RU2806483C1 - Method of assessing immediate results of treatment of patients with disseminated forms of malignant tumors - Google Patents

Abstract

FIELD: medicine; oncology.

SUBSTANCE: invention can be used to assess the sensitivity of a tumor to immuno-oncological medicinal products. The level of spontaneous production of IL-8 (IL-8-KR1, IL-8-M) and IL-10 (IL-10-KR1, IL-10-M) and induced production of IL-12 (IL-12-KP1, IL-12-M) is determined in the patient’s blood (CR) and in the tumor microenvironment (M) before the start of each cycle of treatment with immuno-oncological medicinal products, and the same parameters are also evaluated in the blood (IL-8-KP2, IL-10-KP2, IL -12-KR2) after each treatment cycle. Based on the values obtained, the following coefficients are calculated:

K1= IL-8-M/ IL-8-KP1 - IL-8-M/ IL-8-KP2;

K2 = IL-10-M/ IL-10-KP1 - IL-10-M/ IL-10-KP2;

K3 = IL-12-M/ IL-12-KP2 - IL-12-M/ IL-12-KP1. If the value of at least one of the coefficients is equal to or higher than 1.7, this is regarded as an early sign of progression.

EFFECT: method provides the opportunity to increase the accuracy of assessing the immediate results of treatment of patients with disseminated forms of malignant tumors through the use of immunological parameters of the blood and microenvironment.

1 cl, 8 tbl

Description

The invention relates to medicine, more precisely to oncology, and can find application in the treatment of malignant neoplasms.

Currently, there are three groups of instrumental methods for assessing the immediate results of treatment: 1) methods associated with assessing the linear dimensions of tumor foci; 2) methods for assessing the metabolic response of the tumor; 3) methods related to the volumetry of tumor foci. Each of these can be combined with an assessment of molecular biological factors associated with biological processes associated with carcinogenesis and reflecting the degree of damage to the tumor and its microenvironment (Frontiers in Oncology. 2022; 25 (12): 982983).

The first group includes the WHO criteria, RECIST (Response Evaluation Criteria in Solid Tumors, criteria for assessing the response of solid tumors) and their modifications (Table 1).

Table 1. Comparison of criteria for assessing immediate treatment outcomes: WHO, RECIST v. 1.0, RECIST v. 1.1 WHO RECIST v1.0 RECIST v1.1 Evaluation method Two dimensions. Sum of perpendicular maximum sizes of lesions One dimension. Sum of largest sizes of target lesions Sum of the largest sizes of target lesions and chains of lymph nodes along the short axis Number of lesions assessed All outbreaks Target lesions: no more than 10 in total (no more than 5 per organ) Target lesions: no more than 5 in total (no more than 2 per organ) Full answer Disappearance of all lesions for a period of at least four weeks Disappearance of all lesions for a period of at least four weeks. Short axis chain of lymph nodes less than 10 mm Partial answer ≥ 50% reduction from baseline or nadir for at least four weeks ≥ 30% reduction from baseline or nadir for at least four weeks No new measurable or unmeasurable lesions Stabilization Any variants that do not meet the criteria for progression and partial response Progression ≥ 25% increase from nadir ≥ 20% increase from nadir ≥ 20% increase compared to nadir (at least 5 mm) Appearance of new measurable lesions or progression of unmeasurable lesions Link WHO Handbook for Reporting Results of Cancer Treatment. World Health
Organization Offset Publication No. 48; Geneva (Switzerland), 1979. Cancer 1981; 47:
207-214 J.Natl. Cancer Inst. 2000; 92:205-216 Eur. J. Cancer. 2009; 45: 228-247

RECIST (Response Evaluation Criteria in Solid Tumors) - criteria for assessing the response of solid tumors

To assess tumor response to treatment in 1981. The World Health Organization (WHO) proposed criteria that included four response options: complete response (CR) - complete disappearance of all lesions for a period of at least 4 weeks; partial response (PR) - a decrease in all or individual tumor foci by more than 50% in the absence of progression of the rest; stable disease (SD) - a decrease of less than 50% or an increase of less than 25% in the absence of new lesions; progression (PD) - an increase of more than 25% in one or more lesions, or the appearance of new ones (WHO Handbook for Reporting Results of Cancer Treatment. World Health Organization Offset Publication No. 48; Geneva (Switzerland), 1979. Cancer 1981; 47: 207-214).

In 2000, in connection with the introduction of computed tomography (CT) into widespread clinical practice, the RECIST v criteria were proposed. 1.0, the concepts of measurable, nonmeasurable, and marker lesions were introduced (J. Natl. Cancer Inst. 2000; 92: 205–216).

Measurable lesions are lesions having an axial dimension greater than or equal to 10 cm on helical computed tomography with reconstruction slice thicknesses of 5 mm or greater, or equal to 20 mm on stepwise computed tomography with 10 mm slice thickness.

Not measurable lesions: small (up to 10 mm), bone metastases, leptomeningeal metastases, pericardial and pleural effusions, lesions with cystic or necrotic transformation, lymphogenous carcinomatosis of the skin and lungs.

Marker (target, target) foci - foci of the largest diameter (up to 5 per organ, up to 10 in total), suitable for accurate repeated measurement. The rest are non-marker. RECIST response criteria: complete response (CR) - complete disappearance of all marker lesions; partial response (PR) - a decrease by more than 30% in the sum of the maximum sizes of lesions; stable disease (SD) - a decrease of less than 30% or an increase of less than 20% in the absence of new lesions; progression (PD) - the appearance of a new lesion or an increase in the sum of the largest dimensions by more than 20%.

With the advent of targeted drugs, the nature of tumor dynamics during treatment has changed. In 2009, a modified version of the response criteria appeared - RECIST v 1.1 (Eur. J. Cancer. 2009; 45: 228-247), according to which, when analyzing CT scans, target lesions are identified - lesions of the largest diameter, suitable for accurate repeated measurement.

The emergence of a new generation of drugs aimed at the tumor microenvironment has led to the need to revise previously developed criteria for assessing the immediate results of tumor treatment. The main problems were related to the stabilization and progression of the disease. In cytotoxic chemotherapy, most investigators have not considered stabilization as an indicator of clinical response. At the same time, the mechanism of action of targeted and immunodrugs does not imply rapid cytoreduction, so long-term stabilization indicates the achievement of clinical success (J. Clin. Oncol. 2004; 22: 4442-4445. Annu. Rev. Immunol. 2009; 27: 83- 117).

Immuno-oncological drugs are currently the most intensively studied and promising group of drugs for the treatment of various tumors. Their targets are co-inhibitory molecules and their ligands. Most of the registered drugs are PD-1/PD-L1 inhibitors.

PD-1 is a co-inhibitory molecule that regulates the functions of components of the innate and adaptive immune response. It is expressed on the surface of T-lymphocytes, B-lymphocytes, MF (macrophages), monocytes, DC (dendritic cells). Under physiological conditions, it promotes the formation of tolerance to autoantigens, and in the tumor microenvironment - tumor immunological tolerance (Blood. 2009; 114 (8): 1537-1544). PD-1 ligand (PD-L1) is a transmembrane protein that is expressed on both tumor and immunocompetent cells (T, B - lymphocytes, DC, MF). PD-1/PD-L1 interaction leads to T cell deactivation, T regulatory cell activation, and tumor cell persistence (Immunogenetics. 2018; 70 (2): 73-86). PD-L1 expression on tumor cells is a predictor of favorable prognosis and a marker of sensitivity to chemotherapy. Expression of PD-L1 on lymphocytes in the microenvironment is a marker of sensitivity to blockers of co-inhibitory molecules and an indication for their use.

Currently, five drugs related to PD-1/PD-L1 inhibitors are used in the Russian Federation: Pembrolizumab, Nivolumab, Atezolizumab, Avelumab, Durvalumab.

Pembrolizumab (Keytruda) - IgG ₄ - anti-PD1 antibody. Registered in more than 60 countries around the world. It is used in patients with lung cancer, malignant melanoma and renal cell carcinoma. In some cases, it can be used for locally advanced and metastatic transitional cell cancer of the bladder in the second line after progression on platinum-containing chemotherapy regimens and in the first line if there are contraindications to the use of platinum drugs and the expression of PD-L1 on immune cells.

Atezlizumab (Tecentriq) is a humanized IgG ₁ anti-PD-L1 monoclonal antibody. It is used in patients with malignant melanoma, hepatocellular carcinoma, renal cell carcinoma, locally advanced and metastatic transitional cell carcinoma for the same indications as pembrolizumab. In some cases, by decision of the attending physician and the medical commission for other pathologies.

Nivolumab (Opdivo) is a humanized IgG ₄ anti-PD-1 monoclonal antibody. It is used in patients with malignant melanoma, renal cell and lung cancer.

Durvalumab (Infinzi) is a humanized IgG ₄ anti-PD-L1 monoclonal antibody. It is used in patients with locally advanced and metastatic transitional cell carcinoma in the second line, as well as in lung tumors.

