|Year : 2009 | Volume
| Issue : 1 | Page : 13-16
Oncologists should measure life, not death: A newer perspective of cancer statistics
A Srivastava1, A Sood2
1 Department of Surgery, All India Institute of Medical Sciences, New Delhi, India
2 MBBS Undergraduate, All India Institute of Medical Sciences, New Delhi, India
Department of Surgery, All India Institute of Medical Sciences, New Delhi
Source of Support: None, Conflict of Interest: None
We discuss why we should measure parameters of 'life' and not 'death' in choosing a therapy for cancer and what are the flaws and difficulties in measuring cancer specific deaths. We then submit those parameters which can be measured with high precision and validity in cancer patients and so should be recorded to guide us in deciding the best therapy.
Keywords: Cancer, metastasis, quality of life, recurrence, survival, trials
|How to cite this article:|
Srivastava A, Sood A. Oncologists should measure life, not death: A newer perspective of cancer statistics. Indian J Cancer 2009;46:13-6
| » Introduction|| |
Prof. Michael Baum wrote in his latest book on breast cancer Fast Facts that the "goals of treatment are: to prevent or delay recurrence for as long as possible where there is no evident disease and to relieve symptoms and improve quality of life in patients in whom distant metastases are present at diagnosis."  He was cautious not to use the word 'cure' even for early disease. Prof. Baum has a vast experience of seeing and treating innumerable number of women with this affliction for over half a century. It made me ponder as to why a man of such wide erudition and experience should not claim 'curing' at least some women. Probably, he does not want to take the credit of the good outcome of his operation out of sheer humbleness and humility or perhaps there is no such thing as curing cancer. The word 'cure' is hard to define - is it freedom from all viable /active tumor cells in the body? Or is it one's ability to live for 5 or 10 years (past therapy) without any overt clinico-biochemico-radiological evidence of any known form of the same disease at the primary or distant parts of the body?
| » Do We Know Enough?|| |
But then do we know enough of all the aberrations/disorders resulting from interactions of tumor cells with time (I mean when left for long), with their own 'kith and kin', and with the hostile neighbors (autocrine and paracrine behavior) in an inhibitory or promotional milieu.
May I humbly suggest that the lady who died the other day in the coronary care unit, 15 years after a 'curative Patey's mastectomy', actually had a small (about 20 micron) tumor embolus which on reaching the hostile noisy environment of heart (having lived a peaceful life in the blood sinusoid of spleen - it is another story that those scary Littoral cells kept him dormant) got a good beating by the beating heart, cried and wept so much that in his tears released some sticky matter that Dr Kakkar calls 'tumor induced thromboplastin peptide'.  Hearing so much hue and cry, friendly relatives (the platelets) came to sympathize and gave a real close hug. The next day we did not hear any cry and sob of tumor cells, it was 'all quiet on the cardiac front', you see, the battle was over. The poor forensic pathologist could not see any tumor cells, they were so disfigured after all that thrashing and thumping. Moreover, these dead tumor cells were nicely wrapped by a wreath of red roses of RBC, what a 'floral tribute to the departed soul'. All-in-all it looked very orderly and natural, no wonder the pathologist wrote "death due to natural cause by coronary artery thrombosis."
Of course, death is a natural event that all living beings have to embrace. We are all born to die. In all the Holy scriptures, the Lord tells that "I am the all devouring death, I am the all destroying death that snatches all, and origin of all that shall be born." 
If death is in the hands of God Almighty and not in the hands of mortal surgeons or oncologists, then would it not be prudent to stop counting the number of dead bodies in the cancer graveyard? [Figure 1].
