Learning Multiple Evolutionary Pathways from Cross-Sectional Data

Correspondence of tumor localization with tumor recurrence and cytogenetic progression in meningiomas

OBJECTIVE

Meningiomas are mostly benign tumors that originate from the coverings of the brain and spinal cord. Cytogenetically, they reveal a normal karyotype or, typically, monosomy of chromosome 22. Progression of meningiomas is associated with a non-random pattern of secondary losses of other autosomes. Deletion of the short arm of one chromosome 1 is a decisive step to anaplastic growth in meningiomas.

METHODS

Statistical analyses were performed for the karyotypes of 661 meningiomas with respect to localization, progression, and recurrence of the tumor. A mathematical mixture model estimates typical pathogenetic routes in terms of the accumulation of somatic chromosome changes in tumor cells. The model generates a genetic progression score (GPS) that estimates the prognosis as related to the cytogenetic properties of a given tumor.

RESULTS

In 53 patients, one or several recurrences were documented over the period of observation. This corresponds to a total rate of recurrence of 8.0% after macroscopically complete tumor extirpation. Higher GPS values were shown to be strongly correlated with tumor recurrence (P = 2.9 x 10(-7)). High-risk tumors, both in terms of histology and cytogenetics, are localized much more frequently at the brain surface than at the cranial base (P = 1.2 x 10(-5) for World Health Organization grade and P = 3.3 x 10(-12) for GPS categorization).

CONCLUSION

The tendency of cranial base meningiomas to recur seems to depend on surgical rather than biological reasons. As a quantitative measure, the GPS allows for a more precise assessment of the prognosis of meningiomas than the established categorical cytogenetic markers.

Modelled in vivo HIV Fitness under drug Selective Pressure and Estimated Genetic Barrier Towards Resistance are Predictive for Virological Response

Background

A method has been developed to estimate a fitness landscape experienced by HIV-1 under treatment selective pressure as a function of the genotypic sequence thereby also estimating the genetic barrier to resistance.

Methods

We evaluated the performance of two estimated fitness landscapes (nelfinavir [NFV] and zidovudine [AZT] plus lamivudine [3TC]) to predict week 12 viral load (VL) change for 176 treatment change episodes (TCEs) and probability of week 48 virological failure for 90 TCEs, in treatment experienced patients starting these drugs in combination.

Results

A higher genetic barrier for AZT plus 3TC, (quantified per additional mutation required to develop resistance against these drugs) was associated with a 0.54 (95% confidence interval [CI] 0.30–0.77) larger log10 VL reduction at 12 weeks ( P<0.0001) and a 0.39 (95% CI 0.23–0.66) lower odds of virological failure at 48 weeks ( P=0.0005), in analyses adjusting for the pre-TCE VL and the exact time-lag between the TCE and the date of determining response VL. The strength of these associations was comparable with those seen with expert interpretation systems (Rega, ANRS and HIVDB). A higher genetic barrier to NFV resistance was the only genotypic predictor that tended to be associated with a 0.19 (95% CI 0–0.39) higher log10 VL reduction at 12 weeks ( P=0.05) and a 0.63 (95% CI 0.36–1.09) lower odds of virological failure at 48 weeks ( P=0.10) per additional mutation.

Conclusions

These results suggest that an estimated genetic barrier derived from fitness landscapes may contribute to an improvement of predicted treatment outcome for NFV and this approach should be explored for other drugs.

Stability analysis of mixtures of mutagenetic trees

Background

Mixture models of mutagenetic trees are evolutionary models that capture several pathways of ordered accumulation of genetic events observed in different subsets of patients. They were used to model HIV progression by accumulation of resistance mutations in the viral genome under drug pressure and cancer progression by accumulation of chromosomal aberrations in tumor cells. From the mixture models a genetic progression score (GPS) can be derived that estimates the genetic status of single patients according to the corresponding progression along the tree models. GPS values were shown to have predictive power for estimating drug resistance in HIV or the survival time in cancer. Still, the reliability of the exact values of such complex markers derived from graphical models can be questioned.

Results

In a simulation study, we analyzed various aspects of the stability of estimated mutagenetic trees mixture models. It turned out that the induced probabilistic distributions and the tree topologies are recovered with high precision by an EM-like learning algorithm. However, only for models with just one major model component, also GPS values of single patients can be reliably estimated.

Conclusion

It is encouraging that the estimation process of mutagenetic trees mixture models can be performed with high confidence regarding induced probability distributions and the general shape of the tree topologies. For a model with only one major disease progression process, even genetic progression scores for single patients can be reliably estimated. However, for models with more than one relevant component, alternative measures should be introduced for estimating the stage of disease progression.