Avelumab (Bavenzio) is a humanized IgG ₁ anti-PD-L1 monoclonal antibody. It is used in patients with locally advanced and metastatic transitional cell carcinoma in the second line and melanoma.

To date, several groups of methods are known to determine sensitivity to immuno-oncological drugs. Among them, two groups stand out. The first is associated with the assessment of the expression of the target molecule on the surface of tumor cells or cells of the microenvironment, as well as indicators reflecting the possibility of the formation of neoantigens: mutational load and microsatellite instability. The second is associated with assessing the balance of immunostimulatory and immunosuppressive components in the tumor microenvironment or in the blood.

A predictor of effectiveness and indication for the use of immuno-oncological drugs is the degree of PD-L1 expression on the surface of microenvironmental cells and/or tumor cells. Expression on more than 5% of microenvironmental cells or more than 10% of tumor cells is a predictor of treatment effectiveness (N. Engl. J. Med. 2016; 375 (18): 1767-1778). In some cases, PD-L1 expression is assessed using the 22C3 test system (DAKO PharmaDx), which takes into account the ratio of PD-L1 on tumor cells, lymphocytes and macrophages to the total number of detected tumor cells, multiplied by 100 (CPS, combined positive score). In this case, the predictor of effectiveness is CPS ≥ 10 (Lancet Oncol. 2020;21(1):44-59).

In 2004 - 2005, 200 experts from various specialties (oncologists, immunologists, radiologists) conducted a series of seminars on the use of immunotherapeutic agents in cancer patients, as a result of which a number of conclusions were made: 1) the antitumor effects of immunotherapeutic drugs take longer to realize than chemotherapy drugs; 2) response to immunotherapy without changing drugs can be observed after a period of progression; 3) it is not recommended to change tactics and stop immunotherapy if signs of progression are detected without confirmation during subsequent examinations according to RECIST v 1.1 criteria; 4) the appearance of new small lesions against the background of a decrease in existing ones should be considered as clinically insignificant progression; 5) prolonged stabilization should be considered as an indicator of clinical response (J. Immunother. 2007; 30: 1-15). Based on the recommendations of the working group, immune-specific related response criteria (irRC, immune-specific response criteria) were developed and validated on a group of 487 patients. As with the WHO criteria, the iRC used two measurements, new lesions were not considered a clear sign of progression and required confirmation in subsequent studies (Clin. Cancer Res. 2009; 15 (23): 7412–7420).

Measurements in two perpendicular planes are subject to great variability when different experts are involved. In this regard, Nishini N. et al. proposed modifying the iRC scale by replacing two perpendicular dimensions with one. The irRECIST (immune-related Response Evaluation Criteria in Solid Tumors) scale was created and validated on a group of 57 patients. The authors note the high concordance of both scales (Spearman correlation coefficient - 0.965), as well as better reproducibility of irRECIST (95% confidence interval -16.1%; 5.8%) compared with irRC (95% confidence interval -31.3%; 19.7%; Clin Cancer Res 2013;19(14):3936-3943).

Despite the clear advantages of iRECIST compared to RECIST v.1.1, it has not been validated in large populations and has not been widely used in clinical practice. The majority of studies that assessed the efficacy and safety of modern immunotherapies between 2010 and 2017 used RECIST v.1.1. Its disadvantages clearly emerged after pseudoprogression was described in a number of tumors - a phenomenon in which, during treatment, signs of disease progression appear, confirmed by instrumental methods, and with continued treatment without changing drugs, a partial or complete response is observed (World. J. Urol. 2018;36:1703-1709 J Clin Oncol 2015;33(31):3541-3543).

There is currently no exact explanation for the phenomenon of pseudoprogression. There are several hypotheses in the literature based on the interaction between the tumor and the immune system during treatment. The first is based on the fact that the immune system needs a lot of time to rebuild against the background of stimulation or a decrease in suppressor influences. This leads to clinical effects that are realized later than the interval between examination periods. The second hypothesis is based on the assumption that at the first stage, hyperinfiltration of tumor foci by immune cells is necessary for the clinical effect to occur. In a computed tomography study this is perceived as progression. This hypothesis has been confirmed in several studies where metastases were biopsied during progression. According to different authors, the frequency of pseudoprogression in patients with different nosological forms is from 1.3 to 9.1%, reaching 15 -17% in renal cell carcinoma and melanoma (Frontiers in Oncology. 2022; 25 (12): 982983).

The second problem that radiologists face is the immune-dissociated response, in which some lesions enlarge while others shrink. The frequency of this phenomenon is 3.3 - 9.2%. It may be associated with a different composition of the microenvironment in metastatic foci or with immune-mediated side effects, when conglomerates appear in parenchymal organs, similar in morphology to metastatic foci (sarcoidosis - similar reactions). Immune-dissociated response and pseudoprogression are the most serious obstacles to determining adequate treatment tactics, associated with continuing treatment as usual or prescribing other drugs (Eur. J. Nucl. Med. Mol. Imaging. 2020; 47(5):1158-1167 ).

The third problem that radiologists face when assessing the immediate results of treatment is the phenomenon of hyperprogression (frequency - from 4% to 29%), in which there is a sharp increase in the size and/or number of metastatic foci during the use of immuno-oncological drugs. It is most likely associated with overexpression of targets for immunotherapy on effector rather than immunosuppressive cells. There is no exact algorithm for the differential diagnosis of progression and hyperprogression. At the same time, therapeutic tactics must differ significantly. In case of progression, it is possible to continue treatment as before, taking into account the possibility of pseudo-progression, and in case of hyper-progression, urgent discontinuation of the drug is required (Trends Cancer. 2020; 6(3):181-191).

The RECIST Modification Working Group, which included oncologists, immunologists, radiologists and regulatory representatives, developed and created a validation plan for the iRECIST (Immunotherapy Response Evaluation Criteria in Solid Tumors) score between 2015 and 2016. The core parameters of iRECIST and RECIST are the same, but iRECIST introduced the concepts of “unconfirmed progression” (iUPD) and “confirmed progression” (iCPD) to identify cases of pseudoprogression. iUPD - newly identified signs of disease progression in the form of an increase in the size and/or number of measurable (target and non-target) and non-measurable lesions. To confirm progression, a repeat examination is required, which is carried out 4 to 8 weeks after the initial one. The exception is tumors in which, according to the literature, there is a high frequency of pseudoprogression (renal cell carcinoma, melanoma). In this case, a re-examination can be carried out at a later date (12 - 24 weeks). In case of stabilization (iSD, immunological stabilization), reduction (iPR, immunological partial response) or complete disappearance of newly identified lesions (iCR, immunological complete response), it is recommended to continue treatment as before. In cases where iCPD is detected during re-examination, it is recommended to change treatment tactics.

When using iRECIST, the frequency of examination is usually once every 6 to 12 weeks if the dynamics of the lesions are positive during treatment. In routine clinical practice, no more than five target lesions are assessed. Assessing larger volumes is common in scientific research (Lancet Oncol. 2017;18(3):e143-e152).

Hodi F.S. et al. in 2018 proposed another modified version of the response criteria for those cases when the effectiveness and safety of new generation immunotropic drugs are assessed - imRECIST (immune-Modified Response Evaluation Criteria in Solid Tumors; immunomodified response evaluation criteria in solid tumors; J. Clin. Oncol 2018;36:850-858). The model was created based on the analysis of data obtained when assessing the effectiveness and safety of atezolizumab in patients with lung, bladder, kidney and melanoma cancer.

In terms of lesion scoring criteria, imRECIST is generally similar to other single-measurement scores. Differences relate to the interpretation of the dynamics of measurable and non-measurable lesions, which ultimately leads to a more adequate definition of progression-free survival.

In addition to changes in size, the reversion of measured lesions into non-measured lesions and vice versa is taken into account. The disappearance of non-measurable lesions along with measurable lesions indicates a complete response to treatment. At the same time, their increase does not indicate the progression of the disease as a whole. As with other endpoints, confirmation of progression is required after 4–8 weeks. However, when stabilization, partial or complete response was achieved at re-evaluation, the fact of progression was not taken into account when calculating progression-free survival rates.

When comparing RECIST v 1.1 and imRECIST, the best overall response rate was 2% higher, the disease control rate was 8-13% higher, and the median progression-free survival was 1.5 months longer according to imRECIST criteria. At the same time, the median overall survival was longer when new non-targeted lesions appeared and was lower when using imRECIST criteria. Thus, the imRECIST criteria, compared with the standard ones, are more accurate in assessing long-term treatment outcomes (Table 2).