It is possible that some residual tumor cells produce toxins like free oxygen radicals or DNA adducts that may hasten the development of/or aggravate the pre-existing degenerative conditions like stroke, heart disease, thromboembolic phenomenon, arthritis, senile changes, or even a new cancer.  In the language of causality, the presence of cancer may serve as a 'sufficient cause' for degenerative conditions, in the causal pie model of Rothman.  The above degenerative changes may happen as part of 'natural history of cancer' or as a consequence of the toxic chemoradiation or a major surgery that perturbs the milieu interior . Thus, in the life of a cancer survivor, it may be hard to measure or count all the ill effects of cancer or its therapy. The detection of residual or persistent tumor is a function of the sensitivity of clinician's fingers in palpating a tumor nodule, node, or enlarged liver. It also depends on the resolution of the diagnostic tool employed in searching for subclinical tumor. Thus, modern spiral CT scanner has a higher sensitivity than older forms of CT scanner, and MRI scan is even more superior to spiral CT scan. Tomorrow, if PET scan becomes a routine test in the follow up of cancer patients it will perhaps detect much more recurrences that remain hitherto hidden/dormant by current modes of imaging. Thus, definition of 'disease-free state' is not absolute and will vary accordingly. In that case, the term 'disease-free survival' seems an oxymoron. So, once you get cancer you live with it or some of the aftermath of its interaction with the host tissues or reaction of therapy.
| » Miscalculations in Calculations|| |
The survival is very much dependent on the age of the patient at diagnosis. For a cancer with equal lethality in all ages, the mortality would be higher in the older population because of concomitant illnesses. Some authors try to correct for this by determining the cause of death and analyzing the 'cause specific mortality'. Unfortunately, the determination of the cause of death is not always accurate. Thomas Saphner and coworkers chose to calculate the annual hazard of recurrence as opposed to survival in the group of women entered in to "Eastern Cooperative Oncology Group" (ECOG) studies on breast cancer, to avoid the need of distinguishing death caused by breast cancer as opposed to death due to other causes.  The problem with concept of '5 and 10 year survival rates' is that we measure it by counting the number of dead ones. The overall or all-cause mortality can be measured fairly accurately, if the outcome is available for all the patients with no loss to follow up. However, it is only possible in Great Britain, Europe, or USA where accurate lifelong record on every patient is maintained. The process of identifying underlying causes of death from clinical case histories is of limited reliability and therefore of limited accuracy. 
The precise cancer-specific or cancer-related mortality is impossible to measure, unless we perform autopsies in all the death cases. As illustrated above, even on autopsy, a small focus of cancer may be missed. Even in the presence of histologically demonstrated tumor deposit in say liver, at autopsy, how does one ascertain that it is the cause of death? It may be 'cancer associated' death but not necessarily 'cancer induced' death, for example, one may die of coronary thrombosis while harboring a silent liver deposit. Moreover, the 'art of autopsy' is now dying so the next generation of pathologists will not be adequately skilled to perform a detailed postmortem.  In a recent editorial in The Lancet , the present role of autopsy in the medical practice has been described as follows "Forensic pathologists distinguish 'true' post-mortem appearances from 'artefacts' and other misleading phenomena occurring after death, and accept that even 'true' post-mortem appearances did not always reveal the cause of death. Many doctors regard the autopsy as a crude and largely outmoded procedure. Condemnation of the unauthorized retention of organs and tissue at autopsy threatens to undermine public consent for autopsy. No longer the final arbiter of medical knowledge, the autopsy has lost much of its authority and now has a marginal role in contemporary medical practice." 
The problem gets compounded in countries like India, where most people do not consent for mutilation of their dear ones for religious reasons. Hence 5, 10, 20 year cancer-related mortality data can never be accurate in many parts of the globe.
Thus, overall survival (OS) rate and disease-free survival rate, both are out of the question to be applicable globally. What should we measure then in a cancer case?
As a pragmatic doctor, may I suggest that our endeavour should be to measure the parameters of life, viz. Pain-free life in days/months or years, disability free life (as QUALY or DALY),  as days spent in a garden, or by the lake/ river/ sea. Count the cakes/puddings/ice creams enjoyed. Measure the value time spent with the relatives or grandchildren. Record the days without a crutch/walker or a wheel chair, the number of 'cancer charity walks', or pilgrimage she went to… The list goes on as the life goes on. Ask Shakespeare, Wordsworth, and Tagore to complete this list, not a pessimist oncologist please, who can only count the dead!
The concept of 5, 10 years mortality is so deeply ingrained in our minds (probably, because we are so scared of death) that we rest all our decisions for choosing an operation or a prevention strategy based on mortality. In his masterpiece treatise Breast cancer: Decision to screen , Sir Patrick Forrest laid so much emphasis on mortality data that he chose to offer the screening program only to women between 50-64 years.  Younger and older women were not included simply because there was no reduction in mortality in the screened population.
