Beyond Guidelines: Conventional Wisdom About Strict Adherence Can Be Misleading

Maurie Markman, MD

Editor-in-Chief of OncologyLive

Senior vice president for Clinical Affairs and National Director for Medical Oncology Cancer Treatment Centers of America, Eastern Regional Medical Center

In the classic Hans Christian Andersen children’s story, The Emperor’s New Clothes, two tailors fool the emperor into believing that they have made wonderful new clothes for him when in fact they have created nothing. When he goes outside wearing his imaginary outfit, his subjects are unwilling or unable to acknowledge the truth that the emperor is wearing no clothes, but a little child⎯ unaffected by what the adults all around him fail or refuse to see and oblivious to conventional wisdom on the subject of the monarch⎯pronounces: “But he has nothing on at all!”

Unfortunately, in the opinion of this commentator, the message of this wonderful children’s story can be applied to an important misconception in certain epidemiologic studies in the cancer arena related to conclusions drawn from possible relationships between two independent clinical events. The problem arises when either a study investigator or individuals not involved with the particular study (eg, lay media) confuse what is observed to be a possible association between the events as actually representing a clear direct cause-and-effect relationship between one of the events and the other.

Consider, for a moment, a recent editorial in The New York Times¹ that discussed an abstract presented at the 2013 annual meeting of the Society of Gynecologic Oncology dealing with a population-based study on the management of epithelial ovarian cancer.² The editorial noted that the study found “only 37% (of patients) received the care recommended in guidelines set by the National Comprehensive Cancer Network [NCCN].” It further commented that “this kind of failure is not uncommon in American medicine,” and “in the case of ovarian cancer, the consequences of inadequate care are tragic.”¹

While it is impossible to disagree with this general sentiment regarding the importance of adequate care for all cancer patients, including women with ovarian cancer, and to emphasize the critical relevance of efforts undertaken at the national, regional, and local hospital levels to optimize the quality of cancer care, the question to be addressed in this commentary is what specifically in this particular research project provides solid evidence to support the conclusion that the percentage of patients who did not receive care recommended in NCCN guidelines actually represents “inadequate care.”

In fact, a competent surgical or medical physician caring for a woman with ovarian cancer who utilizes her/ his clinical skills and experiences (the process of clinical judgment) may quite appropriately decide that⎯based on the particular patient’s medical status including her physiologic age, relevant and serious comorbidities, or the extent of her disease⎯it would not be in that individual’s best interest for her to undergo guideline-based therapy. The guidelines call for an attempt at aggressive/ maximal surgical cytoreduction followed by combination platinum-based cytotoxic chemotherapy.

Guideline Adherence and Survival in Epithelial Ovarian Cancer

Assertion

Patients receiving guideline-based therapy experience superior survival compared with patients not treated based on guidelines.

Cause-and-effect explanation

Association explanation

Treatment based on guidelines is the reason for the favorable outcome.

The decision to not employ guidelines is largely based on clinical judgment that particular patients are not appropriate for that strategy.

Role of Comorbidities

Further, no one with an understanding of clinical oncology would question the conclusion that a group of cancer patients with comorbidities would be predicted to have a statistically defined, population-based shorter survival compared with patients with a similar extent of disease (eg, stage, overall tumor volume) but without these comorbidities. Strong support for this statement comes from a recent report involving more than 3000 women with ovarian cancer that found the presence of comorbidity versus no comorbidity was a highly relevant clinical factor for survival in the malignancy (hazard ratio = 3.31), independent of other established features such as stage, histology, performance status, and age.³

In addition, the presence of truly extensive disease, a term that would encompass the magnitude of the tumor burden beyond the often overly simplistic concept of stage solely defined by tumor location, may influence a clinician’s decision to employ a particular strategy such as aggressive surgical cytoreduction. Thus, for example, a patient with stage 3C ovarian cancer (“Patient A”) with several 2-cm to 3-cm masses may be an excellent candidate for this surgery while a different patient with stage 3C ovarian cancer (“Patient B”) with extensive diffuse intra-abdominal carcinomatosis may be better managed by an alternative approach. It is not difficult to answer the question of which of these two stage 3C patients would be predicted to have a superior outcome, statistically speaking.

But what role has the aggressive surgery actually played in the superior outcome of Patient A compared with Patient B versus the unfavorable biological characteristics of their respective cancers? This biology includes critically relevant but also poorly understood features such as the rate of tumor growth and pattern of progression (eg, local vs widespread diffuse infiltrative or systemic), as well as the extent of inherent drug resistance and rapidity of the development of acquired resistance.

