Non-Hodgkin lymphomas (NHL) are a significant problem for modern healthcare systems. According to the International Agency for Research on Cancer of the World Health Organization, as of 2022, NHL ranked 9th among all malignant neoplasms in terms of incidence and 11th in terms of mortality [1]. Moreover, according to forecasts, the incidence will continue to increase in the coming years [2]. In 2024, in Russia, there were 7.83 new cases of NHL per 100,000 population, amounting to 17,000 new patients. Diffuse large B-cell lymphoma (DLBCL) accounts for 30–40% of NHL; follicular lymphoma (FL) accounts for 20–30% [3, 4]. The median age at diagnosis is 58–59 years. Thus, a significant proportion of patients with NHL are still of working age at diagnosis, and increasing their life expectancy and quality of life will affect not only social but also economic indicators.

The provision of cancer care in the Russian Federation is carried out within the framework of the Federal Project “Fight Against Cancer” [5], which is part of the National Project “Long and Active Life” [6]. The goal of the national project is “to increase life expectancy to 78 years by 2030 and to 81 years by 2036, including an accelerated increase in healthy life expectancy” [6]. Overall, the project’s measures are expected to increase average life expectancy by 2.38 years, of which 0.17 years should come from cancer control. Another key performance indicator of the Federal Program is to achieve a “Proportion of persons surviving 5 years or more from the date of diagnosis of a malignant neoplasm” of 67% by 2030 [5]. Thus, increasing the life expectancy of the population is one of the key requirements for the healthcare system as a whole. This fully applies to the oncology service, for which increasing overall patient survival is a key task.

An analysis of the factors that contributed to increased life expectancy in the United States over a 25-year period (1990–2015) showed that at least 35% of the increase was provided by the introduction of biopharmaceutical drugs into widespread medical practice [7]. The most common representatives of this group of drugs in oncohematology are monoclonal antibodies and products derived from them. For example, the addition of the immunobiological drug rituximab, a monoclonal antibody to CD20, to chemotherapy for NHL played a key role in its time [8]. At the same time, a number of studies have described the low efficacy of rituximab-containing therapy regimens in various NHL, including DLBCL and FL, in certain groups of patients at high risk of progression or in cases of relapse/refractoriness to the therapy [9, 10, 11]. Thus, there is a need to continuously expand the range of therapy regimens used through innovative drugs that can significantly increase treatment efficacy. In recent years, the list of such drugs available for use in the Russian Federation has expanded significantly. In particular, the drugs polatuzumab vedotin [16], used in the first [12] and subsequent [13] lines of DLBCL therapy, and obinutuzumab, used in the treatment of FL [14, 15, 17] both in the first and later lines, have been registered and included in the List of Vital and Essential Medicines (VED) in the Russian Federation.

In the event of NHL relapse after the use of immunobiological drugs in the first or second line, only chemotherapy drugs, which have insufficiently high efficacy and poor tolerability, remain available for patients in the third and subsequent lines in current therapeutic practice. In this situation, in addition to the routine rotation of various low‑efficacy chemotherapy regimens in late‑stage NHL, bifunctional (bispecific) antibodies have become increasingly widespread in recent years [18]. Examples include glofitamab, used in the treatment of DLBCL [19], and mosunetuzumab, used in follicular lymphoma [20].

Both drugs are bispecific monoclonal antibodies that bind the glycoprotein CD20 on the surface of B‑cells and the glycoprotein CD3 as part of the T‑cell receptor complex on the surface of T‑lymphocytes [18]. By simultaneously binding CD20 on the surface of B‑cells and CD3 on the surface of T‑cells, these drugs promote the formation of an immunological synapse, followed by potent activation and proliferation of T‑cells, secretion of cytokines, and release of cytolytic proteins, resulting in lysis of CD20‑expressing B‑cells [18].

Since the cost of innovative drugs exceeds that of routine immunochemotherapy, in real‑world practice patients are often treated with the most affordable and less expensive therapy first, and innovative drugs are pushed to later lines of therapy. This does not allow the full potential of innovative therapy to be realized in preserving patient life. Therefore, it is relevant to evaluate the impact of rational, earliest possible use of innovative drugs, supplemented by the use of bispecific antibodies in the third line, on overall patient survival and, consequently, on achieving the target values of the performance indicators of the national project “Long and Active Life”.

A Markov model of progression‑free survival and overall survival was constructed for three lines of therapy. The time horizon was 25 years, with a cycle length of 1 month. Clinical efficacy data were obtained from published randomized controlled trials. For each regimen, Kaplan‑Meier curves of overall survival and progression‑free survival were digitized using specialized software WebPlotDigitizer ver. 5.2. This yielded a two‑dimensional array of “time from study start” – “proportion of patients without event”. Next, individual patient data were reconstructed (pseudo‑dataset creation) using the methodology of Guyot P. et al. [26]. This method was recommended by NICE DSU [27] for constructing partitioned survival models when access to primary clinical trial data is unavailable. Pseudo‑dataset generation was performed in the statistical computing environment R (version 4.4.2) using the R‑Studio interface (version 2024.12.1–563) and the IPDfromKM library [28].

