Relationship Between Response and Dose in Published, Contemporary Phase I Oncology Trials

Authors: Antonious Hazim MD1, Gordon Mills MD, PhD2, Vinay Prasad MD, MPH2,3,4, Alyson Haslam PhD2, and Emerson Y. Chen MD2
View More View Less
  • 1 School of Medicine,
  • | 2 Division of Hematology Oncology, Knight Cancer Institute,
  • | 3 Department of Public Health and Preventive Medicine, and
  • | 4 Center for Health Care Ethics, Oregon Health & Science University, Portland, Oregon.

Background: As progress continues in oncology drug development, this study aimed to examine whether the previously established association between drug dose and efficacy in the era of cytotoxic therapies remains true in today’s phase I dose-escalation oncology trials. Methods: A systematic review of early-phase dose-finding trials of single-agent oncology drugs from 2015 to 2018 was conducted to examine the relationship between drug dose and objective responses. Cancer-specific trials were included if they determined maximum tolerated dose (MTD) and/or recommended phase II dose (RP2D). Data related to the study drug, study design, treatment response, cancer type, dose levels, MTD, and RP2D were all collected. Dose level was categorized into 4 categories (≤40%, 41%–80%, 81%–120%, and >120% of the RP2D) and was further analyzed by class of drug. Results: A total of 175 phase I studies were identified, with a total of 7,330 patients showing a median response rate of 5% (range, 0%–83%) across trials. A total of 93 trials with 2,506 participants had response data corresponding to drug dose level. In this subset, the median response rate was 5% (range, 0%–83%) across trials. Across all participants in this subset, the response rate was 12% (57 of 491) among those in the dose range of ≤40% of RP2D, 17% (95 of 562) among those in 41% to 80% of RP2D, 23% (272 of 1,206) among those in 81% to 120% of RP2D, and 29% (71 of 247) among those in >120% of RP2D (P<.001). The response rate at ≤40% of RP2D for targeted antibody was 5%, 4% for cellular therapy, 19% for immunotherapy, and 21% for small-molecule targeted inhibitors. Conclusions: Whereas our study of published phase I trials continued to show a low response rate consistent with earlier studies, the relationship between response and dose does not always peak at 81% to 120% of RP2D anymore, likely due to the use of novel immunotherapy and targeted agents with distinct efficacy and toxicity patterns.

Phase I cancer trials are designed to investigate drug safety, tolerability, pharmacodynamics, and pharmacokinetics, and if responses are seen, they may provide reassuring information about drug activity. Prior pooled analyses of published phase I trials testing primarily cytotoxic chemotherapies have shown low overall response rates (ORRs), such as a 1991 study showing a 6% response rate across 228 phase I studies,1 and more recent analyses confirming a 4% to 6% response rate.2,3 These studies, which focused on cytotoxic agents,13 also showed that an increasing dose of cytotoxic chemotherapy was associated with improved tumor response and increased toxicity. In one earlier pooled analysis, the maximum response was statistically shown to occur most often at the recommended phase II dose (RP2D), or at approximately 81% to 120% of RP2D.1

Recent advances in drug development have allowed for targeted agents, immunotherapy, and other classes of agents, which, in regard to efficacy, may not correlate well with dose level, toxicity, or biologic framework.47 Thus, the RP2D of these newer classes of agents not only incorporates the maximum tolerated dose (MTD) but also must account for delayed and atypical adverse effects that occur after a course of treatment.8 Although contemporary phase I trial designs have mitigated some inefficiencies related to nonclassic toxicities seen in novel agents,9 it remains unclear whether the maximum response of these newer classes of agents correlates with the recommended dose used for phase II trials.

For these reasons, we conducted a systematic review of recent early-phase dose-finding clinical trials of novel oncology drugs to examine the dose–efficacy relationship. We aimed to determine whether the RP2D in early-phase trials correlates with their preliminary efficacy endpoints in the modern era, in which noncytotoxic agents dominate oncology drug development.

Methods

Overview

We conducted a systematic review of contemporary early-phase dose-finding clinical trials of novel oncology drugs to examine the relationship between investigative drug dose and objective response rate.

Literature Search and Selection

We searched MEDLINE/PubMed for the terms cancer and oncology to identify dose-finding phase I studies from November 15, 2015, through November 15, 2018. The following filters were used: full text, English, humans, clinical trial phase I, and adult aged ≥18 years. The literature search was last accessed on February 2, 2019. Studies were included if they enrolled adult humans aged ≥18 years, investigated cancer treatment (both malignant hematology and oncology), and examined the safety and preliminary efficacy of a proposed drug dose. Data from a minority of dose-expansion cohorts (phase I/II studies) were included for this review. Studies that focused on multidrug combination, radiation, topical, or surgical therapies, or drug delivery were excluded. Studies not accessible by Oregon Health & Science University or with overly incomplete or duplicate data were also excluded (Figure 1).