Table 2. Comparison of criteria for assessing immediate treatment results: irRC, irRECIST, iRECIST, imRECIST irRC, 2009 irRECIST, 2013 iRECIST, 2017 imRECIST, 2018 Evaluation method Two dimensions. Sum of perpendicular maximum sizes of lesions One dimension. Sum of largest sizes of target lesions Number of lesions assessed Target lesions: no more than 10 in total (no more than 5 per organ) Target lesions: no more than 5 in total (no more than 2 per organ) New measurable lesions Progression is not a criterion. The sizes of new lesions are summed up with pre-existing ones Response is assessed using criteria: unconfirmed progression (iUPD) and confirmed progression (iCPD) Progression is not a criterion. The sizes of new lesions are summed up with pre-existing ones New undetectable lesions New ones are not a criterion for progression. Disappearance (together with measurables) - criterion for complete response New ones are not a criterion for progression. Disappearance (together with measurables) - criterion for complete response Full answer Disappearance of measurable and non-measurable lesions Partial answer Reduction in the size of measured lesions by more than 50% compared to the baseline. No new measurable or unmeasurable lesions Reduction in the size of measured lesions by more than 30% compared to the baseline. No new measurable or unmeasurable lesions Stabilization Any variants that do not meet the criteria for progression and partial response Progression Increase in size of measurable lesions ≥ 25% compared to nadir Increase in size of measurable lesions ≥ 20% compared to nadir (minimum 5 mm) Confirmation of progression Yes Yes ¹ Yes ² Yes ³ Link Clin. Cancer Res. 2009; 15 (23): 7412-7420. Clin. Cancer Res. 2013; 19 (14): 3936-3943. Lancet Oncol. 2017; 18(3):e143-e152 J. Clin. Oncol. 2018; 36: 850-858 ¹ Requires confirmation between 4 and 12 weeks after initial progression;
² requires confirmation in the period 4-8 weeks after initial progression;
³ requires confirmation between 4 and 8 weeks after initial progression.
iUPD (immune unconfirmed progressive disease) - unconfirmed progression (immunotherapy);
iCPD (immune unconfirmed progressive disease) - confirmed progression (immunotherapy);
irRECIST (immune-related Response Evaluation Criteria in Solid Tumors) - immunity-related criteria for assessing the response of solid tumors;
iRECIST (Immunotherapy Response Evaluation Criteria in Solid Tumors) - immunotherapeutic criteria for assessing the response of solid tumors;
irRC (immune-specific related response criteria) - immunospecific response criteria;
imRECIST (immune-Modified Response Evaluation Criteria in Solid Tumors) - immunomodified response evaluation criteria in solid tumors

Despite the fact that methods for assessing the effectiveness of direct treatment based on linear measurements are currently standard, and their modern modifications make it possible to take into account the possibility of pseudo-progression, according to the majority of expert radiologists, they have a number of significant drawbacks.

Firstly, in all scales, target lesions are selected, and the decision about the nature of the clinical response (partial response, progression, stabilization) is mainly made on the basis of their dynamics. If there are multiple small and undetectable lesions, their total volume may be greater than the target volume. Therefore, long-term treatment results in this case are determined by the dynamics of the tumor mass, which is not included in the main measurements.

Secondly, the listed methods assume a high degree of subjectivity in the assessment. The boundaries of foci are assessed differently by researchers, especially in cases where they differ significantly in qualifications and experience. Such differences may have a particularly strong impact on the assessment of treatment results in borderline situations, when it is necessary to make a differential diagnosis between stabilization and progression.

Thirdly, the maximum size of the lesions may depend not only on the actual dynamics of the volume of the tumor mass, but also on the position of the patient during the study, as well as on respiratory movements.

Fourthly, measurements of linear dimensions and conclusions about the dynamics of tumor sizes are based on the assumption of the sphericity of metastatic foci and the uniformity of their growth. At the same time, modern research shows that this is not true. Most lesions have an irregular shape and do not grow evenly.

Fifthly, computed tomographic methods, as mentioned above, do not allow a clear distinction between progression and hyperprogression, and the differential diagnosis of pseudoprogression and progression requires a long time (4 - 6 weeks), which leads to a delay in choosing an adequate tactic, and, as a consequence, , worsening long-term treatment results.

There is a known group of methods for assessing the immediate results of tumor treatment associated with molecular visualization of metabolic processes - positron emission tomography (PET, Positron Emission Tomography). When using PET, the distribution of any tumor cell biomarkers or glucose uptake ( ¹⁸ F-FDG - PET) is assessed, which reflects the metabolic rate, and, as a consequence, the degree of tumor aggressiveness (Semin. Nucl. Med. 2007; 37: 400-419 ). When carrying out immunotherapy, the metabolism of tumor cells is disrupted at the first stage, and morphological changes are detected at the second stage. Currently, several methods have been proposed for assessing the immediate results of treatment, based on changes in tumor metabolism during treatment (Table 3).

Table 3. Metabolic criteria for immediate tumor response to treatment Lesion assessment CMR PMR SMD PMD Confirmation of progression New outbreaks EORTC, 1999 (Eur. J. Cancer. 1999; 35(13):1773-1782) Average SUV Disappearance of active lesions SUV reduction 15 - 25% after 1 treatment cycle and >25% with 2 or more treatment cycles SUV increase <25% or increase <15% Increase in SUV >25% or appearance of new lesions - P.D. PERCIST, 2009 (J. Nucl. Med. 2017; 58(9):1421-1428) Most active lesion assessed by SUL Disappearance of active measurable lesions Reduction of the SUL peak by 30% or more (at least 0.8 in absolute units) Failure to meet CMR, PMR, PMD criteria Peak SUL increase >30% (>0.8 in absolute units) - P.D. PECRIT, 2017 (J. Nucl. Med. 2017; 58(9):1421-1428) All FDG uptake lesions Disappearance of all foci and chains of lymph nodes <10 mm Reduction in sum of lesion diameters ≥ 30% Failure to meet PMR, PMD criteria. Maintain SUL change for at least 3 weeks Increase in sum of lesion diameters ≥ 20% - P.D. PERCIMT, 2018 (Eur. J. Nucl. Med. Mol. Imaging. 2018; 45(8):1289-1296) Round lesions with FDG uptake above hepatic or physiological Complete disappearance of active lesions Complete disappearance of some lesions. Lack of new Failure to meet CMR, PMR, PMD criteria ≥ 4 new lesions with functional diameter <1 cm; ≥ 3 new lesions with a functional diameter of 1 1.5 cm; ≥ 3 new lesions with a functional diameter ≥ 1.5 cm; - PD when at least 4 lesions appear imPERCIST5, 2019 (J. Nucl. Med. 2019; 60(3):335-341) Up to 5 lesions on CT with FDG uptake above physiological Complete disappearance of active lesions Reduce the sum of peak SUL values by 30% or more Failure to meet CMR, PMR, PMD criteria Increase in sum of peak SUL values by 30% or more - New lesions are included in the sum of maximum diameters iPERCIST, 2019 (Eur. J. Nucl. Med. Mol. Imaging. 2022;.49(7): 2323-2241) Like PERCIST Disappearance of active measurable lesions Like PERCIST Failure to meet CMR, PMR, PMD criteria ≥ 30% increase in peak SUL or appearance of new lesions UPMD Confirmation required in 4 - 8 weeks CMR (complete metabolic response) - complete metabolic response; PMR (partial metabolic response) - partial metabolic response; SMD (stable metabolic disease) - metabolic stabilization; PD (progressive disease) - progression; PMD (progressive metabolic disease) - metabolic progression; UPMD (unconfirmed PMD) - unconfirmed metabolic progression;
iCPD (immune unconfirmed progressive disease) - confirmed progression (immunotherapy); EORTC (European Organization for Research and Treatment Cancer) - European Organization for Research and Treatment of Cancer; PERCIST (Positron Emission Tomography (PET) Response Criteria in Solid Tumors) - PET criteria for assessing solid tumors; PECRIT (PET/KT Criteria for Early Prediction of Response to Immune Checkpoint Inhibitior Therapy) - PET/CT criteria for early prediction of response to checkpoint inhibitors; PERCIMT (PET Response Evaluation Criteria for immunotherapy) - PET criteria for assessing response to immunotherapy; imRECIST5 (immune-Modified Positron Emission Tomography (PET) Response Criteria in Solid Tumors) - immunomodified PET criteria for assessing the response of solid tumors; iPERCIST (immune Positron Emission Tomography Response Criteria in Solid Tumors) - immunological PET criteria for the assessment of solid tumors. SUV (standardized uptake value) - standardized uptake level; SUL (lean body mass-normalized standardized uptake value) - a standardized unit of absorption in terms of lean body mass.

In 1999, the European Organization for Research and Treatment of Cancer (EORTC) proposed metabolic response criteria based on 18F-FDG uptake and the SUV (standardized uptake value) criterion (Eur. J Cancer 1999;35(13):1773-1782). After 10 years, these criteria were transformed into PERCIST (Positron Emission Tomography (PET) Response Criteria in Solid Tumors; PET criteria for the evaluation of solid tumors), which were based on the use of SUL (lean body mass-normalized standardized uptake value; standardized absorption unit in in terms of lean body mass) (J. Nucl. Med. 2009; 50 (Suppl 1):122S-150S). Both methods have not received widespread clinical use, since correlations with complete and partial response according to computed tomography studies have not been obtained in large groups of patients (Med. 2020; 99(51): e23815).

