Clinical Trials for Rare Lung Diseases: Lessons from Lymphangioleiomyomatosis

Abstract

Lymphangioleiomyomatosis (LAM) is a rare, slowly progressive neoplasm that causes gradual but often life-threatening cystic destruction of the lung. Advances in our understanding of the molecular and cellular pathogenesis have LAM have identified a number of promising targets for testing in therapeutic trials. However, the design, prioritization, organization, and implementation of clinical trials in rare lung diseases poses unique challenges, including geographically disperse populations, sluggish enrollment, off- label drug use, burdensome regulations, and paucity of validated surrogate endpoints.

Overview of Interventional Trials in LAM

Lymphangioleiomyomatosis (LAM) is a debilitating and frequently fatal cystic lung neoplasm of women that is associated with mutations in tuberous sclerosis complex (TSC) genes.¹ LAM affects approximately 30%–40% of female patients with TSC, and a minimum of 2 people per million who do not have TSC, totaling at least 250,000–300,000 affected subjects worldwide. There is no effective treatment for LAM, but a decade of effort from the National Heart Lung and Blood Institute (NHLBI) LAM Registry and Intramural LAM protocol has defined the natural history of the disease and set the stage for trials.² A large body of scientific knowledge gathered over the past 8 years has identified a plethora of molecular targets and promising therapies for LAM, many of which are FDA-approved for other indications and ready for testing (Table 1, Fig. 1).³ These target/drug combinations include, but are not limited to, mTOR(mammalian target of rapamycin)/sirolimus, rheb/atorvastatin, tyrosine kinase receptors/imatinib, estrogen receptor/afimoxifine, aromatase/letrozole, and metalloproteinases/doxycycline (Table 1). Potent inhibitors of mTOR itself and other kinases in the mTOR pathway, such as Akt, PI3K, Rheb, and S6K, are underdevelopment for other diseases and will likely become available over time. The first phase II trial of mTOR inhibition in TSC and LAM was recently published⁴ and a large randomized controlled trial (RCT) of mTOR inhibition in LAM is underway. The latter study, called MILES (Multicenter International LAM Efficacy of Sirolimus Trial), cost 4 million dollars, required the development of an international consortium comprised of 13 sites, and will span over 5 years from start to finish. It has become clear to the LAM scientific and patient communities that building and rebuilding networks for RCTs designed to test each of the many promising therapies for LAM is not an efficient strategy. Trial designs must be tailored to take maximal advantage of the limited number of patients available, and to optimize access to trials for patients at all stages of disease (see discussion below).

FIG. 1.

Signaling pathways and targets for LAM. Binding of ligands to cell surface receptors such as the insulin receptor and platelet-derived growth factor receptor (PDGF) result in phosphorylation of downstream proteins (e.g., IRS 1, PI3K, PDK1 downstream of IRS 1) and ultimately Akt. The estrogen receptor (ER) can also signal through Akt. Akt activates the TSC1/TSC2 complex, which in turn downregulates the activity of mTOR, the mammalian target of rapamycin, through an intermediate protein called Rheb. In the absence of TSC1 or TSC2, the mTORC1 complex (composed of mTOR and raptor and msLT8 (not shown)) is constitutively activated, resulting in inappropriate ribosome biogenesis, protein translation, and cell growth (through downstream effectors S6K and S6, as well as 4EBP1 and eIF4E), and downregulation of autophagy. mTORC2 (composed of mTOR, msLT8, and rictor), which controls cytoskeletal arrangement and cell movement through downstream effectors ROCK and Rho, is also activated in the absence of TSC1 or TSC2. Various steps in this signaling pathway can be blocked by currently available agents, including selective estrogen antagonists, tyrosine kinase inhibitors, angiogenesis inhibitors, lymphangiogenesis inhibitors, farnesyl transferase inhibitors, statins, mTOR inhibitors, metalloproteinase inhibitors, and Rho inhibitors.

Table 1.