Estimation of an in vivo fitness landscape experienced by HIV-1 under drug selective pressure useful for prediction of drug resistance evolution during treatment

MOTIVATION

HIV-1 antiviral resistance is a major cause of antiviral treatment failure. The in vivo fitness landscape experienced by the virus in presence of treatment could in principle be used to determine both the susceptibility of the virus to the treatment and the genetic barrier to resistance. We propose a method to estimate this fitness landscape from cross-sectional clinical genetic sequence data of different subtypes, by reverse engineering the required selective pressure for HIV-1 sequences obtained from treatment naive patients, to evolve towards sequences obtained from treated patients. The method was evaluated for recovering 10 random fictive selective pressures in simulation experiments, and for modeling the selective pressure under treatment with the protease inhibitor nelfinavir.

RESULTS

The estimated fitness function under nelfinavir treatment considered fitness contributions of 114 mutations at 48 sites. Estimated fitness correlated significantly with the in vitro resistance phenotype in 519 matched genotype-phenotype pairs (R(2) = 0.47 (0.41 - 0.54)) and variation in predicted evolution under nelfinavir selective pressure correlated significantly with observed in vivo evolution during nelfinavir treatment for 39 mutations (with FDR = 0.05).

AVAILABILITY

The software is available on request from the authors, and data sets are available from http://jose.med.kuleuven.be/~kdforc0/nfv-fitness-data/.

Application of oncogenetic trees mixtures as a biostatistical model of the clonal cytogenetic evolution of meningiomas

Meningiomas are mostly benign tumors that originate from the coverings of brain and spinal cord. Typically, they reveal a normal karyotype or monosomy for chromosome 22. Rare clinical progression of meningiomas is associated with a nonrandom pattern of secondary losses of other autosomes. Deletion of the short arm of one chromosome 1 appears to be a decisive step for anaplastic growth in meningiomas. We calculated an oncogenetic tree model that estimates the most likely cytogenetic pathways of 661 meningioma patients in terms of accumulation of somatic chromosome changes in tumor cells. The genetic progression score (GPS) estimates the genetic status of a tumor as progression in the corresponding tumor cells along this model. Large GPS values are highly correlated with early recurrence of meningiomas [p < 10(-4)]. This correlation holds even if patients are stratified by WHO grade. We show that tumor location also has an impact on genetic progression. Clinical relevance of the GPS is thus demonstrated with respect to origin, WHO grade and recurrence of the tumor. As a quantitative measure the GPS allows a more precise assessment of the prognosis of meningiomas than categorical cytogenetic markers based on single chromosomal aberrations.

Combining Cross-Sectional and Prospective Data Methods to Improve Transition Parameter Estimation for Characterizing the Accumulation of HIV-1 Drug Resistance Mutations

The order and rate of acquisition of HIV drug resistance mutations have been estimated from longitudinal and cross-sectional data using Markov models and branching trees, respectively. This article proposes methods that make use of both longitudinal and cross-sectional data simultaneously by employing link functions between the two parameter sets. Most functions that link the two parameter sets also depend on the time on treatment before the start of the study-information that may not be available. Nonetheless, under certain assumptions, some link functions eliminate the dependence on time. Using such functions, the two sources of information can be combined to improve the precision of parameter estimation. The method also accommodates error in the link functions from uncertainty in the assumptions required for the links or other reasons. These methods are applied to data from AIDS Clinical Trial Group protocol 398, a randomized comparison of mono- versus dual-protease inhibitor use in heavily treatment experienced HIV patients. Combining the two sources of information allows detection of differences between rates of transition that are not detectable using prospective data alone.

HIV-1 subtype B protease and reverse transcriptase amino acid covariation

Despite the high degree of HIV-1 protease and reverse transcriptase (RT) mutation in the setting of antiretroviral therapy, the spectrum of possible virus variants appears to be limited by patterns of amino acid covariation. We analyzed patterns of amino acid covariation in protease and RT sequences from more than 7,000 persons infected with HIV-1 subtype B viruses obtained from the Stanford HIV Drug Resistance Database (http://hivdb.stanford.edu). In addition, we examined the relationship between conditional probabilities associated with a pair of mutations and the order in which those mutations developed in viruses for which longitudinal sequence data were available. Patterns of RT covariation were dominated by the distinct clustering of Type I and Type II thymidine analog mutations and the Q151M-associated mutations. Patterns of protease covariation were dominated by the clustering of nelfinavir-associated mutations (D30N and N88D), two main groups of protease inhibitor (PI)-resistance mutations associated either with V82A or L90M, and a tight cluster of mutations associated with decreased susceptibility to amprenavir and the most recently approved PI darunavir. Different patterns of covariation were frequently observed for different mutations at the same position including the RT mutations T69D versus T69N, L74V versus L74I, V75I versus V75M, T215F versus T215Y, and K219Q/E versus K219N/R, and the protease mutations M46I versus M46L, I54V versus I54M/L, and N88D versus N88S. Sequence data from persons with correlated mutations in whom earlier sequences were available confirmed that the conditional probabilities associated with correlated mutation pairs could be used to predict the order in which the mutations were likely to have developed. Whereas accessory nucleoside RT inhibitor-resistance mutations nearly always follow primary nucleoside RT inhibitor-resistance mutations, accessory PI-resistance mutations often preceded primary PI-resistance mutations.