In 2017, based on an analysis of the results of treatment of a small group of melanoma patients, it was proposed to combine the RECIST 1.1 and PERCIST criteria, as a result of which the PECRIT response criteria were developed (PET/KT Criteria for Early Prediction of Response to Immune Checkpoint Inhibitior Therapy, PET/CT criteria early prediction of response to checkpoint inhibitors), which demonstrated 100% sensitivity and 93% specificity in the study population (J. Nucl. Med. 2017; 58(9):1421–1428). The sensitivity and specificity of these criteria have not been further evaluated in larger populations.

In 2018, the PERCIMT (PET Response Evaluation Criteria for immunotherapy) criteria were proposed, which are modified EORTC criteria. They were developed on a group of patients with melanoma treated with ipilimumab. Compared with the EORTC, the PERCIMT criteria have better sensitivity (93.6%) and specificity (87.8%). The sensitivity and specificity of these criteria have not been further evaluated in larger populations (Eur. J. Nucl. Med. Mol. Imaging. 2018;45(8):1289-1296).

In 2019, imRECIST5 criteria (immune-Modified Positron Emission Tomography (PET) Response Criteria in Solid Tumors) were proposed, which are based on assessing the dynamics of the sum of peak SULs of five lesions. The method was developed by analyzing the results of treatment of 60 patients with melanoma. It demonstrated good results in assessing early response with subsequent correction of tactics, which led to a significant increase in overall survival rates by 35%. The sensitivity and specificity of these criteria have not been further evaluated in larger populations (J. Nucl. Med. 2019;60(3):335–341).

In 2019, the iPERCIST criteria (immune Positron Emission Tomography Response Criteria in Solid Tumors, immunological PET criteria for the evaluation of solid tumors) were proposed. The fundamental difference from other methods is the need to confirm progression within 4 to 8 weeks after the appearance of new lesions. The method was developed by analyzing the results of treatment of patients with non-small cell lung cancer. The sensitivity and specificity of these criteria have not been further evaluated in larger populations (EJNMMI Res. 2019;9:8).

A common limitation of the metabolic-based immediate response assessment methods being developed is that they target metabolically active tumors that represent only a small percentage of total malignancies. All proposed methods were developed in small groups of melanoma patients, and only iPERCIST was developed in a group with non-small cell lung cancer. These methods have not been validated in other pathologies characterized by less pronounced glucose uptake. Active metabolism is also characteristic of effector cells in the tumor microenvironment. Therefore, a decrease in glucose uptake cannot be clearly interpreted as a positive effect of treatment. In addition, this group of methods does not solve problems with pseudoprogression, hyperprogression and immuno-dissociated response. All these unresolved issues hinder the widespread implementation of metabolic response assessment methods into diagnostic standards (Eur. J. Nucl. Med. Mol. Imaging. 2022;.49(7): 2323-2241).

There is a known group of methods for assessing the immediate results of treatment associated with measuring the volume of tumor foci - volumetry, which can be carried out in automatic, semi-automatic or manual modes. With automatic volumetry, the calculations are performed entirely by computer. In the semi-automatic mode, after the volume calculation stage, manual correction occurs in order to accurately differentiate lesions located near anatomical structures. In manual delineation mode, lesions are identified by the researcher.

Volumetry is a significantly less subjective method compared to linear estimation methods. In addition, it can be used to estimate the half-doubling period of lesion volumes, which is an important prognostic factor. From a technical point of view, the main task of volumetry is the identification of voxels (a volumetric image element containing the value of a raster element in three-dimensional space; an analogue of two-dimensional pixels for three-dimensional space) located at the border of tissues with different x-ray tissue densities (approximately 500 Hounsfield units).

Despite the promise of volumetry, there are a number of technical difficulties that prevent the standardization and widespread implementation of this method in clinical practice.

1. Factors associated with the morphological characteristics of the lesions: the size of the lesions, their density, proximity to anatomical structures. With small linear sizes of lesions (less than 6 mm), the reproducibility of measurements decreases. This is due, first of all, to the ratio of the sizes of the voxel and the focus. Kuhnigk J.M. et al demonstrated that 41% of the tumor volume was contained in 27 voxels. The remaining voxels contain both tumor and normal tissue, making them difficult to assign to one category or another (IEEE Trans. Med. Imaging 2006;25(4):417-434). Differential density of lesions is also a factor that complicates volumetry, especially in cases where the automatic mode is used. Several studies have demonstrated that the variability of two examinations of the same anatomical region in the presence of lesions with uneven X-ray density increases the variability of volumetry by up to 50% (Eur. Radiol. 2015;25(4):1040-1047). Lesions located near the vessels and pleura are much more difficult to differentiate than intraparenchymal lesions. The variability of their assessment is four times higher. In these cases, manual contouring is usually required (Radiology. 2008;248(2):625-631).

2. Patient related factors - respiratory movements and emphysema. The patient's lung volume changes with each breathing movement, even if the breath is held during the examination. The volume estimate can vary from -19% to +26% (Eur. Radiol. 2009;19(4):800-808). Emphysema may also affect the accuracy of tumor lesion volumetry (Radiology. 2013;268(2):563-571).

3. Factors related to the technical aspects of the study: software, number and type of sensors, contrast agent, slice thickness and type of convolution core. Different manufacturers' software used for volumetry is based on different algorithms. De Hoop B. et al demonstrated that when using different reconstruction algorithms, the 95% confidence interval of measurements is: 16 - 22% (Eur. Radiol. 2009; 19(4):800-808). In another study, this figure reached 50% (Acta Radiol. 2014;55(6):691-698). Das M. et al found that the accuracy of volumetry differed significantly when using 16- and 64-slice tomographs (Eur. Radiol. 2007; 17(8):1979-1984). A more recent study found a difference in lesion evaluation (±18.2%) when using 64-slice devices from different manufacturers (Br. J. Radiol. 2013;86(1029):20130160). Honda O. et al demonstrated that the use of different contrast agents is accompanied by a change in the measured volume by 20% or more (Radiology. 2007; 245(3):881-887). These data were later confirmed in another study (Am. J. Roentgenol. 2010;195(1):149-154). Petrou M. co-authors published data according to which the volumes of lesions calculated using thin sections (1.25 mm) are 40% larger than when using thick ones (5 mm. Am. J. Roentgenol. 2007;188(2): 306-312). Similar data were obtained in more recent studies (Radiology. 2008;247(2):400-408. Acad. Radiol. 2013;20(2):173-180). The convolution kernel, used in reconstructing images from the original data, determines the relationship between spatial resolution and noise in the image. Noise limits contrast resolution and thus the ability to differentiate objects, whose attenuation may differ very little from that of the surrounding background. High-resolution convolution kernels improve spatial resolution but also disproportionately increase noise. Soft, or smoothing, kernels lead to a simultaneous reduction in noise and spatial resolution. A number of studies have shown that assessment accuracy and reproducibility in lesion assessment are better when high-resolution kernels are used (Radiology. 2007;245(3):881-887. Eur. Radiol. 2010;20(5):1180-1187).

In modern works devoted to the volumetry of tumor foci, attempts are made to solve two main problems that arise when using volumetry: 1) calculation of the measurement error, which arises due to the influence of a large number of factors listed above; 2) calculation of the cutoff level of volume changes, which allows differentiating stabilization from progression or partial response (World J. Clin. Onol. 2020;11(2):53-73).

Assessing the systematic measurement error in volumetry is a fundamental point in standardizing this method. In all works that present data comparing one- and two-dimensional measurements and volumetry, the measurement error is carried out, and some publications provide a mathematical apparatus for estimating it.

Xie X. et al., in a study with anthropomorphic phantoms on two different tomographs (Sensation 64, Siemens; Brilliance 64, Phillips) demonstrated that the systematic measurement error is less when using the semi-automatic mode (7.6%) compared to the manual mode (26.4 %). At the same time, no significant differences in volumetry between the two devices were detected, with the exception of a slice thickness of 3 mm. The 95% confidence interval for interscanner variability is ±41.6%; ± 18.2%; ± 4.9% for lesions 3, 5 and ≥ 8 mm. The 95% confidence interval of intrascanner variability is ±28.6%; ± 13.4%; ± 2.6% for lesions 3, 5, and ≥ 8 mm (Br. J. Radiol. 2013;86:20130160).