Separate parametric survival modelling for progression‑free survival was then performed for each intervention using the SurvHE library. Parametric modelling used maximum likelihood estimation (MLE). Model fit was assessed using the Akaike information criterion (AIC) and the Bayesian information criterion (BIC), along with visual inspection of the obtained parametric curves. The following distributions were used for parametric models: exponential, Weibull, Gompertz, log‑logistic, log‑normal, and generalized gamma. The best‑fitting distributions were used to model patient survival over the analysis horizon of 25 years.

To estimate patient distribution across lines of therapy, a Markov model was constructed (Fig. 1). The Markov cycle length was 1 month. The model estimated the probability of a patient being in the following states:

Overall survival of NHL patients receiving three lines of therapy was estimated for different combinations of regimens. Two regimen sequences were selected for comparison for FL and DLBCL. The first used regimens with early introduction of innovative therapy in the first line and the use of innovative bispecific antibodies in the third line. The comparator used regimens with non‑optimal use of innovative therapy – in the second line only and without the use of bispecific antibodies. In this case, rituximab‑based therapy was used in the first and third lines (Table 1).

After lines 1–3, progression meant transition to the next line of therapy. After the third line, overall survival after progression on third line was considered. Thus, patient time on lines 1–3 is determined by progression‑free survival. The terminal state in the model is death.

Progression‑free survival data were estimated from the results of randomized controlled trials for each regimen combination. Reference sources for overall and progression‑free survival values are given in Table 1.

Table 1. Therapy regimen alternation schemes by lines taken into account in the analysis

Line of therapyOptimal regimenNon‑optimal regimenDiffuse large B‑cell lymphoma  1st linePola+R-CHP (polatuzumab vedotin + rituximab + cyclophosphamide + doxorubicin) [12]R-CHOP (rituximab + vincristine + cyclophosphamide + doxorubicin) [12]2nd lineR-GemOx (rituximab + gemcitabine + oxaliplatin) [21]Pola-BR (polatuzumab vedotin + bendamustine + rituximab) [13]3rd lineGlofi (glofitamab) [22]R-GemOx (rituximab + gemcitabine + oxaliplatin) [21]Follicular lymphoma  1st lineG-CHOP/Benda/CVP (obinutuzumab + vincristine + cyclophosphamide + doxorubicin + prednisolone / bendamustine / vincristine + cyclophosphamide + prednisolone) [14]R-CHOP/Benda/CVP (rituximab + vincristine + cyclophosphamide + doxorubicin + prednisolone / bendamustine / vincristine + cyclophosphamide + prednisolone) [14]2nd lineR2 (lenalidomide + rituximab) [23]GB (obinutuzumab + bendamustine) [15]3rd lineMosun (mosunetuzumab) [25]R-DHAP (rituximab + dexamethasone + cytarabine + cisplatin) [24]

Fig. 1. Markov model for estimating the probability of a patient being in different states

Overall survival was calculated as the ratio of the sum of the proportion of surviving patients in each Markov cycle to the number of Markov cycles per year. Five‑year survival was defined as the maximum number of living patients across all lines of therapy 5 years from the start of the analysis.

As described in Materials and methods, modelling was performed for the following regimen sequences:

For both DLBCL and FL, two regimen sequences were compared. In the first (optimal) strategy, an innovative drug was used in the first line (polatuzumab vedotin for DLBCL and obinutuzumab for FL), rituximab‑containing therapy in the second line, and bispecific antibodies (glofitamab and mosunetuzumab, respectively) in the third line. The alternative (non‑optimal) strategy consisted of rituximab‑containing therapy in the first line, an innovative drug in the second line (polatuzumab vedotin and obinutuzumab, respectively), and a different rituximab‑containing regimen not previously used in the third line.

The overall survival results for the stated regimen sequences are shown in Figs. 2 and 3. As can be seen, the use of innovative therapy with polatuzumab vedotin in the first line for DLBCL, followed by the bispecific antibody glofitamab, increased overall survival to 13.9 years. This is 4 years or 40.4% higher than the weighted mean survival with the non‑optimal regimen sequence, which was 9.9 years. Five‑year survival (Table 2) also increased by 5.6%, from 68.0% with the non‑optimal regimen to 73.6% with early use of innovative therapy.

