Technical Note: An investigation of polarity effects for wide-angle free-air chambers.

PURPOSE: Wide-angle free-air chambers (WAFACs) are used as primary standard measurement devices for establishing the air-kerma strength of low-energy, low-dose rate brachytherapy seeds. The National Research Council of Canada (NRC) is commissioning a primary standard wide-angle free-air chamber (NRC WAFAC) to serve the calibration needs of Canadian clients. The University of Wisconsin has developed a similar variable-aperture free-air chamber (UW VAFAC) to be used as a research tool. As part of the NRC commissioning, measurements were carried out for both polarities of the applied bias voltage and the resulting effects were observed to be very large. Similar effects were identified with the UW VAFAC. The authors describe the measurements carried out to determine the underlying causes of the polarity effect and the approach used to eliminate it. METHODS: The NRC WAFAC is based on the WAFAC design developed at the National Institute of Standards and Technology in the USA. Charge measurements for (125)I and (241)Am sources were carried out for both negative and positive polarities on the NRC WAFAC and UW VAFAC. Two aperture sizes were also investigated with the UW VAFAC. In addition, measurements on the NRC WAFAC were carried out with a small bias between the collecting electrode and the shield foil at the downstream end of the chamber. To mitigate all of the polarity effects, the downstream surface of the collecting electrode was covered with a thin layer of graphite on both the NRC and UW chambers. RESULTS: Both chamber designs showed a difference of more than 30 % between the charge collected with positive and negative bias voltages for the smallest electrode separation. It was shown for the NRC WAFAC that charge could be collected in the small gap downstream of the collecting volume by applying a voltage between the shield foil and the collecting electrode, even though an insulating foil (Mylar or polyimide film) separated the conducting surface from the small gap region. The unwanted additional current was shown to be proportional to the size of the aperture for the UW VAFAC. The extra ionization produced in the small gap region was eliminated for both chambers by covering the insulating side of the collecting electrode with a grounded conducting layer. CONCLUSIONS: The small gap region downstream of the collecting electrode in the NRC WAFAC and UW VAFAC can serve as an unwanted source of ion current. It is concluded that a residual electric field in the small gap region may lead to ion transport and to charge being trapped on the surface of the foil. The foil then acts as a capacitor with an equal charge, but of opposite sign, being attracted to the conducting surface. Covering the back of the collecting electrode surface with a grounded conducting layer eliminated the polarity effect.

A comparison of rat SPECT images obtained using (99m)Tc derived from 99Mo produced by an electron accelerator with that from a reactor.

Recent shortages of molybdenum-99 ((99)Mo) have led to an examination of alternate production methods that could contribute to a more robust supply. An electron accelerator and the photoneutron reaction were used to produce (99)Mo from which technetium-99m ((99m)Tc) is extracted. SPECT images of rat anatomy obtained using the accelerator-produced (99m)Tc with those obtained using (99m)Tc from a commercial generator were compared. Disks of (100)Mo were irradiated with x-rays produced by a 35 MeV electron beam to generate about 1110 MBq (30 mCi) of (99)Mo per disk. After target dissolution, a NorthStar ARSII unit was used to separate the (99m)Tc, which was subsequently used to tag pharmaceuticals suitable for cardiac and bone imaging. SPECT images were acquired for three rats and compared to images for the same three rats obtained using (99m)Tc from a standard reactor (99)Mo generator. The efficiency of (99)Mo-(99m)Tc separation was typically greater than 90%. This study demonstrated the delivery of (99m)Tc from the end of beam to the end user of approximately 30 h. Images obtained using the heart and bone scanning agents using reactor and linac-produced (99m)Tc were comparable. High-power electron accelerators are an attractive option for producing (99)Mo on a national scale.

Sci-Sat AM: Brachy - 02: Extracting Wair from the 1976 electron beam measurements of domen and lamperti.

The average energy expended by an energetic electron to create an ion pair in air, Wair is an important quantity in radiation dosimetry. The data obtained by Domen and Lamperti using electron beams in the energy range from 15 to 50 MeV can be used to extract a value for Wair if the electron stopping powers of graphite and air are assumed to be known. We use Monte Carlo techniques to reanalyze these data and obtain a new estimate for the value of Wair Using the EGSnrc Monte Carlo and its associated user codes, as well as the best availabl-e stopping power data for graphite, we calculate the perturbation effects due to the calorimeter and ionization chamber and the effect of extrapolating from scattered to plane-parallel beams. Without further adjustments, the extracted values of Wair show a significant trend as the mean electron energy decreases. We show that part of this trend can be attributed to an incorrect value of the density assigned to the graphite absorbers and part to the likelihood that the nominal energy assigned to the low-energy electron beams is not correct. Using all the data, we obtain a value for Wair of 33.84 eV per ion pair with a relative standard uncertainty of 0.4 %. This result serves to complement values obtained using 60 Co γ-rays, for which the value of the mean excitation of graphite contributes significantly to the uncertainty.

