Osimertinib + Savolitinib to Overcome Acquired MET-Mediated Resistance in Epidermal Growth Factor Receptor-Mutated, MET-Amplified Non-Small Cell Lung Cancer: TATTON.

MET-inhibitor and EGFR tyrosine kinase inhibitor (EGFR-TKI) combination therapy could overcome acquired MET-mediated osimertinib resistance. We present the final phase Ib TATTON (NCT02143466) analysis (Part B, n = 138/Part D, n = 42) assessing oral savolitinib 600 mg/300 mg once daily (q.d.) + osimertinib 80 mg q.d. in patients with MET-amplified, EGFR-mutated (EGFRm) advanced non-small cell lung cancer (NSCLC) and progression on prior EGFR-TKI. An acceptable safety profile was observed. In Parts B and D, respectively, objective response rates were 33% to 67% and 62%, and median progression-free survival (PFS) was 5.5 to 11.1 months and 9.0 months. Increased antitumor activity may occur with MET copy number ≥10. EGFRm circulating tumor DNA clearance on treatment predicted longer PFS in patients with detectable baseline ctDNA, while acquired resistance mechanisms to osimertinib + savolitinib were mediated by MET, EGFR, or KRAS alterations. SIGNIFICANCE: The savolitinib + osimertinib combination represents a promising therapy in patients with MET-amplified/overexpressed, EGFRm advanced NSCLC with disease progression on a prior EGFR-TKI. Acquired resistance mechanisms to this combination include those via MET, EGFR, and KRAS. On-treatment ctDNA dynamics can predict clinical outcomes and may provide an opportunity to inform earlier decision-making. This article is highlighted in the In This Issue feature, p. 1.

Comparing Go/No-Go Decision-Making Properties Between Single Arm Phase II Trial Designs in Oncology.

INTRODUCTION: Simon's design has been widely used in oncology to conduct single arm phase II trials and to make Go/No-Go development decision. Other authors have proposed designs with decision-making frameworks that include a third, "Consider" outcome. For results in the Consider zone, a final Go/No-Go development decision must still be made; however it is typically a subjective decision based on the totality of data and the development landscape. Under this framework, the probability of continuing development when the candidate therapy is truly ineffective or the probability of stopping development when the candidate therapy is truly effective is undefined. METHODS: We use a motivating example to compare end of trial decision-making between Simon's two-stage approach and a Multilevel outcome approach. We present the minimum and maximum development decision error probabilities by varying whether candidates that end in the Consider zone would ultimately continue with development or not. RESULTS: The Multilevel approach typically requires fewer patients, but the risk of making an incorrect drug development decision is inflated above the statistically defined Type I and Type II error rates. Compared to a Type I error rate of 20%, the Multilevel trial's maximum probability of moving forward with an ineffective therapy is 22%, 27%, and 36% for Consider zone sizes of 10%, 20%, and 30%, respectively. CONCLUSION: The Multilevel approach provides flexibility in interpreting moderate efficacy results. However, the flexibility is accomplished with a lower sample size and corresponding uncertainty in the trial outcome that increases the risk of incorrect drug development decisions.

Clinical evaluation of the potential drug-drug interactions of savolitinib: Interaction with rifampicin, itraconazole, famotidine or midazolam.

