The randomized, placebo-controlled trial has become the gold standard of medical research today. This development has been accompanied by an ever-increasing emphasis on evidence-based medicine in curricula and guidelines, increasing regulation of pharmaceutical research, and a concomitant effort to produce study designs that are unreservedly accepted by regulatory agencies. Other study designs run the risk of being considered inferior or passed over. At the 28th Swiss Annual Conference on Phytotherapy in Baden, the relevance of cohort studies offering extended insights into the efficacy of mistletoe was emphasized.
According to Dr. rer. nat. Marcus Reif, Berlin, while evidence-based medicine (EbM) is defined as the conscientious, explicit, and judicious use of current best evidence. However, this wording does not imply that randomized controlled trials (RCTs) always provide the current best evidence. For many clinical trials, an RCT is not feasible or is only feasible under ethically unacceptable circumstances. Blinding may also present technically intractable problems, or randomization, which implies a tie between two treatment alternatives, may not be possible because the therapy under investigation is already well established in everyday use. Experience has shown that the willingness of patients to participate in an RCT is very low, especially for such forms of therapy.
Cohort studies as an alternative?
If the study design is equivalent to an RCT, but without the two elements of randomization and placebo administration, they are referred to as treatment or cohort studies. They can be retroactive or prospective in nature. A so-called “retrolective” design is also possible: First, a detailed study protocol similar to an RCT is designed, but the time of enrollment is then shifted into the past based on medical record data, i.e., the patients can be represented in the study exclusively by retrospective data. At least in theory, they can be further documented prospectively. Furthermore, this type of study uses randomization of specific clinics (rather than patients) as a random element. Of course, it remains non-interventional.
The interpretation of the observed differences is problematic, because they do not necessarily have to be due to the therapy itself, but can also be due to other differences in the initial situation of the different therapy groups.
Bias factors can be equally disruptive. Although they also occur in RCTs, they are more pronounced in nonrandomized studies and therefore require urgent attention. The selection bias describes the different allocation, the information bias arises from the different precision of data collection, the attrition bias is due to the different patient presence, and the performance bias includes the treatment differences beyond the actual therapy. If these factors are documented and taken into account in the statistical evaluation, e.g. with multiple regression models, inaccuracies resulting from this can be avoided.
“What is difficult to circumvent in non-randomized studies are the unknown but study-relevant and thus -biased factors. This argument is made again and again when it comes to giving preference to RCTs. Here, too, however, there are methods, primarily developed for economic and social science issues, which allow a more valid assessment of efficacy,” says Dr. Reif.
Retrolective cohort studies on mistletoe therapy
Up to 80% of all tumor patients use Iscador® or other mistletoe extracts at least once during the course of their disease, and it was not uncommon for them to have to ask their doctor about it themselves. In vitro, the anti-tumor effects of mistletoe (inhibition of cell division, induction of apoptosis, activation of immunocompetent cells, selective entrapment of chemotherapeutic agents within cancer cells) are well documented. The clinical effect of mistletoe therapy continues to cause controversy [1, 2].
Two retrolective cohort studies on the efficacy and safety of the mistletoe preparation Iscador® (ISC) were published in 2009 and 2010. They evaluated medical records of colorectal or pancreatic tumor patients from 1993 to 2002 [3, 4]. In both studies, the treatment groups differed not only in the targeted points of comparison (ISC included or not in the treatment regimen) but also in other important prognostic factors, such as age, BMI, presence of risk factors. The data were therefore analyzed using multivariate statistical models.
In the colorectal cancer study, 804 non-metastatic patients were followed up for a median time of 58 and 51 months, respectively; the median duration of mistletoe therapy was 51 months. Slightly more than half of all patients from both groups received adjuvant chemotherapy, and 17% received radiotherapy.
The pancreatic tumor study examined 396 patients over a median period of 15 and 10 months, respectively, and the median mistletoe therapy also lasted 15 months. Approximately 58% of all patients in both groups received chemotherapy, and 11.4% received radiotherapy.
Both studies showed significantly better findings for the defined target parameters (including disease-free and overall survival) under Iscador® therapy than in the control group.
Source: “Cohort studies as a basis for extended knowledge of the efficacy of herbal preparations using mistletoe as an example”, 28th Swiss Annual Conference on Phytotherapy, November 21, 2013, Baden.
Literature:
- Kienle GS, Kiene H: Complementary cancer therapy: a systematic review of prospective clinical trials on anthroposophic mistletoe extracts. Eur J Med Res 2007 Mar 26; 12(3): 103-119.
- Horneber MA, et al: Mistletoe therapy in oncology. Cochrane Database Syst Rev 2008 Apr 16; (2): CD003297. doi: 10.1002/14651858.CD003297.pub2.
- Friedel WE, et al: Systematic evaluation of the clinical effects of supportive mistletoe treatment within chemo- and/or radiotherapy protocols and long-term mistletoe application in nonmetastatic colorectal carcinoma: multicenter, controlled, observational cohort study. J Soc Integr Oncol 2009 Fall; 7(4): 137-145.
- Matthes H, et al.: Molecular mistletoe therapy: friend or foe in established anti-tumor protocols? A multicenter, controlled, retrospective pharmaco-epidemiological study in pancreatic cancer. Curr Mol Med 2010 Jun; 10(4): 430-439.
InFo Oncology & Hematology 2014; (2)1