I very humbly submit to the learned community of oncologists that they should broaden and widen their horizon of thinking and look beyond death in issues of therapy and screening. As a future health minister of my country, I would like to launch population-based screening with the objective of increasing the number of breasts salvaged (i.e. not mastectomized), number of axillae spared (you do not disturb the armpit for early disease, as Prof. Mansel in his robust data from ALMANAC trial, showed a significant reduction of arm morbidity in sentinel node group patients),  number of women spared the toxicity of chemotherapy (<1 cm, good grade, no disease), with cost saving, reduction of radiation-induced morbidity, and overall QUALYS gained. For oral cancer screening, I will be content to see a reduction in the number of mutilating total or near-total glossectomies or deforming hemimandibulectomies and ugly scarring of radical neck dissection. I will not lament even if the screening does not lower the number of cancer-related deaths.
On the first lecture in the epidemiology class at the School of Public Health (University of North Carolina at Chapel Hill), we were taught about the 'web of causation' and referred to the classic article by Melvyn Sussar; What is a cause and how do we know one?  Truth of the matter is that ascertaining a cause of an event (e.g. cause of death) is very difficult. Yet, clinicians are very quick in jumping to the conclusion and incriminating a factor as causal (remember when as a poor registrar you were admonished by the Professor, "the colonic anastomosis leaked because you did not make a covering colostomy."). In fact, epidemiologists like Susser and Rothman teach us that no disease or event results from a single cause. There is a 'necessary cause' working along with a constellation of 'sufficient causes' .  Thus, in a cancer patient, it is very hard to ascertain that the death was due to cancer. The 'all cause mortality' is the only parameter that can be measured with high validity and precision. But, to choose a therapy on the basis of 'all cause mortality' would be unfair to the therapy, as it includes death not influenced by the therapy, for example, accidents, myocardial infarction, and stroke.
If a therapy improves the local control of disease together with QUALY in a randomized trial, we should accept it as safe and effective therapy and use it even if there is no survival advantage. A recent example of such a situation arose with anastrazole trial reported in The Lancet . The ATAC trialists' group noted a reduction in recurrences, reduction in contralateral breast cancers, and improvement in disease-free survival with anastrazole compared to the tamoxifen arm. However, there was no difference in the OS. If a drug controls the growth of cancer cells effectively then it must improve the OS. In this case, anastrazole appears to be more effective in controlling the growth of cancer cells than tamoxifen and yet is not influencing the OS. Why? I feel, it is the problem with the OS and not the anastrazole which is faring well on so many other parameters of effectiveness. Modern day breast cancer experts have already accepted using anastrazole as a better drug in spite of its failure to impact the OS. 
The randomized trials on soft tissue sarcoma at National Cancer Institute of USA had shown that adjuvant radiotherapy along with limb salvage surgery had a local recurrence (LR) of 20% as compared to amputation with no LR. The OS was similar in both groups.  Brennan et al , at the Memorial Sloan-Kettering Cancer Center reported a reduction in LR with brachytherapy together with surgery for soft tissue sarcoma, when compared with surgery alone without influencing the OS. 
The oncology literature is replete with reports where OS is similar in spite of difference in local or systemic recurrence. 
| » What can we Measure with High Validity and Precision?|| |
- Presence of LR,- as a binary variable, Yes/No - and its date of confirmation. Of all the forms of relapse/recurrence of tumor, the LR is perhaps most precisely measured. It is often present close to the surface for tumors of breast, skin, and head and neck. Its nature can be confirmed by cytology or histology and date of onset can often be recorded with high validity. The occurrence of a LR is an expression of the failure of the anticancer therapy. If a cancer therapy is effective in ablating or destroying cancer cells then it must lower the LR. Hence, the local failure should be considered as the most important measure of the effectiveness of a therapy and a determinant factor in choosing a therapy. No cure is conceivable without local control.
- Presence of metastasis to vital organs, viz. lungs, liver, brain, and bones confirmed by CT/MRI/isotope scan with cytological or histological proof where safe and ethical - as a binary variable, Yes/No - and its date.
- Performance score using standard European Oncology Research and Treatment group guidelines.