Comments on 'Cellular response to modulated radiation fields'.

The authors of a recent paper (Claridge Mackonis et al 2007 Phys. Med. Biol. 52 5469-82) measured cell survival in spatially modulated radiation fields. They claim to have identified two new types of radiation-induced bystander effect. We conclude that their claims are not supported by their data.

Comments on 'Ionization chamber volume determination and quality assurance using micro-CT imaging'.

The authors of a recent paper (McNiven et al 2008 Phys. Med. Biol. 53 5029-43) measured the volume of a particular type of a small ionization chamber using CT images. Using four Exradin A1SL chambers, they find that the volume measured using CT imaging is, on average, 4.3% larger than the value derived from the chamber calibration coefficient. Although they point out that the effective chamber volume is defined by electric field lines between the collector and the chamber body, they do not estimate how the mechanical volume might differ from the effective volume. We have used a commercial software package to calculate the electric field in the cavity and we show that the field lines define a volume that is about 11% smaller than the mechanical volume. We also show that the effective volume is very sensitive to small changes in the chamber geometry near the base of the collector. We conclude that simply determining the mechanical volume without careful consideration of the electric field lines within the cavity is not a useful dosimetric technique.

Measurement of multiple scattering of 13 and 20 MeV electrons by thin foils.

To model the transport of electrons through material requires knowledge of how the electrons lose energy and scatter. Theoretical models are used to describe electron energy loss and scatter and these models are supported by a limited amount of measured data. The purpose of this work was to obtain additional data that can be used to test models of electron scattering. Measurements were carried out using 13 and 20 MeV pencil beams of electrons produced by the National Research Council of Canada research accelerator. The electron fluence was measured at several angular positions from 0 degree to 90 degrees for scattering foils of different thicknesses and with atomic numbers ranging from 4 to 79. The angle, theta 1/e at which the fluence has decreased to 1/e of its value on the central axis was used to characterize the distributions. Measured values of theta 1/e ranged from 1.5 degrees to 8 degrees with a typical uncertainty of about 1%. Distributions calculated using the EGSnrc Monte Carlo code were compared to the measured distributions. In general, the calculated distributions are narrower than the measured ones. Typically, the difference between the measured and calculated values of theta 1/e is about 1.5%, with the maximum difference being 4%. The measured and calculated distributions are related through a simple scaling of the angle, indicating that they have the same shape. No significant trends with atomic number were observed.

The effective point of measurement of ionization chambers and the build-up anomaly in MV x-ray beams.

A precision experimental investigation of the effective point of measurement (EPOM) of ion chambers in megavoltage beams has been carried out. A one-dimensional scanning phantom system was developed with an overall accuracy in the positioning of a chamber of better than 0.15 mm. Depth-dose data were acquired for a 25 MV beam from an Elekta Precise linac (field sizes of 10 x 10 cm and 25 x 25 cm) for measurement depths in the range 0.6-6 cm. The results confirmed the Monte Carlo calculations of an earlier theoretical investigation by Kawrakow [Med. Phys. 33, 1829-1839 (2006)] that the standard shift for cylindrical chambers, recommended in dosimetry protocols of -0.6r (where r is the internal radius of the cavity), is incorrect. A wide range of ion chambers were investigated and it was found that errors of up to 1.4 mm could occur for certain chamber designs (although typical errors for common chambers were around 0.5 mm). A comparison between measurements and Monte Carlo simulations showed that once the correct EPOM is used, the details of the linac geometry are correct, and the parameters of the electron beam striking the bremsstrahlung target have been adequately determined, the EGSnrc Monte Carlo package is capable of reproducing the experimental data to 0.2 mm or better. The investigation also confirmed that for the highest accuracy depth-dose curves in megavoltage photon beams one should use a well-guarded parallel-plate ion chamber. Three chamber designs were tested here and found to be satisfactory-the Scanditronix-Wellhöfer NACP-02, PTW Roos and Exradin All.

Sci-Fri PM: Planning-01: Measured electron and x-ray angular distribution data for benchmarking Monte Carlo codes.