AIMS: We investigated savolitinib pharmacokinetics (PK) when administered alone or in combination with rifampicin, itraconazole or famotidine, and investigated midazolam PK when administered with or without savolitinib in healthy males. METHODS: Savolitinib PK was evaluated before/after: rifampicin (600 mg once daily [QD] for 5 days); itraconazole (200 mg QD for 5 days); a single dose of famotidine (40 mg QD) 2 hours before savolitinib. Midazolam PK was evaluated before/after midazolam (1 mg QD) with or without savolitinib (600 mg QD). Each study enrolled 20, 16, 16 and 14 volunteers, respectively. Plasma samples were collected to determine the effect on PK. RESULTS: The geometric mean ratios (GMR, %) (90% confidence intervals [CIs]) for savolitinib alone and in combination for Cmax , AUC respectively, were 45.4 (41.4-49.9), 38.5 (34.2-43.3) in the rifampicin study (n = 18); 105.2 (87.7-126.3), 108.4 (96.3-122.1) in the itraconazole study (n = 16); and 78.8 (67.7-91.7), 87.4 (81.2-94.2) in the famotidine study (n = 16). The GMRs (90% CIs) for midazolam alone and in combination with savolitinib for Cmax , AUC respectively, were 84.1 (70.0-101.0), 96.7 (92.4-101.1) (n = 14). Savolitinib alone or in combination was well tolerated. CONCLUSIONS: Co-dosing of rifampicin significantly reduced exposure to savolitinib vs savolitinib alone; co-dosing of itraconazole or midazolam with savolitinib had no clinically significant effect on savolitinib or midazolam PK, respectively. Co-dosing of famotidine with savolitinib reduced exposure to savolitinib, although this was not considered clinically meaningful. No new savolitinib-related safety findings were observed.

Ceralasertib (AZD6738), an Oral ATR Kinase Inhibitor, in Combination with Carboplatin in Patients with Advanced Solid Tumors: A Phase I Study.

PURPOSE: This study reports the safety, tolerability, MTD, recommended phase II dose (RP2D), pharmacokinetic/pharmacodynamic profile, and preliminary antitumor activity of ceralasertib combined with carboplatin in patients with advanced solid tumors. It also examined exploratory predictive and pharmacodynamic biomarkers. PATIENTS AND METHODS: Eligible patients (n = 36) received a fixed dose of carboplatin (AUC5) with escalating doses of ceralasertib (20 mg twice daily to 60 mg once daily) in 21-day cycles. Sequential and concurrent combination dosing schedules were assessed. RESULTS: Two ceralasertib MTD dose schedules, 20 mg twice daily on days 4-13 and 40 mg once daily on days 1-2, were tolerated with carboplatin AUC5; the latter was declared the RP2D. The most common treatment-emergent adverse events (Common Terminology Criteria for Adverse Events grade ≥3) were anemia (39%), thrombocytopenia (36%), and neutropenia (25%). Dose-limiting toxicities of grade 4 thrombocytopenia (n = 2; including one grade 4 platelet count decreased) and a combination of grade 4 thrombocytopenia and grade 3 neutropenia occurred in 3 patients. Ceralasertib was quickly absorbed (tmax ∼1 hour), with a terminal plasma half-life of 8-11 hours. Upregulation of pRAD50, indicative of ataxia telangiectasia mutated (ATM) activation, was observed in tumor biopsies during ceralasertib treatment. Two patients with absent or low ATM or SLFN11 protein expression achieved confirmed RECIST v1.1 partial responses. Eighteen of 34 (53%) response-evaluable patients had RECIST v1.1 stable disease. CONCLUSIONS: The RP2D for ceralasertib plus carboplatin was established as ceralasertib 40 mg once daily on days 1-2 administered with carboplatin AUC5 every 3 weeks, with pharmacokinetic and pharmacodynamic studies confirming pharmacodynamic modulation and preliminary evidence of antitumor activity observed.

Osimertinib plus savolitinib in patients with EGFR mutation-positive, MET-amplified, non-small-cell lung cancer after progression on EGFR tyrosine kinase inhibitors: interim results from a multicentre, open-label, phase 1b study.