Putative Trial Targets and Agents in LAM

Akt pathway agents

PI3K inhibitors

mTOR inhibitors

sirolimus

everolimus

Rheb (palmitoylation) inhibitors

statins

farnesyl transferase inhibitors

S6 kinase inhibitors

Rho pathway agents

PDGF pathway inhibitors

tyrosine kinase inhibitors

Gleevec

Metalloproteinase inhibitors

batimistat

marimastat

doxycycline

Vascular therapies

angiogenesis inhibitors

bevacizumab

lymphangiogenesis inhibitors

anti-VEGF-D antibody

sorafenib

sunitinib

Hormonal therapies

progestins

selective estrogen antagonists

aromatase inhibitors

GnRh agonists

Cyclin-dependent kinase inhibitors

roscovitine

Cytokine therapies

interferon β

interferon α

Rationale for choosing molecular and cellular targets in LAM

A growing body of evidence indicates that LAM is caused by mutations in either of the known tuberous sclerosis genes, TSC1 or TSC2, which have been shown to cause TSC (Fig. 1).⁵ TSC1 and TSC2 encode hamartin and tuberin, respectively, which control cell growth, survival, and motility by forming a complex that regulates the activity of the mTOR (mammalian target of rapamycin) signaling pathway.^6–8 The mTOR serine/threonine kinase is the catalytically active subunit of two functionally distinct protein complexes, mTORC1 and mTORC2. mTOR is a 'master switch' that integrates input from cellular systems that report on the status and environment of the cell, including signals arriving from cell surface growth factor receptors, oxygen sensors, nutrient sensors, hormones, and energy sensors. In the absence of hamartin or tuberin, mTORC1 is constitutively activated, and through downstream effectors pS6 and 4EBP1 drives inappropriate protein translation, cellular proliferation, migration, and invasion. Progression to frank malignancy is prevented by a pS6 feedback loop which inhibits the activity of IRS-1 and other upstream elements.^9,10 Everolimus and sirolimus are two FDA-approved macrolides currently being tested in TSC and LAM trials (see below) which form a complex with a protein called FKBP12 and allosterically inhibit mTOR.

Some mTOR driven functions are rapamycin-insensitive, suggesting that combination therapies may be required for maximal effect. Direct kinase inhibitors such as Torin are more potent than the currently available mTOR inhibitors and are promising candidates for future trials.^11,12 Inhibitors of many of the other signaling elements in the Akt pathway including PI3K, Akt, Rheb, and S6 kinase, are being developed for cancer and other indications. Estrogen is suspected to play a role in LAM disease initiation and/or progression, based on the profound gender restriction in LAM (although men also have significant amounts of circulating estrogen). Recent evidence suggests that estrogen can signal through Akt to release the hamartin- or tuberin-deficient cell from pS6-mediated feedback inhibition, and through MEK-dependent pathways to promote pulmonary metastasis by enhancing the survival of detached, tuberin-null cells.¹³ Aromatase is an enzyme that can synthesize estrogen by converting testosterone to estradiol and androsteindione to estrone, and is the primary source of estrogen in postmenopausal women and in those with pharmacologically suppressed ovarian function. Multiple therapeutic strategies for the treatment of LAM are currently being considered based on the inhibition of estrogen production or estrogen action, with GnRh agonists, selective estrogen antagonists and aromatase inhibitors. Matrix metalloproteinase (MMP) imbalance involving MMP-2, MMP-9, and TIMP-3 has been described in LAM lesions and may contribute to cyst formation.^14–16 A direct causal link for MMPs in the pathogenesis of lung destruction has not been established, however. In addition, trials of metalloproteinase inhibitors in cancer have shown no clinical efficacy, most likely because the MMPs are most important in the earliest stages of cancer pathogenesis such as early invasion and micrometastasis.¹⁷ Two trials of doxycycline, a tetracycline family antibiotic with putative anti-MMP and anti-angiogenic activity, are currently underway in the United Kingdom and Australia. Genetic analysis of recurrent LAM lesions in the donor allografts of LAM patients who have undergone lung transplantation have revealed that LAM can metastasize; suggesting a novel mechanism for the disease.^18,19 Potential strategies for intervention based on early resection of the primary tumor have been proposed, but the source of metastasis in LAM is controversial.²⁰ Two leading candidates have emerged, angiomyolipomas that are present in about 30% of sporadic LAM patients and 80% of TSC-LAM patients, and the uterus, which has been reported to harbor LAM lesions in multiple case reports.²¹ Consistent with the metastatic theory, cells containing LOH for TSC genes have been isolated from the blood of LAM patients.²² There is also compelling evidence for lymphatic dissemination in LAM, based on clinical and autopsy observations.^23–25 Kumasaka et al. found that LAM cell clusters enveloped by lymphatic endothelial cells are abundant in the lymphatic lumen and populate chylous fluids and axial and supraclavicular lymph nodes in patients with LAM.²⁵ Abundant expression of lymphatic endothelial markers such as podoplanin, VEGFR-3, and VEGF-C within LAM lesions implicates inappropriate lymphangiogenesis in the pathogenesis of LAM.²⁶ VEGF-D, the ligand for VEGFR-3, is abundantly produced by LAM cells, and is elevated over three-fold in the serum of patients with LAM.^27,28 In essence, LAM cells have adopted a strategy for spread which involves generating their own system of 'escape tunnels' through induction of lymphatic development. It is also quite likely that the cystic lung destruction that occurs in LAM is the result of relentless lymphatic remodeling by infiltrating LAM cells. Anti-lymphangiogenic approaches to LAM therapy, with VEGF-D antibodies or VEGFR-2/VEGFR-3 inhibitors, are under consideration. The unregulated growth, lymphatic or vascular spread, and tissue destruction²⁰ that are described above are entirely consistent with classification of LAM as a cancer. The benign histological appearance of LAM cells and very slow progression of the disease certainly place LAM at the far left in the neoplastic spectrum of severity, but the gradual recognition that LAM is a precious model of metastatic cancer due to a single gene mutation has begun to attract interest from the National Cancer Institute and other federal agencies.