Monte Carlo (MC) studies of the output of medical linear accelerators have demonstrated that in-air profiles are useful in the beam commissioning process. A recent investigation of x-ray profiles (Tonkopi et al, Med. Phys 32 (9), 2005) showed very good agreement between measurement and EGSnrc calculations but to achieve this level of agreement the beam linac spot size, energy and angular divergence had to be treated as variables. In this project we carried out measurements and MC calculations for an electron accelerator for which the initial beam parameters are well known. Two sets of investigations were carried out. In the first we measured electron scatter distributions for a range of scattering foils and electron energies of 13 and 20 MeV. The profiles were parameterised and compared to EGSnrc Monte Carlo calculations. It was found that generally the EGSnrc calculations gave agreement with the measurements within 1.5 %. In the second investigation, which is on-going, in-air profiles were obtained for photon beams produced using different targets (from beryllium to lead). Measured angular distributions were obtained using ion chambers with different build-up caps (low and high-Z) and the sensitivity of the data to small changes in geometry (e.g., moving the x-ray target) was investigated. The photon energy fluence was calculated using EGSnrc and preliminary indications are that the measured and calculated distributions agree to better than 5 %. Work supported in part by NIH grant R01 CA104777-01A2.

Twitch and tetanic properties of human thenar motor units paralyzed by chronic spinal cord injury.

Little is known about how human motor units respond to chronic paralysis. Our aim was to record surface electromyographic (EMG) signals, twitch forces, and tetanic forces from paralyzed motor units in the thenar muscles of individuals (n = 12) with chronic (1.5-19 yr) cervical spinal cord injury (SCI). Each motor unit was activated by intraneural stimulation of its motor axon using single pulses and trains of pulses at frequencies between 5 and 100 Hz. Paralyzed motor units (n = 48) had small EMGs and weak tetanic forces (n = 32 units) but strong twitch forces, resulting in half-maximal force being achieved at a median of only 8 Hz. The distributions for cumulative twitch and tetanic forces also separated less for paralyzed units than for control units, indicating that increases in stimulation frequency made a smaller relative contribution to the total force output in paralyzed muscles. Paralysis also induced slowing of conduction velocities, twitch contraction times and EMG durations. However, the elevated ratios between the twitch and the tetanic forces, but not contractile speed, correlated significantly with the extent to which unit force summated in response to different frequencies of stimulation. Despite changes in the absolute values of many electrical and mechanical properties of paralyzed motor units, most of the distributions shifted uniformly relative to those of thenar units obtained from control subjects. Thus human thenar muscles paralyzed by SCI retain a population of motor units with heterogeneous contractile properties because chronic paralysis influenced all of the motor units similarly.

Fatigue properties of human thenar motor units paralysed by chronic spinal cord injury.

Human muscles paralysed chronically by spinal cord injury (SCI) fatigue excessively. Whether these reductions in force reflect a decrease in the fatigue resistance of the motor units is unknown. Our aim was to determine the fatigability of thenar motor units paralysed chronically (10 +/- 2 years) by cervical SCI. Surface electromyographic activity (EMG) and force were recorded from 17 paralysed motor units (n = 7 subjects) in response to intraneural motor axon stimulation (13 pulses at 40 Hz, 1 s(-1) for 2 min). Unit force decreased progressively, reaching 8-60% of initial after 2 min, whereas both the amplitude and area of the first EMG potentials in the trains increased significantly (both P < 0.05). Thus, transmission of neural signals to the sarcolemma was effective and the reduction in force must reflect impaired processes in the muscle fibres. The median fatigue index for paralysed units (0.31), the ratio of the force at 2 min compared to the initial force, was significantly lower than that for units from control subjects (0.85, P < 0.05), but the distribution of fatigue indices for each population had a similar shape (ranges: 0.08-0.60 and 0.41-0.95, respectively). Hence, chronic paralysis did not limit the range of fatigability typically found for thenar units, only its magnitude. These findings suggest that all paralysed units underwent similar reductions in fatigue resistance. After fatigue, paralysed unit forces were reduced at all frequencies (1-100 Hz, P < 0.05). Twitch contraction and half-relaxation times were increased, as was the frequency needed to produce half maximal force (P < 0.05). Thus, stimulation protocols used to produce functional movements in paralysed muscles need to accommodate the significant and rapid fatigue of the motor units.

Commissioning of an NRC-type sealed water calorimeter at METAS using 60Co gamma-rays.