Doo K. et al studied the dependence of systematic error depending on the nature of image processing. In their work, they used phantoms with diameters of 5, 8, 10 and 12 mm, simulating ground glass lesions and dense lesions (100 Hounsfield units more). Various scanning parameters were used: 10, 20, 30 and 50 mAs. Various filters and iterative reconstruction were used to construct the image. The absolute percentage error for all nodes was significantly lower when using iterative reconstruction compared to filters (7.5 ± 4.7%; 7 ± 6.9%, respectively, p < 0.001). The positive effect of iterative reconstruction is more pronounced at small nodes (p<0.001). The absolute percentage error is significantly lower when processing ground glass nodes using iterative reconstruction compared to filters (2.9 ± 2%; 4.5 ± 2%, respectively, p < 0.001), and this difference is more pronounced at the minimum time scanning (11.8 ± 5.9%; 21.3 ± 6.1%, respectively, p < 0.001. Br. J. Radiol. 2014;87:20130644).

Gavrielides M.A. et al. studied the dependence of measurement error on the structure of the assessed lesions. The study used eight synthetic round-shaped phantoms with different sizes and densities of the central and peripheral regions (20 mm/10 mm; 10 mm/5 mm; 100 HU (Hounsfield units) / -630 HU; 10 HU / -630 HU; -630 HU /100 HU; -630 HU /-10 HU). The phantoms were placed in an anthropomorphic phantom, after which the study was carried out with different scanning parameters. The measurement error for slice thicknesses less than 5 mm ranged from -5.1% to 6.6% for the outer layer and from -12.6% to 5.7% for the inner region. According to the model built on the basis of one-dimensional analysis, the magnitude of the measurement error is reliably influenced by: lesion size, slice thickness, involution kernel (J. Med. Imag. 2016;3(1):013504).

Hwang E.J. et al developed and validated a model predicting the malignancy of lesions, including various volumetric parameters. The study included 50 patients (26 men and 24 women) with a mean age of 62.6 years (range 32 to 82 years), who were observed from November 2013 to April 2014. Inclusion criteria were: 1) the presence of metastatic foci in the lungs; 2) computed tomography examination without contrast enhancement; 3) absence of symptoms of pulmonary damage. The study included patients with renal cell cancer (13 people), thyroid cancer (8 people), lung cancer (6 people), colorectal cancer (4 people); with leiomyoma, head and neck tumors, hepatocellular cancer - 3 people each; with melanoma, gall bladder cancer, chondrosarcoma, thymus tumor, endometrial cancer, breast cancer, histiocytoma - 1 person each. In three patients the primary tumor was not identified. Each patient underwent two consecutive scans using a 64-slice tomograph (Brilliance 64, Philips, the Netherlands) on the back with full inspiration. Foci that met the following criteria were subjected to volumetry: 1) solid structures with a diameter of 4 to 15 mm. 2) absence of cavity and calcification; 3) detected during both scans during the study; 4) surgically removable. A total of 111 lesions that met these criteria were analyzed.

During the study, the authors revealed that the accuracy of measurements is influenced by 4 calculated parameters: absolute percentage error (APE), sphericity, surface voxel fraction (SVP), border voxel fraction (AP). The volume of lesions was calculated automatically. Absolute percentage error (APE) was calculated using the formula:

APE=|V ₁ - V ₂ |/ V _m x 100, where V ₁ and V ₂ are the volumes for two consecutive studies, V _m is the average value of the volume.

By the term “sphericity,” the authors mean the degree of proximity of the investigated source to a sphere. It took a value from 0 to 1.

The surface voxel fraction (SVP) is believed by the authors to play a key role in generating errors in the volumetry process. It was calculated using the formula:

SVP = number of surface voxels/total number of segment voxels.

The proportion of border voxels (AP) reflects errors associated with the differentiation of the lesion and surrounding structures. It was calculated using the formula:

AP = number of external voxels > - 400 HU / total number of external voxels. HU - Hounsfield sedimentation units.

To assess the diagnostic value of the model, the authors assessed the sensitivity and specificity of two different methods used to assess the growth of metastatic lesions: 1) 25% change in lesion volume (standard method); 2) individualized assessment using the 95% percentile of APE. Volumetry of indicator lesions was carried out in two consecutive studies. If the increase in volume exceeded the individual variability threshold corresponding to the 95% percentile of APE, the lesion was considered growing (positive test); if less, it was considered stable (negative test).

To validate the model, a group of patients was used that met the following selection criteria: 1) patients underwent pulmonary metastasectomy between January 2013 and October 2015; 2) within a year before surgery, patients underwent computed tomography examination of the lungs; 3) it was technically possible to compare the removed nodes with the X-ray picture. This group included 41 patients: 25 men and 16 women with a mean age of 52 years (range 16 to 77 years). Of the 82 computed tomography studies, 17 (20, 73%) were contrast-enhanced. The average time interval between two computed tomographic studies was 52 days, and between the second study and surgery was 11 days.

After surgical treatment and histological examination of the removed lesions, the results of the studies were classified as a positive gold standard (metastatic lesion) or a negative gold standard (non-tumor tissue). To estimate the 95% percentile of APE using univariate quantile regression, the following parameters were included as independent variables: mean lesion volume (V _m ), voxel size, SVP, AP, sphericity.

As a result of univariate analysis, it was revealed that for estimating the 95% percentile of APE, the significant parameters are: SVP, AP, V _m (Table 4: results of univariate quantile regression for estimating the 95% percentile of the percentage error). When conducting a multivariate analysis (multivariate quantile regression) to estimate the 95% percentile of APE, the significant parameters were SVP and AP - the proportion of border voxels. Regression formula:

95% percentile APE = 37.82 x SVP + 48.6 x AP - 10.87

Validation of the model, carried out after surgery, showed that the developed method has an advantage over the standard method in assessing progressive lesions in the lungs. The developed model allows us to differentiate metastatic and non-metastatic foci with high accuracy (Eur. Radiol. 2017;27:3257-3265).

Currently, there are three known ways to determine the boundaries of changes in the volume of lesions, allowing one to differentiate stabilization from progression or partial response. The first method is based on the assumption that the lesions are spherical. The radius of the focus is calculated as the average value of the radius of the maximum linear size of the focus and perpendicular to it. The volume is calculated using the standard formula: , where V is the volume of the source, r is the radius of the source. Threshold values for progression and partial response are calculated according to RECIST criteria: A 30% reduction in the sum of maximum dimensions (partial response) corresponds to a 65% reduction in tumor volume; an increase in the sum of the maximum dimensions by 20% (progression) corresponds to an increase in tumor volume by 73% (Radiology. 2002; 225:416-419. Acta Radiol. 2004; 11:1355-1360. Radiol. Med. 2006; 111:365-375 ). The second method is associated with the calculation of the effective diameter of the lesion, which is calculated automatically during the volumetry process as the distance between two voxels with the same characteristics (Acta Radiol. 2004; 11:1355-1360). The third method involves the use of special mathematical methods to assess the boundary differentiating stabilization, partial response and progression (Eur. Radiol. 2007;17:2561-2571. Cancer Imaging. 2015;15:17).

Prasad S.R. et al published the results of a study that compared measurements of one, two lesion sizes, and calculated volumetric data to evaluate immediate treatment outcomes. Of the 38 patients included in the study, 12 based on one size measurement and 13 based on two measurements obtained discordant results with respect to volumetry. In 6 patients with a partial response based on one or two measurements after volumetry, the response was classified as stable. In 2 patients with stabilization according to linear measurements, a partial response was stated according to volumetric data. In 4 patients with progression based on linear measurements, the response was changed to stabilization after volumetric assessment. The authors concluded that volumetry had superior classification power compared with linear measurements (Radiology. 2002;225:416-419). Several other groups of researchers later came to similar conclusions (Acta Radiol. 2004; 11:1355-1360. Radiol. Med. 2006; 111:365-375).

Dicken V. et al conducted a study that compared the effectiveness of different methods of measuring lesion size in assessing the immediate results of treatment. The study was conducted on a group of patients with metastatic renal cell cancer (174 people) who received immunotherapy with interferon-α drugs as part of the second phase of clinical trials. A total of 726 computed tomographic studies were performed. During the work, the diameters of the lesions were measured using three algorithms: a) manual measurement of the maximum sizes of the lesions; b) automatic measurement of dimensions; c) estimation of the effective diameter. According to the authors, the use of effective diameter to assess dynamics is characterized by the maximum number of agreements with the participation of two researchers (87.6%) compared to automatic (86.2%) and manual measurements (63.9%). In addition, of all measurement options, only the effective diameter makes it possible to accurately estimate the minimum dynamics of lesion sizes (10% or 1 mm). The authors conclude that it is necessary to integrate this type of measurement into modern scales for assessing immediate treatment outcomes (Acta Radiol. 2004; 11:1355-1360).

Marten K. et al published the results of a study that calculated cutoff values for lesion volume changes using the Bland-Altman method (Stat. Meth. Med. Res. 1999; 8:135-160). It involved 50 patients with various tumors, including esophageal cancer (13 people), non-small cell lung cancer (10 people), colorectal cancer (7 people), melanoma (3 people), transitional cell bladder cancer (3 people) , chondrosarcoma (2 people), renal cell carcinoma (2 people) and one person each with cancer of the stomach, pharynx, breast, histiocytoma, leiomyosarcoma, osteoarcoma, chorionic carcinoma, pleural sarcoma, synovial sarcoma, and sarcoma of unknown etiology, who had 202 metastatic foci were assessed (from 1 to 5 in each patient).

