skip to Main Content

Eric Widera, MD (@ewidera)

In 1824, Jeremy Bentham published the “Book of Fallacies” in which he criticized fifty arguments used in political debate and explained the sinister interests that led politicians to use them. One of these fallacies he describes as the “sham distinction”, now known better as a “distinction without a difference“. This logical fallacy appeals to a distinction between two two things that ultimately cannot be explained or defended in a meaningful way.  When it comes to cancer and non-cancer pain, one really must question why we are drawing a distinction between these two entities and whether it is science or politics that that demands there be a difference.

The Origins of the Distinction

The term “cancer pain” is relatively new, appearing in greater frequency over the last three decades in both medical journals (Figure 1) and in printed books (Figure 2).  Cancer was viewed as the emperor of all maladies, and pain one of its most feared outcomes.   Use of opioids for cancer pain was encouraged as patients were suffering and the prognosis of their malady inevitably led to short term use.

Figure 2: Use of the phrase “cancer pain” in printed media via Google Ngram

You can also see in figure 1 that the term “non-cancer pain” started to take off about two decades later.  As noted by an article published last week by John Peppin and Michael Schulman, “those using the term “chronic noncancer pain” were in two camps: those who felt that opioids should be avoided in patients without cancer, and those who felt they were yet one more tool for the treatment of these patients.”  A good example of this is an article published in 1995 titled “Opioids for chronic pain of non-malignant origin–caring or crippling” in which an evolution of a distinction was discussed but no significant evidence is given to support a difference:

 The world of pain management, therefore, has evolved two very different models of managing pain. A drug treatment model with a strong emphasis on diagnosis for acute pain and cancer pain; and a behavioural, coping skills model for CPNMO [Chronic Pain of Non-Malignant Origin]. It was inevitable that clinicians versed in one model would begin to apply those principles to the management of pain in the other category.

The Logical Fallacy

The problem with this distinction is that there is very little evidence to support a difference between these two broad categories.  As Peppin and Schulman point out:

“These claims are primarily philosophical, rather than medical or physiologic. As mentioned, pain mechanisms do not discriminate between cancer and noncancer pathophysiology. Patients with cancer or those without cancer have essentially identical pain-generating physiologies, and thus the same mechanisms for the development of their pain (eg, inflammatory pain in a cancer patient will be the same physiological process as in a noncancer patient). Further, cancer patients are living longer and their original pain generators become chronic pain in and of themselves, little different from patients without cancer.

Even if you are more of a splitter than a lumper, it still doesn’t make sense when you include what we know about the biology of cancer.   Cancer, despite recent “moonshot” ideas, is not a single entity.   There are numerous types of leukemia, lymphomas, skin cancers, bones cancers, lung cancers… The list can go on and on. The etiology of pain that may result from these cancers also vary: it can be due to muscle damage, bone invasion and central sensitization, nerve impingement or destruction, chemotherapy induced neuropathy, liver capsular expansion, and on and on the list goes.

From a medical and scientific aspect, it makes little sense to have these two broad categories of chronic pain.   But from a political and emotional aspect, it may make ones life a little easier, because who doesn’t want to relieve the pain of a dying cancer patient.  But what about a cancer survivor?   Are they still considered to have cancer pain once the cancer is cured?  What if the pain is due to chemotherapy-induced neuropathy?   Who will fight their battle politically?

The Alternative

While it may feel wise to distinguish the terms cancer and non-cancer pain from each other, continued promotion of this idea is fraught with logical fallacies, and is best defended by emotions and politics rather than the state of science.

The alternative is to abandon the distinction for all of these reasons and use terms that actually help us understand the nature of an individuals pain and how best to manage it, which may or may not include opioids.  As Peppin and Schulman suggest:

Perhaps a more prudent, less charged set of terms would indicate the origin and generator of the pain. Therefore, a patient with chest-wall pain from radiation due to breast cancer would be labeled “chronic pain of breast cancer radiation-treatment origin”. The patient with pain from an advanced spondylolisthesis would be diagnosed with “chronic pain of spondylolisthesis origin”. The goal here is to continue to be patient-focused, relieve their suffering (instead of contributing to it), and help improve their lives. 