Combined use of the cancer and signaling models in designing trials

The explosion in our understanding of the pathogenesis of LAM has provided a robust framework for identifying molecular targets for trials (Fig. 2). Opportunities for intervention suggested by adoption of the cancer paradigm include each of the putative stages of LAM pathogenesis of 1) somatic mutation, 2) proliferation, 3) lymphangiogenesis, 4) dissemination, 5) implantation, 6) infiltration, and 7) angiogenesis and remodeling. Knowledge of the signaling pathways that are deranged in LAM and their role in each of the cellular processes of growth, motility, angiogenesis/lymphangiogenesis, apoptosis, anoikis, and differentiation provides the tools to target each of the seven pathogenic stages above. The cancer model and the signaling models are complementary and should both be considered in choosing endpoints and designing trials.

FIG. 2.

The MILES Trial is being conducted by the NCRR/NIH supported Rare Lung Disease Consortium. The Rare Lung Disease Consortium (RLDC) is composed of the primary site at the University of Cincinnati (co-Directors Bruce Trapnell, M.D. and Frank McCormack, M.D.), 9 other domestic sites, 3 international sites, and a Data and Technology Coordinating Center at the University of South Florida School of Medicine (Director, Jeff Kirscher, M.D.)

For instance, mTOR inhibitors target multiple pathogenic steps including proliferation, motility/dissemination,²⁹ angio/lymphangiogenesis,^31,32 and anoikis/apoptosis,³² and might be reasonable candidates for trials in which improvement in lung function (through tumor apoptosis and reduction of pulmonary interstitial infiltration) is the primary endpoint. On the other hand, a trial based on inhibition of metastasis by blocking MEK pathways may fail if the primary endpoint is lung function measurement, since a lung that is already extensively infiltrated with LAM cells may continue to deteriorate even if further metastases are interrupted. Finally, metalloproteinase inhibitors that block further cyst formation but perhaps not pulmonary parenchymal infiltration might slow the decline in lung function, but would not be predicted to result in a measurable improvement in lung function, gas exchange or exercise tolerance.

Choice of endpoints in LAM trials

Functional endpoints such as exercise tolerance, survival, dyspnea, and quality of life, are the most informative and important in lung disease trials since they have indisputable meaning for patients. Unfortunately, for LAM, these are also less practical endpoints for efficient trials. Six-minute walk testing is often not very sensitive to changes in functional status, since many LAM patients are young, otherwise healthy, and able to walk 1500–2000 feet despite severe structural pulmonary disease. On that background, detection of improvement is difficult or impossible. More formal exercise testing is expensive, labor intensive, not easily standardized, and often limited to specialized centers. Survival as an endpoint in LAM is also suboptimal, since the length of trials that would be required to answer questions of drug efficacy in a disease in which the average survival is in the range of 90% at 10 years is prohibitive. Quality of life, sense of dyspnea, and other questionnaire-based endpoints are valuable adjuncts, but are more subjective in nature and prone to bias, especially when making recommendations about potentially toxic therapies.

To circumvent the difficulties with these functional endpoints, surrogate endpoints of lung function measurements, biomarkers or structural and functional imaging may be considered. Unfortunately, none of these surrogate measures have been validated as a reliable predictor of survival or quality of life in LAM. Lung function measurements suffer from the inherent limitation of significant inter-test and intra-subject variability, as well as relative insensitivity to small changes in lung function. Serum biomarkers such as serum VEGF-D are just now being developed.^27,28,33 Computed tomography-based measurement of cyst volume percent as a potential surrogate endpoint was recently reported.³⁴ There is an urgent need to develop serum or imaging based measures of lung and/or whole body LAM cell burden and validation of these and other biomarkers as surrogate endpoints for disease burden and survival.