The median volume before treatment was 182.22 mm ³ (from 3.16 to 5159.13 mm ³ ) and 124.79 mm ³ (from 0 to 6359.23 mm ³ ) after treatment. The study performed automated volumetry by two technicians. In 7 solid nodes in two series of measurements (1.7%), manual correction was required due to their location near anatomical structures (vessels, pleura). The variability of measurements when comparing the results obtained by both specialists ranged from -17.9 to 13.2% for the first and from -11.1 to 4.8% for the second. The variability for two measurements by one specialist over a short time interval ranged from - 16.9 to 13.2% for the first and from - 21.3% to 3.9% for the second.

During the study, at the first stage, the immediate results of treatment were assessed using various boundaries defining the area of stabilization: from 70% to 20% with a step of 5%. At the second stage, using the Bland-Altman method, the optimal value of the boundaries characterizing the stabilization of the disease was determined.

A total of 400 assessments (2 researchers, 2 studies) of the immediate results of treatment were carried out on 50 patients. Using 70% as a stabilization limit, disease progression was observed in 13 patients, stabilization in 28, partial response in 7, and complete response in 2. As the margin decreased, the number of patients with stabilization decreased, and those with progression and partial response increased. When using the 20% level as a boundary, only 9 patients (a decrease of 67.9%) showed stabilization, 21 (an increase of 61.5%) had progression, and 18 (an increase of 157.1%) had a partial response. When using the cutoff, the 60% and 65% response classifications by both examiners were identical in the two series of measurements.

At the second stage, the researchers calculated the optimal assessment limit (≥ 35%). Moreover, with the number of metastases ≥ 3, greater agreement is achieved in the classification of immediate treatment results (stabilization, progression, partial response) with RECIST (Eur. Radiol. 2007;17:2561-2571).

Beaumont H. et al conducted a study based on data from the Quantitative Imaging Biomarker Alliance (QIBA) and using a multiparametric mathematical model based on the Geary-transformation to evaluate the optimal cutoff for determining stabilization, partial response, and progression with volumetry. Hinkel (Radiology. 2011;258:906-914. Biometrika. 1969;56:635-639. Manage Sci. 1975;21:1338-41. Am Stat. 1993;47:259-64).

The model was developed using 10,000 lesion volumetric samples from the Numerical Biomarker Imaging Alliance database and validated in a real-life retrospective clinical cohort of 10 patients enrolled in multicenter studies with data from 5 or more consecutive CT scans.

The thickness of the scans varied from 1 to 7 mm. The most common scan was performed at a thickness of 5 mm (87%). The study was conducted on 10 tomographs from various manufacturers (Siemens, General Electric, Toshiba). Voxel sizes varied from 0.5 to 1 mm. The median, minimum and maximum values of the linear sizes of lesions in the group were 40 mm, 11 mm and 117 mm, respectively. Volume dynamics were assessed in semi-automatic mode; After the automatic volumetry stage, manual contouring correction was carried out. A total of 426 lesions were assessed. In 68 cases, a partial response to treatment was diagnosed, in 67 - progression; in other cases, stabilization. The mean lesion volume in the study was 29.446 cm ³ , the median was 30.719 cm ³ (range 0.195 cm ³ to 380.976 cm ³ ). The sizes of most lesions varied from 1 to 100 cm ³ .

The limits of stabilization during volumetry calculated according to the model were -35%; +55%. At the same time, the corresponding limits for the effective diameter sizes were -13%; +16%. Progression of the disease was observed when the volume increased by 55% or the effective diameter increased by 16%; partial answer - with a decrease of 35% and 16%, respectively. Model fit was assessed using Monte Carlo simulation (R Foundation for Statistical Computing; 2011, p. 409). In a real clinical group, 97.1% of estimates agreed with the proposed model (Cancer Imaging. 2015;15:17).

Our center has developed a method for assessing the immediate results of treatment of disseminated forms of renal cell cancer, based on the volumetry of metastatic foci in the lungs (Pat. 2751086 Russian Federation, A61B 6/03. Method for assessing the immediate results of treatment of renal cell cancer / O.E. Molchanov , D.N.Maistrenko, D.A.Granov, E.V.Rozengauz, D.V.Nesterov (RF) // Application 10/9/2019; Published 12/28/2020. Application No. 2020143529). In the proposed method, the volume of metastatic lung damage in a patient is calculated using the formulas:

V ₁ = ; V ₂ = ;

where V ₁ is the total volume of the initial metastatic lesion, i is the number of the lesion in the initial study, vi is the volume of the lesion in the initial study, n1 is the number of lesions in the initial study, V ₂ is the total volume of the metastatic lesion after the treatment cycle, j is the number of the lesion in study after the treatment cycle, vj is the volume of the lesion during the study after the treatment cycle, n2 is the number of lesions after the treatment cycle;

select the largest (Vmax) and smallest (Vmin) values from these volumes and calculate the value of the volume functions taking into account the system error using the formulas:

_Fmax = +183/d to + 49);

F _min = +183/d l + 49),

where, _Fmax is the value of the function of the largest total volume of metastatic lesion in two consecutive measurements, F _min is the value of the function of the smallest total volume of metastatic lesion in two consecutive measurements, - lesion volume in the measurement with the largest total volume of metastatic lesion in two consecutive measurements, - volume of the lesion in the dimension with the smallest total volume of metastatic lesion in two consecutive measurements, nmax - number of metastatic foci in the dimension with the largest total volume of metastatic lesion in two consecutive measurements, nmin - number of metastatic foci in the dimension with the smallest total volume of metastatic lesion in two consecutive measurements measurements, d to - diameter of the lesion in the dimension with the largest total volume of metastatic lesion in two consecutive measurements, d l - diameter of the lesion in the dimension with the smallest total volume of metastatic lesion in two consecutive measurements, - cut thickness,

and if F_max-F_min> 0 and V₁ >V₂, then this is regarded as a partial response to treatment if F_max- F_min> 0 and V₁ < V_2,then this is regarded as progression if F_max- F_min< 0 for any ratio V₁ and V₂ - as stabilization of the process.

The method makes it possible to increase the accuracy of assessing the immediate results of treatment of patients with disseminated renal cell cancer by minimizing the error associated with the technical aspects of the study (slice thickness, software, type of tomograph) and eliminating the need to calculate the limit values of volume changes for the differential diagnosis of progression, stabilization and partial answer. The limitations of this method are due to the fact that it has been validated only in a group of patients with renal cell cancer. It allows one to differentiate well between progression and pseudo-progression, but cannot serve as a criterion for correcting tactics if hyper-progression and an immuno-dissociated response are suspected.

Models for assessing immediate treatment outcomes based on lesion volumetry have a number of advantages compared to criteria based on linear measurements. They make it possible to minimize the subjectivity of the assessment (increasing the reproducibility of the result) and evaluate the dynamics of non-spherical foci. But despite this, volume estimation models have a number of disadvantages. Firstly, in the models proposed to date for calculating the limits differentiating stabilization, progression and partial response, there is no assessment of the influence of slice thickness, involution kernel and the software used. Conversely, studies that evaluate technical error use data obtained by recalculating the RECIST criteria to assume that lesions are spherical. This creates conditions for significant errors when performing volumetry and assessing the immediate results of treatment. Secondly, most models take into account the volumes of only a small number of lesions. But even in those few studies where full volumetry with manual correction is carried out, the accuracy of measurements increases sharply with a decrease in the number and an increase in the size of the assessed metastases. This, of course, negatively affects the diagnostic value of this approach. According to a number of authors, including ours, small lesions are characterized by the highest growth rate. Thirdly, the proposed models do not trace the relationship between immediate treatment results and long-term ones in comparison with criteria that use linear measurements, while similar comparisons were made among the latter (imRECIST. Clin. Oncol. 2018; 36: 850-858).

Another group of methods for assessing the immediate results of treatment is associated with the search for molecular biomarkers that would help, with greater accuracy than radiological methods, to differentiate true progression from pseudoprogression, hyperprogression and an immune-dissociated response. The proposed invention also belongs to this group. There are indications in the literature of many molecular biological components that could be used to monitor the treatment of cancer patients. These involve the assessment of histological, serological markers, circulating immune cells, and the use of “liquid biopsy” techniques, which evaluate circulating DNA and/or circulating tumor cells. Most publications are related to the theoretical substantiation of the possibility of using certain markers in cancer patients (Clin. Cancer Res. 2017; 23 (17): 5074 - 5081. J. Hematol. Oncol. 2014: 70).