BACKGROUND: Preclinical data suggest that EGFR tyrosine kinase inhibitors (TKIs) plus MET TKIs are a possible treatment for EGFR mutation-positive lung cancers with MET-driven acquired resistance. Phase 1 safety data of savolitinib (also known as AZD6094, HMPL-504, volitinib), a potent, selective MET TKI, plus osimertinib, a third-generation EGFR TKI, have provided recommended doses for study. Here, we report the assessment of osimertinib plus savolitinib in two global expansion cohorts of the TATTON study. METHODS: In this multi-arm, multicentre, open-label, phase 1b study, we enrolled adult patients (aged ≥18 years) with locally advanced or metastatic, MET-amplified, EGFR mutation-positive non-small-cell lung cancer, who had progressed on EGFR TKIs. We considered two expansion cohorts: parts B and D. Part B consisted of three cohorts of patients: those who had been previously treated with a third-generation EGFR TKI (B1) and those who had not been previously treated with a third-generation EGFR TKI who were either Thr790Met negative (B2) or Thr790Met positive (B3). In part B, patients received oral osimertinib 80 mg and savolitinib 600 mg daily; after a protocol amendment (March 12, 2018), patients who weighed no more than 55 kg received a 300 mg dose of savolitinib. Part D enrolled patients who had not previously received a third-generation EGFR TKI and were Thr790Met negative; these patients received osimertinib 80 mg plus savolitinib 300 mg. Primary endpoints were safety and tolerability, which were assessed in all dosed patients. Secondary endpoints included the proportion of patients who had an objective response per RECIST 1.1 and was assessed in all dosed patients and all patients with centrally confirmed MET amplification. Here, we present an interim analysis with data cutoff on March 29, 2019. This study is registered with ClinicalTrials.gov, NCT02143466. FINDINGS: Between May 26, 2015, and Feb 14, 2019, we enrolled 144 patients into part B and 42 patients into part D. In part B, 138 patients received osimertinib plus savolitinib 600 mg (n=130) or 300 mg (n=8). In part D, 42 patients received osimertinib plus savolitinib 300 mg. 79 (57%) of 138 patients in part B and 16 (38%) of 42 patients in part D had adverse events of grade 3 or worse. 115 (83%) patients in part B and 25 (60%) patients in part D had adverse events possibly related to savolitinib and serious adverse events were reported in 62 (45%) patients in part B and 11 (26%) patients in part D; two adverse events leading to death (acute renal failure and death, cause unknown) were possibly related to treatment in part B. Objective partial responses were observed in 66 (48%; 95% CI 39-56) patients in part B and 23 (64%; 46-79) in part D. INTERPRETATION: The combination of osimertinib and savolitinib has acceptable risk-benefit profile and encouraging antitumour activity in patients with MET-amplified, EGFR mutation-positive, advanced NSCLC, who had disease progression on a previous EGFR TKI. This combination might be a potential treatment option for patients with MET-driven resistance to EGFR TKIs. FUNDING: AstraZeneca.

A Phase Ib Open-Label Multicenter Study of AZD4547 in Patients with Advanced Squamous Cell Lung Cancers.

Purpose: Squamous cell lung cancers (SQCLC) account for 25% of all NSCLCs, yet the prognosis of these patients is poor and treatment options are limited. Amplified FGFR1 is one of the most common oncogenic events in SQCLCs, occurring in approximately 20% of cases. AZD4547 is a potent and selective FGFR1-3 inhibitor with antitumor activity in FGFR1-amplified SQCLC cell lines and patient-derived xenografts.Experimental Design: On the basis of these data, we performed a phase I study of AZD4547 in patients with previously treated stage IV FGFR1-amplified SQCLCs (NCT00979134). FGFR1 amplification (FGFR1:CEP8 ≥ 2) was determined by FISH. The primary endpoint was safety/tolerability. Secondary endpoints included antitumor activity, pharmacokinetics, pharmacodynamics, and molecular analyses.Results: Fifteen FGFR1-amplified patients were treated. The most common related adverse events (AE) were gastrointestinal and dermatologic. Grade ≥3-related AEs occurred in 3 patients (23%). Thirteen patients were evaluable for radiographic response assessment. The overall response rate was 8% (1 PR). Two of 15 patients (13.3%) were progression-free at 12 weeks, and the median overall survival was 4.9 months. Molecular tests, including next-generation sequencing, gene expression analysis, and FGFR1 immunohistochemistry, showed poor correlation between gene amplification and expression, potential genomic modifiers of efficacy, and heterogeneity in 8p11 amplicon.Conclusions: AZD4547 was tolerable at a dosage of 80 mg oral twice a day, with modest antitumor activity. Detailed molecular studies show that these tumors are heterogeneous, with a range of mutational covariates and stark differences in gene expression of the 8p11 amplicon that likely explain the modest efficacy of FGFR inhibition in this disease. Clin Cancer Res; 23(18); 5366-73. ©2017 AACR.

