Tumor heterogeneity on ¹⁸F-FDG-PET/CT for response monitoring in non-small cell lung cancer treated with erlotinib

With advances in molecular research, molecular-targeted agents such as epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI) have emerged for the treatment of (advanced) non-small cell lung cancer (NSCLC). In NSCLC the EGFR is over-expressed or harbours sensitizing mutations (1). Inhibition of this receptor with TKI therapy such as erlotinib blocks the tyrosine kinase domain of the EGFR, thereby inhibiting downstream signaling pathways involved in cell proliferation, angiogenesis, invasion and metastasis and prevention of apoptosis. They can be orally administered and have a relatively favorable toxicity profile and are registered for the treatment of patients with advanced (chemotherapy-refractory) NSCLC (2). Molecular biologic testing of the tumour has become paramount to consider these (often expensive) targeted treatment options.

The probability of response to EGFR-TKIs is considerably higher in patients with EGFR-mutated tumors (3). However, prediction of response is suboptimal by mutation analysis only (4). In addition, EGFR mutation positive NSCLC represents only 10–15% of all NSCLC (3). Thus, the vast majority of NSCLC are so-called “wild-type” for EGFR. For these patients more controversy arises. It is known that several patients without apparent sensitizing EGFR mutations do benefit from erlotinib therapy (5). Biopsy quality, tissue availability, and heterogeneity within the tumor are factors that may hamper molecular analysis for relevant genetic alterations (6).

Efforts have been made to identify patients most likely to respond to EGFR-TKIs, despite the presence of activating mutations. ¹⁸F-fluorodeoxyglucose positron emission tomography acquired together with low dose computed tomography (FDG-PET/CT) has proven its role as a staging modality (7). In addition, several studies demonstrated that FDG-PET/CT is able to predict response to chemotherapy (8,9).

The initial results in response monitoring of EGR-TKI with FDG-PET/CT using SUV_max are promising (10). Early response measured with FDG-PET/CT seems to correlate to histopathological response (11). However, there is on ongoing search for reliable parameters to measure response with FDG-PET/CT (12). There are several methods for measuring the rate and/or total amount of FDG-accumulation in tumors. PET scanners are designed to measure the in vivo radioactivity concentration [kBq/mL], which is directly linked to the FDG-concentration. Typically, however, it is the relative tissue uptake of FDG that is of interest. The two most significant sources of variation that occur in practice are the amount of injected FDG and the patient size. To compensate for these variations, at least to first order, SUV is commonly used as a relative measure of FDG-uptake. However, there is increasing interest in assessing the global and local-regional heterogeneity of FDG-distribution with feature analysis by using a variety of mathematical methods that describe the relationships between the gray-level intensity of pixels or voxels and their position within an image. Initial validation of the measurement of intratumoral heterogeneity on FDG-PET images appears to provide predictive information at pretherapy imaging in a number of solid tumors.

Recently Cook et al. evaluated this issue (13). The aim of their study was to determine if first-order and high-order textural features on FDG-PET images of NSCLC (I) at baseline; (II) at 6 weeks; or (III) the percentage change between baseline and 6 weeks can predict response or survival in patients treated with erlotinib. They assumed that textural features reflecting heterogeneity on FDG-PET images in patients with NSCLC who are being treated with erlotinib are associated with treatment response and survival. To verify this hypothesis they analyzed a population of 47 patients measuring: (I) First-order textural features included standard deviation, skewness, kurtosis, first-order entropy, and first-order uniformity; (II) high-order features, including coarseness, contrast, busyness, and complexity, derived from three-dimensional matrices describing differences between each PET image voxel and its neighbor, were calculated, taking into consideration for each voxel the neighboring voxels in the two adjacent planes. The median OS was 14.1 months. According to CT RECIST at 12 weeks, there were 21 non-responders and 11 responders. Response to erlotinib was associated with reduced heterogeneity (first-order standard deviation, P=0.01; entropy, P=0.001; uniformity, P=0.001). At multivariable analysis, high-order contrast at 6 weeks (P=0.002) and percentage change in first-order entropy (P=0.03) were independently associated with survival. Percentage change in first-order entropy was also independently associated with treatment response (P=0.01). However, in this analysis the texture parameters appeared to be as predictive as the SUV parameters.

Although the evaluation of Cook et al. was limited to a small series of patients the results are promising and it is possible to consider applicability of the methodology in other clinical studies provided that the calculation software of the textural features becomes available after standardization. Reproducibility for ¹⁸F-FDG textural features has been reported to be as good as or even better than the one used for SUV (14). In the work of Cook et al. measurement of all texture parameters showed a good interobserver variability.