Figure 1.
Figure 1.

Flow diagram of phase I trial selection.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 4; 10.6004/jnccn.2019.7375

Data Extracted

For every trial, we collected the treatment response endpoint of every patient at every dose level, if available. Patients were categorized as having achieved complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). We also combined CR and PR over the total number of patients (those with CR, PR, SD, and PD, including patients who were nonevaluable) to assess the ORR. Most response endpoints were classified using RECIST guidelines.10

In addition, for every trial, information was collected related to the investigational therapy, treatment regimen, study design, respective dose levels, MTDs, and final RP2D (if available). We also extracted the number of patients comprising the total trial populations and the respective tumor types, the median duration of response, and the median progression-free survival per study.

Data Analysis

The proportions of ORR, SD, disease control rate (ie, ORR + SD), and PD from all dose-finding trials were graphically compared across dose levels. The dose levels were organized by their relationship to the RP2D. For example, if a patient achieved SD at a dose of 10 mg and the final RP2D in the trial was 100 mg, then this trial participant would be assigned to the 10% category. We categorized this assignment into ≤40%, 41% to 80%, 81% to 120%, and >120% of the RP2D.1 Proportions of ORR were also compared among these categories using the Fisher exact test. Subgroup analyses were conducted in the specific drug class of interest: targeted antibody, immunotherapy, nonantibody targeted inhibitor, and vaccine/cellular therapy. Specific statistical testing was conducted using SAS 9.4 (SAS Institute Inc.); otherwise, all descriptive calculations and figures (including standard error lines) were performed and created using Microsoft Excel.

Results

We retrieved 831 search results that were individually reviewed, of which 175 phase I trials met our eligibility criteria (see Figure 1). These 175 trials (N=7,330) represented >50 tumor types, 7 drug classes, 145 unique drugs, and a median response rate of 5% (supplemental eTables 1–3, available with this article at JNCCN.org). The most common tumor types represented were lung (12%), colorectal (12%), and breast (11%). Among the 175 trials, 1,040 of 7,330 patients (15%) were labeled as having at least PR to therapy despite a median ORR of 5% across trials. Nonantibody targeted inhibitors and targeted antibodies were the most common therapies tested (47% and 24%, respectively). Immunotherapy and vaccine/cellular therapy had the greatest median response rate of all drug classes (15%). The median progression-free survival was 4.2 months, as illustrated by 84 studies that provided these data.

We then identified 93 studies that specifically reported the treatment response endpoint of every patient at every dose level. Among the 3,890 participants, the median response rate was 5% (range, 0%–83%; Table 1). A total of 47 studies involved nonantibody targeted inhibitors, 24 involved targeted antibodies, 7 involved immunotherapy, 6 involved chemotherapy, 5 involved vaccine therapy, 3 involved hormonal therapy, and 1 involved CAR T-cell therapy. The most common cancer types were leukemia (16%), lung (12%), and breast (10%) (Table 1).

Table 1.

Characteristics of Dose-Escalation Trials

Table 1.

Of the 3,890 patients, 2,506 had both response and dose data, among whom the ORR (CR and PR) was 20% (n=495); 1,095 patients (44%) had SD and 916 (37%) had PD (Figure 2). The proportion of patients with response increased with dose (12% in ≤40% of the RP2D, 17% in 41%–80% of the RP2D, 23% in 81%–120% of the RP2D, and 29% in >120% of the RP2D; P<.001). The disease control rate (CR+PR+SD) seen in 1,590 patients also increased with dose (58% in ≤40% of the RP2D, 61% in 41%–80% of the RP2D, 65% in 81%–120% of the RP2D, and 71% in >120% of the RP2D; P<.001) (Figure 2).

Figure 2.
Figure 2.

Response rate for all study drugs (n=2,506).

Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 4; 10.6004/jnccn.2019.7375

Among the 813 patients treated with targeted antibody medications, 128 (16%) had at least CR or PR, 349 (43%) had SD, and 336 (41%) had PD. The proportion of patients with response was 11 of 209 (5%) in ≤40% of the RP2D, 24 of 189 (13%) in 41% to 80% of the RP2D, 86 of 369 (23%) in 81% to 120% of the RP2D, and 7 of 46 (15%) in >120% of the RP2D (P<.01) (Figure 3). The disease control rate was 46% in ≤40% of the RP2D, 59% in 41% to 80% of the RP2D, 64% in 81% to 120% of the RP2D, and 70% in >120% of the RP2D (P<.01).