This Post Has 11 Comments

  1. Eric,

    This is an important post. This has become such a toxic and political issue. We have to get better control of opiates in this country AND we have to make sure patients get the care they need for their pain.

    I had a patient with metastatic CA who had increasing pain need an early refill of 120 short acting opiates while we were increasing the long-acting and the pharmacist refused on Friday as it was "too early" and "he wasn't comfortable with this."

    We in the pall care community need to lead this debate, not be victims of it. Unfortunately, we don't seem to be doing a good job acknowledging that people are reacting to a crisis of drug addiction that the medical community is also responsible for.

    You should consider repackaging this for a journal – very thoughtful.

  2. What is different is not the diagnosis but the likely time and setting of the rest of the life. For a person likely to die within a month or two, reliance on opioids is fairly easy to manage. For a person likely to live for years and to be in many settings (work, home, nursing home, assisted living, hospitals), the situation is very different and not so readily managed. We need to be honest about that.

    1. Yes Lynn those with untreated cancer(the number of people who have terminal cancer are also being denied treatment at greater and greater numbers) pain get to die soon, without stigma and guilt. Those with untreated high impact chronic pain must resort to suicide to end their pain now. Of course those increasingly more common terminal cancer are being forced to commit suicide sooner as well! And it’s not a fact that treatment of pain was responsible for the ever increasing number of drug addicts dying. Indeed the availablity of high quality opioids decreases the deaths of drug addicts as is clearly visible from the ever increasing death toll from forcing drug addicts to the black market!

  3. Thanks Dan (and stop giving me more work!). I couldn't agree more that managing the opioid abuse problem shouldn't necessitate we increase the suffering of those with pain.

    Lynn, thanks for your comments too. Just thinking back to the last decade and a half on how I think about treatment of chronic lower back pain. My thoughts on treatments have become a lot more nuanced than whether the pain is due to intervertebral disc degeneration or cancer. Diagnosis plays a role but also as you mention other factors such as prognosis, setting, and function. Opioids may be an important modality or one that I carries more harm than benefit.

  4. Great topic.
    1) the distinction should stop being used as code for 'opioids ok ' vs 'opioids not ok'. That's bullshit, but it's not like we have anything to replace it with.
    2) I don't find it a very helpful distinction. Chronic vertebral osteomyelitis pain has similar mechanism and severity of pain than a met. Making distinctions between pain syndromes is important, eg, between purely neuralgic syndromes like chemo neuropathy, chronic non-tissue damage pain eg fibromyalgia and some types of chronic abdominal pain and chronic back pain, acute tissue-damage syndromes, chronic tissue-damage syndromes (like active cancer, but also chronic vertebral osteo, etc), and then (these categories are my own they may be wrong) kinda cross-over categories like bone-on-bone DJD, deforming inflammatory arthritides, MS and movement disorder associated pain. I think the more you get away from acute tissue damage pain and pain at EOL, the role of chronic opioid therapy becomes more problematic and fraught, which doesn't mean it's never appropriate (at all, it is sometimes definitely appropriat), and we are very much hobbled by lack of long-term controlled data for this critcally important patient and public health issue.
    3) I appreciate Joan's comments that prognosis matters. It really, really does, and I've found it most helpful over the years to conceptualize this as a 'goals of opioid therapy' consideration. Eg, patient who has progressive cancer which is worsening, expected to live months, etc – our goal is appropriately focused on alleviation of suffering (globally, including physical pain). True also for dying patients with any diagnosis when the goals are alleviation of suffering. But for non-terminally ill patients, and for 'terminally ill' patients who are likely to live for years, the goals of opioid therapy, or of pain therapy, are more complicated, and promoting long-term socially relevant functioning (whatever that means for a patient), alongside safety, are key/top priorities with alleviating suffering less so. It would be nice if we could do all 3, but we often can't, and certainly can't do it solely with opioids/drugs.