However, other aspects must be elucidated. For instance, the clinical resolution of current PET scanners is still in the order of 4 to 5 mm, which means that for relatively small lung tumors the partial volume effect will make it challenging to accurately measure the volume for tumors with a diameter less than 3 cm with low FDG-uptake (15). Assessment of the heterogeneity within the tumor may suffer from this same lack of resolution. Despite the limitation in spatial resolution the measured SUV distribution inside the tumor still (although blurred) contains information about the heterogeneity of the tumor. Statistical methods are therefore necessary for the evaluation of this distribution.

A disadvantage of this approach is that it is not clear what type of heterogeneity is correlating with the tumor response. Information would be vital for future development towards prospective use. Relatively small tumors, like NSCLC, which are evaluated in this study, show a strong correlation between total uptake en the size of the tumor. For instance in a perfectly spherical tumor where the uptake decreases as function of the distance to the center the standard deviation of the distribution of the FGD-values will scale with the size of the tumor. In this case both total (or peak) SUV and standard deviation of the measured SUV will decrease when the tumor shrinks in size. In addition, the noise within the voxels follows a Poisson distribution, which results in noise, scales with the square root of the counts per voxel. Here a pitfall of using SUV emerges. In SUV calculation, the measured counts per voxel are normalized to the injected dose and body weight. This potentially results in comparable SUV values but different counts per voxel. So when statistics are used to characterize the tumor, SUV values can be misleading.

The approach chosen by Cook et al. to evaluate the statistics of the distribution without modeling the change in this distribution inflicted by the therapy appears suboptimal for gaining more insight in this potential interesting technique for evaluation of tumor response.

An important feature of metabolic response monitoring is the possibility to identify patients who will potentially benefit from therapy. In contrast, patients who do not benefit from therapy are only exposed to potential toxicity for a short period of time (16). In addition, the successes of new therapeutic agents have led to increases in health care costs to a level that is now causing a serious financial burden to patients, hospitals and society (17). With early assessment futile use of medication can be avoided, and patients who do not respond to EGFR-TKI’s may switch to other, more effective treatment.

Several limitations may occur with metabolic treatment monitoring and here we can underline two important features. First, FDG-uptake on PET may reflect various tissue reactions, as tumor progression or regression but also senescence, fibrosis formation, and inflammatory reactions as macrophage infiltration (18). The second consideration concerns the response to erlotinib. It can be expected to develop within several weeks, but apoptosis, transition of necrosis into fibrosis, and inflammatory and granulomatous reactions are difficult to quantify (19). Since some spontaneous necrosis exists in most NSCLCs, therapy response can be a combination of a decrease in the total amount of viable tumor cells and/or a decrease of FDG metabolism at cellular level. FDG-PET cannot differentiate between these two types of responses. Moreover FDG-uptake by the tumor is also dependent on perfusion and the rate of clearance by for instance the kidney (20). These factors make that tumor response assessment with absolute uptake of the tumor in total (expressed as SUV max or mean) is difficult to interpret. Other aspects of tumor response like volume or heterogeneity can also be measured using FDG-uptake. A big advantage is that these measured values are not solely dependent on the absolute uptake but take changes within the tumor into account.

In conclusion, Cook et al. have shown that response monitoring using FDG-PET/CT textural features has potential in targeted treatment with erlotinib in NSCLC patients. Patients with substantial decrease of metabolic activity during erlotinib treatment will probably benefit from continued treatment. However, various aspects of the method (quantification tools, cut-off values, etc.) need to be standardized before the software becomes widely available in a similar manner as SUV-measurements. They opened an additional window for innovation but simultaneously a new challenge for molecular hybrid imaging.

Acknowledgements

None.

Footnote

Provenance: This is an invited article commissioned by the Guest Editor Lihua Chen [(Department of Radiology, Taihu Hospital (PLA 101Hospital), Wuxi, China)].

Conflicts of Interest: The authors have no conflicts of interest to declare.