As part of a collaborative project between the National Research Council of Canada (NRC) and the Swiss Federal Office of Metrology and Accreditation (METAS), a sealed water calorimeter was built at NRC and transferred to METAS. The calorimeter is operated at 4 degrees C and uses two thermistor probes in a sealed glass vessel containing high-purity water to measure the radiation-induced temperature rise. The various correction factors have been evaluated and the estimated standard uncertainty on the absorbed dose to water is 0.41%. An extensive set of measurements using 60Co gamma-rays was carried out at NRC and two ionization chambers were calibrated against the absorbed dose determined calorimetrically. The chambers were also calibrated against the NRC standard for absorbed dose. After transferring the calorimeter to METAS, a similar set of measurements was carried out using their 60Co beam and the same two ionization chambers were calibrated against the absorbed dose to water established at METAS. The discrepancy between the three sets of calibration coefficients was smaller than the estimated standard uncertainty of 0.47% on the ratio of any pair of calibration coefficients.

The response of lif thermoluminescence dosemeters to photon beams in the energy range from 30 kV x rays to 60Co gamma rays.

The energy response of standard (TLD-100) and high-sensitivity (TLD-100H) LiF thermoluminescence dosemeters (TLDs) has been studied for photon beams with mean energies from about 25 keV to 1100 keV. Canadian primary standards for air kerma were used to establish the air kerma rates for each of the photon beams. TLDs were mounted in a PMMA holder and the air kerma response was measured as a function of energy. The EGSnrc Monte Carlo code was used to model the TLD holder and calculate the absorbed dose to the TLD chip per unit air kerma for each beam. The measured and calculated results were combined to obtain the intrinsic dose response of the TLD chip. Broadly, our results are consistent with existing data, which show a marked difference in the energy dependence of the two materials. However, the precision of our measurements (standard uncertainty of about 0.6%) has permitted the identification of features that have not been noted before. In particular, the energy dependence of the two materials is quite different in the important energy region delimited by 137Cs and 60Co gamma rays.

The effect of wall thickness on the response of a spherical ionization chamber.

Air-filled ionization chambers are used widely for radiation dosimetry. For some applications it is important to know the effect on the chamber response of photon attenuation and scattering in the chamber walls. Traditionally, the wall effect is determined by measuring the chamber response as a function of wall thickness and extrapolating linearly to zero thickness. We have constructed a spherical graphite chamber with variable wall thickness. The change in the chamber response with wall thickness has been measured in a 137Cs gamma-ray beam. Our data show that the change in response is not linear with wall thickness, in agreement with the theoretical prediction of Bielajew (1990 Med. Phys. 17 583-7). A linear versus non-linear extrapolation of the measured data to zero wall thickness leads to a difference of almost 1% in the estimate of the wall correction factor, Kw. The value of Kw obtained using the non-linear extrapolation is in good agreement with the result obtained using Monte Carlo techniques.

Comparison of measured and Monte Carlo calculated dose distributions from the NRC linac.

We have benchmarked photon beam simulations with the EGS4 user code BEAM [Rogers et al., Med. Phys. 22, 503-524 (1995)] by comparing calculated and measured relative ionization distributions in water from the 10 and 20 MV photon beams of the NRC linac. Unlike previous calculations, the incident electron energy is known independently to 1%, the entire extra-focal radiation is simulated, and electron contamination is accounted for. The full Monte Carlo simulation of the linac includes the electron exit window, target, flattening filter, monitor chambers, collimators, as well as the PMMA walls of the water phantom. Dose distributions are calculated using a modified version of the EGS4 user code DOSXYZ which additionally allows scoring of average energy and energy fluence in the phantom. Dose is converted to ionization by accounting for the (L/rho)water(air) variation in the phantom, calculated in an identical geometry for the realistic beams using a new EGS4 user code, SPRXYZ. The variation of (L/rho)water(air) with depth is a 1.25% correction at 10 MV and a 2% correction at 20 MV. At both energies, the calculated and the measured values of ionization on the central axis in the buildup region agree within 1% of maximum ionization relative to the ionization at 10 cm depth. The agreement is well within statistics elsewhere. The electron contamination contributes 0.35(+/- 0.02) to 1.37(+/- 0.03)% of the maximum dose in the buildup region at 10 MV and 0.26(+/- 0.03) to 3.14(+/- 0.07)% of the maximum dose at 20 MV. The penumbrae at 3 depths in each beam (in g/cm2), 1.99 (dmax, 10 MV only), 3.29 (dmax, 20 MV only), 9.79 and 19.79, agree with ionization chamber measurements to better than 1 mm. Possible causes for the discrepancy between calculations and measurements are analyzed and discussed in detail.