Metabolic Disposition of Osimertinib in Rats, Dogs, and Humans: Insights into a Drug Designed to Bind Covalently to a Cysteine Residue of Epidermal Growth Factor Receptor.

Preclinical and clinical studies were conducted to determine the metabolism and pharmacokinetics of osimertinib and key metabolites AZ5104 and AZ7550. Osimertinib was designed to covalently bind to epidermal growth factor receptors, allowing it to achieve nanomolar cellular potency (Finlay et al., 2014). Covalent binding was observed in incubations of radiolabeled osimertinib with human and rat hepatocytes, human and rat plasma, and human serum albumin. Osimertinib, AZ5104, and AZ7550 were predominantly metabolized by CYP3A. Seven metabolites were detected in human hepatocytes, also observed in rat or dog hepatocytes at similar or higher levels. After oral administration of radiolabeled osimertinib to rats, drug-related material was widely distributed, with the highest radioactivity concentrations measured at 6 hours postdose in most tissues; radioactivity was detectable in 42% of tissues 60 days postdose. Concentrations of [(14)C]-radioactivity in blood were lower than in most tissues. After the administration of a single oral dose of 20 mg of radiolabeled osimertinib to healthy male volunteers, ∼19% of the dose was recovered by 3 days postdose. At 84 days postdose, mean total radioactivity recovery was 14.2% and 67.8% of the dose in urine and feces. The most abundant metabolite identified in feces was AZ5104 (∼6% of dose). Osimertinib accounted for ∼1% of total radioactivity in the plasma of non-small cell lung cancer patients after 22 days of 80-mg osimertinib once-daily treatment; the most abundant circulatory metabolites were AZ7550 and AZ5104 (<10% of total osimertinib-related material). Osimertinib is extensively distributed and metabolized in humans and is eliminated primarily via the fecal route.

Decision-making in early clinical drug development.

This paper illustrates an approach to setting the decision framework for a study in early clinical drug development. It shows how the criteria for a go and a stop decision are calculated based on pre-specified target and lower reference values. The framework can lead to a three-outcome approach by including a consider zone; this could enable smaller studies to be performed in early development, with other information either external to or within the study used to reach a go or stop decision. In this way, Phase I/II trials can be geared towards providing actionable decision-making rather than the traditional focus on statistical significance. The example provided illustrates how the decision criteria were calculated for a Phase II study, including an interim analysis, and how the operating characteristics were assessed to ensure the decision criteria were robust. Copyright © 2016 John Wiley & Sons, Ltd.

AZD9291 in EGFR inhibitor-resistant non-small-cell lung cancer.

