Sequential trials with randomization

Clinical trials, with almost no exception, are longitudinal (Chow and Liu, 2004). This means that data are accumulated sequentially over time. From the perspective outlined so far in the book, the statistical analysis takes place once the number of subjects stated in the study protocol have been enrolled, randomized, and completed their participation in the trial. This approach can be called the Fixed design or fixed sample design approach. Another design of interest in clinical trials is the group sequential design, in which interim analysis plays a crucial role. 

Table 2.3 shows some broad estimates of the failure to achieve efficacy in patient populations for some common classes of drugs. In light of these kinds of numbers, the best that can be done, for example, to get a particular patient s hypertension under control involves a lot of trial and error. Upon initial treatment, only approximately 40 to 50% of patients respond sufficiently to a randomized choice of a single therapeutic drug. However, this can be increased by the use of sequential monotherapy, with the subsequent addition of each major drug class or rotation through the various alternatives until the most effective single therapeutic or combination is identified. ... 

A trial with 240 one-day-old (Ross) chicken was carried out at the Trakya University, Turkey (Liickstadt et al., 2004) where an acid blend was compared against a negative eontrol. Each of the two applications was repeated 12 times, with 10 chickens per replicate in a randomly selected design. The trial period was Ifom 1-35 days of age. The acidifier treatment (a combination of formic and propionic acid and their salts, based on an inorganie sequential release medium) was added at a dosage rate of 3 kg/t feed. The trial feed (based on com and full fat soybean) was applied as starter (0-14 days), grower (15-28 days) and finisher phases (29-35 days). Feed and water was available ad libitum. 

Hydroxamic acid derivatives, which belong to a new class of NO donors, have been shown to inhibit the matrix metalloproteinases (MMPs) [112]. MMPs are a family of zinc-dependent endopeptidases, which play a critical role in multiple steps in the metastatic cascade and facilitate neoangiogenesis. Numerous hydroxamic acids, such as marimastat, have been developed, that bind the zinc atom in the active catalytic domain of MMPs. During a randomized Phase III trial, comparing marimastat with placebo in patients with metastatic breast cancer, marimastat was not associated with an improvement in progression-free survival or overall survival. Other studies also indicated no benefit for MMP inhibitors when used either in combination with chemotherapy or sequentially after first-line chemotherapy in a variety of cancers [113]. Currently, many pharmaceutical companies have suspended clinical development of this kind of agent. 

A German study randomized 98 patients with unresectable hypopharyngeal head and neck carcinoma to sequential or concomitant chemoradiotherapy (64). Patients on the sequential arm received two courses of cisplatin (25 mg/m2 for 5 d) and continuous infusion 5-FU (750 mg/m2 for 5 d) followed by G-CSF for 6 d. The second cycle was repeated on d 14 and then followed by standard radiotherapy. Identical dosages of chemoradiotherapy were used in the concomitant arm but incorporated a 21-d interval between each chemotherapy cycle and G-CSF support. Sequential treatment resulted in a CR of 49% in comparison to the concomitant arm of 57%. At 2-yr, the median survival was improved in the concomitant arm (53% vs 33%). However, one-third of the initial patients enrolled were withdrawn from study due to medical or socioeconomic problems whether this resulted in a significant disparity between the two different treatment arms was not noted. Mature data from this trial have not been published to date. 

A Japanese trial compared sequential delivery of chemotherapy and radiation therapy to concurrent delivery of chemotherapy and radiation (47). Patients were randomized to receive concurrent hyperfractionated radiation therapy (d 2 of cycle 1 of chemotherapy) or to sequential chemotherapy followed after the fourth cycle by hyperfractionated radiation therapy. The radiation dose was45 Gygivenin 1.5 Gy fractions twice daily for a total of 30 fractions in 3 wk. The chemotherapy given was cisplatin and etoposide. The median survival for the concurrent schedule was 29 mo and for the sequential schedule was 19 mo. The 2-yr survival was 50% for the concurrent therapy and 40% for the sequential therapy (47). These results favored concurrent therapy and are the best results to date for patients with LD-SCLC. 