Absorbed-dose beam quality conversion factors for cylindrical chambers in high energy photon beams.

Recent working groups of the AAPM [Almond et al., Med. Phys. 26, 1847 (1999)] and the IAEA (Andreo et al., Draft V.7 of "An International Code of Practice for Dosimetry based on Standards of Absorbed Dose to Water," IAEA, 2000) have described guidelines to base reference dosimetry of high energy photon beams on absorbed dose to water standards. In these protocols use is made of the absorbed-dose beam quality conversion factor, kQ which scales an absorbed-dose calibration factor at the reference quality 60Co to a quality Q, and which is calculated based on state-of-the-art ion chamber theory and data. In this paper we present the measurement and analysis of beam quality conversion factors kQ for cylindrical chambers in high-energy photon beams. At least three chambers of six different types were calibrated against the Canadian primary standard for absorbed dose based on a sealed water calorimeter at 60Co [TPR10(20)=0.572, %dd(10)x=58.4], 10 MV [TPR10(20)=0.682, %dd(10)x=69.6), 20 MV (TPR10(20)=0.758, %dd(10)x= 80.5] and 30 MV [TPR10(20) = 0.794, %dd(10)x= 88.4]. The uncertainty on the calorimetric determination of kQ for a single chamber is typically 0.36% and the overall 1sigma uncertainty on a set of chambers of the same type is typically 0.45%. The maximum deviation between a measured kQ and the TG-51 protocol value is 0.8%. The overall rms deviation between measurement and the TG-51 values, based on 20 chambers at the three energies, is 0.41%. When the effect of a 1 mm PMMA waterproofing sleeve is taken into account in the calculations, the maximum deviation is 1.1% and the overall rms deviation between measurement and calculation 0.48%. When the beam is specified using TPR10(20), and measurements are compared with kQ values calculated using the version of TG-21 with corrected formalism and data, differences are up to 1.6% when no sleeve corrections are taken into account. For the NE2571 and the NE2611A chamber types, for which the most literature data are available, using %dd(10)x, all published data show a spread of 0.4% and 0.6%, respectively, over the entire measurement range, compared to spreads of up to 1.1% for both chambers when the kQ values are expressed as a function of TPR10(20). For the PR06-C chamber no clear preference of beam quality specifier could be identified. When comparing the differences of our kQ measurements and calculations with an analysis in terms of air-kerma protocols with the same underlying calculations but expressed in terms of a compound conversion factor CQ, we observe that a system making use of absorbed-dose calibrations and calculated kQ values, is more accurate than a system based on air-kerma calibrations in combination with calculated CQ (rms deviation of 0.48% versus 0.67%, respectively).

Fricke dosimetry: the difference between G(Fe3+) for 60Co gamma-rays and high-energy x-rays.

A calibration of the Fricke dosimeter is a measurement of epsilon G(Fe3+). Although G(Fe3+) is expected to be approximately energy independent for all low-LET radiation, existing data are not adequate to rule out the possibility of changes of a few per cent with beam quality. When a high-precision Fricke dosimeter, which has been calibrated for one particular low-LET beam quality, is used to measure the absorbed dose for another low-LET beam quality, the accuracy of the absorbed dose measurement is limited by the uncertainty in the value of G(Fe3+). The ratio of G(Fe3+) for high-energy x-rays (20 and 30 MV) to G(Fe3+) for 60Co gamma-rays, G(Fe3+)MV(Co), was measured to be 1.007(+/-0.003) (confidence level of 68%) using two different types of water calorimeter, a stirred-water calorimeter (20 MV) and a sealed-water calorimeter (20, 30 MV). This value is consistent with our calculations based on the LET dependence of G(primary products) and, as well, with published measurements and theoretical treatments of G(Fe3+).

Development and sensitivity to visual impairment of the Low Vision Functional Status Evaluation (LVFSE).

BACKGROUND: Functional limitation because of visual impairment is one of the most important measures of low vision rehabilitation outcome because it represents the true impact of visual impairment upon the daily lives of individuals. The Low Vision Functional Status Evaluation (LVFSE) was designed to evaluate functional limitation because of visual impairment using both observed and self-reported performance on daily activities. METHODS: Performance on 27 daily activities was evaluated for 155 visually impaired subjects. Subjects also provided ratings of task difficulty. RESULTS: Correlations between observed performance and visual acuity, contrast sensitivity, and visual field loss severity were moderate, r = -0.60 to r = 0.34. Correlations between self-reported task performance difficulty and measures of visual clinical state were weaker, r = -0.44 to r = 0.21. CONCLUSIONS: The LVFSE shows promise as a sensitive measure of low vision-related functional status but more work is needed to refine the measure.