Improved prediction of response to antiretroviral combination therapy using the genetic barrier to drug resistance

BACKGROUND

The outcome of antiretroviral combination therapy depends on many factors involving host, virus, and drugs. We investigate prediction of treatment response from the applied drug combination and the genetic constellation of the virus population at baseline. The virus's evolutionary potential for escaping from drug pressure is explored as an additional predictor.

METHODS

We compare different encodings of the viral genotype and antiretroviral regimen including phenotypic and evolutionary information, namely predicted phenotypic drug resistance, activity of the regimen estimated from sequence space search, the genetic barrier to drug resistance, and the genetic progression score. These features were evaluated in the context of different statistical learning procedures applied to the binary classification task of predicting virological response. Classifier performance was evaluated using cross-validation and receiver operating characteristic curves on 6,337 observed treatment change episodes from the Stanford HIV Drug Resistance Database and a large US clinic-based patient population.

RESULTS

We find that the choice of appropriate features affects predictive performance more profoundly than the choice of the statistical learning method. Application of the genetic barrier to drug resistance, which combines phenotypic and evolutionary information, outperformed the genetic progression score, which uses exclusively evolutionary knowledge. The benefit of phenotypic information in predicting virological response was confirmed by using predicted fold changes in drug susceptibility. Moreover, genetic barrier and predicted phenotypic drug resistance were found to be the best encodings across all datasets and statistical learning methods examined.

AVAILABILITY

THEO (THErapy Optimizer), a prototypical implementation of the best performing approach, is freely available for research purposes at http://www.geno2pheno.org.

Hyperdiploidy defines a distinct cytogenetic entity of meningiomas

BACKGROUND

The most common chromosomal aberration found in meningiomas is monosomy 22. Progression and recurrence of meningiomas are usually associated with additional chromosome losses. Rarely, however, meningiomas have strongly hyperdiploid karyotypes with over 50 chromosomes; the objective of this study was to explore the cytogenetic and histopathologic patterns as well as the clinical significance of hyperdiploidy in meningiomas.

METHODS

Within a series of 677 consecutive meningiomas, we identified a subgroup comprising 16 cases that display a strikingly uniform pattern of hyperdiploidy mostly without structural chromosome rearrangements, as shown by banding techniques and, in the single structurally aberrant case, spectral karyotyping.

RESULTS

These meningiomas each have between 50 and 56 chromosomes, with trisomy 12 (14/16 cases), trisomy 20 (13/16 cases), trisomy 5 (12/16 cases), and trisomy 17 (10/16 cases). Histomorphologically, hyperdiploid meningiomas feature a heterogeneous phenotype. However, they are associated with a higher histological grade, and decreased expression of alkaline phosphatase as compared to meningiomas with typical karyotype. In two patients, recurrences were documented and three patients died of disease during the period of observation, indicating a worse prognosis of hyperdiploid than of cytogenetically typical meningiomas.

CONCLUSION

We conclude that hyperdiploidy constitutes a small but clinically relevant entity of biologically aggressive meningiomas, which are cytogenetically distinguishable from the majority of common-type meningiomas.

Evolution on distributive lattices

We consider the directed evolution of a population after an intervention that has significantly altered the underlying fitness landscape. We model the space of genotypes as a distributive lattice; the fitness landscape is a real-valued function on that lattice. The risk of escape from intervention, i.e., the probability that the population develops an escape mutant before extinction, is encoded in the risk polynomial. Tools from algebraic combinatorics are applied to compute the risk polynomial in terms of the fitness landscape. In an application to the development of drug resistance in HIV, we study the risk of viral escape from treatment with the protease inhibitors ritonavir and indinavir.

A mutagenetic tree hidden Markov model for longitudinal clonal HIV sequence data

RNA viruses provide prominent examples of measurably evolving populations. In human immunodeficiency virus (HIV) infection, the development of drug resistance is of particular interest because precise predictions of the outcome of this evolutionary process are a prerequisite for the rational design of antiretroviral treatment protocols. We present a mutagenetic tree hidden Markov model for the analysis of longitudinal clonal sequence data. Using HIV mutation data from clinical trials, we estimate the order and rate of occurrence of seven amino acid changes that are associated with resistance to the reverse transcriptase inhibitor efavirenz.

Computational methods for the design of effective therapies against drug resistant HIV strains

The development of drug resistance is a major obstacle to successful treatment of HIV infection. The extraordinary replication dynamics of HIV facilitates its escape from selective pressure exerted by the human immune system and by combination drug therapy. We have developed several computational methods whose combined use can support the design of optimal antiretroviral therapies based on viral genomic data.