Networks for LAM trials: The Rare Lung Disease Consortium and LAM Foundation Clinics

The MILES study was conducted under the auspices of the NIH/National Center for Research Resources (NCRR) Rare Lung Disease Consortium (RLDC), comprised of 13 sites distributed around the United States, in Canada and Japan, and a Data and Technology Coordinating Center (DTCC) in Florida³⁵ (Fig. 3). The concept behind the RLDC was that networks built for the purpose of trials in one rare disease could be used for multiple rare diseases and serial trials. The RLDC executed trials in LAM, alpha one antitrypsin deficiency, and pulmonary alveolar proteinosis, and has remained intact even as the RLDC funding from the NCRR has ended. Another novel, durable mechanism for conducting trials in rare diseases is to integrate clinical care and research at distributed centers. The LAM Foundation has recently opened LAM Clinics across the United States (Fig. 4). The objectives are to improve clinical outcomes in LAM and to organize the 800 North American LAM patients registered with the Foundation in a manner that facilitates the conduct of trials. LAM clinics are designed to provide state of the art, multidisciplinary care to LAM patients close to their homes. They also serve as incubators for expertise, in which pulmonary physicians, thoracic surgeons, nephrologists, interventional radiologists, and gynecologists become increasing familiar with LAM over time as patients accrue to their Clinic. LAM Clinics that are already open include University of Cincinnati, Columbia University, University of Pennsylvania, University of California Los Angeles, University of California San Francisco, Stanford, University of Michigan, Mayo Clinic, Cleveland Clinic, Tufts, and Medical University of South Carolina. A subset of LAM Clinics that have the interest, resources, and expertise to conduct clinical research, in conjunction with Dr. Joel Moss's program at the NHLBI, form the “LAM Clinic Research Network” (LCRN). The LCRN is focused to maintaining a population of LAM patients ‘at the ready’ for clinical trials, by gathering quality-controlled clinical and research data in a uniform fashion. In this model, patients see LAM experts in the clinics at defined intervals, perhaps every 6 months, for clinical care. During those visits, studies that are part of both clinical care and research trials, including pulmonary function tests and 6-minute walk distance, are obtained in a quality-controlled manner. When a trial becomes available, each patient enrolls with running baseline that can serve as her own control as the intervention is executed. This design lends itself to several efficient and rapid trial designs, including 'N of 1' trials discussed below. Over time, research funding from federal and nonfederal sources will be used for ongoing support of research coordinators, trial visits, and data management. A distributed network for conducting LAM trials has many advantages over the centralized ‘center of excellence’ model, in which patients must be well enough to travel, costs are high, and extent of disruption to work and family life limit enrollment.

FIG. 3.

Choosing candidate trial therapies based on cancer staging targets and signaling targets in LAM. The model of LAM pathogenesis can be divided into six stages, which can be targeted by blocking signaling pathways and cellular processes involved in each step.

FIG. 4.

LAM Foundation Clinics.

International cooperation

In LAM, as with all rare diseases, international cooperation for clinical trials is critical. Patient populations ranging in size from 50 to 250 subjects are organized and ready for trials in the United Kingdom, Australia/New Zealand, Japan, France, Germany, Italy, Spain, Scandinavia, Brazil, and China. These concentrated populations are ideal for conducting phase I/II trials with the multiple candidate therapies that have been proposed for LAM. For phase III multicenter international trials, sponsorship by pharmaceutical companies is important. The burden of international regulations, the nuances of drug import/export, and language barriers make incorporation of international sites into investigator-initiated trials quite difficult.

Rationale for conducting early, staged LAM trials in humans

Given that there are no existing animal models of LAM that mimic the lung disease, in many cases the preferred approach to testing a promising therapy may be to bypass animal experimentation and proceed directly to human trials. This strategy is especially attractive for drug candidates with well-understood safety profiles that are already FDA-approved for other indications. Examples of drugs in this category would include mTOR inhibitors, statins, doxycycline, selective estrogen inhibitors, GnRH agonists, and aromatase inhibitors. Development of biomarkers that are faithful reporters of the burden of LAM cells or the stage of disease will greatly facilitate the conduct of rapid efficient trials that can identify the most promising candidates for larger multicenter trials. This staged approach of rapid, exploratory trials with biomarker endpoints in regionally concentrated populations, followed by multicenter, international, pivotal RCTs with functional endpoints is an attractive strategy for answering meaningful questions and for avoiding overlapping and conflicting trials in this small patient population.

Current LAM Trials

There are currently three categories of interventional trials for LAM underway; a) mTOR inhibitor trials targeting renal angiomyolipomas that include LAM patients, b) mTOR inhibitor treatment trials for LAM, and c) doxycycline trials for LAM.