BACKGROUND: The EGFR T790M mutation is the most common mechanism of drug resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors in patients who have lung cancer with an EGFR mutation (EGFR-mutated lung cancer). In preclinical models, the EGFR inhibitor AZD9291 has been shown to be effective against both EGFR tyrosine kinase inhibitor-sensitizing and T790M resistance mutations. METHODS: We administered AZD9291 at doses of 20 to 240 mg once daily in patients with advanced lung cancer who had radiologically documented disease progression after previous treatment with EGFR tyrosine kinase inhibitors. The study included dose-escalation cohorts and dose-expansion cohorts. In the expansion cohorts, prestudy tumor biopsies were required for central determination of EGFR T790M status. Patients were assessed for safety, pharmacokinetics, and efficacy. RESULTS: A total of 253 patients were treated. Among 31 patients enrolled in the dose-escalation cohorts, no dose-limiting toxic effects occurred at the doses evaluated. An additional 222 patients were treated in five expansion cohorts. The most common all-cause adverse events were diarrhea, rash, nausea, and decreased appetite. The overall objective tumor response rate was 51% (95% confidence interval [CI], 45 to 58). Among 127 patients with centrally confirmed EGFR T790M who could be evaluated for response, the response rate was 61% (95% CI, 52 to 70). In contrast, among 61 patients without centrally detectable EGFR T790M who could be evaluated for response, the response rate was 21% (95% CI, 12 to 34). The median progression-free survival was 9.6 months (95% CI, 8.3 to not reached) in EGFR T790M-positive patients and 2.8 months (95% CI, 2.1 to 4.3) in EGFR T790M-negative patients. CONCLUSIONS: AZD9291 was highly active in patients with lung cancer with the EGFR T790M mutation who had had disease progression during prior therapy with EGFR tyrosine kinase inhibitors. (Funded by AstraZeneca; ClinicalTrials.gov number, NCT01802632.).

Treatment of patients with the hypereosinophilic syndrome with mepolizumab.

BACKGROUND: The hypereosinophilic syndrome is a group of diseases characterized by persistent blood eosinophilia, defined as more than 1500 cells per microliter with end-organ involvement and no recognized secondary cause. Although most patients have a response to corticosteroids, side effects are common and can lead to considerable morbidity. METHODS: We conducted an international, randomized, double-blind, placebo-controlled trial evaluating the safety and efficacy of an anti-interleukin-5 monoclonal antibody, mepolizumab, in patients with the hypereosinophilic syndrome. Patients were negative for the FIP1L1-PDGFRA fusion gene and required prednisone monotherapy, 20 to 60 mg per day, to maintain a stable clinical status and a blood eosinophil count of less than 1000 per microliter. Patients received either intravenous mepolizumab or placebo while the prednisone dose was tapered. The primary end point was the reduction of the prednisone dose to 10 mg or less per day for 8 or more consecutive weeks. RESULTS: The primary end point was reached in 84% of patients in the mepolizumab group, as compared with 43% of patients in the placebo group (hazard ratio, 2.90; 95% confidence interval [CI], 1.59 to 5.26; P<0.001) with no increase in clinical activity of the hypereosinophilic syndrome. A blood eosinophil count of less than 600 per microliter for 8 or more consecutive weeks was achieved in 95% of patients receiving mepolizumab, as compared with 45% of patients receiving placebo (hazard ratio, 3.53; 95% CI, 1.94 to 6.45; P<0.001). Serious adverse events occurred in seven patients receiving mepolizumab (14 events, including one death; mean [+/-SD] duration of exposure, 6.7+/-1.9 months) and in five patients receiving placebo (7 events; mean duration of exposure, 4.3+/-2.6 months). CONCLUSIONS: Our study shows that treatment with mepolizumab, an agent designed to target eosinophils, can result in corticosteroid-sparing for patients negative for FIP1L1-PDGFRA who have the hypereosinophilic syndrome. (ClinicalTrials.gov number, NCT00086658 [ClinicalTrials.gov].).

Pergolide versus levodopa monotherapy in early Parkinson's disease patients: The PELMOPET study.