In the treatment of follicular lymphoma, the non‑optimal regimen sequence provided a weighted mean survival of 17.0 years. The use of obinutuzumab in the first line and mosunetuzumab in the third line increased this by 5.2 years (30.8%) to 22.3 years. Notably, patients receiving mosunetuzumab in the third line had prolonged survival (3.3 years) after completing 3 lines of therapy. Five‑year survival for the considered regimen sequences was 91.2% and 97.3%, respectively (see Table 2). Rational use of innovative therapy added 6.1% to this indicator.

Table 2. Proportion of individuals surviving 5 years or more from the date of diagnosis of malignant neoplasm, %

NosologyTarget value for 2030Non‑optimal regimenOptimal regimenDiffuse large B‑cell lymphoma67.0%68.0%73.6%Follicular lymphoma–91.2%97.3%

Fig. 2. Weighted survival of patients with large B-cell lymphoma who received 3 lines of therapy when comparing different combinations of treatment regimens

Legend (years):Optimal regimen (Pola+R-CHP/R-GemOx/Glofi):

Non‑optimal regimen (R-CHOP/Pola-BR/R-GemOx):

Fig. 3. Weighted mean survival of patients with follicular lymphoma who received 3 lines of therapy when comparing different combinations of therapy regimens

Legend (years):Optimal regimen (*G-CHOP/Benda/CVP → R2 → Mosun*):

Non‑optimal regimen (R-CHOP/Benda/CVP → GB → R-DHAP):

Fig. 4. Weighted survival time of patients with large B-cell lymphoma and follicular lymphoma who received 3 lines of therapy at diagnosis at age 59 years

Legend:

The obtained overall survival values allow us to estimate the impact of using the optimal regimen sequence on the patient’s total life expectancy. At a median diagnostic age of 59 years and with DLBCL and FL proportions of 35% and 25% of total NHL, respectively, the weighted average life expectancy of patients with DLBCL and FL was calculated. In the case of delayed use of innovative therapy, the patient’s life expectancy could reach 71.9 years, which is 1.5 years lower than the average life expectancy in 2023. At the same time, the earliest possible use of innovative drugs and the use of bispecific antibodies in the third line increases life expectancy by 4.5 years to 76.4 years. This value exceeds the life expectancy target for Russia as a whole for 2026 and is only 1.6 years lower than the target for 2030.

Increasing overall life expectancy and five‑year survival from the date of diagnosis are key performance indicators of the National Project “Long and Active Life” and the Federal Project “Fight Against Cancer” implemented within its framework. The analysis showed that rational prescribing of innovative therapy has a significant impact on these indicators. Unfortunately, in current socioeconomic realities, cancer patients often receive not the most effective therapy in the first lines, but the most accessible one, while highly effective regimens are postponed to later lines. It must be taken into account that the later innovative therapy is prescribed, the lower its contribution to overall survival. For example, the analysis shows that when polatuzumab vedotin is used in the first line for DLBCL, it provides a weighted mean progression‑free survival of 11.7 years, or 82.5% of overall survival. When the same drug is used in the second line after a rituximab‑containing regimen, the estimated weighted mean progression‑free survival on the drug is 1.9 years, or 19.6% of overall survival. Thus, delayed use of innovative therapy does not allow its full therapeutic potential to be realized.

The second factor that influenced the achievement of the stated performance indicators is effective therapy in the third line. For example, mosunetuzumab is used as a fixed course and therefore has fixed costs, but it has a significant impact on patient survival after completing third‑line therapy. In the analysis, the total estimated survival after 2 lines of therapy (3rd line plus overall survival after receiving 3rd line) in patients who received mosunetuzumab was 5.9 years, which was 2.3 times higher than the same indicator when rituximab‑containing therapy was used in the third line (2.6 years).

The estimation of overall life expectancy showed that the use of innovative therapy in the first line immediately after diagnosis, as well as the use of bispecific antibodies in later lines, can potentially extend the life of an NHL patient to the level of the average life expectancy in the Russian Federation calculated for the entire population. Thus, rational use of innovative therapy actually allows a patient with oncohematological disease to achieve a life expectancy comparable to the national average.

Thus, the analysis shows that the application of the principles of rational pharmacotherapy in choosing treatment regimens for NHL patients has a significant impact on achieving the performance indicators of public health programs. The introduction of innovative therapy and the earliest possible use of highly effective regimens can increase five‑year survival for DLBCL patients from 68.0% to 73.6%. Overall survival would then increase by 4 years or 40.4% – from 9.9 years to 13.9 years. In FL patients, five‑year survival would increase from 91.2% to 97.3%, and overall survival would increase by 5.2 years or 30.8% – from 17.0 to 22.3 years. At a median diagnostic age of 59 years, the use of the optimal regimen sequence would provide patients with non‑Hodgkin lymphomas a life expectancy comparable to the targets set within the national project “Long and Active Life”.