References

1. Soria JC, Mok TS, Cappuzzo F, et al. EGFR-mutated oncogene-addicted non-small cell lung cancer: current trends and future prospects. Cancer Treat Rev 2012;38:416-30. [Crossref] [PubMed]
2. Rosell R, Carcereny E, Gervais R, et al. Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol 2012;13:239-46. [Crossref] [PubMed]
3. Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497-500. [Crossref] [PubMed]
4. Kawahara A, Yamamoto C, Nakashima K, et al. Molecular diagnosis of activating EGFR mutations in non-small cell lung cancer using mutation-specific antibodies for immunohistochemical analysis. Clin Cancer Res 2010;16:3163-70. [Crossref] [PubMed]
5. Gridelli C, De Marinis F, Di Maio M, et al. Gefitinib as first-line treatment for patients with advanced non-small-cell lung cancer with activating Epidermal Growth Factor Receptor mutation: implications for clinical practice and open issues. Lung Cancer 2011;72:3-8. [Crossref] [PubMed]
6. Masters GA, Temin S, Azzoli CG, et al. Systemic therapy for stage IV non-small-cell lung cancer: American Society of Clinical Oncology Clinical Practice Guideline Update. J Clin Oncol 2015;33:3488-515. [Crossref] [PubMed]
7. Schmidt-Hansen M, Baldwin DR, Hasler E, et al. PET-CT for assessing mediastinal lymph node involvement in patients with suspected resectable non-small cell lung cancer. Cochrane Database Syst Rev 2014;11:CD009519. [PubMed]
8. Dose Schwarz J, Bader M, Jenicke L, et al. Early prediction of response to chemotherapy in metastatic breast cancer using sequential ¹⁸F-FDG PET. J Nucl Med 2005;46:1144-50. [PubMed]
9. Terasawa T, Lau J, Bardet S, et al. Fluorine-18-fluorodeoxyglucose positron emission tomography for interim response assessment of advanced-stage Hodgkin's lymphoma and diffuse large B-cell lymphoma: a systematic review. J Clin Oncol 2009;27:1906-14. [Crossref] [PubMed]
10. van Gool MH, Aukema TS, Hartemink KJ, et al. FDG-PET/CT response evaluation during EGFR-TKI treatment in patients with NSCLC. World J Radiol 2014;6:392-8. [Crossref] [PubMed]
11. van Gool MH, Aukema TS, Schaake EE, et al. Timing of metabolic response monitoring during erlotinib treatment in non-small cell lung cancer. J Nucl Med 2014;55:1081-6. [Crossref] [PubMed]
12. Kahraman D, Holstein A, Scheffler M, et al. Tumor lesion glycolysis and tumor lesion proliferation for response prediction and prognostic differentiation in patients with advanced non-small cell lung cancer treated with erlotinib. Clin Nucl Med 2012;37:1058-64. [Crossref] [PubMed]
13. Cook GJ, O'Brien ME, Siddique M, et al. Non-small cell lung cancer treated with erlotinib: heterogeneity of (18)F-FDG uptake at PET-association with treatment response and prognosis. Radiology 2015;276:883-93. [Crossref] [PubMed]
14. Tixier F, Hatt M, Le Rest CC, et al. Reproducibility of tumor uptake heterogeneity characterization through textural feature analysis in 18F-FDG PET. J Nucl Med 2012;53:693-700. [Crossref] [PubMed]
15. Pawaroo D, Cummings NM, Musonda P, et al. Non-small cell lung carcinoma: accuracy of PET/CT in determining the size of T1 and T2 primary tumors. AJR Am J Roentgenol 2011;196:1176-81. [Crossref] [PubMed]
16. Schaake EE, Kappers I, Codrington HE, et al. Tumor response and toxicity of neoadjuvant erlotinib in patients with early-stage non-small-cell lung cancer. J Clin Oncol 2012;30:2731-8. [Crossref] [PubMed]
17. Meropol NJ, Schrag D, Smith TJ, et al. American Society of Clinical Oncology guidance statement: the cost of cancer care. J Clin Oncol 2009;27:3868-74. [Crossref] [PubMed]
18. Gupta NC, Graeber GM, Bishop HA. Comparative efficacy of positron emission tomography with fluorodeoxyglucose in evaluation of small (<1 cm), intermediate (1 to 3 cm), and large (>3 cm) lymph node lesions. Chest 2000;117:773-8. [Crossref] [PubMed]
19. Lara-Guerra H, Chung CT, Schwock J, et al. Histopathological and immunohistochemical features associated with clinical response to neoadjuvant gefitinib therapy in early stage non-small cell lung cancer. Lung Cancer 2012;76:235-41. [Crossref] [PubMed]
20. Toriihara A, Kitazume Y, Nishida H, et al. Comparison of FDG-PET/CT images between chronic renal failure patients on hemodialysis and controls. Am J Nucl Med Mol Imaging 2015;5:204-11. [PubMed]