Dopamine agonists are used as initial treatment in patients with Parkinson's disease (PD) to reduce incidence and severity of motor complications. This paradigm is based on long-term studies, allowing "rescue" therapy with levodopa. The present strict monotherapy study (PELMOPET, the acronym for the pergolide-versus-L-dopa-monotherapy-and-positron-emission-tomography trial) evaluated the efficacy and safety of pergolide versus levodopa without levodopa "rescue" medication. This multicenter, double-blind, randomized, 3-year trial compared pergolide monotherapy (n=148) with levodopa monotherapy (n=146) in dopamine-naive patients with early PD (Hoehn and Yahr stage 1-2.5). Primary efficacy measures were clinical efficacy, severity and time to onset of motor complications, and disease progression. During the 3 years, severity of motor complications was significantly lower and time to onset of dyskinesia was significantly delayed in the group receiving pergolide (3.23 mg/day) compared with those receiving levodopa (504 mg/day). However, time to onset of motor complications was not longer in patients receiving pergolide after 3 years. Symptomatic relief (assessed by Unified Parkinson's Disease Rating Scale [UPDRS], UPDRS II, and III, Clinical Global Impressions [CGI] severity, and CGI and Patient Global Impressions [PGI] improvement) was significantly greater in patients receiving levodopa. Adverse events led to discontinuation of therapy in 17.6% of pergolide patients and 9.6% of levodopa patients. This is the first study comparing strict monotherapy with a dopamine agonist versus levodopa in previously untreated early PD. In principle, both levodopa and a dopamine agonist such as pergolide seem to be suitable options as initial PD therapy. The choice remains with the treating physician based on the different efficacy and adverse event profiles.

Body composition, IGF-I and IGFBP-3 concentrations as outcome measures in severely GH-deficient (GHD) patients after childhood GH treatment: a comparison with adult onset GHD patients.

If GH therapy of children with GH deficiency (GHD) has been adequate, body composition should be comparable to that of patients who have undergone normal childhood development and become hypopituitary thereafter. To assess this, body composition was determined in 92 patients with childhood onset (CO) GHD, aged 18-30 yr, who had been treated to final height with GH for 8.98 +/- 4.30 yr and had stopped treatment 1.57 +/- 1.20 yr previously, but who remained GHD (assessed by a GH stimulation test and IGF-I values). These were compared with 35 age-matched GH-naïve hypopituitary patients with adult onset (AO) GHD. Lean body mass, fat mass, and total bone mineral content were assessed by dual energy x-ray absorptiometry and corrected for actual height. CO patients were shorter (CO height, -1.18 +/- 1.16 SD score; AO height, -0.38 +/- 1.12 SD score; P < 0.001) and had lower body mass index (CO, 23.19 +/- 5.76 kg/m(2); AO, 28.9 +/- 6.27 kg/m(2); P < 0.001) than the AO group. Although there were gender differences, within genders CO patients had lower lean body mass, fat mass, and bone mineral content (P < 0.001 in all cases) compared with AO patients. Standard deviation scores for IGF-I (CO female, -9.2 +/- 3.1; AO female, -5.2 +/- 2.6; CO male, -6.4 +/- 2.7; AO male, -3.5 +/- 2.3; P < 0.001 within each gender) and IGFBP-3 (CO female, -3.5 +/- 2.5; AO female, -1.7 +/- 1.5; CO male, -2.8 +/- 2.0; AO male, -1.1 +/- 1.6; P < 0.001 within each gender) were significantly lower in the CO group. These results suggest that patients with CO GHD who were treated to final height suffer a significant maturational deficit despite GH replacement during childhood.

Short-term safety and efficacy of human GH replacement therapy in 595 adults with GH deficiency: a comparison of two dosage algorithms.