Unfortunately, none of the seven randomized trials that have compared radiation therapy alone vs neoadjuvant cisplatin-containing chemotherapy plus radiation therapy demonstrated an improvement in overall or disease-free survival with combined-modality therapy (Table 2). Two studies actually demonstrated poorer survival with neoadjuvant chemotherapy. Souhami et al. (15) reported a significantly poorer survival rate with neoadjuvant chemotherapy in a small trial of patients with stage IIIB disease. This outcome was partly due to increased toxicity and poor compliance in patients who received chemotherapy. Another trial of neoadjuvant epirubicin and cisplatin was closed early when interim analysis revealed a significantly higher recurrence rate in the chemotherapy arm (16). These trials fail to provide any evidence that sequential cisplatin-containing chemotherapy and radiation therapy are of benefit. Possible explanations for the disappointing results include the effects of chemotoxicity, altered compliance, and possible accelerated repopulation of resistant clones after neoadjuvant chemotherapy. 

There is a tendency among control and statistics theorists to refer to trial and error as one-variable-at-a-time (OVAT). The results are often treated as if only one variable were controlled at a time. The usual trial, however, involves variation in more than one controlled variable and almost always includes uncontrolled variations. The trial-and-error method is fortunately seldom a random process. The starting cycle is usually based on manufacturers specifications or experience with a similar process and/or material. Trial variations on the starting cycle are then made, sequentially or in parallel, until an acceptable cycle is found or until funds and/or time run out. The best cycle found, in terms of one or a combination of product qualities, is then selected. Because no process can be repeated exactly in all cases, good cure cycles include some flexibility, called a process window, based on equipment limitations and/or experience. 

Hodsman AB, Fraher LJ, Watson PH, Ostbye T, Stitt LW, Adachi JD, Taves DH, Drost D. A randomized controlled trial to compare the efficacy of cyclical parathyroid hormone versus cyclical parathyroid hormone and sequential calcitonin to improve bone mass in postmenopausal women with osteoporosis. J Clin Endocrinol Metab 1997 82(2) 620-8. 

In 65 patients with ischemic stroke treated for 6 weeks with intramuscular Sygen (monosialoganglioside 40 mg/day) in a double-blind, placebo-controlled study, there were no significant differences between the groups (22). A subsequent double-blind, sequential, multicenter, randomized, placebo-controlled trial of two doses of Sygen did not provide convincing evidence of efficacy (23). 

Various companies and investigational groups continue to examine the relative efficacy and safety of different forms of combined postmenopausal treatment. In a randomized, placebo-controlled trial 579 women were treated for 26 cycles with sequential combinations of 17-beta-estradiol 1 mg plus dydrogesterone 5 or 10 mg or 17-beta-estradiol 2 mg with dydrogesterone 10 or 20 mg (27). The effects of these treatments in the 442 women who underwent biopsy were considered satisfactory in terms of... 

Finally, we should note that molecules cannot cover a surface completely. As a first approximation, the random site model [37] assumes that they can be represented by impenetrable spheres, adsorbing sequentially on a uniform surface, in locations selected randomly (see Figure 3.5). If the molecule adsorbing on a new trial position overlaps with a previously adsorbed molecule, it does not stick to the surface otherwise, it is adsorbed. An initially fresh surface covers quickly. 

Fornier, M. N., Seidman, A. D., Theodoulou, M., Moynahan, M. E., Currie, V., Moasser, M., Sklarin, N., Gilewski, T., D Andrea, G., Salvaggio, R., Panageas, K. S., Norton, L. and Hudis, C., Doxorubicin followed by sequential paclitaxel and cyclophos-phamide versus concurrent paclitaxel and cyclophosphamide 5-year results of a phase II randomized trial of adjuvant dose-dense chemotherapy for women with node-positive breast carcinoma, Clin. Cancer Res., 2001(12), 3934, 2001. 

---