How Guidelines Were Formed

Thus, while it is certainly possible that adherence to guideline-based therapy in ovarian cancer may favorably impact outcome, at least for a subset of patients, the population-based finding of an association between such adherence and survival absolutely should not be considered sufficient evidence to conclude that this outcome resulted from the use, or lack of use, of particular management strategies. In the absence of solid evidence- based phase III randomized controlled trial data, the reported association between guideline adherence and survival may occur as much from the observation that patients selected to undergo such treatment, based on the clinical judgment of their treating physicians, have superior baseline clinical and biological features compared with those individuals where the decision is made to not employ this management strategy.Further, it is relevant to acknowledge here that the development of the guidelines for the primary therapy of epithelial ovarian cancer, which call for an initial attempt at maximal surgical cytoreduction prior to the administration of cytotoxic chemotherapy, essentially are based exclusively on both retrospective data and the opinions of specialists in this area rather than on the results of information generated from prospective randomized phase III clinical trials.

In fact, employing the analysis noted above, an equally rational argument (also based on less-thanoptimal evidence) can be advanced that while multiple retrospective reviews have revealed a strong association between survival and aggressive surgical resection that leaves the patient with very small volume or no macroscopic residual disease, the ability to achieve this clinical state may really be little more than a biomarker of the presence of a more favorable cancer biology rather than the cause of the superior clinical outcome.

For example, the most infiltrative (carcinomatosis) and locally extensive (multiple lymph nodes, diaphragmatic involvement) ovarian malignancies are less likely to be completely cytoreduced to a microscopic residual state and more likely to result in an inferior outcome. It is possible the features that ultimately define the ability of an excellent gynecologic cancer surgeon to perform an optimal resection are similar or essentially identical to the previously noted biological factors that control tumor growth and spread, as well as drug resistance.

Keeping Judgment in Equation

One additional point can be made to seriously challenge the conventional wisdom regarding primary surgery in advanced epithelial ovarian cancer. In contrast to the complete absence of solid evidence-based data to support aggressive initial surgery, a landmark phase III study has revealed the therapeutic equivalence of primary chemotherapy (neoadjuvant chemotherapy) followed by surgery and additional cytotoxic therapy, compared with aggressive surgery followed by chemotherapy, with the neoadjuvant approach being found to result in less treatment-related morbidity and mortality.⁴Despite the absence of solid clinical evidence, it is reasonable to inquire if there is really anything wrong with simply accepting the argument that guideline adherence is the best approach for disease management. While for most patients it is reasonable to acknowledge that an attempt at maximal surgical cytoreduction can be undertaken and completed both effectively and safely, it is quite possible that strict adherence to such guidelines or the use of the guidelines to establish a scorecard to define clinical excellence might lead physicians to ignore their own often considerable experience and clinical judgment in defining optimal care on behalf of their patients, and proceed to perform a procedure that under other circumstances they might consider unwise. And can this be considered a positive outcome?

Finally, we return to the story of the emperor’s new clothes and ask the question: What is the real evidence that guideline adherence is the cause of the superior outcome noted in the Society of Gynecologic Oncology abstract, with the opposite outcome being predicted in the absence of guideline adherence? Further, what are the evidence-based data to even support the current guidelines that require in most patients an aggressive attempt at primary surgical cytoreduction?

Is it perhaps time to inquire of the emperor what exactly he is wearing?

References

1. Inadequate treatment of ovarian cancer [editorial]. New York Times. March 14, 2013: A34.
2. Bristow R, Chang J, Ziogas A, et al. NCCN treatment guidelines for ovarian cancer: a population-based validation study of structural and process quality measures. In: Proceedings from the 44th Annual Meeting of the Society of Gynecologic Oncology; March 9-12, 2013; Los Angeles, CA. Abstract 45.
3. Sperling C, Noer MC, Christensen IJ, et al. Comorbidity is an independent prognostic factor for the survival of ovarian cancer: a Danish register-based cohort study from a clinical database [published online ahead of print January 3, 2013]. Gynecol Oncol. 2013;129(1):97-102. doi: 10.1016/j.ygyno.2012.12.039.
4. Vergote I, Tropé CG, Amant F, et al. Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer. N Engl J Med. 2010;363(10):943-953.