The aim of GH replacement therapy in GH-deficient adults is to optimize response with minimum incidence of adverse reactions, but optimal therapy regimens are still to be established. This two-arm parallel study examined effects of two GH dose algorithms in adults with GH deficiency of adult or childhood onset. Patients on low dose (LD; n = 302) received GH at 3 microg/kg per day for 3 months increasing to 6 microg/kg per day for 3 months, and those on conventional dose (CD; n = 293) started on 6 microg/kg per day for 3 months increasing to 12 microg/kg per day for 3 months. The proportion of patients completing therapy was greater for the LD group than the CD group for the first 3 months (93.0% vs. 88.1%; P = 0.037) and overall for the 6 months (90.7% vs. 84.0%; P = 0.013). Both dose groups showed significant increases in lean body mass and decreases in fat mass for all time points. Percent increase in lean body mass was less with LD than CD over the first 3 months (2.43 +/- 4.33 vs. 3.58 +/- 4.69%; P = 0.006) but not overall for the 6-month period (4.38% +/- 5.34% vs. 5.21% +/- 5.99%; P = 0.141). Percent decrease in fat mass was less with LD than CD for the first 3 months (-2.81% +/- 7.81% vs. -5.53% +/- 8.64%; P < 0.001) and overall for the 6-month period (-6.35% +/- 9.42% vs. -9.45% +/- 12.07%; P = 0.006). IGF-I SD score increased less with LD than CD for 0 to 3 and 0 to 6 months, although for IGF-binding protein-3 SD score, there was no significant difference between doses at any time. Arthralgia was the only adverse event that occurred significantly less frequently with LD than with CD. Calculated changes based on gender and onset indicated greater changes in males than females for body composition, but there was little difference in GH-related adverse events between males and females. The lower starting dose with dose titration appeared more favorable, but differences in response between genders and onset of GH deficiency need to be taken into account when setting an individual dose regimen.

Confirmation of severe GH deficiency after final height in patients diagnosed as GH deficient during childhood.

OBJECTIVE: Human GH treatment of patients with childhood-onset (CO) growth hormone deficiency (GHD) ceases when they reach final height; this provides an opportunity to retest GH status in all patients before determining whether GH therapy will be required in adult life. At present, the diagnostic approach to these patients is not fully standardized. This study aimed to characterize a large group of previously GH-treated CO GHD patients and establish their GH status. PATIENTS AND METHODS: The multinational study included 167 patients diagnosed as GH deficient and treated with hGH to final height during childhood. Mean age was 19.2 years and mean height standard deviation score (SDS) was -1.08. Peak serum GH concentrations were determined in standard GH stimulation tests. IGF-I and IGFBP-3 concentrations were determined at a central laboratory and converted to SDS values by reference to a normal population. RESULTS: Using only a peak GH value of less than 3 microg/l (1 mg = 3 U) in stimulation tests as the cut-off, 133 (79.6%) patients would be classed as GH deficient. Using only an IGF-I value less than -2 SDS as the cut-off, 134 (80.2%) patients would be classed as GH deficient. However, by using both criteria there were 120 (71.9%) patients who were definitely severely GH deficient (group 1) and 20 (12.0%) who were not GH deficient (group 2), leaving 14 (8.4%) classed as GH deficient from IGF-I SDS only (group 3) and 13 (7.8%) classed as GH deficient from stimulation test only (group 4). There was no difference between the groups in height SDS or body mass index (BMI), but the GH-deficient patients tended to have been diagnosed at a younger age (group 1, 8.2 +/- 3.9; group 2, 10.0 +/- 4.0; P = 0.052). For patients classed as GH deficient compared with those not GH deficient, the percentage of males was lower (group 1, 64.2%; group 2, 90.0%; P = 0.022) and the percentage with multiple pituitary hormone deficiencies was higher (group 1, 81.7%; group 2, 20.0%; P < 0 .001), with the other two groups being intermediate in each case. Only the group classed as GH deficient by both criteria had a mean IGFBP-3 less than -2 SDS and both IGF-I SDS and IGFBP-3 SDS increased steadily across the four groups. CONCLUSIONS: A high percentage (71.9%) of these childhood-onset GH-deficient patients were still GH deficient in adult life and are likely to require further hGH treatment. While 12.0% could be classed as definitely no longer GH deficient, there are some patients who are intermediate (16.2%) and may be classed as GH deficient by one criterion but not the other. When GH stimulation test results and IGF-I concentration are discordant, the IGFBP-3 level does not establish diagnosis and the hGH treatment requirement of such patients remains a dilemma.