    4) along these lines, I htink we all see patients with metastatic, 'active' cancer and chronic cancer related pain (eg breast, prostate, myeloma) who live for years, often 5+ years – should we have the same approach for these patients as we should have for our end stage cancer pateints who likely have <1y to live

  5. Eric, I'm harkening back to the olden days of the 80's & 90's, when there was a big effort to liberalize the use of opioids for people with cancer-related pain, and to a lesser extent, those with AIDS-associated pain. At the intersection of compassion & fatalism there was a sense that, because so many people at that time died of their cancer, it didn't matter if they became "addicted" (prevalent then, as today, was a belief that use of opioids = addiction). Cancer pain was also the low-hanging fruit for those of us advocating for more use of effective tools for pain in general. [There is still, after all these years, a perception that opioids aren't effective for 'chronic pain.' As you and others have pointed out, pain is mechanism based, not calendar based]. Aside from opioids (and some other medications), one of the big distinctions between treatment of cancer and noncancer-related pain has been the range of treatments used. Cancer pain treatment has had a heavier emphasis on medication, noncancer pain treatment has had a heavier emphasis on a rehabilitative approach [until the pendulum swung too far and "everyone," especially patients and nonspecialists, decided that taking a pill was easier & cheaper than other interventions]. Burton, et al, had a great article, Chronic Pain in the Cancer Survivor: A New Frontier (Pain Medicine, 2007; 8:2, 1899-98). They also pointed out the fallacy of the cancer-noncancer duality and suggested a multimodality approach chronic pain associated with cancer and cancer treatment. If I had to do my career over, I might consider taking a broad rehabilitation approach to most conditions, including pain.

  6. Great article,

    in Italy (and I think all over Europe) rapid onset oral fentanyl (sublingual, transmucosal) is registered for use (and paid for by the helthcare system) exclusively for BTcP (BreakThrough cancer Pain).
    Think how the other patients with BTP (no little "c"), maybe with associated kidney failure, feel when we are only allowed to use oral short acting morphine…

  7. I teach people the "classic" distinction between chronic non-cancer and cancer pain, but then teach them about the spectrum of pain, something like:

    Fibromyalgia—–Chronic LBP —————Traumatic injuries ——Vasculopath —-Cancer
    Fibromyalgia, chronic daily HA, chronic pelvic/abd pains, chronic low back pain (without major structural abnormality) are on the left side. These cause real pain, but there is no obvious physical cause. On the far right is cancer, with severe vascular disease, severe destructive arthritides, traumatic injuries all on that side. For these, the cause of pain is easily apparent. Other pain syndromes may fall somewhere along this spectrum.
    "Kamell's Law on Opioids in Chronic Pain" 🙂 states that "The more obvious the physical source of pain, the higher the likelihood of benefit and lower the risk from use of opioids."
    While caution is advisable with use of opioids for ALL patients, I am much more likely to treat "cancer-spectrum" pain similar to cancer pain including opioids, while avoiding them for patients without obvious sources for physical pain.
    (Again, note that it is not about "real" vs "unreal" pain, this is ONLY a statement about efficacy of opioids.)

  8. Chronic pain in cancer care is not getting addressed very well. Thinking that treating long term pain with opioids may give a dual purpose eg providing relieve to stress related pain and facilitate a bonding/attachment. We need to ask: Are we treating psychosocial pain here???

    The evidence says that long term opioid use in chronic pain patients are only effective in a very small cohort. when patients bring up depression there may be an association with chronic pain and a psychiatric comorbidity…. Long term opioid use can worsen pain over time and is less likely to make functional gains. The odds of recovering from chronic pain is 4 times higher if patients don't use opioids at all.

    My stand on this is following:
    – Continuous, persistent pain should be seen as a disease of the CNS,not just a symptom of illness or injury, a neuro-inflammatory condition (So I would suggest Nsaids as well, perhaps the SR formulation)
    – Emotional states can become the generator of pain ( depression needs to be addressed)
    – Opioids use can be the cause of pain, patients become sensitised to central pain in which they become independent of it generators. Prevention is the key.

    – treat anxiety, treat muscle tension, manage functional goal setting
    – work towards opioid minimisation
    – Screening for psychological distress and dependency
    – use of non opioid treatment modalities
    – use of support groups and self empowerment

    Anne Booms
    Nurse Practitioner Palliative Care/Oncology

Leave a Reply

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

Back To Top