Angiomyolipoma endpoint trials that include LAM patients

CAST (NCT00457808; completed).The Cincinnati Angiomyolipoma Sirolimus Trial was a phase I/II proof-of-principle trial that was designed to determine if sirolimus could reduce the size of angiomyolipomas.⁴ The trial was funded by the National Cancer Institute, the LAM Foundation, and the Tuberous Sclerosis Alliance. Wyeth (Madison, WI) provided the study drug. Patients with measurable fat-containing renal tumors of 1 cm or greater in size were treated for 1 year with doses of sirolimus that are typically used in renal transplant patients to prevent rejection. In the second year of the trial, renal and pulmonary endpoints were followed off therapy. The primary endpoint was change in angiomyolipoma volume, measured by magnetic resonance imaging (MRI), over the course of 1 year. There were 11 patients with LAM in CAST, and spirometry was performed at regular intervals. After 1 year of receiving the drug, angiomyolipoma volume decreased by almost 50%, and airflow measurements of FEV1 and FVC increased by approximately 10% and 15%, respectively. Although the trial was not placebo controlled, we know from the NHLBI LAM Registry that over this interval FEV1 and FVC would have been predicted to decline by about 3%–5%.³⁶ Surprisingly, once the study article was stopped, angiomyolipoma volumes returned to nearly baseline levels over the course of an observation year off therapy. The beneficial effect of sirolimus on lung function appeared to be more durable, however. Although the rate of decline in FEV1 after treatment ended reverted to the rates described in the natural history study conducted by the National Institutes of Health³⁶ (roughly loss of 75 cc of FEV1 per year) after the treatment period ended, FEV1 and FVC were still significantly above baseline at the end of the observation period. Gas trapping declined while subjects were on the drug, but there was no effect on total lung capacity or diffusing capacity for carbon monoxide. There were a number of adverse events, including six hospitalizations while on therapy, suggesting caution in the use of these agents outside of trials. The CAST trial was not powered to answer questions about the efficacy of sirolimus in LAM.

TESSTAL (NCT00490789; active, not recruiting).TESSTAL is a trial of the efficacy and safety of sirolimus for renal angiomyolipomas in patients with TSC and sporadic LAM.³⁷ The TESSTAL trial is phase II study that was very similar in design to the CAST trial, except that all patients receive sirolimus for a period of 2 years instead of one, and measures of cognitive function are obtained. Enrollment of began in 2005, and the trial is expected to conclude in August 2009. Adult men and women with an angiomyolipoma of at least 2 cm in diameter and either TSC or sporadic LAM are eligible. The primary endpoint is also angiomyolipoma size. In a preliminary report in January 2008, the investigators reported reductions in angiomyolipoma size that were similar to what was seen in CAST.⁴ Lung function measurements were only available for four patients, but did not show any definite trends toward improvement. This study is expected to be published in early 2010.

Sirolimus multicenter angiomyolipoma trial (NCT00 126672; active not recruiting). This phase II multicenter open label trial of sirolimus in patients with TSC or LAM who have angiomyolipomas began in 2005 and closed for enrollment in 2009. The eligibility criteria and endpoints of the trial are very similar to CAST. The primary site is in Boston.

RAD001 angiomyolipoma trial (NCT00457964; active, recruiting). This phase II, open label trial began in 2005 and will complete enrollment in August 2009. To qualify, TSC or LAM patients must be between the ages of 18 and 25, and have an angiomyolipoma of 1 cm or greater in size. The primary endpoint is angiomyolipoma size. Patients using continuous supplemental oxygen were excluded. The structure and endpoints of the trial are very similar to CAST. A long-term follow-up study of participants is underway (NCT00792766).

EXIST-2 RAD001 multicenter trial (NCT00790400; active, recruiting). Efficacy and safety of RAD001 in patients 18 years and over with angiomyolipoma associated with either tuberous sclerosis complex or sporadic lymphangioleiomyomatosis (LAM). This is a multicenter, 2:1 randomized, placebo controlled trial of RAD001 (everolimus) that began enrolling in March 2009 and is expected to conclude in January 2014. To be eligible, patients must be 18 years or older, and have an angiomyolipoma of at least 3 cm in diameter. The primary endpoint is time to angiomyolipoma progression measured by MRI, and secondary endpoints include change in skin lesions, renal function, and biomarkers. Patients who have severe lung function impairment are excluded. The trial is 5 years in length. Over 20 sites are planned throughout North America and Europe.

Lung function as primary endpoint trials

MILES (NCT004 14648; active, not recruiting). The Multicenter International LAM Efficacy of Sirolimus (MILES) Trial was designed to determine if treatment with sirolimus leads to an improvement in lung function (FEV1), and whether the benefits of sirolimus outweigh the risks in patients with LAM. MILES was funded by the FDA, the LAM Foundation, Wyeth, the Tuberous Sclerosis Association, and the Adler Foundation. It was designed as a Phase III, intention to treat, randomized, double-blind, placebo controlled study, with 1 year on study drug or placebo and 1 year of observation. The trial opened in December 2006, and completed enrollment in August 2009 with 110 patients registered, and 89 randomized. An interim analysis is scheduled to be executed when 40 patients reach the 1 year point, in mid-November 2009. The final endpoint will occur in August 2010, when all patients complete 1 year on therapy. Enrollment in MILES lagged far behind the original target of 240 patients, but close to the target of 120 patients revised downward in 2007 based on event rates and effect sizes estimated from the CAST trial that was concluding at that time.