- Quality of life or disability adjusted life years on standard validated questionnaire.
| » Conclusion|| |
It is difficult to measure cancer-specific death rates reliably. Moreover, a therapy may significantly control the disease locally without an impact on the death rate or OS. Hence, an agent or intervention should be chosen in the treatment of cancer if it improves the quality of life together with reduction in LR even if it does not significantly increase OS.
| » References|| |
|1.||Baum M , Schipper H. Fast Facts: Breast Cancer. In: "Management of Primary Cancer" Chapter 5. 2 nd ed. Health Press; Oxford: 2002. p. 61-86. |
|2.||Petralia GA, Lemoine NR, Kakkar AK. Clinical Trials Centre for Surgical Sciences, Barts and the London School of Medicine, UK. Mechanisms of disease: the impact of antithrombotic therapy in cancer patients. Nat Clin Pract Oncol 2005;2:356-63. |
|3.||A.C. Bhaktivedanta Swami Prabhupada. Bhagavad-Gita - As it is Chapter 10 verse 34. Bhaktivedanta Book Trust (Pub) Revised edition (March 1, 1997). ISBN-10: 0892131233 |
|4.||Mittra I. Nature Clinical Practice Oncology. Editorial April 2007;4:203. |
|5.||Rothman KJ, Greenland S. In: Modern Epidemiology. 2 nd ed. Causation and Causal Inference. Philadelphia: Lippincott Raven Publishers; 1998. p. 7-28. |
|6.||Saphner T, Tormey DC, Gray R. Annual Hazard Rates of recurrence for Breast Cancer after Primary therapy. J Clin Oncol 1996;14:2738-46. [PUBMED] [FULLTEXT]|
|7.||Mant J, Wilson S, Parry J, Bridge P, Wilson R, Murdoch W, et al . Clinicians did not reliably distinguish between different causes of cardiac death using case histories. J Clin Epidemiol 2006;59:862-7. [PUBMED] [FULLTEXT]|
|8.||Levison DA. Autopsy teaching: a dying art. Editorial, Medical Teacher, 2004:26(4):293-4. O'Grady G. The breakfast club: case study of a teaching-autopsy curriculum. Medical Teacher, 2004;26:377-8. |
|9.||Clark MJ. Autopsy. Lancet 2005;366:1767. |
|10.||Murray C, Lopez A. Alternative projections of mortality and disability by cause 1990-2020: Global Burden of Disease Study. Lancet 1997;349:1498-504. |
|11.||Forrest P. Breast Cancer: The Decision to screen. London: Nuffield Provincial Hospitals Trust (Pub); 1990. |
|12.||Mansel RE, Fallowfield L, Kissin M, Goyal A, Newcombe RG, Dixon JM, et al . Randomized multicenter trial of sentinel node biopsy versus standard axillary treatment in operable breast cancer: The ALMANAC trial. J Natl Cancer Inst 2006;98:599-609. [PUBMED] [FULLTEXT]|
|13.||Susser M. What is a cause and how do we know one? A grammar for pragmatic epidemiology. Am J Epidemiol 1991;133:635-48. |
|14.||ATAC Trialists' Group. Results of the ATAC (Arimedex, Tamoxifen, Alone or in Combination) trial after completion of 5 years' adjuvant treatment for breast cancer. Lancet 2005;365:60-2. [PUBMED] [FULLTEXT]|
|15.||Rosenberg SA, Tepper J, Glatstein E, Costa J, Baker A, Brennan M, et al . The treatment of soft-tissue sarcoma of the extremities: Prospective randomized evaluation of (1) limb-sparing surgery plus radiation therapy compared with amputation and (2) the role of adjuvant chemotherapy. Ann Surg 1982;196:305-15. [PUBMED] [FULLTEXT]|
|16.||Brennan MF, Hilaris B, Shiu MH, Lane J, Magill G, Friedrich C , et al . Local recurrence in adult soft tissue sarcoma. A randomized trial of brachytherapy. Arch Surg 1987;122:1289-93. |
|This article has been cited by|
||Most Breast Cancer Screening Trials Have a Flawed Design
| ||Nishant Gurnani, Anurag Srivastava |
| ||Indian Journal of Surgery. 2011; |
|[VIEW] | [DOI]|