Sunday, October 6, 2019

Part 6 - Why Is Cancer Pain So Special?

by Drew Rosielle (@drosielle)

A Series of Observations on Opioids By a Palliative Doc Who Prescribes A Lot of Opioids But Also Has Questions.

This is the 5th post in a series about opioids, with a focus on how my thinking about opioids has changed over the years. See also:
Part 1 – Introduction, General Disclaimers, Hand-Wringing, and a Hand-Crafted Graph.

Part 2 – We Were Wrong 20 years Ago, Our Current Response to the Opioid Crisis is Wrong, But We Should Still Be Helping Most of our Long-Term Patients Reduce Their Opioid Doses

Part 3 – Opioids Have Ceiling Effects, High-Doses are Rarely Therapeutic, and Another Hand-Crafted Graph

Part 4 – Everything We Were Taught About High Doses Was Wrong, And The Same Hand-Crafted Graph

Part 5 – Why Do We Lump The Non-Cancer Pain Syndromes Together?

This is Part 6, which asks the sort of inverse question to Part 5 – Why Is Cancer Pain So Special?

Why is cancer pain, in so many guidelines and organizational/public policies ‘exempted’ from the dosing and safety guidelines. This is not a rhetorical question – I have not actually ever read a coherent explanation as to why.

One might reasonably assume there’s all this research demonstrating the safety and efficacy of long-term opioid use in cancer patients. In reality, that research is non-existent.

Guidelines like the CDC’s accurately point out that there are essentially no data demonstrating the long term efficacy of opioids for ‘chronic non-cancer pain’ (CNCP). True enough.

But where is the study proving the long-term safety and efficacy of opioids for cancer pain? Where are the well-designed, controlled (or at the very least prospectively designed and closely monitored) studies of opioids for cancer pain which follows patients for longer than 6 months? It’s possible I’ve missed some key studies, but don’t think I have (but correct me in the comments if I have!).

Notably, there are a collection of studies like this one, all involving 6 months or longer of open-label monitoring after a RCT of an opioid for cancer pain. It’s middling quality data involving highly selected patients, and, importantly, for every study like this in the cancer world there’s a similar one like this one in involving CNCP patients. In other words, I don’t see that we are in a situation where we can say we have more convincing, higher quality data for cancer patients than non-cancer patients overall.

And, what exactly do we mean by ‘cancer pain?’ A patient with metastatic breast cancer to bone who has bone pain? I think most of us would consider their pain ‘cancer pain’? What about someone with severely painful multiple myeloma who achieves a long-term remission but their bone pain only partially resolves? Their X-rays and scans continue to show widespread lytic bone lesions, but they don’t have active myeloma. And they may go 5-10 years before relapse. Is that cancer pain?

What about someone with severe, non-healing, osteoradionecrosis of the jaw but no evidence of active oral cancer anymore? Let me tell you, that patient will have severe pain, potentially for life. What about the patient with breast cancer – does it matter she may live over 5 years even with widespread, painful bone metastases? Is that still ‘cancer pain’ (obviously it is), but if so why do we collectively seem to make an important difference between her and the patient who may also live many years with painful, nonhealing osteoradionecrosis? None of this is to mention chemotherapy induced peripheral neuropathies – are those cancer pain? I’m not proposing that I know the answer to these questions – what I’m trying to do however is to make clear that I don’t think we are operating with a set of really fuzzy, imprecise, and ill-defined concepts even when we say something like ‘cancer pain.’

So, I think about all this a lot, and here’s where my current thinking is.

First off, I’ve been trying to find the answer to the question why is cancer pain treated so much differently in these guidelines? Like I intimated above, I don’t actually believe it’s due to there being significantly more/better research data out there establishing the safety and efficacy of opioids for cancer pain than the CNCP syndromes. As I’ve been writing this, I did some rooting around to try to see if I was wrong, and didn’t find anything to change my mind about this. If you want to, dig through this list of Cochrane reviews on the subject, and then try to convince yourself that, in important ways, we have better data for cancer than non-cancer pain syndromes.

Is it because we believe cancer generally hurts more than other pain syndromes? I hope not, that is stupid and wrong.

Is it because we believe patients with cancer are more morally deserving of pain relief than others? Christ, I hope not.

Is it history? That cancer pain has a deep history of decades of advocacy internationally to normalize chronic opioid use for it which has been successful? Maybe, but opioids for CNCP advocacy also has a pretty deep history of advocacy too, and I don’t think this is enough to explain the difference.

Is it because there is something unique about the cause of pain in cancer that makes it reasonably more likely to respond well to opioids long-term than other syndromes? While acknowledging this remains speculative for all of us, I do think this idea at least has some basis in reality, but it only goes so far, and does not justify broad lumping of all the CNCP syndromes together, or all pain-from-cancer syndromes together. That is to say, for at least some of the cancer pain syndromes we encounter, patients, in pretty straight-forward way, have space forming, inflammatory, and infiltrative tumors causing the pain. Ie, the cause of the pain is from clinically obvious, direct, inflammation-heavy, tissue damage and destruction, and, arguably, pain from this etiology (let me call it ‘tissue damage pain’ for short here) responds better long-term to opioids than pain from other causes.