Doxycycline trials for LAM

Small trials in Australia and the United Kingdom (EUDRA CT registration 2007-003745-32) are underway to test the safety and efficacy of doxycycline in patients with LAM. Both studies are randomized, placebo controlled and 2 years in length. The primary endpoint in the UK trial is rate of decline in FEV1, the Australia trial cites multiple endpoints with none identified as primary.

General Principles on Study Design of Clinical Trials in Rare Diseases

Generally, clinical trial designs for rare diseases must meet the same rigorous standards as do designs for trials for diseases that occur with much higher frequency.³⁵ They should ask important scientific questions, reduce bias, and have a high likelihood of obtaining a scientifically acceptable answer. Study questions should be answered with the fewest number of enrolled subjects possible, irrespective of the number of subjects available,³⁵ and enrollment criteria should facilitate broad access and conclusions that can be generalized. However, there are several unique features and alternative study designs that may be considered particularly relevant for consideration in the study of rare diseases.

A randomized controlled trial is generally considered the gold standard in clinical research because intrinsic in its design is the minimization of the impact of confounding variables (both measured and unmeasured). Indeed, one of the main benefits of RCTs over observational studies is the equal allocation of important covariates between the intervention and control arms of a study, provided that there are sufficient numbers of study subjects. Thus, the results of RCTs are often considered as the strongest evidence in testing a hypothesis. However, RCTs are not easily performed and are costly (both time and resources). Recruitment of a sufficient number of patients can be particularly difficult in rare diseases,³⁸ leading to lengthy accrual periods. In addition, randomization requires that the investigator and the subject consider themselves in a state of equipoise (i.e., they believe that the treatment received from either arm of a randomized trial is equivalent unless proven otherwise). This is difficult for participants who want to believe that their treatment will be based upon what is best for them and not the ‘flip of a coin’, and difficult for physicians who also think that they are ethically bound to provide the ‘best’ treatment. Equipoise is also difficult since trials are often developed because an investigator feels that an experimental therapy is better and they wish to test that hypothesis in a rigorous fashion.³⁵ Many subjects are hesitant to enroll in trials if they have a likelihood of being assigned a potentially inferior arm (i.e., have a likelihood of not receiving the experimental therapy) or randomized to a placebo. Several study design techniques have been developed to attempt to address some of these difficulties in performance of classical RCTs.

A study design technique that increases efficiency and deals with the concerns of potential placebo treatment assignment is the use of concurrent controls in which participants can serve as their own control. Such designs are beneficial if there is less within patient variability in a treatment response than there is between-patient variability.³⁵ In such cases, outcome estimates will have less variance and the study design will require less accrual, leading to study efficiency. Some examples of designs that use concurrent controls include cross-over designs and “N-of-1” designs. These study designs do have certain caveats, however. Concurrent controls can only be used in situations where the following conditions are met: there is a relatively rapid response to the intervention; the response abates after the intervention is withdrawn, and the natural history of the severity of the disease does not change appreciably over the study time period.³⁵ These designs may work well for certain chronic diseases, but there are many situations in which the assumptions can not be justified.

In cross-over trials, participants are randomized to a treatment arm for a time period at the end of which the outcome is assessed and then the subjects are ‘crossed over’ to the other treatment. Cross-over designs use the same patients twice and effectively halve the number of patients that must be enrolled. ‘N-of-1’ designs use the same patients a number of times (generally up to 5) and are even more efficient.³⁵ Both the cross-over design and ‘N -of-1’ trials make the same assumption that a washout period is assumed to eliminate the effect of the treatment after the intervention is withdrawn.

Factorial designs are somewhat akin to cross-over designs with respect to increased study efficiency, but differ notably in that they involve a double randomization in which two questions are asked concurrently in the same participant study population. This is as if the investigators are conducting two simultaneous studies in the same patient population with an overall sample size savings of approximately 50% for both. This type of design assumes that there is no interaction between the two treatments (i.e., the biologic effect of the first intervention is not mediated or modified by the second intervention). Once again, this is assumption is often difficult to validate.

Summary: Future Trials in LAM

LAM is uniquely positioned for discovery through clinical trials. The molecular basis of the disease is well-studied and at least partially understood, a number of promising candidate therapies are already FDA approved for other indications and ready for testing, patients are generally young, well-informed, and motivated, and mechanisms to conduct geographically distributed trials are in place. For early phase trials where there is little available clinical evidence of efficacy, a study design that requires fewer subjects and can explore multiple interventions in a high throughput format such as open label phase II studies or the ‘N-of-1’ design is likely preferable. Interventions that show promise in these types of pilot studies should be followed with larger randomized controlled trials. For the later phase studies, consideration may also be given to study designs with increased efficiency (e.g., Factorial designs) so that the limited number of available study subjects can yield the most useful information. In addition, as individual therapies are shown to have an impact on disease course/outcome, there will likely be a need for well- designed studies of combination therapies akin to the chemotherapy model of cancer management.