Figure 3.
Figure 3.

Response rate for targeted antibody (n=813).

Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 4; 10.6004/jnccn.2019.7375

With regard to the 239 patients who received immunotherapy, 52 (22%) had at least CR or PR, 120 (50%) had SD, and 67 (28%) had PD. The proportion of patients with response was 12 of 64 (19%) in ≤40% of the RP2D, 7 of 37 (19%) in 41% to 80% of the RP2D, 31 of 133 (23%) in 81% to 120% of the RP2D, and 2 of 5 (40%) in >120% of the RP2D (P=.59) (Figure 4). The disease control rate was 56% in ≤40% of the RP2D, 81% in 41% to 80% of the RP2D, 78% in 81% to 120% of the RP2D, and 40% in >120% of the RP2D (P<.01).

Figure 4.
Figure 4.

Response rate for immunotherapy (n=239).

Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 4; 10.6004/jnccn.2019.7375

Distributions of nonantibody targeted inhibitor (n=1,220) and vaccine/cellular therapy (n=102) are depicted in Figures 5 and 6, respectively. Among the 1,220 patients who were treated with small molecule targeted agents, 32 of 155 (21%) responded at ≤40% of the RP2D, 61 of 170 (23%) responded at 41% to 80% of the RP2D, 148 of 620 (24%) responded at 81% to 120% of the RP2D, and 60 of 175 (34%) responded at >120% of the RP2D (see Figure 5). Among the 102 patients who were treated with cellular or vaccine therapy, 2 of 49 (4%) responded at ≤40% of the RP2D, 0 of 22 responded at 41% to 80% of the RP2D, 3 of 24 (13%) responded at 81% to 120% of the RP2D, and 0 of 7 responded at >120% of the RP2D (see Figure 6).

Figure 5.
Figure 5.

Response rate for nonantibody-targeted inhibitor (n=1,220).

Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 4; 10.6004/jnccn.2019.7375

Figure 6.
Figure 6.

Response rate for vaccine/cellular therapy (n=102).

Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 4; 10.6004/jnccn.2019.7375

Discussion

Multiple studies have sought to characterize the response rates of phase I trials13; these 3 earlier studies, representing >20,000 patients enrolled in phase I trials from 1970 to 2009, reported an overall objective response of 4% to 7%, despite 1 study showing notable variation from year to year, with a response rate as high as 20% reported.3 We found that the median response rate in contemporary phase I trials in this study remained approximately 5% of 7,330 patients. However, through aggregating the total number of responding patients and dividing by the total number of treated patients, we found that 1,040 patients (15%) had at least a PR. This difference likely occurred because drugs that generate response are more prone to having their trials enlarged with the use of expansion groups and/or increased enrollment. Thus, we found that the probability that a patient would realistically respond to a novel drug in phase I testing was better captured by the ORR reported from each trial rather than the response rate weighted by the sample size of each trial.

There has been variability in earlier reports regarding the relationship between dose–response and outcomes, including some that proposed that low and intermediate doses seem to have the best outcomes.1113 In our study, we found that among 2,506 patients in whom response could be correlated with dose, there was a positive relationship between dose level and total response rate (see Figure 2), with the greatest response rate (29%) occurring at a higher dose level (>120%) than the RP2D but still with notable responses (12%) at ≤40% of the RP2D. These findings are in sharp contrast with those for cytotoxic agents, which normally show the greatest response rate at 81% to 120% of the RP2D and the lowest response rates at ≤40% of the RP2D due to insufficient dose and at >120% possibly due to toxicities.1

The dose–response curves for each drug class all had varying patterns. Interestingly, the targeted antibody class peaked at 81% to 120% of the RP2D, similar to what is expected with cytotoxic agents, but this class also had high proportions of SD across all levels of RP2D (see Figure 3). The dose–response curves for immunotherapy (Figure 4) and nonantibody targeted agents (Figure 5) both had response rates peaking at >120% of the RP2D but still had response rates of 19% and 21%, respectively, observed at ≤40% of the RP2D. High rates of disease progression for vaccine and cellular therapies across all dose levels are observed due to the fact that these technologies are still premature for most cancers. Given these data, selecting an intermediate dose level of a targeted biologic, small-molecule targeted inhibitor, or immunotherapy agent showing promising efficacy for future trials could still provide reliable disease stability, even if true treatment response is seen at higher doses. In addition, a time-to-event endpoint, such as overall survival or progression-free survival, in phase II/III trials is a better determination of treatment benefit for many cancers, because these endpoints account for durable disease control and are often seen in immunotherapy, targeted biologic, and small-molecule targeted inhibitors. A more conservative dose selection may balance the risk/benefit profile of an investigational drug in subsequent trials, especially when some adverse effects are atypical or occur long after the treatment period. However, these hypotheses should be further explored with other systematic reviews that correlate the relationship between phase I trials and subsequent phase II/III trials.

