There is a very interesting study on this topic dealing with the long-term survivors from pancreatic cancer in Finland. Using the national database the authors identified 78(1.3%) >5-year survivors among 5837 patients registered to have pancreatic adenocarcinoma. After review of survivors the authors said that “… in 33 of the 78 cases a histological diagnosis of pancreatic cancer had never been made, and the majority of the remaining 45 patients turned out not to have pancreatic adenocarcinoma after a review. The results suggest that the majority of patients with long-term survival following the diagnosis of pancreatic cancer have never had pancreatic adenocarcinoma.”

Remarkable study ha?

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1968742/

Thank you for sharing this case. But my question aimed at the roots of these phenomena. Does anybody knowa research group/study lookin at the reasons (molecular markers, cell phenomena, tumor genetics etc) in these rare patients? We talk a lot about individualizing cancer therapy and in some way we make progress, but I think we fail to know the reasons why!

Such an example in question is the documented long-term survival of a patient having pancreatic cancer without any undue deleterious or toxic effects. The effective regimen comprised intravenous administration alpha-lipoic acid (an endogenous compound synthesized in humans, and an endogenous coenzyme for several crucial mitochondrial enzyme complexes involved in cellular energy generation and for the endogenous oxidative decarboxylation of alpha-oxo acids and amino acids, and a metabolic antioxidant) and a low-dose opioid antagonist with little toxicity. I have no clue of a study that you are looking for, although discussions have centered on several issues such as tumor size, lymph node status, lymph node metastasis, tumor markers such as HER-2, estrogen and progesterone receptors, carcinogen detoxifying enzyme gene variants, p53, and histology (subtypes) in relation to the prolonged the survival of patients with invasive cancers.

There are also patients with early tumours who do very badly for some reason. For example we resected a T1N0M0 oesophageal cancer 6 months ago. The patient has recently died of wide spread metastases. We consider most patients 'cured' with this kind of histology. Clearly there are things we don't truely understand regarding the biology of some patients' tumours. Biology is king.

Thank you for your answer. It reflects the "dark side" of my question as we also see patients from cancer we usually can tret successful in more than 95%.

There is a very interesting study on this topic dealing with the long-term survivors from pancreatic cancer in Finland. Using the national database the authors identified 78(1.3%) >5-year survivors among 5837 patients registered to have pancreatic adenocarcinoma. After review of survivors the authors said that “… in 33 of the 78 cases a histological diagnosis of pancreatic cancer had never been made, and the majority of the remaining 45 patients turned out not to have pancreatic adenocarcinoma after a review. The results suggest that the majority of patients with long-term survival following the diagnosis of pancreatic cancer have never had pancreatic adenocarcinoma.”

Remarkable study ha?

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1968742/

This really is a very important topic! We should analyse the characteristics of patients doing exceptionally well, or on the ither hand exceptionally bad, in order to find out the nost relevant aspect of tumour behaviour!

Is there any pathologist or molecular biologist willing to do this job? I think amny clinicians would be willing to contribute tumour samples.

@ Alexander Julianov: Thank you for your comment. I know this study. It is remarkable! But in other regions/hospitals the results were different. Perhaps you know the study from M.Trede with 25% longterm survival after pancreatic resection (5 years) in pancreatic cancer. I listened to a lot of discussions about these data and this working group re-examined their resected specimens and had (I know not for sure; it is too long ago; only 1 or 2 revisions of the diagnosis). But in a discussion during a French meeting of abdominal surgery in 90-ties in Strasbourg, I had to learn, that most of these patients with survival of 5 years died in year 6 or 7 because of pancreatic cancer.

Even today the diagnosis of pancreatic cancer can be very demanding for the pathologist! But I posted my question more in general! The reason was I met a lady with gastric cancer and synchronous peritoneal carcinomatosis and then metachronous liver metastases 3 years after first diagnosis and she seems to be free of tumor since 1 year. It is funny, but it is true. I saw more of these patients and I think with our sophisticated methods these patients doing extraodinary well or bad could help us to understand this disease better and to establish perhaps new ways of treating cancer.

Dear Dr. Becker,

I fully agree with your fundamental question: what is the reason for the opposite results (that is to say, prognosis) that we see in what we call the same disease? In other words, who are the patients (and why) on the extreme tails of the Gauss curve? Are they the same disease that is "measured" by the median?

In my opinion, definitly no! The question is that still have no knowledge on how to differentiate these subgroups.