Four areas of research are fundamentally considered:

- Assessment of the biological potential of the tumor:

a. Biomarkers of tumor cells - PD-L1, PD-L2, tumor-associated antigens. Markers of subpopulations of tumor-infiltrating lymphocytes: CD3, CD4, CD8, PD-1, CTLA, CD25, CD45RO, CD11b, CD57, CD68, LAG-3.

b. Genomic markers: neoantigens, microsatellite instability genes. Genome of tumor-infiltrating lymphocytes: deep sequence of T and B receptors.

c. Transcriptomic and epigenomic analysis of tumor cells and tumor-infiltrating lymphocytes.

d. Evaluation of functional markers: Ki-67, granzyme B.

- Study of blood cells and DNA:

a. Tumor biomarkers: long and short DNA chains (ctDNA, ctRNA, miRNA), exosomes. Biomarkers of tumor-infiltrating lymphocytes: miRNA, autoantibodies, cytokines and chemokines, HLA haplotype.

b. Assessment of circulating tumor and immunocompetent cells.

- Flow cytometry:

a. Phenotypic and functional characteristics of immune cells: activated (CD3, CD4, CD8, Ki-67, CTLA-4, PD-1, LAG-3, ICOS, TIM-3); regulatory (CD3, CD4, CD25, FOXP3, CD127, Ki-67, CD45 RA); memory cells (CD45 RO); myeloid cells of suppressor origin (CD14+ HLADR ^low ).

b. Detection and functional characterization of circulating tumor cells (CTC).

- Antigen-specific T cell monitoring: binding, clonal activation, T cell receptor and HLA characterization.

The closest to the proposed method is the method of assessing the immediate results of treatment using the RECIST v 1.1 criteria (Eur. J. Cancer. 2009; 45: 228-247), according to which, when analyzing CT scans, targeted lesions are identified - lesions of the largest diameter, suitable for accurate repeated measurement . In total, no more than five lesions are assessed (no more than two per organ). Target lesions must be at least 10 mm in diameter in organs, at least 10 mm in length in case of bone lesions (lytic or mixed lesions) and at least 15 mm in case of lymph node lesions. The remaining foci are classified as non-targeted. To assess the size of metastases in parenchymal organs and bones, the largest diameter of the lesion in the axial plane is used, and for lymph nodes, the size of the lesion along the short axis.

RECIST v 1.1 overall response categories: complete response (CR), partial response (PR), stable (SD), progression (PD), not evaluable (NE).

The dynamics of target lesions are assessed by calculating the sum of their maximum diameters (and the smallest diameter in case of damage to the lymph nodes):

1. Complete response (CR) - disappearance of all extralymphatic target lesions; reduction of all pathological lymph nodes to a short-axis size of less than 10 mm.

2. Partial response (PR) - a decrease in the sum of the maximum diameters of lesions by at least 30% compared to the primary study.

3. Progression (PD) - an increase in the sum of maximum diameters by at least 20% compared to the smallest sum of maximum diameters (nadir) calculated in any of the studies (including the primary one). The sum of the maximum diameters should increase by at least 5 mm.

4. Stabilization (SD) - absence of dynamics in the sum of maximum diameters, as well as its increase or decrease that does not meet the PR or PD criteria.

Assessment of non-targeted lesions:

1. Complete response (CR) - disappearance of all extralymphatic non-targeted lesions. Reduction of all pathological lymph nodes to a short-axis size of less than 10 mm. Normalization of tumor markers.

2. Incomplete response/lack of progression - persistence of at least one non-target lesion or persistence of tumor markers above normal.

3. Progression (PD) - reliable progression of existing non-targeted lesions.

The overall response to treatment is assessed by comparing the dynamics of targeted and non-targeted lesions (Table 4): assessment of the immediate results of tumor treatment according to RECIST v 1.1 criteria).

Table 4. Assessment of immediate results of tumor treatment according to RECIST v 1.1 criteria. Targeted foci Non-targeted lesions New outbreaks General answer CR CR - CR CR not CR/not PD - PR CR NE - PR PR Not PD or NE - PR SD Not PD or NE - SD P.D. any +/- P.D. Any P.D. +/- P.D. Any Any + P.D.

CR - complete response, PR - partial response, SD - stabilization, PD - progression, NE - cannot be determined.

Thus, the prototype method, like the similar standard methods described above for assessing the effectiveness of direct treatment, based on linear measurements, have a number of disadvantages and do not have sufficient information content and accuracy.

The technical result of the present invention is to increase the accuracy of assessing the immediate results of treatment of patients with disseminated forms of malignant tumors through the use of immunological parameters of the blood and microenvironment. This result is achieved by the fact that in the known method of assessing the immediate results of treatment, including computed tomographic assessment of metastatic foci, according to the invention, Additionally, the level of spontaneous production of IL-8 (IL-8-KR1, IL-8-M) and IL-10 (IL-10-KR1, IL-10) is determined in the patient’s blood (CR) and in the tumor microenvironment (M). -M) and induced production of IL-12 (IL-12-KP1, IL-12-M) before the start of each cycle of treatment with immuno-oncological drugs, and also evaluate the same parameters in the blood (IL-8-KP2, IL-10-KP2 , IL-12-KR2) after each treatment cycle and based on the obtained values, the coefficients are calculated:

K1= IL-8-M/ IL-8-KP1 - IL-8-M/ IL-8-KP2;

K2 = IL-10-M/ IL-10-KP1 - IL-10-M/ IL-10-KP2;

K3 = IL-12-M/ IL-12-KP2 - IL-12-M/ IL-12-KP1;

and if the value of at least one of the coefficients is equal to or higher than 1.7, this is regarded as an early sign of progression.

Having been professionally treating patients with RCC for a number of years, we assessed the long-term and immediate results of treatment of patients, using various evaluation criteria, including RECIST and RECIST v 1.1, various immunotherapy options, as well as fundamental and clinical aspects of the interaction of the tumor and the immune system. At the first stage, we preliminary assessed the dynamics of immunological parameters with various immediate treatment results. Since 2016, immuno-oncological drugs from the group of PD-1 and PD-L1 inhibitors began to appear in clinical practice: Pembrolizumab, Nivolumab, Atezolizumab, Avelumab, Durvalumab. Indications for their use are the expression of target molecules (PD-1/PD-L1) in tumor tissue or, in some cases, even without confirmation of expression, when this is assumed by treatment standards.

We studied the dynamics of immunological parameters in the blood, as well as the characteristics of the tumor microenvironment in patients with different immediate treatment results.

Since 2017, the immune status and content of immunocompetent cells in the tumor microenvironment have been studied in patients with disseminated forms of various tumors (renal cell cancer, bladder cancer, melanoma, breast cancer) in order to assess the prognostic and predictive significance of various parameters. The content of lymphocytes, their subpopulations and cytokines in the blood of patients was assessed. The analyzes were carried out in the immunological laboratory of the Federal State Healthcare Institution VTsERM named after. A.M.Nikiforov Ministry of Emergency Situations of Russia on a laser flow cytometer Cytomics FC 500 (BECKMAN COULTER Inc., USA) using monoclonal antibodies and consumables from BECKMAN COULTER Inc., IMMUNOTECH S.A.S., Protein Contour LLC and Cytokin LLC. The range of studied parameters and their reference intervals are reflected in tables 5 (peripheral blood lymphocyte subpopulations) and 6 (cytokine profile parameters).

Table 5. Subpopulations of peripheral blood lymphocytes Phenotype Lymphocyte subpopulation Reference interval (blood). Relative content (%) Reference interval (blood). Absolute content (x 10 ⁹ /l) CD3+CD16- Mature T lymphocytes 52 - 76 950 - 1800 CD3+CD8+ Cytotoxic lymphocytes 23 - 40 450 - 850 CD3+CD4+ T helper cells 31 - 46 570 - 1100 CD4+CD8+ Double-positive T cells 0.1 - 1.1 5 - 15 CD3-CD16+CD56+ Natural killers 9 - 19 180 - 420 CD3-CD8+; Activated natural killer cells 1.5 - 6 18 - 150 CD16+CD56+HLA DR+ 0 - 5 0 - 120 CD3+CD16+CD56+ TNK - Cells 0.1 - 8 5 - 200 CD3 -CD19+ B lymphocytes 6 - 18 150 - 450 CD4+CD25+CD127-FOXP3 T regulatory cells (Treg) 0.3 - 10 0 - 110 CD3+ HLA DR+ Activated T cells 0 - 5 0 - 120 HLADR+ Activated antigen presenting cells 6 - 22 150 - 550 αβ -T Alpha/beta T cells 60 - 80 925 - 1965 γδ - T Gamma/delta T cells 2 - 7 1 - 115 LIN-HLADR-CD33 ⁺ CD66b ^- CD14 ⁺ CD15 ^- Myeloid-derived suppressor cells 0.2 - 11 0 - 120