As oncology drug development focuses more on novel drug classes, an increasing number of studies are using endpoints such as drug levels, drug metabolism, and preliminary efficacy rather than conventional toxicity endpoints.14 Some studies fail to achieve an MTD or have little rationale for the dose recommended for a phase II trial.15 These findings likely contributed to the nonclassic patterns of dose–response relationship seen in our analysis. More refined toxicity evaluation, a definition of MTD, and a dose selection with respect to drug class and response are all areas of focus for future investigations. These patterns reflect the observation that contemporary dose-finding designs need to be restructured.

Our study has several limitations. First, our analysis is limited to published dose-finding phase I trials. Earlier research has established that not all phase I trials conducted are represented in the literature, nor do all phase I trials report specific dose levels, RP2D, and efficacy data by dose levels.1619 Whether our results would materially change with the addition of unpublished trials remains unknown but warrants exploration.

Second, our study was limited to phase I studies that only tested a single agent. There are phase I studies that include novel therapies in combination with other agents. We also note that some phase I studies have tested different dose schedules rather than dose levels.

Third, our results reflect unique agents studied during our stated time period and may not apply to other highly promising targets or newer agents. Similarly, our analysis tended to favor targeted agents because this class of drug type was more readily studied during our stated time frame, likely because of the decreased focus on cytotoxic chemotherapy. We would expect a shift to occur as more phase I studies further expand on CAR T-cell therapy and immunotherapy.

Fourth, most studies did not provide toxicity data for individual patient results. Although these data were not the main focus of our study, it would be interesting to know how dose–response levels correlated, if at all, with toxicity. Future studies could also systematically measure how well toxicities are captured in a phase I study compared with later-phase trials, and how the RP2D in phase I trials translates into actual doses used in phase II and III trials.

Fifth, we did not have access to all patient data for each phase I study. Our hope is that future phase I studies will make patient data readily available for systematic analyses.

Finally, our search strategy was broad, limited to one search engine, and involved individual analysis of >800 articles. We may have missed some phase I trials during our search time frame. For these reasons, we again encourage future investigators to add to our analyses as new phase I trials become available.

Conclusions

This study of published phase I oncology trials shows a modest response rate among all tested agents (∼5%), in line with results of earlier studies, and the relationship between response and dose may vary by class of agent, with novel nonantibody targeted inhibitors and immunotherapies especially deviating from a classic, positive relationship between dose level and treatment response. These results should be confirmed and further studied with a determination of the risk/benefit balance to better inform optimal dose selection for subsequent trials.

References

  • 1.

    Von Hoff DD, Turner J. Response rates, duration of response, and dose response effects in phase I studies of antineoplastics. Invest New Drugs 1991;9:115122.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2.

    Subbiah IM, Wheler JJ, Hess KR, et al.. Outcomes of patients ≥65 years old with advanced cancer treated on phase I trials at MD Anderson Cancer Center. Int J Cancer 2017;140:208215.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3.

    Horstmann E, McCabe MS, Grochow L, et al.. Risks and benefits of phase 1 oncology trials, 1991 through 2002. N Engl J Med 2005;352:895904.

  • 4.

    Wong KM, Capasso A, Eckhardt SG. The changing landscape of phase I trials in oncology. Nat Rev Clin Oncol 2016;13:106117.

  • 5.

    Postel-Vinay S, Aspeslagh S, Lanoy E, et al.. Challenges of phase 1 clinical trials evaluating immune checkpoint-targeted antibodies. Ann Oncol 2016;27:214224.

  • 6.

    Le Tourneau C, Lee JJ, Siu LL. Dose escalation methods in phase I cancer clinical trials. J Natl Cancer Inst 2009;101:708720.

  • 7.

    Parulekar WR, Eisenhauer EA. Phase I trial design for solid tumor studies of targeted, non-cytotoxic agents: theory and practice. J Natl Cancer Inst 2004;96:990997.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8.

    Postel-Vinay S, Collette L, Paoletti X, et al.. Towards new methods for the determination of dose limiting toxicities and the assessment of the recommended dose for further studies of molecularly targeted agents—dose-limiting toxicity and toxicity assessment recommendation group for early trials of targeted therapies, an European Organisation for Research and Treatment of Cancer-led study. Eur J Cancer 2014;50:20402049.

