable patient, and pulm/IR won't touch it (for whatever reason) OR patient reasonably refuses biopsy, I will offer SBRT.

For me, this usually comes up with 1) patients with severe emphysema, 2) nodules that are right on the diaphragm, or 3) nodules near vascular/central structures --with 3) I am probably using 8-15 fractions

4) nodule surrounded by ridiculous bullae. Have had a few of those

 

[RadOncDoc21]

She's being simmed Monday. Outcome here was never in doubt, just curious about the nomogram. Rounding up in the case, it was 69%, so I trust it.

Does she have any family members, friends, hell any pets with nodules?

 

[thecarbonionangle]

Just hit it hard brotha. You can do it

 

[wandering star]

In the context of screening the "nomogram" is codified into LUNG RADS:

https://www.acr.org/-/media/ACR/Files/RADS/Lung-RADS/LungRADSAssessmentCategoriesv1-1.pdf

The Fleischner Society guidelines are also relevant:

https://pubs.rsna.org/doi/10.1148/rg.2018180017

I think the current LUNG RADS is sourced from a larger dataset than the NEJM paper that was linked at the beginning of this discussion, and LUNG RADS includes probabilities of malignancy. Obviously in the context of prior cancer you are no longer in the screening population, and the probability of malignancy needs to updated with this prior knowledge. Details like "what cancer did the patient have" and "how likely is it to recur in the lung" are going to be the main questions. You might imagine that this would change the prediction A LOT and that it would be hard to power a nomogram to give appropriate confidence intervals unless you lump disparate categories together (e.g., are you going to have a separate category for stage I lung cancer 3 years ago, what if it was 10 years ago, what about breast cancer, what about ...) so this falls into the realm of clinical judgment. A simplified 5-point nomogram based on a few hundred patients might do more harm than good here because there is a lot of nuance and you need to do a lot of lumping to make a nomogram.

LUNG RADS is based essentially on size alone (and PET avidity if available). Some subspecialized radiologists are able to do better than size alone by using features of the tumor, its location, etc, but the vocabulary and intuition here has not been standardized. As previously discussed PET and growth over time are reasonable next options.

 

[Ray D. Ayshun]

In the context of screening the "nomogram" is codified into LUNG RADS:

https://www.acr.org/-/media/ACR/Files/RADS/Lung-RADS/LungRADSAssessmentCategoriesv1-1.pdf

The Fleischner Society guidelines are also relevant:

https://pubs.rsna.org/doi/10.1148/rg.2018180017

I think the current LUNG RADS is sourced from a larger dataset than the NEJM paper that was linked at the beginning of this discussion, and LUNG RADS includes probabilities of malignancy. Obviously in the context of prior cancer you are no longer in the screening population, and the probability of malignancy needs to updated with this prior knowledge. Details like "what cancer did the patient have" and "how likely is it to recur in the lung" are going to be the main questions. You might imagine that this would change the prediction A LOT and that it would be hard to power a nomogram to give appropriate confidence intervals unless you lump disparate categories together (e.g., are you going to have a separate category for stage I lung cancer 3 years ago, what if it was 10 years ago, what about breast cancer, what about ...) so this falls into the realm of clinical judgment. A simplified 5-point nomogram based on a few hundred patients might do more harm than good here because there is a lot of nuance and you need to do a lot of lumping to make a nomogram.

LUNG RADS is based essentially on size alone (and PET avidity if available). Some subspecialized radiologists are able to do better than size alone by using features of the tumor, its location, etc, but the vocabulary and intuition here has not been standardized. As previously discussed PET and growth over time are reasonable next options.

This is the nodule characteristic factor, but it treats usain bolt's nodule the same as my 75 yo f on o2 with 100 py history.

 

[radiation]

Integrating genomic features for non-invasive early lung cancer detection - Nature

Circulating tumour DNA in blood is analysed to identify genomic features that distinguish early-stage lung cancer patients from risk-matched controls, and these are integrated into a machine-learning method for blood-based lung cancer screening.

www.nature.com

Some day soon we will be able to diagnose these with liquid biopsy

(a real rarity btw -Nature paper by a rad onc actually doing meaningful research)