Table 6. Cytokine profile parameters Cytokine Reference interval Spontaneous production Induced production Serum concentration Interleukin-1β (IL-1, pg/ml) 0 - 50 1000 - 5000 0 - 50 Interleukin-2 (IL-2, pg/ml) 0 - 5 10 - 100 0 Interleukin-4 (IL-4, pg/ml) 0 - 50 100 - 400 0 - 50 Interleukin-6 (IL-6, pg/ml) 0 - 50 1000 - 3000 0 - 50 Interleukin-8 (IL-8, pg/ml) 0 - 100 1000 - 5000 0 - 50 Interleukin-10 (IL-10, pg/ml) 0 - 50 100 -400 0 - 50 Interleukin-12(IL-12, pg/ml) 0 - 50 100 - 600 0 - 50 Interferon-α (IFN-α, pg/ml) 0 - 50 100 - 500 0 - 50 Interferon-γ (IFN-γ, pg/ml) 0 - 50 1000 - 5000 0 - 50 Tumor necrosis factor-α (TNF-α, pg/ml) 0 - 50 500 - 1500 0 - 50 Transforming growth factor-β (TGF-β, pg/ml) 0 - 50 500 - 1000 0 - 20

The level of lymphocytes and their subpopulations was assessed in peripheral blood. Cytokine profile analysis was carried out in two stages. At the first stage, serum concentration was determined. The second involved a study in cell culture, where their potential was determined by assessing spontaneous and induced production of cytokines. The assessment of lymphocytes, their subpopulations and cytokine profiles was carried out at the same time of day to avoid the influence of circadian fluctuations.

In the tumor microenvironment, the immunological profile was assessed once, which is due to the availability of material for analysis. Reference intervals for determining the concentration of immunocompetent cells, as well as the concentration, spontaneous and induced production of cytokines in the microenvironment currently do not exist.

Tumor material with a volume of about 10 mm ³ was placed in two bottles, one of which contained the DMEM-F12 nutrient medium to determine spontaneous production, and the other contained the same nutrient medium with a complex of polyclonal activators (phytohemagglutinin 4 μg/ml, concanavalin A 4 μg /ml, lipopolysaccharide 2 µg/ml). Next, incubation occurred at 37°C for 72 hours. Tumor cells were pelleted by centrifugation at 2000 rpm for 15 min. After sedimentation of cells, the concentration of cytokines and growth factors was determined using the enzyme immunoassay method.

The algorithm for developing a method for assessing the immediate results of treatment of patients with disseminated forms of tumors included three stages. The values of the indicators were studied simultaneously with the evaluation of the results according to RECIST 1.1. The conclusion about the nature of the response to treatment was confirmed during long-term observation of patients. If the results of direct assessment subsequently turned out to be erroneous, a comparison of the dynamics of immunological parameters was carried out taking into account the newly obtained data. At the first stage, a univariate analysis was carried out, where all immunological parameters that were assessed in the tumor microenvironment before treatment, as well as in peripheral blood before treatment and after each cycle were included as variables (Tables 5, 6). At the second stage, a multivariate analysis was carried out, in which mutually influencing components were excluded. At the third stage, coefficients reflecting the ratio of intratumoral cytokine production and production in the patient’s lymphocyte culture were included in the multivariate analysis. The values of the coefficients K1, K2, K2 were determined by the methods of limiting parameter values and the accuracy of the model. The results of univariate and multivariate analysis are shown in Table 7.

SP - spontaneous production; IP - induced production.

To develop the method, data from 79 patients with different nosological forms were used, including renal cell carcinoma (17 patients), melanoma (18 patients), triple-negative breast cancer (21 patients), and invasive bladder cancer (23 patients). For validation, a group of clinically followed patients (77 people) with the same nosological forms was used: renal cell (22 patients), melanoma (12 patients), triple negative breast cancer (17 patients), invasive bladder cancer (26 patients) ( table 8). The validation algorithm included assessment of immediate treatment outcomes using RECIST 1.1 criteria; assessment of the immediate results of treatment using the developed method, comparison of the results of both methods taking into account data from subsequent clinical observation (Table 8).

Table 8. Discrepancies between the assessment of immediate treatment results according to RECIST 1.1 criteria and using the developed method Groups Response to treatment RECIST 1.1.,
N Error, K (%) Designed
method, N Error, K (%) Development, 79 people
(448 definition of immediate result) RCC,
17 people P.D. 23 6 (26.1%) 46 - SD + PR + CR 84 31 (36.9%) 61 - RMP, 23 people P.D. 37 12 (32.4%) 48 - SD + PR + CR 79 14 (17.7%) 68 - Melanoma, 18 people P.D. 32 8 (25%) 44 - SD + PR + CR 78 26 (33.3%) 66 - TNBC, 21 people P.D. 28 9 (32.1%) 37 - SD + PR + CR 87 22(25.3%) 78 - Validation, 77 people (288 determinations of immediate results) RCC, 22 people P.D. 21 6 (28.6%) 29 1(3.4%) SD + PR + CR 78 12(15.4%) 60 2(3.3%) RMP, 26 people P.D. 24 7(29.2%) 37 3(8.1%) SD + PR + CR 65 9(13.8%) 52 3(5.8%) Melanoma, 12 people P.D. 10 2(20%) 19 0 SD + PR + CR 39 7(17.9%) thirty 1(3.3%) TNBC, 17 people P.D. 15 4(26.7%) 23 1(4.3%) SD + PR + CR 36 11(30.6%) 28 0

CR - complete response, PR - partial response, SD - stabilization, PD - progression, RCC - renal cell carcinoma, RBC - bladder cancer, TNBC - triple negative breast cancer, N - number of assessments of immediate treatment results, K - number errors.

In the validation group (77 people, 288 studies), the frequency of errors in assessing immediate treatment outcomes using RECIST 1.1 criteria. amounted to 20.1%; in the subgroup with disease progression - 27.1%; in the subgroup with an objective response (the sum of complete, partial responses and stabilization) - 17.9%. The frequency of errors in assessing the immediate results of treatment using the developed method was 3.8% in the group as a whole; in the subgroup with disease progression - 4.6%; in the subgroup with an objective response -3.3%, which is significantly less compared to RECIST v 1.1.

The essence of the method is as follows.

For a patient with a disseminated tumor, before treatment in the tumor microenvironment (M) and in the blood (CR), the level of spontaneous production of IL-8 (IL-8-CR1, IL-8-M) and IL-10 (IL-10-CR1 , IL-10-M) and induced production of IL-12 (IL-12-KP1, IL-12-M), after each cycle of treatment and evaluate the same parameters in the blood (IL-8-KP2, IL-10-KP2 , IL-12-KP2), then the coefficients K1 = IL-8-M/ IL-8-KP1 - IL-8-M/ IL-8-KP2 are calculated from the obtained values; K2 = IL-10-M/ IL-10-KP1 - IL-10-M/ IL-10-KP2; K3 = IL-10-M/ IL-10-KP2 - IL-10-M/ IL-10-KP1. If at least one of the coefficients is greater than 1.7, then this is regarded as an early sign of progression. In parallel, the patient’s immediate treatment results are assessed using the RECIST v criteria. 1.1.

Next, the patient receives treatment according to accepted standards. The data obtained using the proposed method can be taken into account when adjusting treatment tactics according to the decision of the attending physician or medical commission.

The method has a number of significant advantages compared to known analogues:

1. Allows more accurate assessment of the immediate results of treatment of cancer patients with immuno-oncological drugs.

2. Based on the use of modern achievements of molecular biology and cellular immunology.

3. Allows you to increase the effectiveness of treatment by choosing the most adequate treatment option.

A method for assessing the immediate results of treatment of patients with disseminated forms of tumors was developed in the group of molecular biological prognosis and individualization of treatment of the Federal State Budgetary Institution "Russian Scientific Center for Radiological and Chemical Technologies named after. acad. A.M. Granov" of the Ministry of Health of the Russian Federation and has now undergone clinical testing in 156 patients with a positive result.

Claims (5)

A method for assessing the sensitivity of a tumor to immuno-oncological drugs, characterized in that additionally the level of spontaneous production of IL-8 (IL-8-KP1, IL-8-M) and IL-8 is determined in the patient’s blood (CR) and in the tumor microenvironment (M). 10 (IL-10-KP1, IL-10-M) and induced production of IL-12 (IL-12-KP1, IL-12-M) before the start of each cycle of treatment with immuno-oncological drugs, and also evaluate the same parameters in the blood ( IL-8-KR2, IL-10-KR2, IL-12-KR2) after each treatment cycle and the obtained values are used to calculate the coefficients: K1= IL-8-M/ IL-8-KP1 - IL-8-M/ IL-8-KP2; K2 = IL-10-M/ IL-10-KP1 - IL-10-M/ IL-10-KP2; K3 = IL-12-M/ IL-12-KP2 - IL-12-M/ IL-12-KP1; and if the value of at least one of the coefficients is equal to or higher than 1.7, this is regarded as an early sign of progression.