In a population based survey of LAM patients regarding attitude toward participation in trials, the primary reasons cited for enrollment in clinical trials were access to information, access to the drug, closer monitoring and altruism toward other LAM patients.³⁹ The most frequent reasons given for not enrolling in trials were failure to meet the enrollment criteria, stable disease, and risks of side effects from the drug. Less often cited but still important reasons for not enrolling were inclusion of a placebo arm, access to the drug off label or through concurrent open label trials, fear of toxicity, and inconvenience or cost. These motivations for and against trial participation differ considerably from those of trials for more common diseases, and must be taken into account when designing trials. In so far as is possible, studies in rare disease should be designed to be as inclusive as possible, to take full advantage of the limited numbers of patients available, and to produce results that can be generalized to the population has a whole.

Footnotes

Author Disclosure Statement

Drs. Kinder and McCormack have no conflicts of interests or financial ties to disclose.

References

McCormack

. Lymphangioleiomyomatosis: A clinical update. Chest 1, 2008; 33:507–516.

Ryu

, Moss

, Beck

et al. The NHLBI lymphangioleiomyomatosis registry: Characteristics of 230 patients at enrollment. Am J Respir Crit Care Med, 2006; 173:105–111.

Juvet

, McCormack

, Kwiatkowski

, Downey

. Molecular pathogenesis of lymphangioleiomyomatosis: Lessons learned from orphans. Am J Respir Cell Mol Biol, 2007; 36:398–408.

Bissler

, McCormack

, Young

et al. Sirolimus for angiomyolipoma in tuberous sclerosis complex or lymphangioleiomyomatosis. N Engl J Med, 2008; 358:140–151.

Crino

, Nathanson

, Henske

. The tuberous sclerosis complex. N Engl J Med, 2006; 355:1345–1356.

Ito

, Rubin

. gigas, a Drosophila homolog of tuberous sclerosis gene product-2, regulates the cell cycle. Cell, 1999; 96:529–539.

Potter

, Huang

, Xu

. Drosophila Tsc1 functions with Tsc2 to antagonize insulin signaling in regulating cell growth, cell proliferation, and organ size. Cell, 2001; 105:357–368.

Tapon

, Ito

, Dickson

, Treisman

, Hariharan

. The Drosophila tuberous sclerosis complex gene homologs restrict cell growth and cell proliferation. Cell, 2001; 105:345–355.

, Henske

. Estrogen-induced activation of mammalian target of rapamycin is mediated via tuberin and the small GTPase ras homologue enriched in brain. Cancer Res, 2006; 66:9461–9466.

10.

, Astrinidis

, Howard

, Henske

. Estradiol and tamoxifen stimulate LAM-associated angiomyolipoma cell growth and activate both genomic and nongenomic signaling pathways. Am J Physiol Lung Cell Mol Physiol, 2004; 286:L694–L700.

11.

Thoreen

, Kang

, Chang

et al. An ATP-competitive mammalian target of rapamycin inhibitor reveals rapamycin-resistant functions of mTORC 1. J Biol Chem, 2009; 284:8023–8032.

12.

Thoreen

, Sabatini

. Rapamycin inhibits mTORC1, but not completely. Autophagy, 2009; 5:725–726.

13.

, Robb

, Morrison

et al. Estrogen promotes the survival and pulmonary metastasis of tuberin-null cells. Proc Natl Acad Sci USA, 2009; 106:2635–2640.

14.

Zhe

, Yang

, Jakkaraju

, Schuger

. Tissue inhibitor of metalloproteinase-3 downregulation in lymphangioleiomyomatosis: Potential consequence of abnormal serum response factor expression. Am J Respir Cell Mol Biol, 2003; 28:504–511.

15.

Hayashi

, Fleming

, Stetler–Stevenson

et al. Immunohistochemical study of matrix metalloproteinases (MMPs) and their tissue inhibitors (TIMPs) in pulmonary lymphangioleiomyomatosis. Hum Pathol, 1997; 28:1071–1078.

16.

Matsui

, Takeda

, Yu

, Travis

, Moss

, Ferrans

. Role for activation of matrix metalloproteinases in the pathogenesis of pulmonary lymphangioleiomyomatosis. Arch Pathol Lab Med, 2000; 124:267–275.

17.

Zucker

, Cao

, Chen

. Critical appraisal of the use of matrix metalloproteinase inhibitors in cancer treatment. Oncogene, 2000; 19:6642–6650.

18.