While I don’t think I can cite you clear research data to prove this, I myself basically believe this/accept it to be true: tissue damage pain responds pretty well to opioids, long-term. Obviously, for many CNCP syndromes there is not this ‘tissue damage’ etiology. However, there are many syndromes, ones I’d note that you see a lot if you care for people with serious illness, that do involve tissue damage – sickle cell disease, some types of chronic wound pain, some types of osteonecrosis, ischemic limb pain, possibly pain from some of the inflammatory arthritidies, etc. (Referring to ‘cancer pain’ scenarios I mentioned above, I don’t myself know how to really categorize the chronic bone pain in patients who have achieved remission for multiple myeloma. Is that ‘tissue damage pain’ as I’ve defined it? I’m not sure.)

So, if this is the reason we treat cancer pain as special, my basic request to the world is that we actually say what we mean, and replace cancer pain with something like ‘pain from active tissue damage.’ It’s a godawful mouthful I know, undoubtedly there’s a more euphonious term one can use, but hopefully you get my point. Maybe we should coin a new term: “inflammalgia?”

Finally, is cancer pain treated special because patients with cancer are assumed to be terminally ill? At the end of the day, I think this is actually the predominant reason cancer pain is exempted from all these guidelines. It’s the same reason ‘palliative care’ patients are exempted – it’s because the people who write these guidelines think we’re only caring for patients near the end of life.

To an extent, I do think the poor prognosis-basis for the difference has some merit. If someone has a short prognosis (I’m not going to define what that means but I know it when I see it), I think it’s a pretty uncontroversial statement that having a short prognosis markedly changes the benefit:risk ratio when it comes to opioids. While preservation of function is typically a goal until a patient is really in the final stages of life, for patients with terminal illness and severe pain the simple goal of aggressively relieving suffering is often paramount, and opioids do a really good job of that, especially in the short term. Additionally, the risks to a patient of a lot of the long-term pernicious effects of opioids are less if time is short.

Makes sense to me, but I worry about an unexpected effect of exempting cancer pain from these guidelines and discussions: it gives some people the sense that opioids are unquestionably indicated for patients with cancer and we don’t really have to worry about pain. I’ve seen many situations in which a patient clearly has a dangerous OUD, who has pain and cancer, and people just keep on prescribing. Or, worse, a patient is clearly diverting and people just keep prescribing because 'they have cancer and we have to treat the pain, right?' You encounter the attitude also of 'well they’re dying so who really cares', which makes me want to vomit, as if addiction is a pleasurable thing for people and a great time for a family and not anything other than a tremendous source of total suffering, although I’m happy to say that lately it’s mostly been patients saying that to me ("I have cancer you have to give me meds. Who cares I’m dying anyway"), not my colleagues.

Fundamentally, I’m not really arguing that the policies or guidelines should be rewritten to lump cancer pain and CNCP together. Selfishly, we already waste a ton of time in my clinic begging to get opioids covered for my patients as it is, even for dying cancer patients, and I don’t want anything to happen to make that harder. Mostly I’m arguing the great split between cancer and noncancer pain is harmful because it encourages fuzzy thinking about risk in cancer patients, and lumps too many unrelated non-cancer pain syndromes together, in ways that are really harmful for people, and encourages clinicians and policy makers to think about, eg, a 60 year old patient with severe degenerative arthritis who can’t get out of bed due to pain with a 60 year old patient with depression, anxiety, a history of alcohol use disorder, and ill-defined chronic abdominal pain and headaches in the same category.

It encourages us to simplify what is complex, and our patients deserve better.

Thanks for listening,

Part 1 – Introduction, General Disclaimers, Hand-Wringing, and a Hand-Crafted Graph.
Part 2 – We Were Wrong 20 years Ago, Our Current Response to the Opioid Crisis is Wrong, But We Should Still Be Helping Most of our Long-Term Patients Reduce Their Opioid Doses
Part 3 – Opioids Have Ceiling Effects, High-Doses are Rarely Therapeutic, and Another Hand-Crafted Graph
Part 4 – Everything We Were Taught About High Doses Was Wrong, And The Same Hand-Crafted Graph
Part 5 – Why Do We Lump The Non-Cancer Pain Syndromes Together?
Part 6 - Why Is Cancer Pain So Special?

For more Pallimed posts about opioids.
For more Pallimed posts by Dr. Rosielle click here.

Drew Rosielle, MD is a palliative care physician at the University of Minnesota Health in Minnesota. He founded Pallimed in 2005. You can occasionally find him on Twitter at @drosielle.

Pallimed | Blogger Template adapted from Mash2 by Bloggermint