There is a very elegante discussion in this link http://www.cancernetwork.com/bladder-cancer/bladder-cancer-patient-everolimus-has-complete-response?GUID=D82D71D5-4B8F-4980-9929-BE3485A76729&rememberme=1&ts=14032014#sthash.qE6Zylxv.dpuf, addressing the case of a patient with bladder cancer with an exceptional response to mTOR plus anti-angiogenic inhibition, exploring the molecular basis of that particular response and how understanding the genetic profile of this ever adapting disease will help us not only to better understand this disease but obtaining the best possible results with the treatment

As one of those unexpected survivors, with a prognosis of weeks and too far gone for any available treatments to be of use (6+years ago), I have had an interest in this too. Like others, I suspect there may be an explanation at the molecular level which taxonomies of cancer are moving to. There's a similar request for such research as you hope for by MacAdam in 2003 ("Spontaneous Regression").  There is also a Dr David Solit in the USA doing similar work: Looking for lessons in cancer’s ‘miracle’ responders (http://www.bdlive.co.za/life/health/2013/09/18/looking-for-lessons-in-cancers-miracle-responders) who would probably welcome your approach.

Sometimes tumors are not tumors but infecious diseases. HSV can induce glioblasoma and astrocytoma, CMV and HPV correlate with glioblastoma, only as opportunists? Trichomonas correlates with Prostata-CA. Treponema induces many cancers, Leptospira leukemia or colorectal cancer, and Borrelia can induce cancer. In Pubmed you find many publications. But do these agents induce real cancer?

Not in one case (publication submitted): A B-CLL patient with a correct diagnosed pathology on bone marrow biopsy, suffered additionally from a Borrelia infection. The leukemia disappeared completely after the borreliosis had been treated successfully.

Thus, Borrelia infection appeares as Leukemia.

Any B-CLL patient should be treated as Borrelia case first.

And the long term survivors should be tested for a number of infectious diseases.

FWIW, I can't cite any study but I would venture a novice guess of one or more of

(1) misdiagnosis (wasn't really cancer),

(2) few DNA mutations with the driving mutation being compensated for after diagnosis either epigenetically (e.g, fix bad malnutrition) or controllable with an appropriate inhibitor drug,

(3) the cancer does respond to treatment (of whatever kind) and doesn't have supporting mutations (e.g., p53/TP53, or maybe ones enabling metastasis after surgery or radiation to remove/destroy the primary tumor) enabling a resistant variant of the cancer to emerge (i.e., feedback-loop compensation or natural selection survival "adaptation") for an exceptionally long time.

I would think that the possible reasons for why a patient might survive cancer unexpectedly long are as numerous and varied as the range of mutations/rearrangements and subsequent variants found in cancers, i.e., countless.  Discovering why a particular kind of cancer cell behaves like cancer and escapes the body's natural control mechanisms seems an easier challenge.  (Many "driving mutations" and "supporting mutations" have already been identified.)

Thanks for taking up this issue! I have also experience of patients who are alive despite very bad prognosis.

Studies showing diagnostic problems (like the Finish study) expose managerial reasons of unusual survival.

Studies showing biological reasons have been difficult to find, at least for me.

When I do tests for individualization of anticancer treatment, I see huge differences in responses of different patients/tumors/tumor cells, even if the histological diagnostic is similar. If to take into account that we may expect more than 100 molecular profiles of a single type of cancer, e.g. IDC for breast cancer, individual differences in molecular profiles of tumors are certainly among key reasons.

I am sure that individualization of molecular profiles of tumors (genomic, RNAs and especially proteins) will lead to more publications of unusual survival...

With best regards,

Serhiy Souchelnytskyi

Whilst some excellent outcome patients are definitely misdiagnosed (recall the rate of alternative biopsy among patients with clinical metastatic cancer is about 3%) all of us in routine practice have experience with exceptional cases eg a patient of mine with gallbladder cancer biopsy proven fungating out of the biliary drainage tubes given RT in a palliative schedule (20Gy/5fractions)  has a CR, never given any other treatment, dies of complications of poor biliary drainage (cirrhosis,sepsis) 6 years later, no clinical/radiological suggestion at any stage of recurrence. 

I think it would be very informative to develop a study of such exceptional patients, undertake deep analysis of tumor  and host factors as this may give us some interesting 'left of lield' insights

A wrong diagnosis is an obvious possibility, but long-survivors certainly deserve more attention. For exampple in breast cancer we see both kind of "outliers": excellent outcome of patients with relatively advanced disease and terrible clinical course in patients with small tumours. The available biological characteristics (particularly c-erb) seem to explain only a small part of these discrepancies.

I think putting together a collaborative database would be a good idea, then we would need a (molecular)pathologist willing to study these tumour samples.