Continuous and discontinuous changes in the unit cell of HIV-1 reverse transcriptase crystals on dehydration.

A crystal form of HIV-1 reverse transcriptase (RT) complexed with inhibitors showed diffraction to a high-resolution limit of 3.7 A. Instability in the unit-cell dimensions of these crystals was observed during soaking experiments, but the range of this variability and consequent change in lattice order was revealed by a chance observation of dehydration. Deliberately induced dehydration results in crystals having a variety of unit cells, the best-ordered of which show diffraction to a minimum Bragg spacing of 2.2 A. In order to understand the molecular basis for this phenomenon, the initial observation of dehydration, the data sets from dehydrated crystals, the crystal packing and the domain conformation of RT are analysed in detail here. This analysis reveals that the crystals undergo remarkable changes following a variety of possible dehydration pathways: some changes occur gradually whilst others are abrupt and require significant domain rearrangements. Comparison of domain arrangements in different crystal forms gives insight into the flexibility of RT which, in turn, may reflect the internal motions allowing this therapeutically important enzyme to fulfill its biological function.

3'-Azido-3'-deoxythymidine drug resistance mutations in HIV-1 reverse transcriptase can induce long range conformational changes.

HIV reverse transcriptase (RT) is one of the main targets for the action of anti-AIDS drugs. Many of these drugs [e.g., 3'-azido-3'-deoxythymidine (AZT) and 2',3'-dideoxyinosine (ddI)] are analogues of the nucleoside substrates used by the HIV RT. One of the main problems in anti-HIV therapy is the selection of a mutant virus with reduced drug sensitivity. Drug resistance in HIV is generated for nucleoside analogue inhibitors by mutations in HIV RT. However, most of these mutations are situated some distance from the polymerase active site, giving rise to questions concerning the mechanism of resistance. To understand the possible structural bases for this, the crystal structures of AZT- and ddI-resistant RTs have been determined. For the ddI-resistant RT with a mutation at residue 74, no significant conformational changes were observed for the p66 subunit. In contrast, for the AZT-resistant RT (RTMC) bearing four mutations, two of these (at 215 and 219) give rise to a conformational change that propagates to the active site aspartate residues. Thus, these drug resistance mutations produce an effect at the RT polymerase site mediated simply by the protein. It is likely that such long-range effects could represent a common mechanism for generating drug resistance in other systems.

Unique features in the structure of the complex between HIV-1 reverse transcriptase and the bis(heteroaryl)piperazine (BHAP) U-90152 explain resistance mutations for this nonnucleoside inhibitor.

The viral reverse transcriptase (RT) provides an attractive target in the search for anti-HIV therapies. The nonnucleoside inhibitors (NNIs) are a diverse set of compounds (usually HIV-1 specific) that function by distorting the polymerase active site upon binding in a nearby pocket. Despite being potent and of generally low toxicity, their clinical use has been limited by rapid selection for resistant viral populations. The 2.65-A resolution structure of the complex between HIV-1 RT and the bis(heteroaryl)piperazine (BHAP) NNI, 1-(5-methanesulfonamido-1H-indol-2-yl-carbonyl)-4- [3-(1-methyl-ethylamino) pyridinyl] piperazine (U-90152), reveals the inhibitor conformation and bound water molecules. The bulky U-90152 molecule occupies the same pocket as other NNIs, but the complex is stabilized quite differently, in particular by hydrogen bonding to the main chain of Lys-103 and extensive hydrophobic contacts with Pro-236. These interactions rationalize observed resistance mutations, notably Pro-236-Leu, which occurs characteristically for BHAPs. When bound, part of U-90152 protrudes into the solvent creating a channel between Pro-236 and the polypeptide segments 225-226 and 105-106, giving the first clear evidence of the entry mode for NNIs. The structure allows prediction of binding modes for related inhibitors [(altrylamino)piperidine-BHAPs] and suggests changes to U-90152, such as the addition of a 6 amino group to the pyridine ring, which may make binding more resilient to mutations in the RT. The observation of novel hydrogen bonding to the protein main chain may provide lessons for the improvement of quite different inhibitors.