    • Search Google Scholar
    • Export Citation
  • 9.

    Le Tourneau C, Gan HK, Razak AR, et al.. Efficiency of new dose escalation designs in dose-finding phase I trials of molecularly targeted agents. PLoS One 2012;7:e51039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Eisenhauer EA, Therasse P, Bogaerts J, et al.. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228247.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11.

    Gupta S, Hunsberger S, Boerner SA, et al.. Meta-analysis of the relationship between dose and benefit in phase I targeted agent trials. J Natl Cancer Inst 2012;104:18601866.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12.

    Moreno García V, Olmos D, Gomez-Roca C, et al.. Dose-response relationship in phase I clinical trials: a European Drug Development Network (EDDN) collaboration study. Clin Cancer Res 2014;20:56635671.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Jain RK, Lee JJ, Hong D, et al.. Phase I oncology studies: evidence that in the era of targeted therapies patients on lower doses do not fare worse. Clin Cancer Res 2010;16:12891297.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14.

    Hansen AR, Cook N, Amir E, et al.. Determinants of the recommended phase 2 dose of molecular targeted agents. Cancer 2017;123:14091415.

  • 15.

    Viala M, Vinches M, Alexandre M, et al.. Strategies for clinical development of monoclonal antibodies beyond first-in-human trials: tested doses and rationale for dose selection. Br J Cancer 2018;118:679697.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16.

    van den Bogert CA, Souverein PC, Brekelmans CTM, et al.. [Non-publication is common among phase 1, single-center, not prospectively registered, or early terminated clinical drug trials]. Ned Tijdschr Geneeskd 2017;161:D1498.

    • Search Google Scholar
    • Export Citation
  • 17.

    Chapman PB, Liu NJ, Zhou Q, et al.. Time to publication of oncology trials and why some trials are never published. PLoS One 2017;12:e0184025.

  • 18.

    Decullier E, Chan AW, Chapuis F. Inadequate dissemination of phase I trials: a retrospective cohort study. PLoS Med 2009;6:e1000034.

  • 19.

    Massey PR, Wang R, Prasad V, et al.. Assessing the eventual publication of clinical trial abstracts submitted to a large annual oncology meeting. Oncologist 2016;21:261268.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation

Submitted July 1, 2019; accepted for publication October 29, 2019.

Previous presentation: Data from this manuscript was deemed publication-only at the 2019 ASCO Annual Meeting; May 31–June 4, 2019; Chicago, Illinois. Abstract e14583.

Disclosures: Dr. Mills has disclosed that he is a scientific advisor for AstraZeneca, Catena Pharmaceuticals, Critical Outcome Technologies, ImmunoMET, Ionis, Medimmune, Nuevolution, Pfizer, Precision Medicine, Signalchem Lifesciences, Symphogen, Takeda/Millennium Pharmaceuticals, and Tarveda; has stock options for Catena Pharmaceuticals, ImmunoMet, SignalChem, Spindle Top Ventures, and Tarveda; has received grant/research support from Abbvie, Adelson AstraZeneca, Critical Outcomes Technology, Illumina, Ionis, Immunomet, Karus Therapeutics, Pfizer, Nanostring, Takeda/Millennium Pharmaceuticals, and Tesaro; and has licensed technology to Nanostring and Myriad Genetics. Dr. Prasad has disclosed that he receives grant/research support from Arnold Ventures; is a consultant for UnitedHealthcare; has received honoraria from Medscape; and that his podcast is supported by Patreon. The remaining authors have disclosed that they have not received any financial consideration from any person or organization to support the preparation, analysis, results, or discussion of this article.

Funding: This study was funded by the Knight Cancer Institute and the Laura and John Arnold Foundation.

Correspondence: Emerson Y. Chen, MD, Division of Hematology Oncology, Oregon Health & Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239. Email: cheem@ohsu.edu

Supplementary Materials

  • View in gallery

    Flow diagram of phase I trial selection.

  • View in gallery

    Response rate for all study drugs (n=2,506).

    Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

  • View in gallery

    Response rate for targeted antibody (n=813).

    Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

  • View in gallery

    Response rate for immunotherapy (n=239).

    Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

  • View in gallery

    Response rate for nonantibody-targeted inhibitor (n=1,220).

    Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

  • View in gallery

    Response rate for vaccine/cellular therapy (n=102).

    Abbreviations: CR, complete response; PR, partial response; PD, progressive disease, RP2D, recommended phase II dose; SD, stable disease.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 1792 457 56
PDF Downloads 597 267 27
EPUB Downloads 0 0 0