Bittmann

, Dose

, Muller

, Dienemann

, Vogelmeir

. Lymphagioleiomyomatosis: Recurrence after single lung transplantations. Hum Pathol, 1997; 26:1420–1423.

19.

Karbowniczek

, Astrinidis

, Balsara

et al. Recurrent lymphangiomyomatosis after transplantation: Genetic analyses reveal a metastatic mechanism. Am J Respir Crit Care Med, 2003; 167:976–982.

20.

Henske

. Metastasis of benign tumor cells in tuberous sclerosis complex. Genes Chromosomes Cancer, 2003; 38:376–381.

21.

Longacre

, Hendrickson

, Kapp

, Teng

. Lymphangioleiomyomatosis of the uterus simulating high-stage endometrial stromal sarcoma. Gynecol Oncol, 1996; 63:404–410.

22.

Crooks

, Pacheco–Rodriguez

, DeCastro

et al. Molecular and genetic analysis of disseminated neoplastic cells in lymphangioleiomyomatosis. Proc Natl Acad Sci USA, 2004; 101:17462–17467.

23.

Berner

, Franzen

, Heilo

. Fine needle aspiration cytology as a diagnostic approach to lymphangioleiomyomatosis. A case report. Acta Cytol, 1997; 41:877–879.

24.

Kamitani

, Yabuuchi

, Soeda

et al. A case of lymphangioleiomyomatosis affecting the supraclavicular lymph nodes. J Comput Assist Tomogr, 2006; 30:279–282.

25.

Kumasaka

, Seyama

, Mitani

et al. Lymphangiogenesis-mediated shedding of LAM cell clusters as a mechanism for dissemination in lymphangioleiomyomatosis. Am J Surg Pathol, 2005; 29:1356–1366.

26.

Kumasaka

, Seyama

, Mitani

et al. Lymphangiogenesis in lymphangioleiomyomatosis: Its implication in the progression of lymphangioleiomyomatosis. Am J Surg Pathol, 2004; 28:1007–1016.

27.

Seyama

, Kumasaka

, Souma

et al. Vascular endothelial growth factor-D is increased in serum of patients with lymphangioleiomyomatosis. Lymphat Res Biol, 2006; 4:143–152.

28.

Young

, Inoue

, McCormack

. Diagnostic potential of serum VEGF-D for lymphangioleiomyomatosis. N Engl J Med, 2008; 358:199–200.

29.

Goncharova

, Goncharov

, Lim

, Noonan

, Krymskaya

. Modulation of cell migration and invasiveness by tumor suppressor TSC2 in lymphangioleiomyomatosis. Am J Respir Cell Mol Biol, 2006; 34:473–480.

30.

Huber

, Bruns

, Schmid

et al. Inhibition of the mammalian target of rapamycin impedes lymphangiogenesis. Kidney Int, 2007; 71:771–777.

31.

Kobayashi

, Kishimoto

, Kamata

, Otsuka

, Miyazaki

, Ishikura

. Rapamycin, a specific inhibitor of the mammalian target of rapamycin, suppresses lymphangiogenesis and lymphatic metastasis. Cancer Sci, 2007; 98:726–733.

32.

Kenerson

, Dundon

, Yeung

. Effects of rapamycin in the Eker rat model of tuberous sclerosis complex. Pediatr Res, 2005; 57:67–75.

33.

Glasgow

, Avila

, Lin

, Stylianou

, Moss

. Serum vascular endothelial growth factor-D levels in patients with lymphangioleiomyomatosis reflect lymphatic involvement. Chest, 2009; 135:1293–1300.

34.

Schmithorst

, Altes

, Young

et al. Automated algorithm for quantifying the extent of cystic change on volumetric chest CT: Initial results in lymphangioleiomyomatosis. AJR Am J Roentgenol, 2009; 192:1037–1044.

35.

Griggs

, Batshaw

, Dunkle

et al. Clinical research for rare disease: Opportunities, challenges, and solutions. Mol Genet Metab, 2009; 96:20–26.

36.

Taveira–DaSilva

, Steagall

, Moss

. Lymphangioleiomyomatosis. Cancer Control, 2006; 13:276–285.

37.

Davies

, Johnson

, Tattersfield

et al. Sirolimus therapy in tuberous sclerosis or sporadic lymphangioleiomyomatosis. N Engl J Med, 2008; 358:200–203.

38.

Tashkin

, Elashoff

, Clements

et al. Cyclophosphamide versus placebo in scleroderma lung disease. N Engl J Med, 2006; 354:2655–2666.

39.

Kinder

, Sherman

, Young

, Hagaman

, Opresco

, Byrnes

, McCormack

. Predictors for clinical trial participation in the rare lung disease lymphangioleiomyomatosis. Respir Med, 2009 December 3Epub ahead of print.